Immunologic Approaches to the Classification and Management of Lymphomas and Leukemias

Cancer Treatment and Research

WILLIAM L MCGUIRE, series editor

Livingston RB (ed): Lung Cancer 1. 1981. ISBN 90-247-2394-9. Humphrey G Bennett, Dehner LP, Grindey GB, Acton RT (eds): Pediatric Oncology 1. 1981.

ISBN 90-247-2408-2. DeCosse 11, Sherlock P (eds): Gastrointestinal Cancer 1. 1981. ISBN 90-247-2461-9. Bennett 1M (ed): Lymphomas 1, including Hodgkin's Disease. 1981. ISBN 90-247-2479-1. Bloomfield CD (ed): Adult Leukemias 1. 1982. ISBN 90-247-2478-3. Paulson DF (ed): Genitourinary Cancer 1. 1982. ISBN 90-247-2480-5. Muggia FM (ed): Cancer Chemotherapy 1. ISBN 90-247-2713-8. Humphrey G Bennett, Grindey GB (eds): Pancreatic Tumors in Children. 1982.

ISBN 90-247-2702-2. Costanzi 11 (ed): Malignant Melanoma 1. 1983. ISBN 90-247-2706-5. Griffiths CT, Fuller AF (eds): Gynecologic Oncology. 1983. ISBN 0-89838-555-5. Greco AF (ed): Biology and Management of Lung Cancer. 1983. ISBN 0-89838-554-7. Walker MD (ed): Oncology of the Nervous System. 1983. ISBN 0-89838-567-9. Higby Dl (ed): Supportive Care in Cancer Therapy. 1983. ISBN 0-89838-569-5. Herberman RB (ed): Basic and Clinical Tumor Immunology. 1983. ISBN 0-89838-579-2. Baker LH (ed): Soft Tissue Sarcomas. 1983. ISBN 0-89838-584-9. Bennett 1M (ed): Controversies in the Management of Lymphomas. 1983. ISBN 0-89838-586-5. Humphrey G Bennett, Grindey GB (eds): Adrenal and Endocrine Tumors in Children. 1983.

ISBN 0-89838-590-3. DeCosse 11, Sherlock P (eds): Clinical Management of Gastrointestinal Cancer. 1984.

ISBN 0-89838-601-2. Catalona WI, RatliffTL (eds): Urologic Oncology. 1984. ISBN 0-89838-628-4. Santen RJ, Manni A (eds): Diagnosis and Management of Endocrine-Related Tumors. 1984.

ISBN 0-89838-636-5. Costanzi 11 (ed): Clinical Management of Malignant Melanoma. 1984. ISBN 0-89838-656-X. Wolf GT (ed): Head and Neck Oncology. 1984. ISBN 0-89838-657-8. Alberts DS, Surwit EA (eds): Ovarian Cancer. 1985. ISBN 0-89838-676-4. Muggia FM (ed): Experimental and Clinical Progress in Cancer Chemotherapy. 1985.

ISBN 0-89838-679-9. Higby DJ (ed): The Cancer Patient and Supportive Care. 1985. ISBN 0-89838-690-X. Bloomfield CD (ed): Chronic and Acute Leukemias in Adults. 1985. ISBN 0-89838-702-7. Herberman RB (ed): Cancer Immunology: Innovative Approaches to Therapy. 1986.

ISBN 0-89838-757-4. Hansen HH (ed): Lung Cancer: Basic and Clinical Aspects. 1986. ISBN 0-89838-763-9. Pinedo HM, Verweij 1 (eds): Clinical Management of Soft Tissue Sarcomas. 1986.

ISBN 0-89838-808-2. Higby DJ (ed): Issues in Supportive Care of Cancer Patients. 1986. ISBN 0-89838-816-3. Surwit EA, Alberts DS (eds): Cervix Cancer. 1987. ISBN 0-89838-822-8. lacobs C (ed): Cancers of the Head and Neck. 1987. ISBN 0-89838-825-2. MacDonald IS (ed): Gastrointestinal Oncology. 1987. ISBN 0-89838-829-5. RatliffTL, Catalona WI (eds): Genitourinary Cancer. 1987. ISBN 0-89838-830-9. Nathanson L (ed): Basic and Clinical Aspects of Malignant Melanoma. 1987.

ISBN 0-89838-856-2. Muggia FM (ed): Concepts, Clinical Developments, and Therapeutic Advances in Cancer

Chemotherapy. 1987. ISBN 0-89838-879-5. Frankel AE (cd): Immunotoxins. 1988. ISBN 0-89838-984-4. Bennett 1M, Foon KA (eds): Immunologic Approaches to the Classification and Management

of Lymphomas and Leukemias. 1988. ISBN 0-89838-355-2.

Immunologic Approaches to the Classification and Management of Lymphomas and Leukemias

edited by

JOHN M. BENNETT University of Rochester Cancer Center Rochester, New York 14642 USA

and

KENNETH A. FOON Department of Internal Medicine University of Michigan Ann Arbor, Michigan 48109 USA

..... 1988 KLUWER ACADEMIC PUBLISHERS •• BOSTON / DORDRECHT / LANCASTER ."

Distributors

for North America: Kluwer Academic Publishers, 101 Philip Drive, As­sinippi Park, Norwell, MA 02061, USA for the UK and Ireland: Kluwer Academic Publishers, Falcon House, Queen Square, Lancaster LAI lRN, UK for all other countries: Kluwer Academic Publishers Group, Distribution Centre, P.O. Box 322, 3300 AH Dordrecht, The Netherlands

Library of Congress Cataloging in Publication Data

Immunologic approaches to the classification and managment of lymphomas and leukemias I edited by John M. Bennett and Kenneth A. Foon.

p. cm. - (Cancer treatment and research) Includes bibliographies and index. ISBN-13: 978-14612-89654 e-ISBN-13: 978-1-4613-1713-5 DOl: 10.1007/978-1-4613-1713-5

1. Lymphomas-Immunological aspects. 2. Leukemia-Immunological aspects. 3. Lymphomas-Classification. 4. Leukemia-Classification. 5. Immunodiagnosis. 6. Antibodies, Monoclonal-Therapeutic use. I. Bennett, John M., 1933- II. Foon, Kenneth A. III. Series.

[DNLM: 1. Leukemia-classification. 2. Leukemia-therapy. 3. Lymphoma-classification. 4. Lymphoma-therapy. WI CA693/WH 525 1325] RC280.L9L953 1988 616.99'442079-dc19 87-31255 DNLM/DLC for Library of Congress

Copyright

© 1988 by Kluwer Academic Publishers Softcover reprint of the hardcover 1 st edition 1988 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, mechanical, photocopying, recording, or otherwise, without the prior written permission of the publishers, Kluwer Academic Publishers, 101 Philip Drive, Assinippi Park, Norwell, MA 02061, USA.

Table of Contents

Foreword to the Series Vll

Preface ix

List of Contributors Xl

1. Immunologic approaches to the classification of lymphomas and lymphoid leukemias K.A. FOON, R.P. GALE, and R.F. TODD III 1

2. Immunologic approaches to the classification of non-Hodgkin's lymphomas T.M. GROGAN, C.M. SPIER, L.C. RICHTER, and C.S. RANGEL 31

3. Detection of central nervous system involvement in patients with leukemia or non-Hodgkin's lymphoma by immunological marker analysis of cerebrospinal fluid cells H. HOOIJKAAS, H.l. ADRIAANSEN, and 1.1.M. VAN DONGEN 149

4. Detection of residual disease in acute leukemia using immunological markers D.H. RYAN and 1.1.M. VAN DONGEN 173

5. Radioimmunoscintigraphy of lymphoma with monoclonal antibodies 1.A. CARRASQUILLO and S.M. LARSON 209

6. Radiolabeled antibodies in Hodgkin's disease S.E. ORDER 223

7. Interferon therapy for lymphoproliferative disorders M.S. ROTH, P.A. BUNN, and K.A. FOON 231

VI

8. Monoclonal antibody therapy of lymphomas and leukemia M.S. KAMINSKI and K.A. FOON 253

9. Autologous bone marrow transplantation in acute leukemia and lymphoma following ex vivo treatment with monoclonal antibodies and complement A.S . FREEDMAN, T. TAKVORIAN, L.M. NADLER, K.C. ANDERSON, S.E. SALLAN, and J . RITZ 265

Index 285

Cancer Treatment and Research

Foreword

Where do you begin to look for a recent, authoritative article on the diagnosis or management of a particular malignancy? The few general on­cology textbooks are generally out of date. Single papers in specialized journals are informative but seldom comprehensive; these are more often preliminary reports on a very limited number of patients. Certain general journals frequently publish good indepth reviews of cancer topics, and published symposium lectures are often the best overviews available. Un­fortunately, these reviews and supplements appear sporadically, and the reader can never be sure when a topic of special interest will be covered.

Cancer Treatment and Research is a series of authoritative volumes which aim to meet this need. It is an attempt to establish a critical mass of oncology literature covering virtually all oncology topics, revised frequently to keep the coverage up to date, easily available on a single library shelf or by a single personal subscription.

We have approached the problem in the following fashion. First, by dividing the oncology literature into specific subdivisions such as lung cancer, genitourinary cancer, pediatric oncology, etc. Second, by asking eminent authorities in each of these areas to edit a volume on the specific topic on an annual or biannual basis. Each topic and tumor type is covered in a volume appearing frequently and predictably, discussing current diagnosis, staging, markers, all forms of treatment modalities, basic biology, and more.

In Cancer Treatment and Research, we have an outstanding group of editors, each having made a major commitment to bring to this new series the very best literature in his or her field. Kluwer Academic Publishers has made an equally major commitment to the rapid publication of high quality books, and world-wide distribution.

Where can you go to find quickly a recent authoritative article on any major oncology problem? We hope that Cancer Treatment and Research provides an answer.

WILLIAM L. MCGUIRE

Series Editor

Preface

Within the past decade advances in immunology have led to the identifica­tion of precise stages of lymphocyte differentiation: of both Band T subsets. Classifications of lymphoma, strictly on an immunologic basis, now parallel the more traditional morphologic schemes. Molecular probes that identify immunoglobulin gene rearrangement and T-cell receptor genes are readily available. Highly specific monoclonal antibodies, defining cell surface anti­gens, can be utilized for 'marrow purging' prior to autologous transplantation and for specific therapy, either alone or coupled with drugs, toxins, or radio­nuclides. In this volume we have addressed major diagnostic and therapeutic issues that confront investigators in the 1980s. We have focused on disorders of lymphocytes because of the explosive knowledge that has developed in this important arena of oncology.

In Chapter 1, Foon, Gale, and Todd provide the latest information on the monoclonality of lymphomas and leukemias of lymphoid origin. The nomenclature refers to clusters of differentiation (CD) whenever possible. The tables are of major help in providing investigators with guidelines for classification purposes. Chapter 2, by Grogan and coworkers at Arizona, represents a monumental and successful effort to define all of the non­Hodgkin's lymphomas by immunophenotype. It serves as a valuable re­source for experimental and clinical pathologists.

Hooijkaas and associates in Chapter 3, have provided an excellent over­view of the power of detection of CNS involvement by neoplastic cells in lymphomas and leukemias. Their discussion includes the traditional morpho­logic approaches (cytocentrifuge), immunologic marker analysis, cytogenetics, flow cytometry, and gene rearrangements. The problem of overinterpreta­tion of CALLA + cells is emphasized. TdT reactivity appears to be an important, if not necessary, ingredient of CNS involvement with lymphoid cells. Clearly, morphologic techniques have great limitation in being able to detect significant numbers of residual malignant cells in marrow aspirates. Can immunologic markers be utilized in 'early detection'? In Chapter 4, Ryan and van Dongen present data to suggest that detection of a leukemic cell burden of 109 cells may be possible. Assay systems include TdT and CD 10 (CALLA). Of interest is a comparison of techniques in both Rochester

x

and Rotterdam suggesting that a double immunoflourescence assay may be more sensitive than a flow cytometry assay.

In Chapter 5, Carrasquillo and Larson from NIH have brought their considerable radiological skills together with an informative treatise on a new diagnostic field, radioimmunoscintigraphy (RIS). The most widely used antibody has been TIOl for imaging T-cell lymphomas with a high specifi­city. The potential for using therapeutic isotopes (1 131 ) is discussed. Intra­venous as well as intralymphatic injections are presented.

The second half of our volume begins with a compact presentation by Professor Order (Chapter 6) on the first successful treatment of Hodgkin's disease with radiolabeled antibodies. A new concept of a 'biologic window' is presented. Utilizing 30 mCi of 1-131 antiferritin, 40% of refractory patients responded. Potential successor treatments with 90-Yttrium antiferritin and autologous marrow rescue are presented.

In Chapter 7, a concise review of the status of interferon treatment for lymphoproliferative disorders is presented by Roth, Foon, and Bunn. In addition, the recent success of interferons in certain myeloproliferative dis­eases is highlighted. Considerable information is provided on mechanisms of action of the various interferon species.

Kaminski and Foon, in Chapter 8, elaborate the several different types of monoclonal antibodies that have had demonstrable activity in clinical trials. Anti-idiotype as well as more general antibodies are discussed, including the problems and toxicities associated with such therapy.

One final chapter (9), by Freedman and coworkers from the Dana-Farber Cancer Institute, provides an excellent overview of the current status of autologous bone marrow transplantation with in vitro treatment of bone marrow with 'J5 and J2' monoclonal antibodies. Survival data indicate an approximate 30% disease-free survival at five years, similar to allogeneic results. Similar studies with anti-Bl antibody for relapsed non-Hodgkin's lymphoma patients are presented as well. Preliminary results for the treat­ment of refractory AML appear promising.

John M. Bennett Kenneth A. Foon

editors

List of contributors

ANDERSON, KC, Dana-Farber Cancer Institute , 44 Binney Street, Boston, MA 02115

ADRIAANSEN, HJ, Department of Immunology, Erasmus University Rotterdam, PO Box 1738, 3000 DR Rotterdam, The Netherlands

BUNN, PA, Division of Medical Oncology, University of Colorado Health Science Center, 4200 East 9th Avenue, Denver, CO 80439

CARRASQUILLO, JA, Bldg 10, Room lC490, National Institutes of Health, 9000 Rockville Pike, Bethesda, MD 20205

FOON, KA, Division of Clinical Immunology, State University of New York at Buffalo, Roswell Park Memorial Institute, 666 Elm Street, Buffalo, NY 14263

FREEDMAN, AS, Dana-Farber Cancer Institute, 44 Binney Street, Boston, MA 02115

GALE, RP, University of California, Center for Health Sciences, Los Angeles, CA 90024

GROGAN, TM, Department of Pathology, University of Arizona, Tucson, AZ 85724

HOOIJKAAS, H, Department of Immunology, Academic Hospital Rotter­dam Dijkzigt, Dr. Molewaterplein 40, 3015 GD Rotterdam, The Netherlands

KAMINSKI, M, Division Hematology/Oncology, University of Michigan, Ann Arbor, MI 48109

LARSON, SM, Nuclear Medicine Department, Clinical Center, National Institutes of Health, Bethesda, MD 20205

NADLER, LM, Dana-Farber Cancer Institute, 44 Binney Street, Boston, MA 02115

ORDER, SE, The Johns Hopkins Oncology Center, Department of Radia­tion Oncology, Baltimore, MD 21205

RANGEL, CS, Department of Pathology, University of Arizona, Tucson, AZ 85724

RICHTER, LC, University of Arizona, Pathology Department, Tucson, AZ 85724

RITZ, J, Dana-Farber Cancer Institute, Division Tumor Immunology, 44 Binney Street, Boston, MA 02115

XII

ROTH, MS, Simpson Memorial Institute, University of Michigan, Ann Arbor, MI 48109

RYAN, DH, Pathology/Lab Medicine, University of Rochester Medical Center, 601 Elmwood Avenue, Rochester, NY 14642

SALLAN, SE, Dana-Farber Cancer Institute, 44 Binney Street, Boston, MA 02115

SPIER, CM, Department of Pathology, University of Arizona, Tucson, AZ 85724

TAKVORIAN, T, Dana-Farber Cancer Institute, 44 Binney Street, Boston, MA 02115

TODD, RF III, Simpson Memorial Institute, 102 Observatory, Ann Arbor, MI48109

VAN DONGEN, JJM, Department of Immunology, Erasmus University Rotterdam, PO Box 1738, 3000 DR Rotterdam, The Netherlands

Immunologic Approaches to the Classification and Management of Lymphomas and Leukemias

1. Immunologic approaches to the classification of lymphomas and lymphoid leukemias

Kenneth A. Foon, Robe rt P. Gale, and Robert F. Todd , III

Recent advances in immunology have led to important insights into lympho­cyte differentiation and the cellular origin of lymphoma and lymphoid leu­kemia. It is now possible to precisely define stages of human lymphocyte differentiation utilizing highly specifi c monoclonal antibodies that defi ne cell surface antigens and molecul ar probes that ident ify rea rrangement of im­munoglobulin and T-cell receptor genes. These can be combined with more traditional cell markers such as surface membrane (Smlg) and cytoplasmic immunoglobulin (Clg) on B-Iymphocytes , sheep erythrocyte receptors on T­lymphocytes, and cytochemical stains. Advances in the classification of lymphoma and the lymphoid leukemias and their importance in our under­standing normal lymphoid differentiation and therapeu tic implications are summarized in this chapter.

Cell mar kers

8-lymphocytes

B-l ymphocytes are usually ident ified by the presence of Smlg. Progenitors of B-Iymphocytes , com monly referred to as pre-B-cells are present in feta l liver and normal bone marrow; the cells display cytoplasmic !-l-heavy chain (e

l,)

but lack in tracytoplasmic light chain and Smlg. B-Iymphocytes and pre-B­lymphocytes may also have receptors for the thi rd component of comple­ment (e'3) and for the Fc portion of IgG. Fe and C'3 receptors are not specific fo r the B-cell lineage and are found in other cells such as monocytes and some nonhematopoietic cells. Si milarly, histocompat ibility-related anti­gens (la o r HLA-DR) are also found on the surface of B-cells but are not unique to them [1- 4]. Pl asma ce lls are the most mature B-Iymphocytes; they lack detectable Sm lg but have e lg. Unli kc the e lg fou nd in prc-B­Iymphocytcs, e lg in plasma cel ls includes both heavy and light chai ns.

A numbcr of heteroa ntisera and , more recen tly, monoclonal ant ibod ies that idcn ti fy B-cell-associated an tigens have been described (table I) (4- 16]. For a more detai led description of these antibodies, see reference 17. Where

Bmn~tI. J.M. ,,,,d Foon. K.A .. (cds.). Immull%gic Approaches 10 the Classijicatioll and Management of Lymphomas a"d Leukemias . © 1988 KI" ... .., Academic f'ublishu3. ISBN97S-/·1<l/2-891'Jj-4. All rights rcurved.

Tab

le 1

. M

onoc

lona

l an

tibo

dies

rea

ctiv

e w

ith

hum

an B

lym

phoc

ytes

Ant

ibod

y P

atte

rn o

f re

acti

vity

M

olec

ular

wei

ght

Clu

ster

R

efer

ence

(s

ubcl

ass)

o

f an

tige

n (k

d)

desi

gnat

ion

anti

-Bl,

ant

i-L

eu 1

6 B

-lym

phoc

ytes

, mal

igna

nt B

-cel

ls

35

CD

20

8 an

ti-B

2 B

-lym

phoc

ytes

, m

alig

nant

B-c

ells

(re

cept

or f

or

140

CD

21

9,1

0

Eps

tein

-Bar

r vi

rus

and

C3d

) an

ti-B

4, a

nti-

Leu

12

B-l

ymph

ocyt

es,

mal

igna

nt B

-cel

ls

40

CD

19

7 B

A-l

B

-lym

phoc

ytes

, gr

anul

ocyt

es,

mal

igna

nt B

-cel

ls

45

,55

,65

C

D24

11

, 12

F

MC

I B

-lym

phoc

ytes

, m

alig

nant

B-c

ells

N

R

NA

13

F

MC

7 <

50%

B-l

ymph

ocyt

es, s

ome

mal

igna

nt B

-cel

ls

NR

N

A

14

anti

-J5,

BA

-3

Gra

nulo

cyte

s, m

ost

non-

T-A

LL

, B

urki

tt's

lym

phom

a,

100

CD

10

4-6

fo

llicu

lar

lym

phom

a, s

ome

lym

phob

last

ic l

ymph

oma,

an

d T

-AL

L.

anti

-PC

A-l

and

Pl

asm

a ce

lls,

mal

igna

nt p

lasm

a ce

lls,

wea

kly

on

NR

N

A

15

anti

-PC

A-2

m

onoc

ytes

and

gra

nulo

cyte

s an

ti-P

C-l

Pl

asm

a ce

lls,

mal

igna

nt p

lasm

a ce

lls

28

NA

16

NR

, no

t re

port

ed;

NA

, no

t ap

plic

able

. T

he a

nti-

B s

erie

s an

ti P

C-I

, an

d an

ti P

CA

-l a

re a

vail

able

thr

ough

Cou

lter

Im

mun

olog

y, H

iale

ah,

Flo

rida

; B

A-l

thr

ough

Hyb

rite

ch I

nc, S

an D

iego

, C

alif

orni

a; a

nd t

he O

KB

ser

ies

thro

ugh

Ort

ho S

yste

m,

Inc,

Rar

itan

, N

ew J

erse

y; L

eu s

erie

s th

roug

h B

ecto

n-D

icki

nson

Co,

Mou

ntai

nvie

w,

Cal

ifor

nia.

tv

3

applicable, the nomenclature and clusters of differentiation (CD) defined by the Second International Workshop on Human Leukocyte Differentiation Antigens are shown [18, 19).

T-lymphocytes

T-Iymphocytes were initially identified by their ability to spontaneously bind sheep erythrocytes (E-rosette). T-Iymphocytes also react with T-cell-specific antisera and anti-T-cell monoclonal antibodies, which may also be used to identify T-lymphocytes and have proven to be more sensitive and discri­minatory (table 2) [20-31). Many of these antibodies react with immature T-cells; others react with more mature T-cells. Some of these antibodies identify antigens found on all T-cells, whereas others occur only on T-cell subsets.

A summary of the more frequently referenced antibodies useful for the classification of lymphoma and lymphoid leukemia, and their cluster desig­nations is shown in table 3.

Immunoglobulin and T-cell receptor genes

Recombinant DNA technology has provided important insights into anti­body diversity and antigen-specific T-cell receptors [32). Immunogloblins are composed of heavy chains and kappa and lambda light chains, encoded by genes on chromosome 14, 2, or 22 , respectively [33-36). Immunoglo­bulin genes are encoded by discontinuous segments of DNA [37, 38). At one point in development, a potential antibody-producing cell must productively rearrange variable, diversity, and joining genes (VOJ) which are then linked to the constant region locus. Immunoglobulin gene rearrangements are hierarchical; [t heavy chain rearrangements precede light chain rearrange­ments; kappa light chain rearrangement precedes lambda light chains re­arrangement [38). These rearrangements can be detected by Southern blot analyses of DNA from B-cells using appropriately radiolabeled heavy or light chain probes . Heavy chain rearrangements have been identified in non­B-cells; light chain rearrangements appear to be restricted to B-cells [38, 39). Clonal rearrangements of light chain genes are therefore an extremely sensitive tool to identify B-ceIl malignancies.

The antigen-specific T-cell receptor is a heterodimer formed by a 40-50 kilodalton (kd) a subunit (Ta), and a 40-45 kd ~ subunit (T~) [40). It is associated with three 20-25 kd peptide chains identified by the T3 mono­clonal antibody [40). Recently, cDNA clones to the T~ and Ta receptors have been isolated [41-43). The human T~ receptor gene has been localized to chromosome 7 [44) and the human Ta receptor gene to chromosome 14 [45).

T~ gene rearrangements have been detected in malignant human T-cells by Southern blotting [46-48). This technique can detect as few as 1 % tumor

Tab

le 2

. M

onoc

lona

l an

tibo

dies

rea

ctiv

e w

ith

hum

an T

lym

phoc

ytes

Ant

ibod

y P

atte

rn o

f re

acti

vity

M

olec

ular

wei

ght

Clu

ster

R

efer

ence

(s

ubcl

ass)

of

ant

igen

(kd

) de

sign

atio

n

OK

T6,

NA

1I34

, an

ti-L

eu 6

T

hym

ocyt

es

45

CD

! 21

O

KT

tl,

anti

-Ttl

, an

ti-L

eu 5

P

an-T

-Iym

phoc

yte

(E-r

ecep

tor)

4

0-5

0

CD

2 23

,24

OK

T3,

ant

i-T

3, a

nti-

Leu

4

Pan

-T-l

ymph

ocyt

e (m

itog

enic

) 2

0,2

0,2

5

CD

3 21

O

KT

4, a

nti-

T4,

ant

i-L

eu 3

T

-hel

per/

indu

cer

55

CD

4 21

,22

OK

Tt,

ant

i-T

t, a

nti-

Leu

1,

Pan

-T-I

ymph

ocyt

e, p

an-t

hym

ocyt

e 65

C

D5

21

,25

,26

T

tO!

3A

l, a

nti-

Leu

9 (

4H9)

, W

Tl

Pan-

T -l

ymph

ocyt

e 40

C

D7

27

-29

O

KT

5, O

KT

S, a

nti-

TS,

T

-cyt

otox

ic/s

uppr

esso

r 32

-43

CD

S 22

an

ti-L

eu 2

an

ti-T

ac

Inte

rleu

kin-

2 re

cept

or

55

CD

25

31

Ant

i-T

OI

Sub

set

of T

-ind

ucer

cel

ls

NR

N

A

31

OK

T9,

5E9

Thy

moc

ytes

, ly

mph

obla

sts,

mon

ocyt

es

90

NA

20

,21

(ant

i tra

nsfe

rrin

) O

KT

tO

Thy

moc

ytes

45

N

A

20

NR

, no

t re

port

ed;

NA

, no

t ap

plic

able

. T

he O

KT

ser

ies

of a

ntib

odie

s ar

e av

aila

ble

thro

ugh

Ort

ho S

yste

ms,

Inc

, R

arit

an,

New

Jer

sey;

Leu

ser

ies

thro

ugh

Bec

ton-

Dic

kins

on C

o, M

ount

ainv

iew

, C

alif

orni

a; a

nti-

T t

hrou

gh C

oult

er I

mm

unol

ogy,

Hia

leah

, F

lori

da;

and

TtO

l th

roug

h H

ybri

tech

, In

c, S

an D

iego

, C

alif

orni

a.

.j::

.

5

Table 3. Frequently referenced antibodies with cluster designations

Cluster designation

CDI CD2 CD3 CD4 CDS CD7 CD8 COlO CD 19 CD20 CD2l CD24 CD2S

Antibody

OKT6, anti-Leu 6, NA1I34 OKTll, anti-Tll, anti-Leu 5,9.6 OKT3, anti-T3, anti-Leu 4, UCHT-l OKT4, anti-T4, anti-Leu 3 OKTl, anti-Tl, anti-Leu 1, 10.2, TlOl anti-Leu-9, 3Al, WTl, 4A OKTS, OKT8, anti-T8, anti-Leu 2 anti-JS, BA-3, anti-CALLA anti-B4, anti-Leu 12 anti-Bl, anti-Leu 16 anti-B2 BA-l anti-Tac

cells in a mixed cell population [46]; it is a sensitive diagnostic marker for T­cell diseases. Interestingly, rearrangements of the T~ antigen receptor are reported in 25% of patients with non-T-ALL [49] and in a small proportion of leukemic B-cells [50]. This is similar to the rearrangement of immunoglo­bulin heavy chain genes in approximately 10% of the leukemic T-cell popu­lation studied [39].

Classification of the lymphoid leukemias and lymphomas

Acute lymphoblastic leukemia

Acute lymphoblastic leukemia (ALL) is heterogeneous. The first surface markers used to differentiate subclasses of ALL were E-rosettes [51, 52], which identify a T-cell subclass (15%-20% of cases), and SmIg, which identifies a B-cell subset « 5% of cases). Both T-cell and B-cell subgroups have an unfavorable prognosis [53, 54]. The next important advance in identifying ALL was development of an antiserum to the common acute lymphoblastic leukemia antigen (CALLA) [55]. CALLA or CDlO reactivity identified a non-B, non-T subclass of ALL patients (approximately 70% of cases) with a more favorable prognosis than T-ALL, B-ALL, or non-B, non-T-ALL without CDlO [54]. Other markers such as Ia antigen were commonly found on non-T-ALL and could help differentiate non-T from T-ALL [56]. By testing for cytoplasmic fA heavy chain, a subset designated pre-B-ALL has been identified [57]. Except for the presence of cytoplas­mic fA, this subset expresses the same surface markers as the CDlO-positive form of non-T-ALL; it appears, however, to have a less favorable prognosis [58].

6

With the development of monoclonal antibodies, it became evident that the T-cell subset of ALL was heterogeneous [59-61). More recently, studies of immunoglobulin gene rearrangements and monoclonal antibodies that identify B-cell-associated antigens have demonstrated that most cases of non-T-ALL derive from the B-cell lineage [62, 63]. We review these data and present a new classification for ALL based on these recent observations.

Non- T-acute lymphoblastic leukemia

Two important areas of research have prompted a reassessment of non-T­ALL. First, monoclonal antibodies that recognize B-cell-associated antigens have been identified; many are present on non-T-ALL cells . The most specific of these antibodies is probably CD19, which is present on 95% of cases of non-T-ALL [11, 63]. Second, clonal rearrangements of immuno­globulin genes provide strong evidence for the B-cell lineage of most cases of non-T-ALL [62, 63].

Although la antigen is present on most non-T-ALL, and CDlO is present on 75% of cases of non-T-ALL, these antigens are also identified on ap­proximately 10% of cases of T-ALL. Therefore, B-cell-associated antigens (table 1), which are not identified on T-ALL cells, are the most useful in distinguishing non-T-ALL. The CD19 and CD20 antigens are model antigens for this discussion .

Less than 5% of cases of ALL express Smlg (usually IgM); these cells are typically classified as B-ALL. These cells generally express other B-cell antigens, including CD20, CD19, and la. B-ALL in children is probably a leukemic phase of non-Hodgkin's or Burkitt's lymphoma [52, 53]. Another marker that identifies a subset of non-T-ALL is cytoplasmic II (Cll) heavy chain; x and A light chains and Smlg are typically absent [57]. These cells are considered pre-B-cells. As indicated, most cases of non-T-ALL involve pre­B-cells; thus Cll is useful in determining the level of differentiation of pre­B-cells. Pre-B-cells that synthesize II heavy chain are the most mature cells of this group .

Nadler and coworkers [63] recently classified 138 patients with non-T­ALL on the basis of monoclonal antibodies and immunoglobulin gene re­arrangements. They divided these cases into four major subgroups. The first subgroup was la antigen-positive, representing 5% of cases. Another sub­group expressed the Ia and B4 antigens, representing 15% of cases. The third subgroup expressed the la, CD19, and CDlO antigens, comprising one third of the cases. Finally, one half of the cases of non-T-ALL were la, CDlO, CD19, and CD20 positive. The fourth group was further subdivided into cases with and without Cll. We propose that cases with Cll be placed in a separate group (group V) assuming that they are more mature. The final and most differeniated group, group VI , represents Smlg-positive B­ALL (table 4).

7

Table 4. Classification of non-T-ALL

Antigens Cytoplasmic ~ Surface membrane

la CD19 CDlO CD20 immunoglobulin

Group I + Group II + + Group III + + + Group IV + + + + Group V + + + + + Group VI + + +/- + +

T-acute lymphoblastic leukemia

T-acute lymphoblastic leukemia (T-ALL) represents 15% to 25% of cases of ALL. Clinical features associated with T-ALL include a high blast cell count, predominance of male patients, older patients (15-20 years), and mediastinal masses. T-ALL was originally identified by E-rosette formation. The most sensitive marker for T-ALL is probably CD7. This antigen is present on most thymocytes and T-cells but not on non-T-ALL or B-ceIl lymphomas or leukemias [27-29]. In a study of 23 patients with T-ALL, all cases expressed the CD7 antigen [2S]. Interestingly, CD7 antibodies react with a small proportion of cases that appear to be myeloid leukemias [64]. In addition, an unusually high incidence of CDlO, la, and CD24 expression has been reported in adults with T-ALL [65]. Recently, rearrangement of the TI3 receptor gene in cases of T-ALL has been reported [46-50].

Further subclassification of T-ALL is controversial. Reinherz and col­leagues proposed a subclassification for T-ALL according to the level of thymic differentiation [59]. Several elements of their subclassification of T­ALL have been confirmed; others are controversial. The most primitive thymocytes, referred to as early or stage I thymocytes, react with T9 and TlO antibodies and account for approximately 10% of the thymic cells. In their study, Reinherz and coworkers reported that most T-ALL cells express antigens found on early thymocytes. The next level of thymic differentia­tion, which includes the majority of thymocytes, is referred to as common or stage II. These cells lose T9, retain TIO, and acquire CD1, CD4, and CDS antigens. Approximately 20% of cases of T-ALL express this phenotype. Mature stage III thymocytes no longer express T6 but segregate into CD4 or CDS subsets similar to peripheral blood T-Iymphocytes. Only rarely did Reinherz et al. find T-ALL cells with the phenotype of mature thymocytes or circulating T-Iymphocytes. In a more recent study, Roper and coworkers [61] confirmed many of the findings reported by Reinherz but reported some major differences. In this study, only one third of the T-ALL patients had

8

the phenotype of early or stage I thymocytes ; most had the phenotype of either intermediate or late stage thymocytes. In table 5 we summarize these data and propose a scheme for the classification of T·ALL. The common marker for all of the subgroups is CD7. Nearly all cells also express CD5, and most express CD2 that identifies the E·rosette receptor.

Although Roper and coworkers [61] studied clinical correlations among these three groups of T·ALL, they found no unige clinical features among the subgroups and no differences in remission duration or survival. How­ever, the groups were too small for statistically valid conclusions. Presently we believe it useful to subclassify T-ALL using this system so that data from a number of institutions can be analyzed for clinical correlations between the subgroups of T-ALL.

Non-Hodgkin's lymphoma

The non-Hodgkin's lymphomas are a diverse group of neoplasms whose pathologic classification are controversial. It is even more difficult to cor­relate pathologic classification with immunologic classification. There are, however, a number of immunologic patterns that emerge, and we will attempt to place them within the non-Hodgkin's lymphoma working classi­fication [66] as well as the Rappaport classification [67].

Follicular or nodular lymphomas

The follicular or nodular lymphomas most likely represent neoplastic proli­feration of lymph node-derived follicular center B-lymphocytes. The cell type may be a small cleaved cell (nodular lymphocytic poorly differentiated lymphoma by the Rappaport classification), mixed small cleaved and cleaved or noncleaved large cell (nodular mixed), or predominantly large cell (nodular histiocytic). The first two cell types fall within the working classi­fication as low-grade lymphoma, whereas the latter cell type as an inter­mediate-grade lymphoma. While the predominantly small cleaved cell will almost always express high density monoclonal Smlg, larger cells may be

Table 5. Classification of T-ALL

Antigens

CD7* CDS CD2 CD3 CD4 CDS CD!

Group I + + (90%) + (7S%) Group II + + + + (2S%) + (90%) + (90%) + Group III + + + + +/- t +/- t

* Found on virtually all T-ALL-cells. t No longer simultaneous expression of CD4 and CDS, as found in Group II .

9

Smlg negative [68, 69]. However, the small cleaved and large cells will routinely express la, C019, and C020 antigens and will often express the C021 antigen [69]. Interestingly, more than half of these cases will also express COlO [69, 70]. Follicular lymphoma cells may be found in the patient's peripheral blood as a 'leukemic' phase of the disease (formerly referred to as lymphosarcoma cell leukemia). These cells can usually be differentiated from chronic lymphocytic leukemia (CLL) cells, as they may express COlO, which is not expressed on CLL cells; they do not express the COS pan-T-antigen found on CLL cells; and they generally will have a low percentage of mouse erythrocyte rosette formation (see below) [71, 72].

Malignant lymphoma, small lymphocytic

Malignant lymphoma, small lymphocytic (diffuse lymphocytic well differen­tiated lymphoma in the Rappaport classification) is a low-grade malignancy, and some cases may be identical to CLL. Also included within this sub­classification are the plasmacytoid lymphocytic subgroups with and with­out an IgM monoclonal gammopathy; some of these cases are similar to Waldenstrom's macroglobulinemia (described below). Surface markers on these small lymphocytic cells include low intensity Smlg, mouse erythrocyte receptors, C'3, and receptors for the Fc portion of IgG and la, COI9, C020, C021, CD24, and other B-cell antigens. These features are similar to CLL and the cells also express the COS pan-T-antigen.

Malignant lymphoma, diffuse small cleaved cell, and diffuse mixed small and large cell

Malignant lymphoma, diffuse small cleaved cell (diffuse lymphocytic poorly differentiated lymphoma in the Rappaport classification) is an intermediate prognostic group. The cells are B-lymphocytes that (similar to follicular lymphoma cells) usually display large amounts of monoclonal Smlg. Unlike follicular lymphoma cells, however, they do not usually express COlO; [70]. Similar to follicular lymphoma cells, they do not express the COS antigen as do cells from most small lymphocytic lymphomas and CLL. However, all of these cell types have in common the expression of la, COI9, C020, C021, and other B-cell antigens [69].

The diffuse mixed small and large cell (diffuse mixed lymphocytic-histio­cytic) lymphomas have not been extensively studied but are most likely predominantly B-cell diseases. They are also considered an intermediate­grade prognostic group.

Malignant lymphoma, diffuse large cell and large cell immunoblastic

In the working classification, the diffuse large cell lymphomas are considered within the intermediate prognostic group, whereas large cell immunoblas-

10

tic lymphoma is a high-grade malignancy. By the Rappaport classification, both of these cell types would be described as histiocytic. This is clearly a misdesignation since SO% -90% of cases represent clonal expansions of malignant B-cell [73]. A high percentage of these cells express T9 and no antigens [6S]. Fifty-seven cases of diffuse large cell lymphoma were recently studied and divided into the following subgroups: 1) COI9, C020, and SmIg positive; C02I negative (50%); 2) COI9, C020, SmIg, and C021 positive (30%); 3) COI9, C020, and C02I positive; Smlg negative (10%); 4) C020 and SmIg positive; C019 and C021 negative (10%) [74]. These data suggest that most of these lymphoma's represent the malignant counterpart of B-cells at the midstage of differentiation. Ten to 20% of cases are T-cell lineage; 2% are derived from the monocyte-myeloid lineage. Recently, clonal rearrangement of the T~ receptor has been described in patients with T-derived non-Hodgkin's lymphoma [47, 4S].

Malignant lymphoma, lymphoblastic

Malignant lymphoma or lymphoblastic lymphoma is a high-grade malignancy. The nuclear membrane is characteristically deeply subdivided, imparting either a lobulated (convoluted) appearance or a fine linear (non convoluted) subdivision in a round nucleus. Lymphoblastic lymphoma represents approx­imately one third of the cases of non-Hodgkin's lymphomas in children and 5% of cases in adults. The disease is more prevalent in males; these patients often have a mediastinal mass. In some cases, the disease may evolve into a leukemic phase morphologically indistinguishable from T-ALL. The malig­nant cells are T-cells, form E-rosettes, react with T-cell antisera [75-77], and have rearrangements of the T~ receptor [7S]. Studies with monoclonal antibodies have demonstrated marked heterogeneity. Lymphoblastic lym­phoma cells differ from T-ALL in that the cells rarely express the surface markers common to immature thymocytes (group I) [79]; phenotypes are equally divided among group II and group III T-ALL. Interestingly, in 40% of cases the cells are reported to express COW; COW expression is less common in T-ALL (10%) [70].

Malignant lymphoma, small noncleaved cell

This category includes Burkitt's lymphoma and other lymphomas previously designated undifferentiated non-Burkitt type (high-grade). Burkitt cells from peripheral blood and bone marrow are usually classified as L3 by the French-American-British (FAB) criteria [SO]. Most cases of Burkitt's lym­phoma from Africa are associated with the Epstein-Barr virus (EBV); these are endemic. Most non-African cases (non endemic) are EBV negative [SI]. Chromosomal abnormalities involving chromosome S (carrying the onco­gene c-myc) and either 2, 14, or 22 occur in virtually all cases of endemic and nonendemic Burkitt's lymphoma [S2]. These are designated t(2;S),

11

t(8;14), and t(8;22), respectively. Usually the light chain class expressed on these cells is correlated with the translocation, i.e., x in t(2;8) and f... in t(8;22). African Burkitt's lymphoma cells have C'3 and receptors for the Fc portion of IgG, in addition to the EBV receptor. American Burkitt's lym­phoma cells do not [81]. Phenotyping of cell lines derived from patients with undifferentiated lymphoma of the Burkitt's and non-Burkitt's type have demonstrated heterogeneity [83]. These studies suggest that Burkitt cells follow a divergent pathway of B-cell evolution, as they are all TdT negative (unlike early B-cell non-T-AIl). The most primitive of the Burkitt cell lines are la and Bl positive and mayor may not express COW. Maturation was evident in other Burkitt cell lines by the expression of C~, surface mem­brane IgM, and/or IgM secretion. Some of these Burkitt cell lines also expressed C025.

Peripheral T-cell lymphoma

Peripheral T-cell lymphoma would usually be classified as malignant lym­phoma, large cell immunoblastic (high-grade) under the working formulation . However, this tumor has unique features and will be described separately. The term peripheral T-cell lymphoma is used to distinguish it from lympho­blastic lymphoma of presumed thymic origin. Peripheral T-cell lymphomas are thought to derive from peripheral T-lymphocytes in lymph nodes and other non lymphoid sites. These lymphomas comprise a broad spectrum of morphologic types of lymphocytes. In all instances, the cells have T-cell markers admixed with epithelioid histiocytes, plasma cells, eosinophils, and vascular hypertrophy. Clinically, peripheral T-cell lymphoma is character­ized by generalized lymphadenopathy , weight loss, and a high incidence of pulmonary involvement [84]. Surface markers are usually, but not always, characteristic of mature T-helper cells [85], including C04 helper-associated antigen C03, C02, and C05 pan-T-antigens. Rearrangement of the T~ receptor has been reported [78].

Ty lymphoproliferative disease

Ty lymphocytes are a subset of T-lymphocytes with receptors for the Fc portion of IgG. A high proportion of normal Ty lymphocytes are large granular lymphocytes. These cells are thought to be responsible for natural killer (NK) and antibody-dependent cell-mediated cytotoxicity. A Iympho­proliferative disorder made up of predominantly Ty lymphocytes has been described; we refer to this as chronic Ty lymphoproliferative disease [86]. Typically, patients are elderly, males, with increased Ty lymphocytes infil­trating the bone marrow and spleen [86, 87]. Although the disease is not rapidly progressive, neutropenia and recurrent infections are common. Most patients do not require chemotherapy. Variants of this disease, including a more aggressive form, have also been described [88]. Clonal chromosomal

12

abnormalities [89] as well as clonal rearrangement of the T~ receptor, have been reported [57, 90]. Cells from chronic Ty lymphoproliferative disease usually contain acid phosphatase and ~-glucuronidase and express the pan­T-antigens CD2 and CD3, the suppressor-associated antigens CD8, and the NK-associated antigens Leu 7 (HNK-1).

Cutaneous T-cell lymphoma (mycosis fungoides, Sezary cell leukemia)

Skin lesions are the most prominent feature of patients with cutaneous T-cell lymphoma [91]. Lesions vary from limited plaques to diffuse plaques, tumors, and generalized erythroderma. Rare patients with limited plaque disease and less than 50% with generalized plaques and tumors have extra­cutaneous disease detected by light microscopy evaluation of peripheral blood and lymph nodes. Special studies , including cytogenetic analysis and elec­tron microscopy, indicate blood involvement in over half of the patients with limited plaque disease and most patients with generalized plaques and skin tumors [92] . Analysis of the T~ receptor rearrangement will likely reveal a higher proportion of cases with blood involvement.

The malignant cells in this disorder are characterized by a cerebriform nucleus. In the skin the cells are referred to as mycosis fungoides cells and, in the peripheral blood as Sezary cells. Sezary and mycosis cells form E­rosettes, react with T-antisera and anti-T-monoclonal antibodies [93], and have clonal rearrangements of the T~ receptors [54, 55, 85 , 94]. In most cases, the cells express the phenotype associated with normal helper/inducer T-lymphocytes (CD3, CD4, CD5 positive) [95] and function as helper T-Iymphocytes in in vitro assays [96].

Adult T-cell leukemiallymphoma

Adult T-cell leukemia/lymphoma is associated with a human retrovirus des­ignated human T-cell leukemia/lymphoma virus-1 (HTL V -1) [97]. Virtually all patients tested have antibodies to HTL V-I [98]. Patients with this disease have been identified primarily in Japan, the United States, and the Carib­bean. In the United States, the patients are young (median age 33 years), predominantly black, and born in the Southeast [98]. Common clinical features include a rapid onset of symptoms, with rapidly progressive cutane­ous lesions and hypercalcemia. Skin lesions are variable and include small and large discrete or confluent nodules, or nonspecific plaques, papules, or patches. Patients have increased bony turnover with abnormal bone scans, elevated alkaline phosphatase, and may have lytic bone lesions [99]. Lym­phocytosis is common, and circulating malignant cells are present in low numbers in most patients. Peripheral lymphadenopathy is common, with retroperitoneal and hilar involvement in approximately 50% of cases . Bone marrow, gastrointestinal, pulmonary, leptomeningeal , and hepatic involve­ment are somewhat less common (20%-50%). Response to combination

13

chemotherapy is prompt and often complete, but duration of response is short (median 13 months). Opportunistic infections are extremely common in these patients.

The typical malignant circulating cells have moderately condensed nu­clear chromatin, inconspicuous nucleoli, and a markedly irregular nuclear contour in which the nucleus is divided into several lobes [100]. These cells typically express the phenotype of helper/inducer T-lymphocytes [101] and the C025 antigen that identifies the IL-2 receptor [102]. Variability in the expression of C02 and C03 have been reported [101]. Clonal rearrange­ments of the T~ receptor are identified in cells from patients with adult T-cell leukemia/lymphoma [48, 50, 78, 82]. The leukemic cells are reported to suppress B-cell Ig secretion [103] by a complex mechanism involving induction of suppressor cells following activation of normal suppressor cell precursors [101].

Chronic lymphocytic leukemia and prolymphocytic leukemia

Chronic lymphocytic leukemia is a monoclonal proliferation of SmIg­positive B-lymphocytes [104, 105). Clonality of CLL has been demonstrated by expression of a single Ig light chain, either 'X or A, on the cell surface membrane [106]. More sophisticated techniques have confirmed clonality by showing unique immunoglobulin idiotype specificities [107], a single pattern of glucose-6-phosphate dehydrogenase activity [108], clonal chromosome abnormalities [109], or immunoglobulin gene rearrangement [47]. The malignant B-cell involved in CLL is an intermediately differentiated cell. The cell appears frozen in differentiation and does not mature to the final stage of B-cell development, the mature plasma cell. However, recent data have demonstrated that in vitro treatment of these cells with phorbol esters or pokeweed mitogen can induce differentiation into mature immunoglo­bulin-secreting plasma cells [110]. Under certain circumstances , CLL cells stimulated in vitro with phorbol esters differentiate into cells with cyto­plasmic protrusions and other characteristics of hairy cell leukemia [111].

The B-lymphocyte characteristic of CLL displays a relatively small amount of Smlg, and this has been used to distinguish CLL from the leu­kemic phase of nodular and diffuse lymphocytic lymphomas and from pro­lymphocytic leukemia, where the cells generally display considerably more Smlg [112]. Immunoglobulin isotype analyses indicate that most CLL dis­playa single heavy chain class; typically, It or It and o. Less commonly, y, u, or no heavy chain determinant is found. CLL cells display either 'X or A light chains but never both. B-CLL-cells display receptors for mouse erythro­cytes, a feature characteristic of immature B-lymphocytes [113]. The cells also have the receptor for the Fc portion of IgG and complement, with a relative increase of C'3d receptors (CR2) over C'3b receptors (CRl); this is typical of immature B-cells [114] . B-CLL cells display several antigens, including Ia and human B-cell antigens such as COI9, C020, C021, and

14

CD24. One unanticipated finding was that B-CLL cells display the CDS antigen previously thought to be restricted to T-Iymphocytes. This antigen was first recognized by using heteroantisera [115] and later with the T101 and equivalent monoclonal antibodies [32]. The precise meaning of this anomalous expression of a T-cell antigen is unclear; although a normal B­cell counterpart has been reported in human tonsil lymph nodes [116], and stimulation in vitro of normal B-cells with phorbol ester may induce expres­sion of this antigen [117]. The T01 antigen, reported to define the inducer of suppression within the T-helper subset, was identified on 60 of 75 B-CLL patients' cells [118].

Rearrangement of immunogloblin heavy and light chains has been re­ported as expected in B-CLL cells, however, rearrangement of the TB receptor has also been reported in approximately 10% of cases of B-CLL [50]. This is analogous to the reported TB rearrangement in non-T (pre-B)­ALL, and again emphasizes that immunoglobulin and TB receptor re­arrangement alone are not adequate to assign lineage.

In 3% to 10% of patients with CLL, the disease may evolve into a diffuse histiocytic lymphoma (Richter's syndrome). This may be associated with loss of the T01 antigen [118]. Most data suggest that this evolution involves transformed follicular center B-cells, rather than histiocytes or macro­phages. Some transformations represent evolution of the malignant clone, with expression of the same monoclonal immunoglobulin and karyotypic abnormality present in the original CLL clone [119]. In other cases, the lymphoma cells have different markers and immunoglobulin gene rearrange­ments than the original CLL-cells; these cases probably represent the con­comitant development of a B-cell lymphoma or a histiocytic malignancy in patients with CLL [120, 121].

Pro lymphocytic leukemia (PL) is related to CLL and is also likely to be derived from cells from the medullary cords of the lymph node. Immuno­globulin gene rearrangements of heavy and light chains have been reported [122]. Patients with PL generally have extremely high blast counts and splenomegaly, but lack significant lymphadenopathy. Prolymphoblasts are likely activated cells and appear morphologically immature with a fine lacy nuclear chromatin and one to two nucleoli; they may contain intracyto­plasmiC granules. These cells generally have higher density SmIg than CLL cells; they have Ia and CD20 antigens and may form rosettes with mouse erythrocytes [72]. PL cells from 14 consecutive patients reacted with the FMC7 monoclonal antibody that recognizes an antigen found on one half of normal B-lymphocytes, while cells from only 5 of 20 patients with CLL reacted with this antibody [7].

Approximately 5% of cases of CLL and PL result in a malignant prolifera­tion of T-cells rather than B-cells. These cells react with T-antisera and anti­T-monoclonal antibodies reflecting the phenotypes of mature T-Iymphocytes; they lack SmIg and other B-cell markers [123, 124]. Many of these patients have diffuse organ and skin involvement [123].

15

Hairy cell leukemia

Hairy cell leukemia (leukemic reticuloendotheliosis) is characterized by in­vasion of the bone marrow and spleen by morphologically distinct mononu­clear cells with 'hairy' cytoplasmic projections [125]. These cells usually contain an isoenzyme of acid phosphatase (isoenzyme 5) that is resistant to tartrate; this isoenzyme is not unique to hairy cells. Surface markers of hairy cells are most consistent with a monoclonal proliferation of B-lymphocytes [126]. Smlg with a single light chain is frequently identified [126], as are B­cell-associated antigens. Interestingly, the PCA-1 antigen (but not the PC-1 antigen) , typically on plasma cells, is identified on hairy cells; these data suggest that hairy cells may be preplasma cells [127]. Perhaps the most convincing evidence for the B-cell origin of hairy cells comes from studies of immunoglobulin genes which indicate clonal rearrangement of heavy chain genes and at least one light chain [128, 129]. Most cases of hairy cellieuke­mia demonstrate the C025 antigen (IL-2 receptor) typically identified on select T-cell malignancies and activated T-cells [129].

Myeloma and related disorders

The malignant B-cells of Waldenstrom's macroglobulinemia, heavy chain disease, and multiple myeloma represent a further step in the maturation of medullary cord B-cells [104]. Like CLL cells, cells from patients with Waldenstrom's macroglobulinemia express Smlg and la, C019, and C020 antigens [11]. Unlike CLL cells, however, these cells express the PCA-1 antigen and do not express the C021 antigen nor rosette with mouse erythro­cytes [72]. The plasma cell and its malignant counterpart, the myeloma cell, represent the most differentiated B-lymphocytes. These cells synthesize large quantities of immunoglobulin and have Clg, but usually lack Smlg and the la, C019, C020, and C021 antigens [130]. Plasma cells and myeloma cells, like other mature B-lymphocytes, usually lack COW, but a recent study has suggested that rare cases of COW-positive myeloma represent an aggressive subtype with a poor prognosis [131] . Interestingly, plasma cells and myeloma cells stain intensely with the OKTlO monoclonal antibody, as well as the anti-PCA-1 and anti-PC-1 antibodies [132, 133].

Correlates of cellular differentiation with lymphoid malignancies

Substantial data suggest that the phenotypes of most leukemia cells are not unique but reflect characteristics of normal cells. None of the surface markers reviewed are leukemia specific; all can be identified on normal as well as malignant cells. Most of the monocyte, granulocyte, and lymphocyte anti­gens are found on mature and immature cells. However, COlO, C024, and RFB-1 are expressed primarily on immature bone marrow cells. This obser­vation is consistent with the phenotypes of leukemic cells, since the COW

16

and CD24 antigens are present on primitive leukemia cells (ALL and lym­phoid blast crisis of chronic myelogenous leukemia cells) but only rarely on more mature leukemia or lymphoma cells. Distribution of the reactivity of the monoclonal T-antibodies is likewise consistent with this hypothesis. The most primitive thymocyte markers, OKT9 and OKTlO, are found on most T-ALL cells, while CD3, C04, and C08, which are found on mature thymocytes and circulating T-lymphocytes, are more often identified on more mature T-cell leukemias.

A proposed scheme of normal lymphoid differentiation is presented in figure 1. This scheme is based on the concept that the phenotype of normal lymphoid cells at each level of differentiation can be detected from the phenotype of its malignant counterpart. Although some malignant cells may have an aberrant phenotype, the data presented suggest that most malignant lymphoid cells reflect the phenotype of a normal lymphocyte. The proposed phenotype of the progenitor B-lymphocyte probably has the same surface markers as the group I non-T-ALL and represents the earliest identifiable B-cell. This cell expresses the la antigen, but no other B-cell-associated antigens. The next level of B-ceIl differentiation coincides with the group II non-T-ALL; the C019 antigen is expressed in addition to la and heavy but not light chain immunoglobulin genes are rearranged. At the next level of B-cell differentiation, the cells express COW and light chain gene re­arrangements occurs; this coincides with group III non-T-ALL. With fur­ther B-cell differentiation, the C020 antigen is expressed, followed by cyto­plasmic f.! and then Smlg. At the next level of B-cell differentiation, the early B-cell acquires the C021 antigen and the receptor for mouse erythro­cytes; both Smlg and Clg are present. Most CLL-cells and malignant lym­phoma, small lymphocytic type cells, express the phenotype of intermediate B-lymphocytes. The cells express complement, the receptor for the Fc portion of IgG, COS, in addition to the surface markers identified on more primitive B-cells. At this level of differentiation, there is low-density Smlg. The maturing B-cells express high-density Smlg (lgM, IgG, or IgA) without C021 or mouse erythrocyte receptors. The malignant counterparts of the mature B-ceIl are the follicular small cleaved and large cell lymphomas and the diffuse small cleaved and large cell lymphomas and PL-cells. At the next step of maturation, the plasmacytoid B-cell secretes Ig, usually of the IgM

Figure 1. Schematic representation of human lymphoid differentiation and related lymphoid malignancies. ALL, acute lymphoblastic leukemia; CLL, chronic lymphocytic leukemia; SL, malignant lymphoma, small lymphocytic; FSC, malignant lymphoma, follicular small cleaved cell ; FLg, malignant lymphoma, follicular large cell ; DSC, malignant lymphoma, diffuse small cleaved cell; DLg, malignant lymphoma , diffuse large cell; PL, prolymphocytic leukemia; LL, lymphoblastic lymphoma; Ty-LPD, Ty-Iymphoproliferative disease; ATL, acute T-cell leuke­mia/lymphoma; CTCL, cutaneous T-cell lymphoma; PTCL, peripheral T-cell lymphoma; TdT, terminal deoxynucleotidyl transferase; H, heavy chain; L, light chain; 0, germ-line configura­tion; R, rearranged gene: T~, clonal rearrangement of the T~ receptor; MR, mouse rosette; CR, complement receptor.

Gro

up I

no

n·T

All

G

rou

p II

oo

n·T

ALL

G

rovp

lll

non-

T A

LL

Gro

up

IV

non·

T A

ll

Gro

up V

no

n·1

ALL

~------~~------~

Gro

up V

I oo

o-T

ALL

f--H

airy

Ce

lllu

uke

mia-4

FSC

,Flg

O

SC,O

Lg

PI

Wal

dens

trom

's

Mac

rogl

obul

inem

ia

"'"

-".

. ','

C

OlO

.

".:

01

0

CO

lO

_ _

te0

10

•

•.

• •.

• C

019

o·,

C0

19

••

~C019

.~

C01

9 tt'

.C0

19

••.

' •• C

019

•. 1 ....•

... C

O.1

9 .

_ R

.

_ ".

_ ~.c~

··'c~-_:.

·:C~ ~

RL ~.

: C

D20

H,

' H

_

H

HI'L

Ig

021

":"

C02

1 'Ii

; T

1 ~

/161

Fe

±CD

5 Fe

~!:i:

"_~":

/' ±

e0

21

/"' •

• "".

. p"

.. p"

.. p"

.. p"

.. E.

"y·.

I"."

".di

"~.

"",,,

.•

• j"ym,hO

id "\ C~:,

", I T.A."

\t0

7

Tl~CD5

.•

T9~\

.C02

~

E"'Thy~y

" \

PI""

",,,,

S"m

\ •. ~

___

E""h

''''''''

:. ,.

..

,~. _

Gra

nulo

cyte

s-m

oroo

cyte

s

~\>,.,

., --..

. P

late

lets

Mye

lokl

Ste

m C

ell

Fig

ure

1.

Gro

up II

I T A

LL

II

C07

T-C

LL

1;-C

PO

G,o

"" T

All

TlO

g05

.

CD

a .

: Til

C

02

C07

··.··

.··c····

C03

.'

C05

II ~

:r 02

±F

c C

07

=

•• -~-

~=

C02

...

_,.

. C

O.

Com

mon

Thy

moc

y1e

Gro

up

UlT

ALL

II

C07

Mat

ure-

T (s

uppr

esso

r)

T·C

ll.Jl

. A

Tl

CIC

L

PT

el

II C

02

I ~CD2

~OUCD5

tFc

CD

7 C

DS

CD

4 ',

~,.~

.... _

C0

3

CD4~CD3

Mat

ure

Thym

ocyt

e M

atur

e-T

(h

elpe

r)

PIii!

lsm

acyt

oid·

B

Muh

iple

M

yelo

ma

Pla

sma

Cel

l

iF

......

-.J

IS

subclass, and expresses new surface membrane antigens including OKTlO, PCA -1, and PC-I. It has recently been demonstrated that hairy cell leukemia falls somewhere between the mature B-cell and plasma B-cell; they also express the PCA-l antigen [127]. The plasma cell, the most differentiated B-lymphocyte, expresses the same phenotype as myeloma cells. Although these cells have CIg and produce immunoglobulin, and express OKTlO, PC-I, and PCA-l, they lose other surface membrane markers including Smlg, la, and B-cell antigens.

T-cell differentiation follows a distinct pathway. Early thymocytes (stage I) express CDS, CD7, T9, no, and often C02 (sheep erythrocyte receptor); this phenotype probably represents the malignant counterpart of group 1 T-ALL cells. The common thymocyte (stage II) no longer expresses T9; it gains the COl antigen and simultaneously expresses the helper-associated (CD4) and suppressor-associated antigens (CDS). This cell clearly rearranges the T~ gene, confirming that T~ rearrangement precedes surface membrane expression of the T3-Ti complex. This cell corresponds to the phenotype of group II T-ALL and some lymphoblastic lymphomas. Subsequently, the cells lose either the helper-associated or suppressor-associated antigens. This is equivalent to group III T-ALL or some cases of lymphoblastic lymphoma. In the final stage of maturation, the suppressor-associated cell (CDS) may express the receptor for the Fc portion of IgG, as well as the surface markers previously attributed to the mature thymocyte (including the T3-Ti complex). This coincides with the phenotype of some T-CLL cells and chronic Ty lymphoproliferative disease. The helper-associated mature T­lymphocyte, on the other hand, may express the receptor for the Fc portion of IgM and coincides with the phenotypes of some T-CLL, adult T-cellieukemial lymphoma, cutaneous T-cell lymphoma, and peripheral T-cell lymphoma.

Classification of Hodgkin's disease

Hodgkin's disease (HD) is a malignant neoplasm of uncertain cellular origin characterized by the appearance of distinctive binucleate or multinucleate giant cells (Reed-Sternberg cells, RSe) and their mononuclear variants (Hodgkin's cells, He) [134]. The malignant nature of this disease is suggested by cytogenetic studies that have shown a clonal distribution of chromosomal aneuploidy [135-137]. Considerable debate has arisen as to what constitutes the malignant cell of HD. However, most investigators now agree that the RSCs or HCs (a subset constituting a minute fraction of the tumor mass) represent the neoplastic cell population [13S]. The normal cellular counter­part from which RSCs and HCs arise has not yet been identified [13S].

Investigators have used morphology (light and electron microscopy), cell culture, and immunohistochemistry in an attempt to characterize the nature of the RSC and He. Based on these observations, it has been argued that HD arises from the T-lymphoid [139, 140], B-lymphoid [141-151], or mye-

19

loid-macrophage lineages [134, 152-156]. Although this controversy is unre­solved, the application of immunologic marker analysis has contributed to our further understanding of the disease, and several general statements can be made. First, with few exceptions, most observers have failed to detect the uniform expression of T-cell surface markers (as defined by polyclonal and monoclonal reagents) [151, 157-162] by RSCs or HCs, suggesting that these cells are not of T-lymphocyte origin. RSCs and HCs have been shown to express the Tac antigen (IL-2 receptor) [163], and in two reports these cells were found to be T9 positive (transferrin receptor) [155, 159]. Neither receptor-associated marker is restricted to he T-cell lineage [129, 164]. Second, although the detection of SmIg or CIg in RSCs and HCs would favor a B-cell origin [141, 142, 145], the expression of these determinants is often polyclonal [144, 165-169], which suggests that immunoglobulin is absorbed onto RSC and HC cells rather than being synthesized by the malignant cell [165]. There are no convincing data that RSC or HC produce immunoglobulin. Immunologic staining of RSC and HC for the expression of B-cell differentiation antigens has produced conflicting results [157, 158, 170]. In an interesting case of B-cell HD [171], a patient with nodular sclerosing HD developed a terminal leukemic phase. The circulating HCs expressed the CD19 and CD20 antigens and had cytoplasmic f.! heavy chains and a clonal rearrangement of heavy and light chains (consistent with a B-cell origin). Substantial data, however, favor a myeloid/macrophage ori­gin for HD [155, 156, 165, 166, 169, 172-174]. This conclusion is based on the demonstration of nonspecific esterase (NSE) and acid phosphatase; a-1-antitrypsin, a-1-antichymotrypsin, muramidase; lectin-binding properties; and the variable expression of Fc and C3 receptors on RSC and HC cells. While short-term cell lines believed to be derived from RSCs demonstrate weak phagocytic activity and one line was reported to synthesize IL-1 [172, 174], other established cell lines have not uniformly shown these activities [175]. In most instances RCS and HC do not react with antibodies to monocytes [141, 157-161, 170, 175]. In one report a substantial number of biopsy specimens contained RSC and HC positive for markers characteristic of late granulocytic maturation (TU5, TU6, TU9) [158, 176]. In one estab­lished cell line, the cells phagocytized latex and ink particles, were non­specific esterase-positive, expressed antigenic determinants distinctive for cells of the monocyte-macrophage lineage, and were capable of specific antigen presentation to immune T-cells. Based on Ia expression [141, 157, 161, 170] and characteristic cytochemical features, other authors have sug­gested that the cell of origin for HD is a reticulum cell (either a dendritic cell or an interdigitating reticulum cell) [161, 177]. Finally, some data suggest that the RSC and HC represent a subset of activated lymphoid cells of either T-lymphoid or B-lymphoid origin. This conclusion is based on an immuno­logic analysis in which RSC and HC uniformly expressed the Ki-l marker (35/35 biopsy specimens of all histological subtypes), as defined by a mono­clonal antibody raised by immunization against an established HD cell line

20

[178]. Among normal cells, Ki-1is expressed by T-lymphocytes and B­lymphocytes activated in vitro by a variety of stimuli that also induce IL-2 receptor expression [162]. In situ staining of biopsy specimens from non­neoplastic and reactive tissues demonstrated Ki-1 expression by a population of normal perifollicular lymphoid cells (lymph node and spleen) and variable degrees of expression by abnormal lymphoid cells in cases of angioimmuno­blastic lymphadenopathy and lymphatoid papulosis. Among 290 cases of non-Hodgkin's lymphoma, Ki-1 expression was observed in 19 cases of peripheral T-cell lymphoma and in 45 cases of diffuse large cell lymphoma (including 35 specimens expressing T-cell surface markers and 7 bearing B­cell antigens). These results suggest that Ki-1 is a lymphoid activation antigen which identifies a group of large lymphoid cells in normal and neoplastic tissues (including RSC) that remain poorly characterized. An­other monoclonal reagent, HeFi-1, is similar if not identical to Ki-1 [179].

A number of investigators have studied specimens from HD patients for immunoglobulin and T-cell receptor gene rearrangement. In one study, four of eight different HD biopsy specimens demonstrated a faint rearrangement of the T~ receptor [180]. No rearrangement of the Ig genes was detected in any of the samples. In another study, minor clonal populations (~ 1%), as demonstrated by immunoglobulin heavy chain gene rearrangement, were detected in 3 of 18 cases of HD. No cases of T~ rearrangement were demonstrated, and cases where high percentages of RSCs were present did not demonstrate, any rearrangements [181]. In another study, cases were selected solely on the basis of high RSC content, and at least one immuno­globulin gene was found to be rearranged in a clonal manner, raising the possibility that it occurred in the RSC [182]. It appears clear from these studies that rearrangement of either Ig or T~ receptor genes may occur in biopsy specimens from HD patients. There is little concordance as to which genes rearranged between these studies and minimal supportive evidence that the rearrangements are actually in the RSC population.

In summary, the cellular origin for HD remains unclear. Although Ia and T9 antigen staining of RSC and HC cells have been reported, Ki-1 antigen expression may prove to be the most useful immunologic marker for this disease.

RSC and HC constitute only a small proportion of cells within the tissue of Hodgkin's disease. Recent efforts are directed toward characterizing the remaining cells and has been recently reviewed [183]. Using in situ tech­niques, it appears that lymphoid tissues involved with HD are heterogeneous in immunohistologic make-up; some cases demonstrate numerous T-lym­phocytes with few B-cells, while others exhibit prominent follicles of poly­clonal B-lymphocytes and only small numbers of T-cells within these follicles. In two studies [160, 183], these B-cell-rich cases were of the lymphocyte­predominant type. In specimens containing T-lymphocytes, RSC and HC tend to appear in areas of heaviest T-cell infiltration, suggesting a relation­ship. Many cells within areas of T-cell infiltration are Ia or TlO positive; this

21

suggests that these T-lymphocytes are activated. Several investigators have demonstrated that most HD-associated T-lymphocytes are of the helper cell subset [157, 160, 183-185]. Genotyping for immunoglobulin and T-cell receptor rearrangements may lead to a better understanding of the cellular origin of this disease, but current data have not been conclusive.

Conclusion

The exciting advances in molecular biology and hybridoma technology over the past ten years have led to major advances in our understanding of the cellular origin of lymphoma and leukemia and will likely lead to a better understanding of the etiology of these diseases. Utilizing these techniques it is now possible to more accurately diagnose and classify these disorders; these data may also have therapeutic implications. It is also possible to use molecular probes to detect minimal residual disease.

References

1. Evans RI, Faldetta TJ, Humphreys RE, Pratt OM, Yunis EJ, Schlossman SF: Peripheral human T cells sensitized in mixed leukocyte culture synthesize and express la-like anti­gens. J Exp Med 148:1440, 1978.

2. Fu SM, Chiorazzi N, Wang CY, Montazeri G, Kunkel HG, Ko HS , Gottlieb AB: Ia­bearing T lymphocytes in man. Their identifieation and role in the generation of allogenic helper activity. J Exp Med 148:1423, 1978.

3. Winchester RJ, Ross GO, Jarowski CJ, Wang CY, Halper J, Broxmeyer HE: Expression of la-like antigens on human granulocytes during early phases of differentiation. Proc Natl Acad Sci USA 74:4012, 1977.

4. Ritz J, Pesando JM, Notis-McConarty J, Lazarus H, Schlossman SF: A monoclonal antibody to human acute lymphoblastic leukemia antigen. Nature 283:583, 1980.

5. LeBien TW, BOlle DR, Bradley G, Kersey J: Antibody affinity may influence antigenic modulation of the common acute lymphoblastic leukemia antigen in vitro. J Immunol 129:2287, 1982.

6. Greaves MF, Hariri G, Newman RA, Sutherland DR, Ritter MA , Ritz J : Selective expression of the common acute lymphoblastic leukemia (gp100) antigen on immature lymphoid cells and their malignant counterparts. Blood 61 :628, 1983.

7. Nadler LM, Anderson KC, Marti G, Bates M, Park E, Daley JF, Schlossman SF: B4, a human B lymphocyte-associated antigen expressed on normal mitogen-activated, and malignant B lymphocytes. J Immunol 131:244, 1983 .

8. Stashenko P, Nadler LM, Hardy R , Schlossman SF: Characterization of a human B lymphocyte specific antigen. J Immunol 125:1678, 1980.

9. Nadler LM, Stashenko P, Hardy R, Van Agthoven A, Terhorst C, Schlossman SF: Characterization of human B cell-specific antigen (B2) distinct from B1. J Immunol 126:1941, 1981.

10. Nadler LM, Boyd AW, Park E, Anderson KC, Siaughenhoupt B, Thorley-Lawson OA, Schlossman SF: The n cell restricted glycoprotein (B2) is the receptor for Epstein-Barr virus (EBV). Blood 64:78a (abstract), 1984.

11. Abramson CS, Kersey JH, LeBien TW: A monoclonal antibody (BA-I) reactive with cells of human B lymphocyte lineage. J Immunol 126:83, 1981.

22

12. Pirruccello Sl, LeBien TW: Monoclonal antibody BA-l recognizcs a novel human leuko­cyte cell surface sialoglycoprotein complex. 1 Immunol 134:3962, 1985.

13. Brooks DA, Beckman I, Bradley l, McNamara Pl, Thomas ME, Zola H: Human lym­phocyte markers defined by antibodies derived from somatic cell hybrids. I A hybridoma secreting antibody against a marker specific for human B lymphocytes. Clin Exp Immunol 39:477, 1980.

14. Brooks DA, Beckman GR, Bradley l, McNamara Pl, Thomas M, Zola H: Human lymphocyte markers defined by antibody derived from somatic cell hybrids. IV. A mono­clonal antibody reacting specifically with a subpopulation of human B lymphocytes. J Immunol 126:1373, 1981.

15. Anderson KC, Park EK, Bates MP, Leonard RC, Hardy R, Schlossman SF, Nadler LM: Antigens on human plasma cells identified by monoclonal antibodies. 1 Immunol 130: 1132, 1983.

16. Anderson KC, Bates MP, Slaughenhoupt B, Schlossman SF, Nadler LM: A monoclonal antibody with reactivity restricted to normal and neoplastic plasma cells. 1 Immunol 132:3172, 1984.

17. Foon KA, Todd RF III: Immunologic classification of leukemia and lymphoma. Blood 69:1-31, 1986.

18. Nadler LM: B cell/leukemia panel workshop: Summary and comments. In: Reinherz EL, Haynes BF, Nadler LM, Bernstein ED (eds): Leucocyte Typing, New York, Springer­Verlag 1986, p 3.

19. Haynes BF: Summary of T cell studies performed during the Second International Work­shop and Conference on Human Leukocyte Differentiation Antigens. In: Reinherz EL, Haynes BF, Nadler LM, Bernstein ID (eds): Leucocyte typing, New York, Springer­Verlag, 1986, p 1.

20. Reinherz EL, Schlossman SF: The characterization and function of human immunore­gulatory T lymphocyte subsets. Immunol Today 2:69, 1981.

21. Kung PC, Goldstein G, Reinherz EL, Schlossman SF: Monoclonal antibodies defining distinctive human T cell surface antigens. Science 206:347, 1979.

22. Reinherz EL, Schlossman SF: Regulation of the immune response-inducer and suppressor T-Iymphocyte subsets in human beings. N Engl 1 Med 303:370, 1980.

23. Van Wauwe J, Goossens l, Decock W, Kung P, Goldstein G: Suppression of human T­cell mitogenesis and E-rosette formation by the monoclonal antibody OKTIIA. 1 Im­munol 44:865, 1981.

24. Howard FD, Ledbetter lA, Wong J, Bieber CP, Stinson EB, Herzenberg LA: A human T lymphocyte differentiation marker defined by monoclonal antibodies that block E-rosette formation. 1 Immunol 126:2117, 1981.

25. Engleman EG, Warnke R, Fox RI, Diley J, Benike Cl, Levy R: Studies of a human T lymphocyte antigen recognized by a monoclonal antibody. Proc Natl Acad Sci USA 78:1791,1981.

26. Royston I, Majda lA, Baird SM, Meserve BL, Griffiths lC: Human T Cell antigens defined by monoclonal antibodies: The 65,000-dalton antigen of T cells (T65) is also found on chronic lymphocytic leukemia cells bearing surface immunoglobulin. J Immunol 125: 725, 1980.

27. Haynes BF, Mann DL, Hemler ME, Schroer lA, Shelhamer JH, Eisenbart GS, Strominger lL, Thomas CA, Mostowski HS, Fauci AS: Characterization of a monoclonal antibody that defines an immunoregulatory T cell subset for immunoglobulin synthesis in humans. Proc Natl Acad Sci USA 77:2914, 1980.

28. Link M, Warnke R, Finlay J, Amylon M, Miller R, Dilley J, Levy R: A single monoclonal antibody identifies T-cell lineage of childhood lymphoid malignancies. Blood 62:722, 1983.

29. Vodinelich L, Tax W, Bai Y, Pegram S, Capal P, Greaves MF: A monoclonal antibody (WTI) for detecting leukemias of T-cell precursors (T-ALL). Blood 62:1108,1983.

30. Uchiyama T, Broder S, Waldmann TA: A monoclonal antibody (anti-Tac) reactive with activated and functionally mature human T cells. I. Production of anti-Tac monoclonal

23

antibody and distribution of Tac( +) cells. 1 Immunol 126:1393, 1981. 31. Reinherz EL, Morimoto C, Fitzgerald KA , Hussey RE , Daley lF, Schlossman SF:

Heterogeneity of human T4+ inducer T cells defined by a monoclonal antibody that delineates two functional subpopulations . 1 Immunol 128:463, 1982.

32. Leder P: The genetics of antibody diversity. Sci Amer 246:102, 1982. 33. Croce CM, Shander M, Martinis T, Cicurel L, D'Ancona GG, Dolby TW, Koprowski H:

Chromosomal location of the human immunoglobulin heavy chain genes. Proc Natl Acad Sci USA 76:3416, 1979.

34. Malcolm S, Barton P, Murphy C, Ferguson-Smith MA, Bentley DL, Rabbitts TH: Localization of human immunoglobulin x light chain variable region genes to the short arm of chromosome 2 by in situ hybridization. Proc Natl Acad Sci USA 79:4957, 1982.

35. McBride OW, Hieter PA, Hollis GF, Swan D, Otey MC, Leder P: Chromosomal location of human kappa and lambda immunoglobulin light chain constant region genes, 1 Exp Med 155:1480, 1982.

36. Kirsch IR, Morton CC, Nakahara K, Leder P: Human immunoglobulin heavy chain genes map to a region of translocations in malignant B lymphocytes. Science 216:301, 1982.

37. Seidman JG, Leder P: The arrangement and rearrangement of antibody genes. Nature 276:790, 1978.

38. Korsmeyer Sl: Hierarchy of immunoglobulin gene rearrangements in B-cellleukemias. In: Waldmann TA, (moderator): Molecular genetic analyses of human lymphoid neoplasms: Immunoglobulin genes and the c-myc oncogene. Ann Intern Med 102: 497 , 1985 , pp 499-502.

39. Kitchingham GR, Rovigatti U, Mauer AM, Melvin S, Murphy SB, Stass S: Rearrange­ment of immunoglobulin chain genes in T cell acute lymphoblastic leukemia. Blood 65:725, 1985.

40. Acuto 0, Reinherz, EL: The human T-cell receptor structure and function. N Engl 1 Med 312: 1100, 1985.

41. Hedrick SM, Cohen Dr, Nielsen EA, Davis MN: Isolation of cDNA clones encoding T cell specific membrane associated proteins. Nature 308:149, 1984.

42. Hedrick SM, Nielsen EA, Kavalen 1, Cohen Dr, Davis MN: Sequence relationships between putative T cell receptor polypeptide and immunoglobulins. Nature 308: 153, 1984.

43. Yanagi Y, Yoshikai Y, Leggett K, Clark SP, Aleksander I, Mak TW: A human T cell­specific cDNA clone encodes a protein having extensive homology to immunoglobulin chains. Nature 308: 145, 1984.

44. Isobe M, Erikson 1, Emanuel BS, Nowell PC, Croce CM: Location of gene for ~ subunit of human T-cel receptor at band 7q35, a region prone to rearrangements of T-cells. Science 228:580, 1985.

45. 10nes C, Morse HG, Kao F-T, Carbone A, Palmer E: Human T-cell receptor a-chain genes: Location on chromosome 14. Science 228:83, 1985.

46. Minden MD, Toyonaga B, Ha K, Yanagi Y, Chin B, Gelfand E, Mak T: Somatic re­arrangement fo T-cell antigen receptor gene in human T-cell malignancies. Proc Natl Acad Sci USA 82:1224, 1985.

47. O'Connor NTJ, Wainscoat lS, Weatherall Dl, Gatter KC, Feller AC, Isaacson P, Jones D, Lennert K, Pallensen G, Ramsey A, Stein H, Wright DH, Mason DY. Rearrangement of the T-cell-receptor ~ gene in the diagnosis of lymphoproliferative disorders. Lancet 1:1295,1985.

48. Flug F, Pelicci P-G, Bonetti F, Knowles DM II, Dalla-Favera R: T-cell receptor gene rearrangements as markers of lineage and clonality in T cell neoplasma. Proc Natl Acad Sci 82:3460, 1985.

49. Tawa A, Hozumi N, Minden M, Mak TW, Gelfand EW: Rearrangement of the T-cell receptor ~-chain gene in non-T-cell, non-B-cell acute lymphoblastic leukemia of child­hood. N Engl 1 Med 313: 1033, 1985.

50. Waldmann T A, Davis MM, Bongiovanni KF, Korsmeyer SJ: Rearrangements of genes for the antigen receptor on T cells as markers of lineage and clonality in human lymphoid

24

neoplasms. N Engl J Med 313:776, 1985. 51. Sen L, Borella L: Clinical importance of Iymphoblasts with T markers in childhood acute

leukemia. N Engl J Med 292:828, 1975. 52. Brouet JC, Valensi F, Daniel MT, Flandrin G, Preud'homme JL, Seligmann M: Immuno­

logical classification of acute lymphoblastic leukemias: Evaluation of its clinical signi­ficance in a hundred patients. Br J Haematol 33:319, 1976.

53. Flandrin G, Brouet Jc, Daniel MT, Preud'homme JL: Acute leukemia with Burkitt's tumor cells: A study of six cases with special reference to lymphocyte surface markers. Blood 45:813, 1975.

54. Chessels JM, Hardisty RM, Rapson NT, Greaves MF: Acute lymphoblastic leukemia in children: Classification and prognosis. Lancet 2:1307, 1977.

55. Greaves MF, Brown G, Rapson NT, Lister TA: Antisera to acute lymphoblastic leukemia cells. Clin Immunol Immunopathol 4:67, 1975.

56. Foon KA, Billing RJ, Terasaki P, Cline MJ: Immunologic classification of acute lympho­blastic leukemia: Implications for normal lymphoid maturation. Blood 56:1120, 1980.

57. Vogler LB, Crist WM, Bockman DE, Pearl ER, Lawton AR, Cooper MD: Pre-B Cell leukemia. A new phenotype of childhood lymphoblastic leukemia. N Engl J Med 298:972, 1978.

58. Crist W, Boyett J, Roper M, Pullen J, Metzgar R, Van Eys J, Ragab A, Starling K, Vietti T, Cooper M: Pre-B cell leukemia responds poorly to treatment: A pediatric oncology group study. Blood 63:707, 1984.

59. Reinherz EL, Kung PC, Goldstein G, Levey RH, Schlossman SF: Discrete stages of human intrathymic differentiation. Analysis of normal thymocytes and leukemic Iympho­blasts of T-cell lineage. Proc Nat! Acad Sci (USA) 77:1588, 1980.

60. Schroff RW, Foon KA, Billing RJ, Fahey JL: Immunologic classification of lymphocytic leukemia based on monoclonal antibody-defined cell surface antigens. Blood 59:207,1982.

61. Roper M, Crist WM, Metzgar R, Ragab AH, Smith S, Starling K. Pullen J, Leventhal B, Bartolucci AA, Cooper MD: Monoclonal antibody characterization of surface antigens in childhood T-cell lymphoid malignancies. Blood 61:830, 1983.

62. Korsmeyer SJ, Arnold A, Bakhshi A, Ravetch JV, Siebenlist U, Hieter PA, Sharrow SO, LeBien TW, Kersey JH, Pop lack DG, Leder P, Waldmann TA: Immunoglobulin gene rearrangement and cell surface antigen expression in acute lymphocytic leukemias of T cell and B cell precursor origins. J Clin Invest 71:301, 1983.

63. Nadler LM, Korsmeyer SJ, Anderson KC, Boyd AW, Siaughenhoupt B, Park E, Jensen J, Coral F, Mayer RJ, Sallan SE, Ritz J, Schlossman SF: B cell origin of non-T cell acute lymphoblastic leukemia. A model for discrete stages of neoplastic and normal pre-B cell differentiation. J Clin Invest 74:332, 1984.

64. Chan LC, Pegram SM, Greaves MF: Contribution of immunophenotype to the classifica­tion of differential diagnosis of acute leukemia. Lancet 1 :475, 1985.

65. Sobol RE, Royston I, LeBien TW, Minowada J, Anderson K, Davey FR, Cuttner J, Schiffer C, Ellison RR, Bloomfield CD: Adult acute lymphoblastic leukemia phenotypes defined by monoclonal antibodies. Blood 65:730, 1985.

66. The non-Hodgkin's lymphoma pathologic classification project, National Cancer Institute sponsored study of classification of non-Hodgkin's lymphomas. Summary and description of a working formulation for clinical usage. Cancer 49:2112, 1982.

67. Rappaport H, Winter WJ, Hicks EB: Follicular lymphoma. A re-evaluation of its position in the scheme of malignant lymphoma, based on a survey of 253 cases. Cancer 9:792, 1956.

68. Aisenberg AC, Wilkes BM, Harris NL: Monoclonal antibody studies in non-Hodgkin's lymphoma. Blood 61:469, 1983.

69. Anderson KC, Bates MP, Siaughenhoupt BL, Pinkus GS, Schlossman SF, Nadler LM: Expression of human B cell-associated antigens on leukemias and lymphomas: A model of human B cell differentiation. Blood 63: 1424, 1984.

70. Ritz J, Nadler LM, Bhan AK, Notis-McConarty J, Pesando J, Schlossman SF: Expression

25

of common acute lymphoblastic leukemia antigen (CALLA) by lymphomas of B-cell and T-cell lineage . Blood 58:648 , 1981.

71. Koziner B, Filippa DA, Mertelsmann R, Gupta S, Clarkson B, Good RA, Siegel FP: Characterization of malignant lymphoma in leukemia phase by multiple differentiation markers on leukemic cells. Correlation with clinical features and conventional morpho­logy. Am 1 Med 63:556, 1977.

72. Koziner B, Kempin S, Passe S, Gee T, Good RA, Clarkson BD: Characterization of B-cell leukemias: A tentative immunomorphological scheme. Blood 56:815, 1980.

73. Horning Sl, Doggert RS, Warnke RA, Dorfman RF, Cox , RS, Levy R: Clinical relevance of immunologic phenotype in diffuse large cell lymphoma. Blood 63:1209, 1984.

74. Freedman AS, Boyd AW, Anderson KC, Fisher DC, Pinkus GS, Schlossman SF, Nadler LM: Immunologic heterogeneity of diffuse large cell lymphoma. Blood 65 :630, 1985.

75. Nathwani BN, Kim H, Rappaport H: Malignant lymphoma lymphoblastic. Cancer 38:964, 1976.

76. Rosen Pl, Feinstein DI, Pattengale PK, Tindle BH, Williams AH, Cain Ml , Bonorris lB, Parker lW, Lukes Rl : Convoluted lymphoma in adults, a clinicopathologic entity . Ann Intern Med 89:319, 1978.

77. laffe ES, Berard CW: Lymphoblastic lymphoma: A term rekindled with a new precision. Ann Intern Med 89:415, 1978. .

78. Bertness Y, Kirsch I, Hollis G , lohnson B, Bunn PA lr: T-cell receptor gene rearrange­ments as clinical markers of human T-cell lymphomas. N Engl 1 Med 313:534, 1985 .

79 . Bernard A, Boumsell L, Reinherz EL, Nadler LM, Ritz 1, Coppin H, Richard Y, Yalensi F, Dausset 1, Flandrin G, Lemerle 1, Schlossman SF: Cell surface characterization of malignant T cells from lymphoblastic lymphoma using monoclonal antibodies: Evidence for phenotype differences between malignant T cells from patients with acute lympho­blastic leukemia and lymphoblastic lymphoma. Blood 57:1105, 1981.

80. Bennett 1M, Catovsky D, Daniel M, Flandrin G , Galton DA, Gralnick HR, Sultan C: Proposals for the classification of the acute leukemias. Br 1 Haematol 33:451, 1976.

81. Magrath I: Lymphocyte precursors and neoplastic counterparts in vivo and in vitro. In: Berard CW (moderator): A multidisciplinary approach to non-Hodgkin's lymphomas. Ann Intern Med 94:218, 1981, p 226.

82. Aisenberg AC: New genetics of Burkitt's lymphoma and other non-Hodgkin's lym­phomas. Am J Med 77:1083, 1984.

83. Benjamin D, Bazar LS, lacobson Rl: Identification of stages of maturation arrest in Burkitt 's lymphoma and induction of differentiation with teleocidin: Implications for the origin of Burkitt's lymphoma and a new classification. Blood 66:185, 1985 (abstract).

84. Waldron lA, Leech JH, Glide AD, Flexner 1M, Collins RD: malignant lymphoma of peripheral T-lymphocyte origin. Cancer 40:1604, 1977.

85. Green lP, York lC, Consar lB , Mitchell RT, Flexner 1M, Collins RD, Stein RS: Peripheral T-cell lymphoma: A clinicopathologic study of 42 cases. J Clin Oncol 2:788, 1984.

86. Reynolds CW, Foon KA: Ty-lymphoproliferative disease and related disorders in man and experimental animals: A review of the clinical, cellular and functional characteristics. Blood 64:1146, 1984.

87. Newland AC, Catovsky D, Linch D, Cawley lC, Beverly P, Sam Miguel IF, Gordon­Smith EC, Blecher TE , Shahriari S, Yaradi S: Chronic T cell lymphocytosis: A review of 21 cases . Br 1 Haematol 58:433 , 1984.

88. Kadin ME, Kamoun N, Lamberg 1: Erythrophagocytic Ty lymphoma. A clinical patho­logic entity resembling malignant histiocytosis. N Engl 1 Med 304:648, 1981.

89. Loughran TP, Kadin ME, Starkebaum G , Abkowitz lL, Clark EA, Disteche C, Lum LG, Slichter Sl: Leukemia of large granular lymphocytes: Association with clonal chromo­somal abnormalities and autoimmune neutropenia, thrombocytopenia, and hemolytic anemia. Ann Intern Med 102:169, 1985.

90. Aisenberg AC, Krontiris , TG, Mak TW, Wilkes BM: Rearrangement of the gene for the

26

beta chain ot the T-cell receptor in T-cell chronic lymphocytic leukemia and related disorders. N Engl J Med 313:530, 1985.

91. Schein P: Clinical features. In: Lutzner M (moderator): Cutaneous T-cell lymphomas, the Sezary syndrome, mycosis fungoides and related disorders. Ann Intern Med 83:534, 1975, p 542.

92. Bunn PA, Huberman MS, Whang-Peng J, Schechter GP, Guccion JG, Matthews MJ, Gazdar AF, Dunnick R, Fischmann B, Ihde DC, Cohen MH, Fossieck B, Minna JD: Prospective staging evaluation of patients with cutaneous T-cell lymphoma: Demonstra­tion of a high frequency of extracutaneous dissemination. Ann Intern Med 93:223, 1980.

93. Brouet JC, Flandrin G, Scligmann M: Indications of the thymus-derived nature of the proliferating cells in six patients with Sezary's syndrome. N Engl J Med 289:341,1973.

94. Weiss LM, Hu E, Wood GS, Moulds C, Cleary ML, Warnke R, Sklar J: Clonal re­arrangements of T-ccll receptor genes in mycosis fungoides and dermatopathic lympha­denopathy. N Engl J Med 313:539, 1985.

95. Kung PC, Berger CL, Goldstein G, Logerto P, Edelson RL: Cutaneous T cell lymphoma: Characterization by monoclonal antibodies. Blood 57:261, 1981.

96. Broder S, Edelson RL, Lutzner MA, Nelson DL, MacDermott RP, Durm ME, Goldman CK, Meade BD, Waldmann TA: The Sezary syndrome. A malignant proliferation of helper T cells. J Clin Invest 58:1297, 1976.

97. Poiesz BJ, Ruscetti FW, Gazdar AF, Bunn PA, Minna JD, Gallo RC: Detection and isolation of type C retrovirus particles from fresh and cultured lymphocytes of a patient with cutaneous T cell lymphoma. Proc Natl Acad Sci USA 77:7415, 1980.

98. Bunn PA Schechter GP, Jaffe E, Blayney D, Young RC, Matthews MJ, Blattner W, Broder S, Robert-Guroff M, Gallo RC: Clinical course of retrovirus-associated adult T-cell lymphoma in the United States. N Engl J Mcd 309: 257, 1983.

99. Bunn PA: Clinical features. In: Broder S (moderator): T-cell Iymphoproliferative syn­drome associated with human T-cell leukemia/lymphoma virus. Ann Intern Med 100:543, 1984, p 543.

100. Jaffe E: Histopathology. In : Broder S (moderator): T-cell Iymphoproliferative syndrome associated with human T-cell leukemia/lymphoma virus. Ann Intern Med 100:543, 1984, p 546.

101. Morimoto C, Matsuyama T, Oshige C, Tanaka H, Hercend T, Reinherz EL, Schlossman SF: Functional and phenotypic studies of Japanese adult T cell leukemia cells. ] Clin Invest 75:836, 1985.

102. Tsudo M, Uchiyama T, Uchino H, Yodoi J: Failure of regulation of Tac antigen-TCGF receptor in adult T-cell leukemia cells by the anti-Tac monoclonal antibody. Blood 62:1014, 1983.

103. Uchiyama T, Sagawa K, Takatsuki K, Uchino H: Effect of adult T-cell leukemia cells on pokeweed mitogen-induced normal B-cell differentiation. C1in Immunol Immunopathol 10:24, 1978.

104. Jaffe ES: Non-Hodgkin's lymphomas as neoplasms of the immune system. In: Berard CW (moderator): A multidisciplinary approach to non-Hodgkin's lymphoma. Ann Intern Med 94:218, 1980, p 222.

105. Gale RP, Foon KA: Chronic lymphocytic leukemia: Recent advances in biology and treatment. Ann Intern Med 103:101 , 1985.

106. Preud'homme] , Seligmann M: Surface bound immunoglobulins as a cell marker in human Iymphoproliferative diseases. Blood 40:777, 1972.

107. Giardina SL, Schroff RW, Woodhouse CS, Abrams PG, Rager HC, Kipps TJ, Morgan AC, Foon KA: The generation of monoclonal anti-idiotype antibodics to human B cell­derived leukemias and lymphomas. ] Immunol 135:653, 1985.

108. Fialkow P, Najfeld V, Reddy A, Singer J, Steinmann L: Chronic lymphocytic leukemia: Origin in a committed B-Iymphocyte progenitor. Lancet 2:444 , 1978.

109. Han T, Ozer H , Bloom M, Sagawa K, Minowada J: The presence of monoclonal cyto­plasmic immunoglobulins in leukemic B cells from patients with chronic lymphocytic

27

leukemia. Blood 59:435, 1982. 110. Totterman TH, Nilsson K, Sundstrom C: Phorbol ester-induced differentiation of chronic

lymphocytic leukemia cells. Nature 288:176, 1980. 111. Caligaris-Cappio F, Pizzolo G, Chilosi M, Bergui L, Semenzato G, Tesio L, Morittu L,

Malavasi F, Gobbi M, Schwarting R, Campana D, Janossy G: Phorbol ester induces abnormal chronic lymphocytic leukemia cells to express features of hairy cell leukemia. Blood 66: 1035, 1985.

112. Aisenberg A, Wilkes BM: Relevance of surface markers in chronic lymphocytic leukemia to acute lymphocytic leukemia. Cancer Res 41:4810, 1981.

113. Gupta S, Pahwa R, O'Reilly R , Good RA , Siegal FP: Ontogeny of lymphocyte sub­populations in human fetal liver. Proc Natl Acad Sci USA 73:919, 1976.

114. Ross GD, Polky MJ: Specificity of human lymphocyte complement receptors. J Exp Med 141:1163, 1975.

115. Foon KA , Billing RJ, Terasaki PI: Dual Band T markers in acute and chronic lympho­cytic leukemia. Blood 55: 16, 1980.

116. Caligaris CF, Gobbi M, Bojill M, Janossy G : Infrequent normal B lymphocytes express features of B-chronic lymphocytic leukemia. J Exp Med 155:623, 1982.

117. Miller RA , Gralow J: The induction of Leu-I antigen expression in human malignant and normal B cells by phorbol myristic acetate (PMA). J Immunol 133:3408, 1984.

118. Patrick CW, Libnoch JA, Milson T, Kortright KH, Keller RH: Common variant of B cell chronic lymphocytic leukemia is defined by the monoclonal antibody TO\. Blood 64(abst): 195a, 1984.

119. Brouet JC, Preud'homme JL, Seligmann M, Bernard J: Blast cells with monoclonal surface immunoglobulin in two cases of acute blast crisis supervening on chronic lympho­cytic leukemia. Br Med J 4:23, 1973.

120. Splinter TAW, Born-van Noorloos AA, van Heerde P: CLL and diffuse histiocytic lymphoma in one patient: Clonal proliferation of two different B cells. Scand J Haematol 20:19, 1978.

121. van Dongen JJM, Hoojkaas H, Michels JJ, Grosueld G, de Klein A, van der Kwast TH, Prins MEF, Abels J, Hageijer A: Richter's syndrome with different immunoglobulin light chains and different heavy chain gene rearrangements. Blood 64:571, 1984.

122. Melo JV , Foroni L, Babapulle B, Lewis DS, Bennett M, Robinson D, Parreira L, Luzzatto L, Catovsky D: Prolymphocytic leukemia of B cell type: Rearranged immuno­globulin (lg) genes with defective Ig production. Blood 66:391 , 1985.

123. Brouet JC, Flandrin G, Sasportes M, Preud'homme JL, Seligmann M: Chronic lympho­cytic leukemia of T-cell origin. Immunologic and clinical evaluation in eleven patients. Lancet 2:890, 1975.

124. Catovsky D, Okos A , Wilts haw F, Galetto J, Galton DAB, Stathopoulos G: Prolympho­cytic leukemia of Band T cell type. Lancet 2:232, 1973.

125. Bouroncle BA, Wiseman BK, Doan CA: Leukemic reticuloendotheliosis. Blood 13:609, 1958.

126. Jansen J, LeBien TW, Kersey JH: The phenotype of the neoplastic cells of hairy cell leukemia studies with monoclonal antibodies. Blood 59:609, 1982.

127. Anderson KC, Boyd AW, Fisher DC, Leslie D, Schlossman SF, Nadler LM: Hairy cell leukemia : A tumor of pre-plasma cells. Blood 65:620, 1985.

128. Cleary ML, Wood GS, Warnke R, Chao J, Sklar J: Immunoglobulin gene rearrangements in hairy cell leukemia. Blood 64:99, 1984.

129. Korsmeyer SJ, Greene WC, Cossman J, Hsu S-M, Jensen JP, Neckers LM, Marshall SL, Bakhshi A, Depper JM, Leonard WJ, Jaffe ES, Waldmann TA: Rearrangement and expression of immunoglobulin genes and expression of Tac antigen in hairy cell leukemia. Proc Nat! Acad Sci USA 80:4522, 1983.

130. Halper J , Fu SM, Wang CY, Winchester R, Kunkel HG: Patterns of expression of human ' la-like' antigens during the terminal stages of B cell development. J Immunol 120:1480, 1978.

28

131. Durie BGM, Grogan TM: CALLA-positive myeloma: An aggressive subtype with poor survival. Blood 66:229, 1985.

132. Bhan AK, Nadler LM, Stashenko P, McCluskey RT, Schlossman SF: Stages of B cell differentiation in human lymphoid tissue. J Exp Med 154:737, 1981.

133. Foon KA, Schroff RW, Fligiel S: Monoclonal antibody characterization of plasmacytoma cells associated with T lymphocytes with a suppressor phenotype. Clin Immunol Immuno­pathol 25:59, 1982.

134. Lukes RJ, Butler 11: The pathology and nomenclature of Hodgkin's disease. Cancer Res 26:1063, 1966.

135. Kaplan HS: Hodgkin's Disease (2nd ed). Cambridge MA, Harvard University Press, 1980.

136. Whitelaw DM: Chromosome complement of lymph node cells in Hodgkin's disease. Can Med Assoc J 101:74, 1969.

137. Seif GSF, Spriggs AI: Chromosome change in Hodgkin's disease. J Natl Cancer Inst 39:557, 1967.

138. Hansen NE, Karle H: The elusive Hodgkin cell. Scand J Haematol 26:353, 1981. 139. DeVita VT: Lymphocyte reactivity in Hodgkin's disease: A lymphocyte civil war. N Engl

J Med 289:801, 1973. 140. Biniaminov M, Ramot B: Possible T lymphocyte origin of Reed-Sternberg cells. Lancet

1 :368, 1974. 141. Leech J: Immunoglobulin-positive Reed-Sternberg cells in Hodgkin's disease. Lancet

1:265, 1973. 142. Garvin AJ, Spicer SS, Parmley RT, Munster AM: Immunohistochemical demonstration of

IgG in Reed-Sternberg and other cells in Hodgkin's disease. J Exp Med 139:1077, 1974. 143. Kadin ME, Newcom SR, Gold SB, Stites DP: Origin of Hodgkin's cell. Lancet 2:167,

1974. 144. Taylor CR: The nature of Reed-Sternberg cells and other malignant reticulum cells.

Lancet 2:802, 1974. 145. Taylor CR: An immunohistological study of follicular lymphoma, reticular cell sarcoma

and Hodgkin's disease. Eur J Cancer 12:61, 1976. 146. Taylor CR: Upon the nature of Hodgkin's disease and the Reed-Sternberg cell. Recent

Results Cancer Res 1:214, 1978. 147. Stein H, Papadimitriou CS, Bouman H, Lennert K, Fuchs J: Demonstration of immuno­

globulin production by tumor cells in non-Hodgkin's and Hodgkin's malignant lymphomas and its significance for their classification. Recent Results Cancer Res 1:158, 1978.

148. Anagnostou D, Parker JW, Taylor CR, Tindle BH, Lukes RJ: Lacunar cells of nodular sclerosing Hodgkin's disease. Cancer 39:1032, 1977.

149. Stuart AE, Williams ARW, Habeshaw JA: Rosetting and other reactions of the Reed­Sternberg cell. J Pat hoi 122:81, 1977.

150. Stuart AE, Dewar AE: Properties of anti-hairy cell serum. Br J Haematol 41:163, 1979. 151. Stuart AE, Jackson E, Morris CS: The reaction of xenogenic and monoclonal antisera

with Reed-Sternberg cells. J Pathol 137:129, 1982. 152. Brooks JS: Leukophagocytosis by Sternberg-Reed cells in Hodgkin's disease. N Engl J

Med 300:1115, 1979. 153. Resnick GD, Nachman RL: Reed-Sternberg cells in Hodgkin's disease contain fibronec­

tin. Blood 57:339, 1981. 154. Katz DR: The macrophage in Hodgkin's disease. J Pathol 133:145, 1981. 155. Olsson L: On the natural biology of the malignant cells in Hodgkin's disease. Int J

Radiation Oncology Bioi Phys 11:37-48, 1985. 156. Olsson L, Behnke 0, Pleibel N, D'Amore F, Werdelin 0, Fry KE, Kaplan HS: Establish­

ment and characterization of a cloned giant cell line from a patient with Hodgkin's disease. JNCI 73:809-830, 1984.

157. Dorreen MS, Habeshaw JA, Stansfeld AG, Wrigley PF, Lister TA: Characteristics of Sternberg-Reed and related cells in Hodgkin's disease: An immunohistological study. Br J

29

Haematol 49:469, 1984. 158. Stein H, Gerdes J , Schwab U, Lemke H , Mason DY, Ziegler A, Schienle W, Diehl V:

Identification of Hodgkin and Sternberg-Reed cells as a unique cell type derived from a newly-detected small-cell population. Int J Cancer 30:445, 1982.

159. Watanabe S, Shimosata Y, Sata Y, Shimoyama M, Tobinai K, Mikata A: Immunologic characterization of Reed-Sternberg cells and other cell-components in lymph nodes with Hodgkin's disease. Acta Pat hoi Jpn 34:241, 1984.

160. Poppema S, Bahn AK, Reinherz EL, Posner MR, Schlossman SF: In situ immunologic characterization of cellular constituents in lymph nodes and spleens involved by Hodgkin's disease . Blood 59:226, 1982.

161. Kadin .ME: Possible origin of the Reed-Sternberg cell from an interdigitating reticulum cell. Cancer Treat Rep 66:601, 1982.

162. Stein H, Mason DY, Gerdes J, O'Connor N, Wainscoat J, Pallesen G, Gatter K, Falini B, Delsol G , Lemke H, Schwarting R , Lennert K: The expression of the Hodgkin 's disease associated antigen Ki-l in reactive and neoplastic lymphoid tissues: Evidence that Reed­Sternberg cells and histiocytic malignancies are derived from activated lymphoid cells. Blood 66:848, 1985.

163. Pizzolo G, Chilosi M, Sememzato G, Caligaris-Cappio F, Fiore-Donati L, Perona G, Janossy G: Immunohistological analysis of Tac antigen expression of tissues involved by Hodgkin 's disease . Br J Cancer 50:415, 1984.

164. Sutherland R, Delia D , Schneider C, Newman R, Kemshead J, Greaves MF: A ubiquit­ous membrane glycoprotein defined by monoclonal antibody (OKT9) is the transferring receptor. In: Knapp W (ed): Leukemia Markers, New York, Academic Press, 1981, p 157.

165. Kadin ME, Stites DP, Levy R, Warnke R: Exogenous immunoglobulin and the macro­phage origin of Reed-Sternberg cells in Hodgkin's disease. N Engl J Med 299: 1208, 1978.

166. Mir R , Kahn LB: Immunohistochemistry of Hodgkin 's disease. Cancer 52:2064, 1983. 167. Landaas TO, Godal T, Halvorsen TB: Characterization of immunoglobulin in Hodgkin

cells. Int J Cancer 20:717, 1977. 168. Papadimitriou CS, Stein H, Lennert K: The complexity of immunohistochemical staining

pattern of Hodgkin and Sternberg-Reed cells-demonstration of immunoglobulin, albumin, u[ antichymotrypsin and lysozyme. Int J Cancer 21 :531, 1978.

169. Payne SV, Wright DH, Jones KJ, Judd MA: Macrophage origin of Reed-Sternberg cells: An immunohistochemical study. J Clin Pathol 35:159, 1982.

170. Stuart AE, Volsen SG, Zola H: The reactivity of Reed-Sternberg cells with monoclonal antisera at thin section and ultrastructural levels. Pathology 141:71, 1983.

171. Linch DC, Jones HM, Berliner N, MacLennen K, O'Flynn K, Huens ER, Kay LA, Goff K: Hodgkin-cell leukaemia of B-cell origin. Lancet 1:78, 1985.

172. Ford RJ, Mehta S, Davis F , Maizel AL: Growth factors in Hodgkin's disease. Cancer Treat Rep 66:633 , 1982.

173. Strauchen JA: Lectin receptors as markers of lymphoid cells. II. Reed-Sternberg cells share lectin-binding properties of monocyte-macrophages . Am J Pathol 116:370, 1984.

174. Kaplan HS, Gartner S: 'Sternberg-Reed' giant cells of Hodgkin's disease: Cultivation in vitro , heterotransplantation, and characterization as neoplastic macrophages. Int J Cancer 19:511 , 1977.

175. Diehl V, Kirchner HH, Burrichter H , Stein H, Fonatsch C, Gerdes J, Schaadt M, Heit W, Uchanska-Ziegler B, Ziegler A, Heintz F, Sueno K: Characteristics of Hodgkin's disease­derived cell lines. Cancer Treat Rep 66:615, 1982.

176. Stein H, Uchanska-Ziegler B, Gerdes J , Ziegler A, Wernet P: Hodgkin and Sternberg­Reed cells contain antigens specific to late cells of granulopoiesis. Int J Cancer 29:283, 1982.

177. Beckstead JH, Warnke R, Bainton DF: Histochemistry of Hodgkin's disease. Cancer Treat Rep 66:609, 1982.

178. Schwab U, Stein H, Gerdes J, Lemke H, Kirchner H, Schaadt M, Diehl V: Production of

30

a monoclonal antibody specific for Hodgkin and Sternberg-Reed cells of Hodgkin's disease and a subset of normal lymphoid cells. Nature 299:65, 1982.

179. Hecht TT, Longo DL, Cossman, J, Bolen JB, Hsu S-M, Israel M, Fisher RI: Production and characterization of a monoclonal antibody that binds to Reed-Sternberg cells. J Immunol 134:4231, 1985.

180. Griesser H, Feller A, Lennert K, Minden M, Mak TW: Rearrangement of the ~ chain of the T cell antigen receptor and immunoglobulin genes in lymphoproliferative disorders. J Clin Invest 78:1179-1184, 1986.

18]. Knowles DM, II, Neri A, Pelicci PG, Burke JS, Wu A, Winberg CD, Sheibani K, Dalla­Favera R: Immunoglobulin and T-cell receptor ~-chain gene rearrangement analysis of Hodgkin's disease: Implications for lineage determination and differential diagnosis. Proc Nat! Acad Sci 83:7942-7946.

182. Weiss LM, Strickler JG, Hu E, Warnke RA, Sklar J. Immunoglobulin gene rearrange­ments in Hodgkin's disease. Human Pathol 17: 1009-1014, 1986.

183. Abdulaziz Z, Mason DY, Stein H, Gatter KC, Nash JR: An immunohistological study of the cellular constituents of Hodgkin's disease using a monoclonal antibody panel: Histo­pathology 8:1, 1984.

]84. Morris CS, Stuart AE: Reed-Sternberg/lymphocyte rosette: Lymphocyte subpopulations as defined by monoclonal antibodies. J Clin Pat hoi 37:767, 1984.

185. Aisenberg AC, Wilkes BN: Lymph node T cells in Hodgkin's disease: Analysis of suspen­sions with monoclonal antibody and rosetting techniques. Blood 59:552, 1982.

2. Immunologic approaches to the classification of non-Hodgkin's lymphomas

Thomas M. Grogan, Catherine M. Spier, Lynne C. Richter, and Catherine S. Rangel

Background

Morphologic considerations

This chapter employs the working formulation (WF) of the non-Hodgkin's lymphomas as a histologic base, because its ten morphologic categorizations are well defined and of established clinical utility, identifying prognostic gradations relevant to therapy [1]. Using this firm histologic base, the im­munologic characteristics within each category are emphasized. As this chapter repeatedly demonstrates, many immunologic subtypes exist within certain WF categories, and these cannot always be predicted by morphologic classification [2]. Figure 1A details the WF categorizations relevant to the previously widely used Rappaport classification [3]. Figure 1B details some closely related entities to be discussed. Some of these represent lymphoma­related diseases, others represent newer categorizations derived from im­munologic studies (e.g., peripheral T-cell lymphoma).

It is emphasized that integration of morphologic and immunologic find­ings is pivotal to the proper diagnosis of lymphomas. Our newly acquired technology (e.g., monoclonal antibodies, flow cytometry) sometimes pro­duces as yet inexplicable results. This chapter suggests the latter is being systematically reduced. Nonetheless, in the circumstance of unresolved dilemmas, the traditional histologic diagnosis takes precedence over immunotyping.

Fundamentally, the diagnosis of malignant lymphoma is determined by the precepts of established histologic criteria [1, 3]. Immunotyping, as an independent exercise, serves the purpose of identifying the lineage of the cells involved. That is, having determined the histologic magnitude of the problem, immunotyping reveals the relevant chemistry of the involved cells. Almost invariably, this means assessing at once the phenotype of the neo-

This chapter is gratefully dedicated to Costan Berard, Gist Farr, Stephen Jones, Jack Layton, Henry Rappaport, and Roger Warnke.

BellI/eli, I .M. and Foon. K.A., (eds.), Immunologic Approaches 10 rhe Classificarion and Managemem of Lymphomas and Leukemias. © 1988 Kluwer Academic Publishers. ISBN978-1-4612-8965-4. All rig/us reserved.

32

J:'igure IA.

W:>rking Fonnulation

Low-grade lyrrphana Malianant l~, small l~t1c

Consistent with crL PlaslMcytoid

Malignant lyrTl'lhclna, follicular smell cleaved cell Malignant l~, follicular, · mixed, Sl"Mll

cleaved and large cell

Intermedlate~rarle l~ Malianant lyrmhana, follicular large cell Maliqnant l~, diffu..o;e small cleaved cell Haliqnant lyrl1)tona, diffuse mixed, small-

and larae cell Ppithelioid ceU CCJI1lX)nent

Malignant l~. diffuse la('ge cell Sclerosi~

Hiqh-qrade lymphana Maliqnant lymphana, large cell, inmJnoblastic

Plasrnacytoid P.pithelioid cell corponent

Malignant lyrnptona, 1 ytrptohlast ic Convoluted CP 11 NG'OConvoluted cell

Mal iQnant lyYl1)hcrna, small nonc1eaved cell Rucki tt I 5 Burkitt's-like

Figure lB.

Mditional Entities

Diffuse l~ic, well differentiated ",ith and without plaarMCytoid features

Nodular l~tic, p:')Orly differentiated Nodular mixad, l~tlc and histiocytic

NorIular histiocytic Oiffuse lyrrrhocytic, JXX)rly differentiated Diffuse mix", l~tic and histiocytic

Diffuse histiocytic

Diffuse histiocytic

Diffuse lyrrcb::lblastic with am without convolutions

Burkitt's and diffuse undifferentiated (non­Burkitt's)

""antle zone lymphanaj"rntennediate" lymphocvtic l'(ll1)hana Plasmacytcma P-1ycosis funooides Signet-drg cell lymphoma Peripheral T-cell lymphaM

F.nithelioid cell ccrnponent With aberrant mvelonoiesis Sirrolatinq maliqnant histiocytosis

HTI..V-associated mal ignant lymphanas Compos i te ll'1"phcrna

Figure I . (A) Classification of malignant lymphomas based on the working formulation classifi­cation. All major categories are listed; some minor modifiers are excluded (e.g., fibrosis). (8) Additional types of malignant lymphoma which are not part of the working formulation classification but are to be discussed.

plastic cells and the host response cells. In this complex phenotypic circum­stance, morphology remains critical. For example, the large blastic cells judged to be malignant may be of one phenotype, while the small lymphoid cells judged to be reactive may be of another phenotype (see figure 65). Immunotyping does not suspend morphologic or medical judgment; it en­hances it at the level of additional, adjunctive chemical detail.

Immunologic principles

General comments. As neoplasms of the immune system, lymphomas are best understood biologically by studying their immunologic properties. In

Figure 2. Description of the monoclonal antibodies and cytochemical stains used in this chapter. Included are the antibody subclass, the specificity of each antibody in the immune system, its cluster designation (CD), and the molecular weight of the antigen. The vendor's name and address are also given at the bottom of the figure.

oil ;:: ~ ~

An

libo

dyl

En

zym

e (

Su

b-c

lass

)

an.;-

kapp

a (I

gG

l)

nti-

Iam

bd

a

(Ig

Gl)

I"n.

;-lg

M

(1g

G')

!ant

;-lg

G

(IgG

2b)

~nti

-lgA

l (I

gG3)

!an.

;-lg

A2

(1gG

2b)

I<n.

;-lg

D

(1gG

3)

~b-B

l (1

gG2b

)

i"nb

-B2

(IgM

)

~n';-B4

(Ig

Gl)

~nti

-l12

(l

gG

l)

~n';

-l1'

(I

gG2B

)

~nti

-lI6

(1

gG1)

!'nb

-CR

II

(IgG

2a)

~nb-

PCA-

l (l

yG2

a)

nli-

Leu

-1

(IgG

2a)

!'n

He

u-2

a

(lg

Gl)

nli-

Leu-

3ab

(1gG

1)

nti-le~

(lg

G')

nti-

leu

-Sb

(I

gG2a

)

ntt

-le

u-6

(1

gG2b

)

!l"t

i-le

u-7

(l

gM

)

an

Ha

u-g

(I

gG2a

)

nti

-le

u-1

1b

(I

gM)

-

Spe

cifiC

ity

Abb

<ev

MW

of A

g C

D

Ve

nd

or

An

tib

od

yfE

nzy

me

(S

ub

-cla

ss)

Spo

c;t;c

;,y

-B

CE

ll P

AN

EL

GR

AN

UlO

CY

TE

/HfS

TIO

CY

TE

PA

NE

L

light

cha

in ~K

K

B.D

. a

nti

·Le

u-M

I (I

gM)

Mon

os.lg

ranu

locy

1n

lMl

ligh

t ch

ain

IgA

I

B.D

. a

nti-

leu

-M3

(1

gG2b

) M

onos

.lmec

:roph

r.ges

lM

3

heav

y ch

ain

10M

M

B

.D.

an

li-L

eu

-M5

(1

gG2b

) M

on

ol.

/ma

cro

ph

-oe

s lI

A5

heav

y ch

ain

IgG

G

B

.D.

MY

-7

(1gG

1)

Mye

Iocy

t ••

MY

·7

subc

lass

spe

cifiC

9A

1 A

I B

.D.

But

yrat

e E

ster

ase

Mon

os.lh

istio

cyte

A

NB

E

subc

lass

sp

eci

fic 1

9A2

A2

B.D

. C

hlor

oace

l ...

G

ranu

locy

te

NA

SDCE

he

avy

ch

ain

100

0

B.D

. S

udan

Bte

c:k

B

Gra

nulo

cyte

S

BB

Pan

B c

ell

Bl

3SK

C

D2(

C

ou

lt.r

C3d

re

tep

tor

(rp

) B

2 ,.

OK

C

O2

' C

oulte

r P

RO

LIF

ER

AT

ION

/AC

TIV

AT

ION

PA

NE

L

Pan

8 c

ell

B4

40

&8

OK

C

Ol

Cou

lter

_n

-2

rp

(Ig

GI)

(lg

G2

a I

nter

leuk

in-2

rp

Il-2

Pan

8 c

e41

112

95K

C

Ol

B.D

. an

ti-H

LA

-DR

(II

) (t

gG2a

) H

LA-D

R (

Ia)

la

Pa

n e

cell

l1'

1 SO

l<

CD

22

B.D

. an

ti-Ir

ansf

erri

n rp

(1g

G2a

) ra

ns""

'in r

ecep

tor

TR

F

Pan

B c

ell

11

6

35K

C

D2

( B

.D.

"'_

Ag

(Ig

GI)

rolif

erat

ive

Ag

Ki-8

1

C3d

rp

CR

II ,.

SK

C

02

1

B.D

.

Pla

sma

cell

asso

c. A

g P

CA

-'

Co

ulte

r

T C

ELL

PA

NE

L

MIS

CE

LLA

NE

OU

S P

AN

EL

Pan

T c

ell

11

67K

C

D5

B

.D

Inti

-Tdt

(1

gG)

erm

ina'

Tr.

.,sl

era

se

Tdt

T c

yto

l0xi

c/su

pp

resI

Or

l2

32

64

3K

C

D6

B

.D.

an

ti-C

alla

(1

gG1

)(Ig

G2a

) C

omm

on A

ll A

g C

alla

T h

elpe

rlin

duce

r l3

SS

K C

o.

B.D

an

ti-O

RC

(I

gM)

Den

driti

c A

.. ic

urum

ce

ll D

RC

Pan

T

l'

22-2

8K

C0

3a

B.D

. d

Uu

oo

Cy

Io C

om

(Ig

GI)

.u

cocy

Ie c

om

mo

n A

g.

lC

E-r

oset

te r

ecep

tor

l5

'5-S

OI<

C

O2

B.D

. an

ti-Le

u-8

(1gG

2a)

&B

_l2

&l3

LB

Th

ymo

cyte

s/la

ng

em

an

s l6

'91

< C

D'

B.D

. a

nti-

CR

3 r

p (1

gG2a

) e

R3

racl

lMo

r lI

5

NK

/K c

atls

l7

"'

11O

K

B.D

. an

ti-H

PC

A-1

(1

gG1)

H

uman

Pro

geni

tor

cell

Ag

HP

CA

-l

Pan

T

19

'OK

C

D7

B

.D.

anti-

Cyt

oker

at;n

(1

gG2a

) yt

ok ..

. atin

8.1

8.1

9 C

K

NK

/neu

trop

hils

l1

1

SO

-7O

K

CD

IS B

.D.

Co

ntr

ol

~%BSA;n P

BS

C

on.

-

B.O

.::B

ecto

n-O

icki

nson

P.O

. 8

0x

7375

, M

ou

nta

in V

iew

, C

aWfo

mia

M0

09

C

ou

ller=

Co

utt

er

Imm

tM"lO

logy

"0

We

st 2

0th

St .

....

, H

iale

ah,

Flo

rid

a 3

30

10

O

ako

::O

ako

Cor

pora

tion

22 N

orth

Milp

u 5

1 ...

. , S

an

ta B

arba

ra,

Ca

lifo

rnia

93

10

3

Ga

mm

a::

Ga

mm

a B

iolo

gca

ls 3

70

0 M

ang.

.-n

Ad,

Hou

ston

, T

exa

s n

09

2

P.L

Bio

che

mic

als

::P

ha

rma

cia

po

l B

ioch

em

cals

, In

c. 1

03

7 W

est

McK

inle

y A

ve,

Milw

auke

e, W

iSC

Ons

in 5

32

05

K::

Kik

Jdat

tons

MW

of A

g C

D

CD

IS

55

__

C

D2S

28-_

18

01«8

01<

CO

lO

CD

11

311,4

3.501

<

Ve

nd

or

B.D

.

B.D

.

B.D

.

Cou

lter

B.D

.&C

ou.

B.D

.

B.D

.

Da

ko

P.L

B;o

Dtk

o &

BD

Da

ko

Dak

o

B.D

.

B.D

.

B.D

.

B.D

.

Ga

mm

a

v:>

v:>

34

recent years, the depth of our immunologic understanding of lymphomas has been greatly enhanced by new reagents (e.g., monoclonal antibodies), new techniques (e.g., tissue section immunohistochemistry and immuno­fluorescence), and new machines (e.g., cytocentrifuge, epifluorescence microscopes, and flow cytometers [4-6]. Presently, the most fruitful method of study uses a battery of monoclonal antibodies (40 in our laboratory) to elucidate the lineage of lymphomas. Using both cell suspension and tissue section methods with a full battery of monoclonal antibodies, virtually every lymphoid lesion should demonstrate surface markers, with rare exceptions [7], provided fresh, viable, unfixed tissue is employed as detailed below.

Presently, the variety of monoclonal antibodies commercially available is great and appears ever expanding. This chapter does not account for this huge diversity of reagents. Rather, it provides the perspective gained from the use of 40 monoclonal antibodies on 400 lymphomas typed at the U niver­sity of Arizona lymphoma phenotyping laboratory. Figure 2 gives the de­scriptive qualities of these 40 monoclonal antibodies (MoAB). Besides giving the primary antigenic specificity of each antibody, we indicate, where known, the recent cluster designation (CD) so that comparability to other MoAB may be noted [8]. You will observe our use of more than one MoAB for a single cluster designation (e.g., we use both anti-B] and Leu 16 to detect CD20). This duplication is considered useful on two accounts: 1) when in agreement, it provides confirmation of results, 2) when in disagreement, it is a reminder that different MoAB to the same antigen may vary in binding characteristics (e.g., avidity) [7]. Figure 2 also identifies the immunoglobulin subtype/isotype of MoAB, which sometimes is relevant to patterns of non­specific staining as detailed below.

immunologic phenotyping. As demonstrated throughout this chapter, lym­phomas studied with MoAB have a wide range of antigenic expression, which can be related to the specific phase of lymphoid cell ontogeny from which the neoplastic cell derives [4-6]. As illustrated in figures 3 and 4, each lymphoma has an immunotypic profile that relates to its state of B-ceJl and T-cell differentiation .

In figure 3, we may find B-ceJllymphomas which are of immature (pre-B, pre-pre-B-type), intermediate type (B-cell), 'activated' immunoblastic forms

Figure 3. B-cell phenotypes. Diagrammatic representation of the rangc of B-cell antigenic expression in both normal B-cell development (ontogeny) and in the B-cell neoplasms derivcd from each stage of ontogeny. Abbreviations: AUL, acute undifferentiated leukemia; C-all, common acute lymphoblastic leukemia; pre-B-all, pre-B-cell acute lymphoblastic leukemia; B­ali, B-cell acute lymphoblastic leukemia; BL, Burkitt's/Burkitt's-like leukemia/lymphoma; SLL, small lymphocytic lymphoma; CLL, chronic lymphocytic leukemia; SCL, small cleaved cell lymphoma; LCL, Large cell lymphoma; WM, Waldenstr6m's macroglobulinemia; HCL , hairy cell leukemia; myeloma , multiple myeloma; MR, mouse rosette . See figure 2 for descriptions of antibodies.

35

Figure 3.

36

Figure 4.

37

(see large cell lymphoma), and mature plasmacytoid forms (plasmacytoma, myeloma [4-6]. Immature B-cell forms are more likely to express terminal deoxynucleotidyl transferase (Tdt) , common acute lymphocytic leukemia antigen (CALLA), and cytoplasmic mu; whereas activated B-cell forms frequently lack surface immunoglobulin expression; while mature forms usually express plasma cell (PC) and plasma cell associated (PCA) antigens. Note the co expression of some non-B-cell antigens among B-cell neoplasms: Leu 1 in small lymphocytic lymphoma [9], Leu M5 (histiocytic antigen) in hairy cell leukemia [10].

In normal tissues, the B-cells are usually polyclonal or polytypic, with kappa, lambda, gamma, mu, delta, and alpha immunoglobulin chain-bearing cells admixed [11, 12]. Kappa usually outnumbers lambda by 2: 1 or 3: 1. In neoplastic B-celllymphomas of mid phase and late phase, there is frequently light chain restriction and usually heavy chain restriction, indicating a mono­clonal B-cell proliferation consistent with neoplasia [13].

Similarly, T-cell lymphomas may be immature (thymic-like) or mature (post-thymic or peripheral T-cell type) (see figure 4) [4-6]. The latter may be further divided into mature, nonactivated, or activated types (figure 4). Immature T-cell lymphomas frequently coexpress helper (Leu 3) and T­suppressor/cytotoxic antigens (Leu 2) along with Leu 6 and Tdt [4-6]. Mature T-cell neoplasms frequently express one functional subset antigen (Leu 2 or 3) to the exclusion of the other [4-6]. Mature activated T­lymphomas possess Ia and other activation antigens (transferrin receptor) in abundance [14, 15]. Surprisingly, mature T-cell lymphomas may lose subset antigens and pan-T-antigen in an aberrant idiosyncratic manner, indicating a 'novel' phenotype related to neoplasia [14, 15].

In normal tissues, the only substantial source of immature T-cells is the thymic cortex [4-6]. Hence, finding Tdt and simultaneous Leu 2/3/6 expres­sion outside of the thymus is highly indicative of an immature T-cell malig­nancy [16]. Outside of the thymus, most normal peripheral T-cells have subset status with T-helper cells generally outnumbering T-suppressor/ cytotoxic cells by a ratio of 3: 1 to 4: 1 [4-6, 11, 12]. In contrast to this normal situation, most mature T-cell neoplasms possess one subset antigen to the exclusion of the other, suggesting monoclonality [15]. Monoclonality among T-cells is particularly difficult to judge, since some reactive condi­tions may show a remarkable preponderance of one subset without neo­plastic outcome [17]. Recently, judgment of T-cell monoclonality has been greatly aided by assessment of T-antigen receptor genes to establish clonal rearrangements [18]. More recently, MoAB Leu 8, directed at an immuno-

Figure 4. T-cell phenotypes. Diagrammatic representation of the range of T-cell antigenic expression in both normal T-cell ontogeny and the T-cell neoplasms derived from the develop­mental phases. See figure 2 for descriptions of antibodies. Abbreviations: ALL, acute lympho­blastic leukemia; LBL, lymphoblastic lymphoma.

38

regulatory subset of T-helper and T-suppressor/cytotoxic cells, promises to offer phenotypic evidence of monoclonality [19-21]' Specifically, the Leu 8 antigen is normally found as a 'hodge-podge' of subsets (e.g., Leu 3+8+, Leu 3+8- ) in an inflammatory infiltrate, while a T-cell neoplasm has either Leu 8+ cells alone or Leu 8- cells alone, indicating monoclonality [19-21]. Further evidence of T-cell neoplasia may come from aberrant, novel T-cell phenotypes. These suggest a neoplastic process , since comparable aberrant phenotypes are not usual in inflammatory conditions [22].

Methods.

General comments. Both suspension and tissue section methods of immune assessment are necessary for full categorization of lymphomas. The suspen­sion method allows both strict quantitation of antigen-bearing cells and 'sizing' of cells. It also allows immunologic assessment of malignant cells that naturally occur as a suspension, for example, in a malignant pleural effusion. Tissue section immunohistochemistry and immunofluorescence methods allow delineation of the topography of the lesion (e.g. , immuno­architecture). Immunoarchitectural analysis allows discrimination between neoplastic cells in follicular lymphoma and surrounding host response cells [23]. A confusing admixture of neoplastic and reactive cells may occur when suspension methods are used, but this pitfall is avoided with tissue section methods. Full immunophenotypic profile may be determined by using either serial tissue sections, or serial cytocentrifuge slides and/or flow cytometry coplots.

Tissue handling. Critical to obtaining a high degree of antigen detection are the methods of tissue handling [7]. For suspension work, we use freshly separated, unfixed cells suspended in RPM!. For tissue section work, tissue is snap-frozen in isopentane, quenched in liquid nitrogen, and then stored at -70°C [24]. Serial cryostat sections are then made for immunohistochemical staining. Some flow cytometry studies are also performed on frozen cells. In all cases, the viability of the tissue is of primary concern, since nonviable cells may passively and nonspecifically absorb the staining reagents. The use of frozen material for tissue section work is critical to antigen preservation [7, 24, 25]. Although it involves more time initially, snap-freezing results in fewer antigen negative cases compared to paraffin sections. Snap-frozen tissue, unlike paraffin sections, fully preserves antigenic expression, leading to greater assurance of immunotyping [7, 24, 25].

Paraffin sections are most frequently employed to study plasmacytic lesions. In this circumstance, cytoplasmic immunoglobulin (CIg) preserva­tion seems to benefit from fixation processes and is usually superior to frozen section determinations [7]. Clg detection is best in B5 fixative and less reliable in formalin fixation, with some evidence that formalin fixation may eventually degrade CIg fixation over time [25].

39

Detection methods. A great variety of detection reagents are available, including fluorochromes, enzymes like immunoperoxidase and alkaline phosphatase, and metallic conjugates like immunogold. The flow cytometry studies reported in this chapter employed indirect immunofluorescence with the fluorochrome, allowing laser detection [26], while most cytocentrifuge and tissue section studies employed immunoperoxidase methods. Our labor­atory in particular favors a three stage biotin-avidin immunoperoxidase system [14, 24].

Monoclonal versus polyclonal antisera. We generally favor monoclonal anti­bodies over polyclonal heterosera because of the added specificity and con­stancy of source provided by MoAB. Nonetheless, monoclonals may have their drawbacks, many of which are noted below. One particular drawback relates to the high specificity of MoAB. Because MoAB may detect such a restricted amino acid sequence (epitope), they may fail to detect the highly variable forms (polymorphisms) of certain antigens. The lambda immuno­globulin light chain is a case in point [27]. Monoclonal antilambda light chain may be unreactive in certain lambda-bearing neoplasms because the single malignant clone is an antigenic variant undetected by this particular MoAB. In this circumstance, 'old-fashioned' heterosera with their hetero­geneity of antilambda antibodies will more reliably detect lambda [27]. Thus, we frequently employ heterosera to detect surface Ig when MoAB are unreactive.

In another example, the T-helper lymphocytes of some patients may not mark with conventional OKT4 antibody directed at T-helper antigen [28]. Although these patients at first appear to lack T-helper cells, they have no clinical evidence of immunodeficiency. This paradox is explained by finding that their T-helper cells are identified by OKT4A and Leu 3A antibodies to T-helper antigen. It appears that the OKT4, OKT4A and Leu 3A antibodies each bind to different epitopes (amino acid sequences) on the complex T-helper cell molecule, and certain individuals lack, on a congenital basis, the epitope recognized by OKT4. In spite of this deficiency, they have fully functional T-helper cells [28]. Once again, the overly high specificity of a single MoAB may give false impressions.

Tissue section interpretation. The principles of tissue section immunotyping are revealed in figure 5. These sections from a splenic B-cell lymphoma show the immunotopographic relationship of the neoplastic B-cells (B1 +) to the reactive T-cells (Leu 1 +) in the peri arteriolar lymphoid sheath region. This method allows marriage of immunologic features with micro anatomic features [29]. When this same neoplasm is studied in serial sections (see figure 6), we reveal a monoclonal (kappa, mu restricted) immunoglobulin­bearing B-cell proliferation, indicating neoplasia. The worksheet in figure 6 shows the full phenotype of this B-cell neoplasm with expected pan-B antigen expression as well as common acute lymphocytic leukemia antigen

40

Figure 5. Immunotypographic relationship of neoplastic B-cells (B1 +) to reactive T-cells (Ll +, Leu 1 +) in a splenic lymphoma. (Grogan TM, Bever F , Jolley CS , et al: Immunoarchitecture of splenic lymphomas. Diag Immun 3:126-132, 1985.)

(CALLA) and leukocyte common antigen expression. The capacity to generate 40 antibody studies at once in 40 serial sections is largely facilitated by using the batch method of staining developed by Bindl and Warnke [30] .

Ig expression in tissue section produces complex patterns, and the means of immunoglobulin detection varies with its cellular location (see figure 7). When cytoplasmic, Ig detection is best with paraffin embedded material using heterosera (figure 7 A) [7]. In the surface location, Ig detection is best with frozen sections using MoAB (figure 7B) [7]. When the cells in question are pre-B-cells , their scant amount of cytoplasmic mu is best detected by cytocentrifuge slide preparations (see figure 7C) [31, 32].

Cell suspension immunotyping. Suspension typing has been strongly in­fluenced by machines (epifluorescence microscope, cytocentrifuge, and flow cytometer) . The advantages of these machines are revealed in figures 8-10. In figure 8, a cytocentrifuge preparation studied for immunoglobulin reveals the advantage of combining high power microscopy with immunoperoxidase technique. In figure 8A delicate surface Ig staining decorates lymphoid cells from a case of small lymphocytic lymphoma, while a plasma cell reveals more substantial cytoplasmic Ig staining. In figure 8B lymphoma cells of the pre-B-cell type are shown expressing scant, patchy cytoplasmic mu [31, 32]. Figure 9 compares flow cytometric and cytocentrifuge immunoperoxidase methods, revealing similar results. Each method has its advantages . Flow

41

1-CeII_ T.()oI_ ......... JAc.-.. + ~ 0 Lou , P.,.T 0 1I..a -.z

'-- Lou 2 T~ .- '- ~ .- IQIM Lou 311> T ....... ~ 0 TN' T_ .... IgG

0 Lou • ...... T "N - -All

fIlA · Lou&> E_ .... fIlA, 0 Lou I ~ ep , --IgD Lou 7

HI<.'( _ 0 T ..

T ..... _ .- 8 . - Lou' ...... T ..... Call ~AU) .- B. 0 Lou lib ---- ... DAC ~--.- e. p- IIIIOc.I_ ..... LC ~. eo.-AII ..... Lou '2 p- O IN. ....-.~ 0 Lout ..... Lou ,. p- O IN. ....-.~ 0 Lou '$ ~.'-*" .- Lou 'I - 0 IN. ....-...,...""'- 0 ~, "'-""'colAO ... ~. C3dr ""'.7 ~ 0 ()( ~ fICA·, -"'00II ANIlE ....-.- 0 eo.-

Figure 6. Splenic B-cell lymphoma phenotype. Top: B-cell phenotype in serial frozen sections. Bottom: Worksheet depiction showing full B-cell phenotype and other panel results.

cytometry, using a laser to count fluorescent tagged cells, can quantitate positivity very rapidly and graphically depict the results [26]. Cytocentrifuge preparations preserve the morphologic features of all cells, positive and negative. The two methods complement each other and demonstrate the increasing flexibility in determining immunophenotypes of lymphomas with advances in technology. In figure 10, the use of flow cytometry coplots is illustrated. The phenotypic profile of a case of T-cell lymphoproliferative disease with a predominance of large granular lymphocytes is shown. The laser generated profile reveals coexpression of Leu 2, 4, and 7, suggesting a proliferation of T-cells expressing both T-suppressor/cytotoxic and natural killer cell antigens. Flow cytometry cop lots allow quantitative and qualita­tive inference of elaborate coexpressions resolving very complex cases. Both cell sorting and double labeling studies will further enhance this line of analysis.

Fig

ure

7.

Imm

unog

lobu

lin

(Ig)

det

ecti

on.

(A)

Cyt

opla

smic

Ig

dete

ctio

n in

pla

sma

cells

usi

ng

para

ffin

em

bedd

ed t

issu

e an

d he

tero

sera

. (B

) S

urfa

ce I

g de

tect

ion

usin

g fr

ozen

sec

tion

s an

d M

OA

b. (

C)

Cyt

opla

smic

mu

dete

ctio

n, a

s fo

und

in p

re-B

-cel

ls,

usin

g cy

toce

ntri

fuge

pre

para

­ti

ons

and

anti

mu

anti

body

.

.j:>

. tv

43

A B Figure 8. Cell suspension Ig detection. (A) Surface Ig decorates lymphoid cells from a small lymphocytic lymphoma, while a plasma cell reveals more substantial cytoplasmic Ig staining. (B) Pre-B-cells expressing scant, patchy cytoplasmic mu.

Immunologic diagnosis ('the bottom line'). This chapter emphasizes the application of a battery of monoclonal antibodies to determine immunologic phenotype. This includes panels directed at B-cells and T-cells, monocytes/ histiocytes/granulocytes, proliferation/activation antigens, and miscellaneous antigens (figures 2, 6) . Previous study has established that the battery gener­ated profile is able to offset the undue influence of one aberrant or false marker [14, 31]. The batteries also allow discernment of odd and idiosyn­cratic phenotypes, as found in peripheral T-cell lymphomas [14, 15]. Much as with a liver profile or a panel of antibodies in HLA typing, greater as­surance comes from discerning a pattern of antigenic expression [31].

The end result of phenotypic analysis using such panels is revealed in figure 11. Once a phenotype is derived, it may be compared to phenotypes from patients with the same diagnosis. The results, then, become part of a systematic matrix of results. As revealed in figure 11, in the example of small Iymphocytif lymphoma (SLL) , there appears to be a 'fingerprint' which may be considered characteristic of SLL: Coexpression of Leu 1 and pan-B (Bl) antigens [9]. Thus, if a patient has Leu lIBl coexpression we may place him within the SLL categorization. At this level, immunotyping is beginning to achieve status as an independent diagnostic variable.

44

Figure 9.

::J Q)

....J •

45

a. Leu 11 7%

l Leu 7 72%

L - , Leu 2 89%

l - Leu 3 7%

A leu 4 99%

L. __________ 81 0%

.k __________ Control

Figure /0. Flow cytometry cop lot of T-cell Iymphoproliferative disorder. This laser generated profile reveals coexpression of Leu 2, 4, and 7, suggesting a proliferation of T-cells, expressing both T -suppressor/cytotoxic and natural killer cell antigens.

If the notion that LlfBl coexpression is unique to SLL is challenged, and other lymphomas (all 400 Arizona cases) are searched for comparable phenotypes, three other groups, as shown in figure 11, are found: 1) mantle zone lymphoma, which is closely related to SLL; 2) a rare case of large cell lymphoma; 3) a rare case of small cleaved cell lymphoma [7]. Clearly, marker profiles are becoming systematic, achieving "fingerprint" status and becoming subject to computer and even mathematical analysis.

Limitations and pitfalls. As mentioned above, a number of factors may limit immunologic assessment. Probably the most significant is tissue handling. A delayed specimen may be a nonviable specimen , resulting in nonspecific absorption of staining agents [7]. In suspension, a dye exclusion test can exclude this possibility. However, this is not an option in tissue sections. The next most common error lies with freezing artifacts due to improper tissue freezing, usually due to crystal formation related to slow freezing. Freezing at -150°C in isopentane quenched in liquid nitrogen is ideal and should obviate this problem [24].

Occasionally, particularly in spleen and marrow immunotyping, there may be nonspecific staining due to the pseudoperoxidase activity of the hemoglobin in background erythrocytes [33, 34]. This problem is greatly

Figure 9. Flow cytometric (left) and cell suspension immunoperoxidase (right) analysis of lymphoblastic lymphoma cells.

46

B·CeIl Merkers T Cell Merkers

Cese Surfece Ig

61 L14CRII 64 L12 k A M GAD TRF la L1 L2 L3 L4 L5 L7 L9

SMALL LYMPHOCYTIC LYMPHOMA

1 + + + + + 2 + + + + + 3 + + + + + 4 + + + + + 5 + ± + + + 6 + ± + + + 7 ± ± + + + 8 + ± ± + + 9 + ± 0 ± +

10 + + + + + 11 + + + + +

12 + + + + + 13 + + + + + 14 + + + + + 15 + ± + + + 16 + + 0 + +

17 + + 0 + + 18 ± 0 0 0 0

19 0 0 0 0 0

0++ 0 0 0 0++ 0 0 0 o + ± 0 0 0 000100 + 0 + 0 0 ± + 0 + 0 0 0 + 0 0 ± 0 0 0++ 0 0 0 + 0 0 + 0 0 + 0 + 0 0 0 0++ 0 0 +

+ 0 + 0 0 + 0++ 0 0 ± 0++ 0 0 0 0++ 0 0 0

1+000

+ 0 + 0 0 0 + 0 + 0 0 0

000000

+ + ± + o + + + + + o + + + + + o + I + ± +

± + + + o +

+ 0 0 0 0 0 0 + 0 0 0 0 0 0 + 0 0 0 0 0 0 + 0 0 0 0 0 0 + 0 0 0 0 0 0 + 0 0 0 0 0 0 + 0 0 0 0 0 0 + 0 0 0 0 0 0 + 0 0 0 0 0 0 + 0 0 0 0 0 0 + 0 0 0 0 0 0

± 0 0 0 0 0 0 ± 0 0 0 0 0 0 ± 0 0 0 0 0 0

+±±oooooo ±+±OOOOOO

0+0000000 0+0000000

o+ooo+oo±

MANTLE ZONE L YMPHOMA/INTERMEDIATE CELL LYMPHOMA

20 + + + + + 21 + + + + + 22 + + + + + 23++0++ 24 + + + + +

0++00+ ++ +0+0000+ 0++000 ++ +00+00 ++ +0+000 ++

Leu 1 + LARGE CELL LYMPHOMA

25 + + + + + 26++0++ 27 + ± + + ± 28 + + ND + +

+ 0 + 0 0 + + 0 + 0 0 0 + 0 + 0 0 + 0++ 0 0 0

o + + + + +

ND+

+ 0 0 0 0 0 0 + 0 0 0 0 0 0 + 0 0 0 0 0 0 + 0 0 0 0 0 0 0000000

+ 0 0 0 0 0 0 + 0 0 0 0 0 0 + 0 0 0 0 0 0 + 0 0 0 0 0 0

Figure 11. Tabulation of small lymphocytic lymphoma (SLL) phenotypes . Note that 16 of 18 SLL reveal BI/Leu 1 coexpression. Among other lymphomas , only mantle zone and inter­mediate lymphocytic lymphoma reveal this coexpression with high incidence (4 of 5 cases). Rare « 3%) cases of large cell lymphoma (shown) and follicular lymphoma (not shown) had similar BlILeu 1 coexpressions. Human Pathol, 17:1126-1136, 1986.

47

alleviated by incubation in 0.01 % phenylhydrazine HCL solution prior to immunologic staining. This step facilitates the reduction of background staining, although it also may reduce the intensity of antigen detection by the conjugated antibody ]33, 34]. The endogenous peroxidase activity found in monocytes, histiocytes, granulocytes, mast cells, and eosinophils may create interpretive problems. This may be countered in two ways: 1) com­parison to the histiocytic or myeloid cell pattern detected with anti-Leu Ml, M3, M5, and MY7 or by 2) use of a weak hydrogen peroxide solution prior to MoAB application, although, once again, large amounts decrease antigen detection [33, 34].

Interpretation of staining for immunoglobulin [Ig] may prove especially difficult for several reasons. As illustrated in figure 12, some tissues may contain background interstitial immunoglobulin which may be difficult to discern from faint cellular staining [34, 35]. These cases may be more readily interpreted by attending to several factors: 1) in tissue, initial incubation of tissue section with cacodylate buffer to remove interstitial Ig; 2) in cell suspensions, multiple washings in RPM I and resuspension, as in figure 12; 3) simple microscopic observation [34, 35]. With regard to the latter, as pic­tured in figure 12, one notes that the background staining is irregular, not usually decorating cell surfaces; whereas, true surface Ig appears as a series of darkly stained interconnected rings.

Difficulty with Ig interpretation may also come from passive absorption of Ig by macrophages [7]. This pattern is perceived by comparison with the macrophage specific MoAB pattern. It has been noted that immunoglobulin findings tend to be most equivocal in poorly handled and poorly frozen tissue - a caution for proper and speedy handling of tissue [7]. Reagent factors may also confound Ig detection. The failure of some monoclonal antibodies to detect Ig chain polymorphism was mentioned above; con­sequently it is necessary to use polyclonal heterosera in Ig negative cases to ensure true Ig negativity [27]. Since some MoAB may experience non­specific staining related to their immunoglobulin class (e.g., an IgM MoAB may stain nonspecifically, while an IgG MoAB does not or vice versa), it is also necessary to use isotype matched control antibodies [7] (see figure 2).

Some nonspecific staining may be related to the organ system studied and the detection systems used [36, 37]. In figure 13A a liver biopsy is illustrated, showing nonspecific staining due to endogenous B-vitamin activity in the hepa­tocytes [35, 36]. The endogenous biotin is binding the avidin-horseradish­peroxidase reagent resulting in nonspecific hepatocyte staining. This endogenous activity may be obviated by using avidin/biotin blocking sera [36], as shown in figure 13B. Blocking then allows detection of the under­lying infiltrate, in this case sinusoidal hairy cell leukemia cells (figure 13B).

Occasionally, interpretation may be confounded due to the expression of Fc receptors on neoplastic cells (see figure 14) [38]. In this illustrated exam­ple, note from the cytometry coplots that the cells simultaneously express low intensity markers from multiple lineages, including myeloid, monocytic,

48

Ka a Lambda Figure 12. Interstitial immunoglobulin (Ig). Kappa light chains are found irregularly through­out this lymphoma section reflected in background Ig staining. Note this kappa Ig does not decorate cell surfaces (see suspension below). In contrast, specific lambda light chain staining decorates the lymphoma cell surfaces, appearing as a series of darkly stained interconnecting rings in tissue sections and as dark cells in suspension. (Spier CM, Grogan TM, Ficldcr K, et al: Immunophenotypes in 'well differentiated' lymphoproliferative disorders, with emphasis on small lymphocytic lymphoma. Human Pathol, 17:1126-1136, 1986.

T-cell, and B-ceiL However, notice the absence of reactivity with Leu 7. Checking with figure 2, note that all the reactive antibodies are of IgG subtype; whereas Leu 7 is of IgM subtype. This isotype pattern coupled with the low intensity of most markers suggests the presence of an Fc receptor to IgG. Confirmation of this possibility comes from two additional findings: 1) an anti-B2 was also negative, representing a second IgM monoclonal with

Figure 13. Liver biopsy showing (A) endogenous biotin activity and (B) its successful elimina­tion by blocking with avidin and then biotin blocking sera. (Verdi CJ, Grogan TM, Protell R, et al: Liver biopsy immunotyping to characterize lymphoid malignancies. Hepatology 6:6-13, 1986.)

49

Figure 13.

50

l

I

"

L leu70%

• HPCA197%

.l ________ 82 7%

L. leuM568%

MY454%

2 la 98%

leu344%

8453%

Control Figure 14. Flow cytometry coplot depicting multilineage staining due to surface Fc receptors.

negativity: 2) incubation of the cells in rabbit serum at 37°C for 30 minutes resulted in a loss of low intensity nonspecific staining [7]. Finally, after incubation, strong monocytic markers were found to persist. Monocytes are well described to express Fc receptors, so this pattern is in keeping with the monocytic lineage of these neoplastic cells [39].

It is important to appreciate that not all surface markers are lineage specific: Leu 1, 3, 6, and Leu M5 are good examples. The pan-T-marker, Leu 1, is also found in some B-cell neoplasms (e.g., small lymphocytic lymphoma and mantle zone lymphoma) [9, 40]. The combination of Leu 1 and pan-B-antigens and complement receptors identifies a fingerprint for this tumor [8]. Leu 3 (T-helper antigen) also reacts with macrophages and dendritic cells [41]. The latter T-helper antigen expression is faint but nonetheless may cause interpretive difficulty. Intriguingly, Leu 3 is estab­lished as the viral receptor for HTLV-III [42]. It has been surmised that T­helper expression on T-cells and histiocytes facilitates the entry of the lethal virus into these cells [42] . In the case of the T-helper cells, cell death ensues, understandably followed by immunodeficiency. In the case of the Leu 3+ macrophage and dendritic cell, the HTLV-III virus entry results in loss of this key cell pivotal to B-cell clonal expansion and may explain the follicle lysis seen in AIDS, as well as the associated B-cell deficiencies [43]. Finally, macrophage infection with HTL V -III may facilitate viral persistence in the lungs of AIDS patients [44].

In the case of Leu 6, this immature cortical thymocyte T-cell antigen is also expressed or is cross-reactive on Langerhans cells in the skin [45]. Leu

51

M5 is a macrophage specific antigen coexpressed with pan-B antigens on hairy cell leukemia cells (HCL) [10]. The coexpression of Leu M5 and pan­B-antigens appear to be a useful fingerprint to identify HCL [10].

In some cases of immature lymphoma/leukemia there may be multiple coexpressed B-cell and T-cell antigens, suggesting a rare, nonsensical bi­phenotypic lineage [46, 47]. This appears to represent lineage ambiguity or lineage infidelity - a concept which challenges the view that all lymphoid neoplasms derive from a discrete phase of B-cell or T-cell ontogeny [45, 46].

As revealed in figure 15, some monoclonal antibodies to lymphoid markers may detect the same amino acid sequence in cells of nonlymphoid lineage: Leu 4 on neurons, as well as T-cells [48]; Leu 7 on prostate cells [49] and Ewing's sarcoma cells [50, 51], as well as NK cells; OKMI and Leu M1 on carcinoma cells, as well as monocytes [52, 53]; CALLA on bile canaliculi and intestinal mucosa , as well as leukemic cells [54, 55]; B2 on keratino­cytes, as well as B-cells [56]. In some cases these may represent fortuitous cross-reactions, but in others, they may represent the same antigen site (e.g., epitope) . Some of these cross-reactions have proven intriguing. For example in figure 15A B2 (C3dr) staining is shown on keratinocytes, whereas it is normally associated with B-cells. Since C3dr is known to be the receptor for the Epstein-Barr (EB) virus, it is now speculated that this receptor may serve as the entry point of EB virus into the body [57, 58]. Specifically, C3dr expression in upper airway passages has been linked with EB virus entry and the causation of nasopharyngeal carcinoma [56]. The finding of CALLA expression in the bile canaliculi has also proven clinically relevant, since passive serotherapy with anti-CALLA linked to toxins could theoreti­cally wipe out liver secretory function [55]. This is a reminder that many cross-reactions remain to be explored. In the case of Leu 4 on neurons, several intriguing possibilities are raised. Perhaps this molecule may serve as a link between the chemistry of immune function and that of brain function. Perhaps Leu 4 or its equivalent is the cerebral viral receptor in certain dementing viral illnesses like AIDS-related dementia [59].

A few limitations of flow cytometry should also be appreciated. Firstly, in the usual indirect IF procedures, it is wise to recall that it is specifically surface antigen, not cytoplasmic antigen, which is detected [26, 46]. Hence by standard technique the pre-B-cells shown in figures 7, 8, and 71 might not be detected by the usual cytometry methods. Secondly, epithelial cells and stromal cells may not go into suspension. In tumors with an epithelial component (e.g., lymphocyte thymoma (see figures 54 and 55)) there may be a detection failure due to sampling error. Thirdly, the topography of the lesion is lost with admixture of neoplastic and host response cells.

ReprodUcibility. In spite of the numerous potential pitfall r and limitations, a recently completed double-blind study between institutions immunopheno­typing non-Hodgkin's malignant lymphomas indicates substantial reproduci­bility of antigenic determinations (> 93% among 1388 determinations) [60].

52

c ., . .

D,.' • . ~ .'-

Figurel5. Cross-reactivity of monoclonal antibodies in non lymphoid tissues. (A) Skin biopsy showing epithelial staining with C]dr. Cross-reactivity of antiCalla antibody with (B) hepatic bile canaliculi and (C), (D) surface of gastric epithelium.

A southeastern cancer study group has also established a greater than 90% success rate for immunotyping of transported lymphoma material [61]. Two additional studies similarly verify the stability of transported specimens [62, 63]. Recent collaborative lymphoma studies indicate a significant problem of reproducibility of histologic diagnosis between institutions; agreement ranged between 14% and 77% for various subtypes [64, 65]. In contrast, the studies above document a high agree of immunotypic reproducibility (> 90%), suggesting the promise of immunophenotyping in improving the diagnostic assessment of lymphomas [60].

53

Small lymphocytic lymphoma

General comments

As the only diffuse low-grade lymphoma in the working formulation classi­fication of malignant lymphomas, small lymphocytic lymphoma (SLL) is characterized by prolonged survival with or without treatment [1, 4].

The histologic hallmark of the typical case of SLL is a monotony of small, apparently mature, lymphocytes that flood the lymphoid tissue , replacing all normal features (figure 16A). Vaguely nodular aggregates of large cells, termed pseudofollicular growth centers (PFGCs) (figure 16B), are occa­sionally found in SLL; these are not known to have prognostic significance [9, 66]. Relevant to prognosis is the finding of plasmacytoid differentiation within SLL proliferations [66].

Immunologic considerations

The majority (> 95%) of SLLs are of B-cell origin [66]. B-SLLs express pan-B-antigens (Bl, B4, L12, L14, L16), surface immunoglobulin, and com­plement receptors (C3dr), placing them at the midpoint of B-cell maturation (figure 17) [9]. Characteristically, immunoglobulin expression is weak or even nondetectable [67]; alternatively, polyclonal interstitial immunoglo­bulins derived from reactive cells may be found to a greater degree than neoplastic cell surface Ig [9]. This difficulty of Ig assessment may be over­come by using a preliminary wash with a cacodylate buffer, which removes obscuring interstitial immunoglobulin and allows detection of the surface immunoglobulins [35]. The use of mononuclear cell suspensions is another effective method, since the repeated washings eliminate interstitial immuno­globulin (figures 12A and 12B) [9].

A well described finding in B-SLL, although not completely explained, is the coexpression of the pan-T-antigen Leu 1 with pan-B-antigens (figure 17) [40, 68-70]. When coupled with Ig expression, especially of both mu and delta heavy chains and C3dr (B2 ) expression, the coexpression of Leu 1 and pan-B-antigens can be considered the hallmark immunophenotype in SLL (figures 11, 17, 18) [9]. However, exceptions occur: SLL may occasionally not coexpress Leu 1, and other malignant lymphomas (e.g., MZL) may occasionally do so (figure 11) [68, 61, 71]. In regard to the B-SLLs that do not coexpress Leu 1, this lack appears to correlate with the plasmacytic 'differentiation' observed in those cases (figure 11) [9]. It appears that as the cells switch, or are stimulated, to cytoplasmic immunoglobulin production the ability to co express Leu 1 is lost [9, 71].

Immunologic studies of the PFGCs of SLL demonstrate a higher proli­ferative or Ki-67 activity relative to the surrounding areas, consistent with the perception that these represent growth centers (figure 19) [9] . PFGC also show evidence of activation, as evidenced by the strong staining for the

54

Figure 16.

55

Figure 17. Typical B-cell SLL immunophenotype. (Spier eM, Grogan TM, Fielder K, et al: Immunophenotypes in 'well differentiated' lymphoproliferative disorders , with emphasis on small lymphocytic lymphoma. Human Pathol, 17, 1126-1136, 1986.)

Figure 16. Small lymphocytic lymphoma, (SLL). The typical morphologic findings in SLL area shown at left. Pseudo follicular growth centers in SLL are shown at right.

56

• •• • • 84 - 84% .. ~ • • •

Figure 18. Coexpression of Leu 1 and pan-B (B4)-antigen in SLL demonstrated by flow cytometry.

transferrin receptor (TRF). Most PFGCs do not show any relation to residual germinal centers (GC), as evidenced by lack of staining for DRC. Rare PFGCs with DRC positivity may be found, suggesting a temporal relationship between PFGCs and GCs. This uncommon association may relate to the natural history of the disease, with a continual dropout of DRC accompanying PFGC growth [9].

Plasmacytic 'differentiation' in SLL (figure 20) has been related to poorer prognosis so that the delineation of a plasmacytic component may have clinical relevance [66]. However, it should not be assumed that the plasma­cytic component is necessarily neoplastic, since the plasmacytic component in some SLL cases has proven to be polyclonal, perhaps representing host response to neoplasia [9]. As mentioned above, these polyclonal plasma cells may secrete Ig in an interstitial pattern, obscuring surface Ig delinea­tion. The surface Ig expression, then, is best detected after a cacodylate wash or suspension wash, while CIg in plasma cells is best detected in fixed paraffin embedded tissue [7].

T-cell variants of SLL which are morphological\y indistinguishable from B-SLL have been described, but are uncommon, comprising 1%-5% of al\ SLL as discussed in the peripheral T-cel1 lymphoma section (figure 66) [72].

Histologic transformation of SLL

SLL may transform to a diffuse large cell lymphoma, with sudden clinical deterioration and foreshortened survival [4]. Some cases of diffuse large cell

•

. '~"'~'" '. , , \. .'<* ·~tt •

. ' , ;.

" ,

57

B

"

," "

"

c Figure 19. Immunophenotypic study of pseudofollicular growth centers (PFGCs) in SLL. (A) Histologic appearance of PFGCs, (8) ORC antibody showing demonstration of ORC within PFGCs . (C) Ki-67 antibody showing cell proliferation within a PFGC. (0) TRF antibody showing greater activation within PFGCs.

lymphoma have the immunophenotypic characteristics of SLL, that is, a B­cell neoplasm with coexpression of Leu 1, suggesting a derivation of this LCL from SLL [9].

Differential diagnosis of small lymphocytic lymphoma

1. Lymphocyte predominant (LP) Hodgkin's disease, diffuse subtype, may be easily confused with SLL upon cursory examination, especially since the small lymphocytes of either disorder may appear identical. Furthermore, Reed-Sternberg-like cells may be found in either disease [73]. However, the presence of Land H variants and mummified cells are present in LP Hodgkin's disease; PFGCs are present in SLL. The clinical and immunologic findings

58

Figure 20. Surface versus cytoplasmic Ig staining in plasmacytic variant of SLL. (A) plasma­cytic component in SLL, noted histologically. Surface staining for kappa (B I) and lambda (CI) light chain immunoglobulins, in frozen tissue sections, indicating a lambda+ monoclonal population. Cytoplasmic immunoglobulin detection on formalin-fixed, paraffin-embedded tissuc showed absent kappa (B2) and abundant lambda (C2) light chains. (Spier CM, Grogan TM, Fielder K, et al: Immunophenotypes in 'well differentiated' lymphoproliferative disorders, with emphasis on small lymphocytic lymphoma. Human Pathol, 17, 1126-1136, 1986.)

in the two diseases vary widely from each other and are less likely to be confused [73]. 2. Both mantle zone lymphoma and 'intermediate' lymphocytic lymphoma (MZLlIDL) may be confused with SLL, since both are comprised of small lymphoid cells with slight nuclear irregularity [74-77]. Immunologically,

59

these two lymphomas demonstrate a kinship to SLL, complete with the coexpression of Leu 1 and pan-B-antigens (figure 69). Clinically, it is not uncommon to find widespread disease at diagnosis [75]. Unlike most B-SLL, mantle zone lymphoma shows a propensity for involving the gastrointestinal tract, especially the duodenum [78]. Like SLL, these lymphomas have a prolonged course and survival [74]. See following mantle zone discussion for further details. 3. The diffuse form of small cleaved cell lymphoma (DSCL) may be mis­taken, on occasion, for SLL. However, close histologic examination reveal­ing small angular lymphoid cells will soon lead to the correct diagnosis. Immunologically, a rare case of DSCL may cause confusion if it coexpresses Leu 1, but the strong surface immunoglobulin expression is quite inconsis­tent with SLL [67, 69]. Clinically, the disease may be widespread in each entity, but survival in DSCL is shorter [1]. 4. Chronic lymphocytic leukemia (CLL) may be especially problematic since the cell morphology, appearance in tissue, and ability to transform to ag­gressive, large-cell disease, show marked similarities [66, 79]. While CLL show initial peripheral blood involvement, this may also occur in later stages of SLL [SO]. Each demonstrates prolonged survival, although the natural history of CLL may be slightly shorter [SO]. One immunologic test useful in distinguishing the two diseases is the mouse rosette assay (Mr) [81]; in CLL the value is characteristically greater than 60%, while in SLL it is less than 24% [81]. It is conjectured that the ability of the malignant CLL cell to form a rosette with mouse red blood cells correlates with its ability to circulate in the peripheral blood [81].

Follicular lymphomas

General comments

This common form of malignant lymphoma, comprising some 50% of adult non-Hodgkin's lymphocytic lymphomas, is represented in the working for­mulation as three cytologic variants of a single disease [1, 4]. Although the utility of these subset categorizations has been questioned by some [1], the three variants of follicular lymphoma retain separate status in the working formulation because of their differences in morphology, natural history, response to therapy, and median survival [1, 4, 82, 83].

The histologic basis of this distinction is revealed in figure 21. In all three variants, the characteristic architectural feature is the occurrence of neoplas­tic nodules throughout the lymph node with very little intervening cortical lymphoid tissue (figure 21A) [S4]. Cytologically, the three subtypes contain a variable proportion of small angular and large blastic lymphoid cells. The most common variant, comprised of a predominance of small angular lym­phoid cells, is called follicular small cleaved cell lymphoma (FSCL) (figure

60

Figure 21.

61

21B). At the other extreme, with a predominance of large cells, is follicular large cell lymphoma (FLCL) (figure 21C). Follicular mixed lymphoma (FML) (figure 21D) is an admixture of the latter two components [4, 82, 83]. Although as pictured this seems to be a straightforward scheme, consider­able debate exists as to the boundaries separating these subsets, with sub­stantial variability in diagnostic reproducibility among pathologists [64, 65].

Immunologic assessment of follicular lymphoma has delineated coexpres­sion of B-cell antigens and complement (C3dr) receptors, in close association with dendritic reticulum cells (DRC) , signifying a neoplasm derived from normal germinal center B-cell counterparts [10, 85-89]. Although some have considered all follicular lymphomas to be immunologically alike [65], recent study of activation and proliferation antigens suggests immunologic categorizations corresponding to the three cytologic variants [25, 90].

Follicular small cleaved cell lymphoma

Among the cytologic subsets of follicular lymphoma (FL), follicular small cleaved cell lymphoma (FSCL) demonstrates the most conspicuous and constant coexpression of surface immunoglobulin (Slg) and C3dr (B2) (figure 3) [11, 85-89]. Like the other FL subsets, there is constant pan-B-cell antigen expression (B1, B4, and Leu 12, 14, 16) as expected in a neoplasm of B-cell lineage [87, 89]. Since the follicular neoplastic cells occur in close association with dendritic reticulum cells (DRC), a germinal center origin is apparent [86, 88, 89]. The high degree of Slg and B2 coexpression help place FSCL in the mid phase of B-cell differentiation (figure 3) [88] . In contrast with FML, FSCL is both morphologically and immunologically homogene­ous. Compared to FML and FLCL, there is a lower proliferative index (Ki-67) and less expression of activation antigens (TRF, la, IL-2) [29,90]. There is variable expression of B-cell maturation antigens (e.g., IgD and CALLA), suggesting different FSCL may derive from slightly different points of B-cell ontogeny [29, 89, 91].

Figure 22 illustrates the conspicuous immunologic aberrancy of FSCL compared to normal reactive lymph node germinal centers. Themonotypic pattern of light chain restricted immunoglobulin expression documented therein contrasts strikingly with the usual polyclonal immunoglobulin ex­pression in reactive nodes [11, 13, 29, 86-89]. This monotypia signifies replacement of the nodal follicles by a single B-cell clone [13]. The strong immunoglobulin expression illustrated (kappa+, mu+, delta+) is characteris­tically associated with small cleaved cell lymphoma, as opposed to small lymphocytic lymphoma which has faint surface immunoglobulin expression

Figure 21. Histologic appearance of follicular lymphomas. (A) Neoplastic follicles replacing normal lymph node architecture. (B) Higher power view of follicular small cleaved cell malignant lymphoma. (C) Detail of follicular large cell lymphoma . (D) Follicular mixed cell lymphoma.

62

Figure 22. Comparison of follicular lymphoma and reactive lymph node immunophenotypes.

[67]. As illustrated, follicular lymphoma cells, like normal B-cells, can ex­press more than one heavy chain, although usually only one heavy chain is found. Delta expression is more common among FSCL than FML or FLCL [89]. Heavy chain switching (gamma and alpha expression) is more common among FML and FLCL (figure 28) [89]. Interestingly, the IgD coexpression commonly found in FSCL has been described as a B-cell marker relevant to prognosis [91]. As illustrated in figure 22, IgD expression is normally found in mantle zone, not follicular, cells, so its expression in FSCL is an immuno­logic aberration, contrary to the view that follicular lymphomas are simply clonal expansions of normal follicular counterparts.

Besides the type of immunoglobulin, it is also the pattern of immunoglo­bulin staining within neoplastic follicles which specifically differs from that in reactive follicles. In reactive follicles there is scant cellular and abundant intercellular immunoglobulin, whereas in follicular lymphoma the neoplastic cells strongly express immunoglobulin in a definite cellular, as opposed to intercellular, pattern [86, 92, 93]. In this regard, note the absence of poly-

63

typic extracellular immunoglobulin in the neoplastic nodule (e.g., lambda, gamma, and alpha) in figure 22. Once again, follicular lymphoma involves aberrations beyond just a monoclonal B-cell proliferation.

As illustrated in figure 22, dendritic reticulum cell (DRC) localization in follicular lymphoma corresponds closely to that in reactive germinal centers. This persistence of DRC in follicular lymphoma indicates preservation of this key cellular component even in neoplasia [12, 29, 89, 94]. Besides suggesting the germinal center origin for follicular lymphomas, DRC persis­tence leads to the speculation that DRC may playa role in sustaining or driving B-cell clonal expansion or DRC may effect homing of neoplastic B-cells throughout the body [94]. DRC may chemically define the domain in which clonal expansion and homing occur [12, 94]. DRC persistence in follicular lymphoma runs counter to our view that a neoplasm represents autonomous growth independent of adjacent cellular control. The aberrancy in follicular lymphoma may involve accessory antigen presenting cells as well as B-cells.

Assessment of T-cell antigens in follicular lymphomas (FL) demonstrates a considerable range of T-cell activity [29, 87, 88, 95]. As illustrated in figure 22, the localization, number, and proportion of T-helper (TH) and T­suppressor/cytotoxic (Ts/c) cells may not differ from those in physiologically reactive nodes. Yet in other cases of follicular lymphoma, reduced numbers of interfollicular T -cells are recorded, as well as decreased T H/T s ratios [29, 87, 95]. Two studies have emphasized the common occurrence of increased numbers of T-suppressor/cytotoxic (Leu 2+) cells within neoplastic nodules [87, 95]. The prominence of intrafollicular suppressor/cytotoxic cells in FL contrasts noticeably with the paucity of such cells in reactive follicles and may reflect a host response to the neoplasm (e.g., tumor suppression) [87, 95]. Interestingly, the number of persisting T-cells within neoplastic follicles is higher for FSCL and FML than FLCL [87]. Since the majority of inter­follicular T-cells in FL are of T-helper type, they may facilitate the neoplastic B-cell proliferation rather than act as a host defense against it [95].

As with DRC persistence, the T-cell data raises doubt about the auton­omy of FL monoclonal B-cell proliferations and raises consideration of possible immunoregulatory influence by T-cells [95]. It may be that immuno­regulatory T-cells are part of the aberrancy of FL; they could block the normal differentiation of follicular B-cells and prevent their achieving late secretory status. The validity of this immunoregulatory T-cell blockade of FL B-cells is supported by the experiments of Braziel et al. [96], who were able to overcome such a block in vitro and induce Ig secretion by follicular lymphomas by removing autologous T-cells and providing allogeneic normal T -cells and phorbol esters. Evidence for the immunoregulatory responsive­ness of FL also comes from the finding of their response to B-cell growth factors [97]. Host immunoregulatory mechanisms may be pivotal in any understanding or effective treatment of FI

Ki-67 is a mouse monoclonal antibody which identifies a human nuclear

64

antigen present in proliferating cells and not resting cells [90, 98]. By delineating cells throughout the growth cycle, Ki-67 can reveal the true kinetic characteristics of tumors. In figure 22 a striking degree of prolifera­tion in the reactive germinal center is documented, reflecting this as a site of B-cell clonal expansion [90]. Note that the FSCL in figure 22 has consider­ably fewer proliferative cells relative to the reactive germinal center. This indolent, subnormal proliferative rate corresponds with the indolent nature of FSCL, with a median survival of 7.4 years [1]. It is paradoxical that FSCL as a neoplastic entity appears to grow more slowly than physiologic germinal centers. The indolent nature of FSCL from the perspective of Ki-67 corre­lates with the faint/moderate activation antigen (e.g., TRF, IL-2, Ia) expres­sion of these cells. Thus, FSCL appears not only to be indolent, but also arrested, in a nonactivated, largely resting stage of mid B-cell development [88].

Although the small cleaved lymphoid cells of FSCL are indolent and resting, they surprisingly demonstrate the usual widespread migratory capa­city of the normal B-cell. As demonstrated in figures 23, 24, and 25 this single neoplastic B-cell clone commonly involves the spleen, bone marrow, peripheral blood, and liver simultaneously. As shown, the FSCL cells selec­tively involve the B-cell zones of the splenic white pulp [29], the paratra­becula of the marrow [34], and the portal zones of the liver [37]. This indicates that one clone takes over almost as a form of 'immunologic stran­gulation'. This represents either multifocal neoplastic transformation or homing of neoplastic cells to the B-cell zones of these organs. With regard to the latter, homing, the presence of extranodal DRC in FL may be relevant (figure 25). DRC, normally absent or scant in physiologic liver, has only been found in pathologic liver states, most notably associated with FL (figure 25) [36, 37]. This raises unanswered questions. Might DRC precede FL cells to the liver and lead to FL homing to the liver, or might DRC migration follow B-cell spread to the liver? In either case, DRC directed B­cell clonal expansion appears pivotal to the spread of FL. Once again FL does not appear to be the autonomous clonal expansion expected in malig­nancy [29].

Follicular mixed cell lymphoma

Although follicular mixed cell lymphoma (FML) demonstrates consistent pan-B-antigen expression (B 1+' B4 +, LI2 +, LI4 +, LI6 +) like FSCL, it differs by having less SIg and C3dr expression and more proliferative activity (Ki-67) and activation antigen expression (TRF, IL2, Ia) than FSCL [29, 88, 90,93,98-101].

The substantial range of SIg expression in FML is illustrated in figure 26. As shown, some FML manifest monotypic light and heavy chain restricted SIg expression, as with FSCL [29, 99, 100]. However, in contrast, with the very rare report of SIg- FSCL [23, 87], SIg- neoplastic nodules are common

65

Figure 23. Comparison of follicular small cleaved cell lymphoma (top) and normal spleen immuno-types (bottom).

among FML (e.g ., > 50% in two studies) [88, 93]. Since these SIg- FML express pan-B-antigens, there is little doubt as to their B-cell lineage [88, 93]. This SIg- state probably corresponds with the transition to an immuno­blastic phenotype (figure 3), indicating an immunotype similar to large cell lymphoma (LCL) . It has previously been speculated that Slg- FML with a LCL-like phenotype may be more likely to progress to large cell lymphoma,

66

. . . . .. • .. .

~ , .. .. . , - ," .... t .· .. ....

Figure 24. Follicular small cleaved cell lymphoma involving the bone marrow.

L4

may be more responsive to combination chemotherapy, and may have a more substantial prospect for cure [88].

Cell suspension studies reveal that even FML of monotypic SIg+ type have conspicuously fewer Ig-bearing cells than SIg+ FSCL [88, 93, 99, 100]. As revealed in figure 27, compared to reactive lymph nodes (NL) , FSCL typically have greater than 70% SIg+ cells, whereas FML typically have fewer than 30% SIg+ cells [88, 99-101]. In keeping with the mixed (small and large, resting and activated) neoplastic populations in FML, there -ap­pears to be a frequent admixture of SIg+ and SIg- cells. Clearly, FML is immunologically more complex than FSCL, including a broader range of B-cell differentiation. It is this range of B-cell differentiation and range of cytology that particularly characterizes FML and contrasts with the cytolo­gic and immunologic homogeneity of FSCL [88].

Figure 25. Follicular small cleaved cell lymphoma involving the liver.

67

Figure 25.

68

Figure 26. The immunologic variants of follicular mixed cell lymphoma , SIg+ variant at top; SIg- variant at bottom. (From Grogan TM, Hicks MJ, Jolley CS, et al: Identification of two major 8-cell forms of nodular mixed lymphoma. Lab Invest 51:504-514, 1984.)

0) H

100

75

U) 50 ~

69

FSCL FML LCL NL Figure 27. Graphic depiction of the percentage of surface immunoglobulin (sIg) positive cells in follicular lymphomas. Reactive lymph nodes (NL, far right) .

FSCL appears to represent clonal expansion of a B-cell clone frozen or blocked at a single stage of B-cell differentiation (mid B-cell). FML, in contrast, demonstrates several B-cell stages in simultaneity, indicating some staggered maturation [88]. Indeed, in FML the phenomenon of apparent zonal maturation within neoplastic nodules has been described as zonal Ig and C3dr expression. Focal variable heavy chain expression may also occur in FML. Note in figure 28 that IgA2 is expressed in a single FML neoplastic nodule, while the others lack A2 expression, indicating heavy chain switch­ing in one nodule and not the others. Switching to IgG and IgA is more commonly noted in FML and FLCL than in FSCL [89].

In contrast with FSCL, FML have more variable expression of C3dr (B2) [88, 102]. In general, C3dr/B2 expression appears to mirror SIg expression, with SIg- FML nodules lacking C3dr expression and SIg+ FML and FSCL cells usually demonstrating C3dr expression [88]. Since C3dr is expressed early in antigen activation and then lost in subsequent B-cell maturation (see figure 3), the SIg- C3dr FML phenotype would appear to arise from a more mature B-cell than SIg+C3dr+ FMLlFSCL [103]. Figure 29 illustrates this range of C3dr expression in FL. Since C3dr is known to play a role in immunoregulation of B-cells, the loss of C3dr expression in FML may herald a loss in B-cell immunoregulation, facilitating transition to clinically more aggressive diffuse or large cell forms of lymphoma [104]. There may be a

70

.~

010: Ilnly one of

Figure 29. Range of C3dr (B2) expression in follicular mixed lymphoma. (A) and (B) show the strong parallel expression of both Bl and B2 in a reactive lymph node. (C) and (D) show similar coexpression in a sIg+ FML. (E) and (F) show the loss of B2 in a sIg- FML. (From Grogan TM, Hicks MJ, Jolley CS, et al. Identification of two major B-cell forms of nodular mixed lymphoma. Lab Invest 51:504-514, 1984).

71

direct relationship between loss of C3dr expression and FML's propensity to histologic transformation.

Follicular large cell lymphoma

This rare form of follicular lymphoma is characterized by a predominance of large blastic cells, an abundance of mitotic figures, and frequent conversion to diffuse large cell lymphoma [4]. Although historically referred to as a histiocytic lesion, modern assessment indicates derivation from large 'acti­vated' B-cells [4]. Because of its rarity, there are few published accounts of modern immunologic assessment of this entity, and prolonged comment is precluded. Some follicular large cell lymphomas (FLCL) express mono­clonal surface immunoglobulin; others, in keeping with their activated state, have a SIg -, pan-B +, Ia +, TRF+ phenotype [25, 87, 93]. As illustrated in figures 30 and 31, FLCL differ from FSCL and FML by their higher Ki-67 proliferative index and high degree of TRF and Ia expression. As illustrated in these figures, FLCL are the most proliferative and activated of the FL, entirely consistent with their more aggressive clinical behavior [25, 87, 93].

Follicular lymphoma variants

Signet ring cell lymphoma. This rare lymphoma appears to occupy a well defined niche in B-cell neoplasia [105]. To date, most are variants of FSCL, although a few large cell variants are described [105]. The signet ring formation, as in figure 32, results from abnormal accumulation of monotypic immunoglobulin within tumor cells [105]. Compared to usual FSCL, these appear to be FL with further evolution to Ig production, but still short of secretory status as described in some FL with neoplastic plasma cells [106]. In signet ring B-cell FSCL, aberrations of surface or internal membrane recycling, as proposed by Grogan et al. [107], might account for the 'Ig constipation' of these cells. Clinically, in keeping with the usual FSCL histology, the patients have an indolent course.

Although signet ring cells in lymphoma were once equated unequivocally with B-cell lineage, there are recent descriptions of signet ring lymphomas of T-cell lineage (see figure 33) [107, 108]. While the distinction between signet-B and T-cell lymphomas might seem academic, signet-T-celllympho­mas as a form of peripheral T-cell lymphoma are clinically more aggressive than FSCL signet-B [107]. Like signet-B, signet-T cell lymphoma probably develops its characteristic signet ring appearance from abnormal membrane recycling [107].

Follicular lymphoma with histologic transformation. Follicular lymphoma usually begins as an indolent disease. Subsequently, a more aggressive course commonly ensues, with histologic transformation from a follicular to diffuse and/or small to large cell type [4, 109]. Most transformed FL become

72

, .' iI,·· , . .

, "

. :

Figure 30,

73

large cell lymphomas [4, 109]. In cases with immunologic study, the im­munoglobulin isotype of pretransformation and posttransformation tissues has been the same, indicating common lineage [110] with some differences in activation antigens [25] (see also figure 74).

Occasionally there is blastic transformation of FSCL with accompanying leukemic involvement and brief survival [111, 112]. Blastic or atypical forms of FSCL sometimes occur as a primary form of FL [111].

Follicular variants of Burkitt's and Burkitt's-like lymphoma. Both primary and secondary forms of these variants have been described. Descriptions of the primary follicular presentation of Burkitt's [113] and Burkitt's-like lym­phoma [114] are pivotal in appreciating the likely germinal center origin of these two diseases [113, 114]. The rarity of these variants probably reflects their high kinetic index with quick transition to diffuse form. Rarely, folli­cular variants of Burkitt's-like lymphoma have been described as a form of blastic transformation complicating longstanding indolent FSCL (see figure 60) [114].

Differential diagnosis of follicular lymphoma

Benign disorders. On occasion both atypical follicular lymphoid hyperplasia [115] and extranodal 'pseudolymphomas' [116] can be difficult to distinguish from FL. Since both may cause large, even frightening masses, they may appear unequivocally malignant clinically. While the sheer extent of these masses and their degree of lymphoid effacement may raise concern, immuno­logic assessment reassures that the process is reactive and not neoplastic [115, 116]. As illustrated in figure 34, follicular hyperplasia demonstrates polytypic Ig expression, in contrast with monotypic Ig in FL. Furthermore, in reactive conditions, polytypic immunoglobulins stain the DRC network in a dendritic pattern; whereas in FL monotypic Ig, staining is restricted to the small round follicular B-cells in a 'cellular' pattern, with no staining of the DRC network by other Ig isotypes [11, 89, 93]. In the immunologic defini­tion of follicular B-cell neoplasia, much rests on establishing light and heavy chain Ig restriction [13]. The lack of immunoglobulin staining of the dendritic network also suggests lymphoma [93]. Also, as mentioned in the discussion of FML and FLCL, the neoplastic nodules in these conditions are frequently Ig negative, a finding not associated with physiologic lymph node function [11, 93, 10]. Therefore, the absence of follicular Ig may also be taken as evidence of follicular neoplasia. Since rare reports of presumed 'reactive' lymphoid infiltrates with temporary monotypia are described [117] and oc­casional biclonal lymphomas are described [118], the principle of equating monotypic Ig expression to unequivocal neoplasia remains open to question.

Figure 30. Immunotype of a typical follicular large cell lymphoma.

....

.'

.'

" , ".

. .. ' .

....

"

. : .. '

... i·

""0

'

I"

.: .

'

. ,',

.. ' '.

I

. '.' ... ~

• r

.',

"'" ....

... .

", I'

, ,

.'0:-

."

~

• ,-

I •

• J

•

"

.

-..

. ' '.

"

.~

~ ..

A

"

.. ".'

)

. ,

"

~

"

•• ~.

;.

'.

" 'I"

~""

' •

" '.

j,

.' .•. 4-

'!.: •

..•. :

.. ,

:.

. "'\

1 .

.. ;

....

.... t

;.0:,

_.

,'~ '.

' f..

) •. ;

....

J _

"......

. .

, :'1

; .,.

,. \'

~ .. '

',' ::

':~;

at.:."

: .

........

' .. _o

"~"

" •

<I ...

"0

.

....

~

•

I ~:':'

" :~~. "

~ ..

.. :

:.. ..

~ ... ,

.,

,

..' ~

, ~.

. ~""'

" ,.

..

. :~; ~ '.

,; ~'.

,.;r-.

. ~;-'; :

'~;h~

i .. , :

.: '

'''': .. '

.. ' '

:"'~

~~t: i(:.

.:, 1

, :,.

.\1, ..

.... .

."'

. J

•• .,,, S~

.~. ~

'''''''

' •

I ..

;'

,"

'"''''

_....

~ '., ~r.

r ..

'

. t

•.

06

'.

' •

• '.~,

: ~.'

:.,~;

r. ~"

I ..:.

-.,'\

'.

\ '~

... t .

... '

j ...

' ....

. I,

:1.'

' '"

tJ·. ;.:

.,~, ;

!'l;~

' "

' •• (

.1.. . .,

'....

.....

~'

,~,

. #

"';

' \.

: ,:;

!'I-~':',

-.,::

' .

•• ..

·l

, ..

.. ,

.• I"

: I.

~~ •

.I'

., ..

\,

. " ':.'

" ',

..

I J

:-.

, , .

':: ;

. . \

'

. , . f.

," " 8

'~ J.)o

" .~

.

'

" , ,

~ . ".

' , .,.

.-:-.~

: .f

. ~ ....

, -:-. ,

, "

~,

,-..

. '

. .....

Fig

ure

31.

The

ran

ge o

f ce

ll pr

olif

erat

ion

amon

g fo

llicu

lar

lym

phom

as a

s re

veal

ed b

y K

i-67.

(A

) F

SC

L;

(B)

FM

xL;

(C)

FL

CL

.

, .

, .. \

.. ' .. "

'C

-..I

"""

Fig

ure

32.

His

tolo

gic

and

imm

unol

ogic

fea

ture

s of

a s

igne

t-ri

ng c

ell

lym

phom

a, B

-cel

l ty

pe.

Ant

i-Ig

G (

righ

t) s

how

s re

acti

vity

wit

h cy

topl

asm

ic I

g.

-..J

U

l

76

Figure 33. Signet-ring cell lymphoma, T-cell type. T-helper antigen (right) shows heavy accu­mulation within the cells.

This is a caution that immunologic assessment simply defines the chemistry of the proliferating cells and that integration with histologic findings and clinical circumstances remains paramount.

Malignant disorders. True follicular lymphoma may occasionally be con­fused with the pseudofollicular pattern found in some other forms of B-cell and T-cell lymphoma. In the case of small lymphocytic lymphoma (SLL) 'pseudofollicular growth centers' have caused difficulty (see preceding) [66]. In the case of SLL, co-expression of Leu 1 is a useful discriminant, although occasional FL with Leu 1 + expression have been described [68]. More useful would be the typical absence of DRC and the prominent TRF expression in SLL pseudofollicles [9]. In contrast, FSCL cells are prominently associated with DRC and express scant TRF [94].

Pseudofollicular T-cell lymphomas usually take the form of high grade lymphoblastic lymphomas (LBL) which invade tissue planes (e.g., breast lobules, tonsillar crypts) to give the false histologic impression of nodularity [119]. As illustrated in figure 35, this histologic mimicry of FL can be

Figure 34. Reactive hyperplasia, lymph node. (A) and (8) show the striking follicular prolifera­tion. (C) Stain for lambda and (D) kappa light chains arc both positive, indicating the polyclonal nature of the proliferation. (Gay R, Fielder K, Grogan TM, et al: Quinidine-induced reactive lymphadenopathy. Am J Med, 82:143-145,1987.)

77

Figure 34.

78

Figure 35. Pseudofollicular T-cell lymphoblastic lymphoma, breast. Top right: Tdt activity confirms the immature T-cell lineage of the infiltrate . (Schwartz JE, Grogan TM, Hicks MJ , et al: Pseudonodular T-cell lymphoblastic lymphoma. Am J Med 77:947-949 , 1984.)

impressive. However, immunologic assessment readily distinguishes LBL with strong T-antigen expression [119].

Nodular, lymphocyte predominant Hodgkin's disease (NLPHD) can be confused with FSCL because of the similarity of pattern [73] . This confusion is compounded by the revelation that the large polylobated neoplastic cells in LPHD may be of B-cell lineage [120, 121]. The histologic distinction be­tween NLPHD and FSCL is generally straightforward, with the distinctive small angular FSCL cells readily distinguished from the round and regular small reactive cells in NLPHD. The neoplastic cells in NLPHD comprise the large irregular 'L & H' cells, 'mummified' cells, and Reed-Sternberg cells. The isolated pattern of B-antigen expression in large polylobated cells should allow immunologic distinction from FSCL with more consistent small cell monotypic Ig expression. Large polylobated NLPHD cells have been vari­ably described as polytypic and/or monotypic Ig-bearing [120, 121].

Figure 36. Diffuse small cleaved cell lymphoma. (A) Histologic appearance. (8) Ki-67 staining. (C) Negative reaction for kappa. (D) Positive reaction for lambda light chain.

79

... .. ~

"!

~ o •

, .. . . ... ....

.. . ~ r'

,. . ~ ....... ....

, .' . . .. o . ~.

,I. . ~

" , -\

... ~o

# . • '," , . . .- ~ . '

~ .. - • : ~

. o. \0. <

.. ~

I

" . ' . ' ~ . .,

. / ' ~

.. .. .. ,

(. .( r r-

.... -" :1, -.J[ \.

Figure 36.

80

Diffuse lymphomas

General comments

The diffuse lymphomas (excepting SLL) show a greater aggressiveness than their follicular counterparts [1]. Each of these categories is histologically diverse, and non-neoplastic elements, such as epithelioid histiocytes and sclerosis, are common components [1]. Extranodal presentation is more frequently found in the diffuse lymphomas compared to the follicular types [122]. Survival in diffuse lymphomas is generally shorter than in follicular lymphomas [1, 82]; it also varies by subtype, with small cleaved cell diffuse lymphomas having a longer survival than the diffuse mixed and diffuse large cell lymphomas, in turn [1].

The cells in diffuse small cleaved cell lymphoma (DSCL) are character­ized by their small size and irregularity of their nuclear contour [1]. This type represents the diffuse counterpart of follicular small cleaved cell lym­phoma; atypical cases of SCL showing nuclei with blastoid features are known to occur [111]. The mitotic activity among DSCL is generally low but can be variable (figures 36-38). The histologic separation of lymphoblastic lymphoma from DSCL is partially based on the very high mitotic rate in LBL [1].

Diffuse mixed cell lymphoma (DMxL) shows an admixture of small cleaved and large blastoid lymphoid cells [1, 123] with a moderately in­creased mitotic rate compared to DSCL. DMx has been further subdivided by the morphology of the small and large cells [123]. One peculiar variant of DMxL is characterized by an abundance of epithelioid histiocytes along with pleomorphic large cells which resemble Reed-Sternberg cells [124]. De­scribed by Karl Lennert in 1968, this variant was initially believed to be a peculiar form of Hodgkin's disease [125]. However, the atypicality of the small lymphoid cells, along with the clinical findings of widespread disease in older patients, soon brought this entity into the non-Hodgkin's lymphoma category [124].

Diffuse large cell lymphomas (DLCL), which comprise approximately one quarter of all malignant lymphomas, are considered either intermediate­grade or high-grade malignancies in the working formulation, depending on their subtype [1]. They show morphologic heterogeneity: Cleaved, non­cleaved and immunoblastic cell types are recognized among others. The immunoblastic type and the 'Lennert's' type (i.e., with a high content of epithelioid histiocytes) are considered high-grade lesions because of their aggressive clinical course (see figure lA) [1]. It should be noted, however,

Figure 37. Follicular and diffuse small cleaved cell lymphoma. Left: The vague nodularity present in histologic sections is overshadowed by the largely diffuse pattern of cell infiltration. Upper right: focal and scattered anti-CRz activity. Lower right: Ki-67 staining. The asterisk (*) indicates the same area of follicular CR2 activity above.

81

'. '

.',

'! . ';:."-

... : ... : ' .. :.,:",:.~.",,~ . . "

. ~ ... ,,'

... . . : : ... , ."

Figure 37.

82

Figure 38.

83

that the differences in survival between the intermediate and high-grade large cell lymphomas are minimal, hence all LCL should probably be con­sidered high-grade malignancies [25].

Immunologic descriptions

Each of the follicular lymphomas may eventuate in its diffuse counterpart; the small cleaved and mixed cell types may also transform to diffuse large cell lymphomas [126]. Significantly, however, ORC staining is not found in these lymphomas, suggesting they may have escaped immunoregulation [94]. Con­comitantly, the proliferation marker Ki-67 is readily found, in keeping with the higher grade status of diffuse lymphomas relative to follicular lympho­mas [90]. Beyond their possible follicular origin, each of the diffuse lympho­mas must also be considered a malignancy in its own right, with their own complex immunologic signatures. Among the IBL subset of LCL, there is frequent documentation of previous abnormal immune status (e.g., angio­immunoblastic lymphadenopathy [AILO] preceding T-IBL) [127-130].

Diffuse small cleaved cell malignant lymphoma. The usual B-cell-diffuse small cleaved cell malignant lymphoma (OSCL) shows an origin at the midpoint of B-cell differentiation (figure 3) [2, 6]. Characteristically, im­munoglobulin and pan-B-antigen expression are strong (see figure 36) [11, 29 , 67, 87, 89]. Oendritic reticulum cells (ORCs), which have a high affinity for immune complexes and playa role in B-cell 'homing' and clonal expan­sion, are found in intimate association with follicular lymphomas [94]. In contrast, there is a consistent absence of ORC in OSCL, indicating a fundamental aberrancy [89]. It appears that when ORCs' are lost, organiza­tion of lymphoid cells into a follicular architecture does not occur. Perhaps other subtle abnormalities of immune regulation occur with loss of ORC; the OSCL cells may then grow with fewer constraints than in the follicular lymphomas.

This reciprocal relationship between ORCs and cell proliferation is re­vealed by cases of follicular lymphomas showing transformation to diffuse disease, as demonstrated in figure 37. This case reveals greater cell proli­feration in areas of ORC loss. Whether neoplastic proliferation affects the loss of the ORCs, or vice versa is still unknown.

Some atypical OSCL with 'blastoid' appearance have been described [111]. Some of these are now more specifically characterized as the lympho­blastic variant of lymphoma (see lymphoblastic lymphomas) [131 , 132]; others have proven to be of B-cell lineage [16, 31]. Figure 38 shows a

Figure 38. Diffuse small cleaved cell lymphoma, atypical 'blastoid' variant. (A) Note the fine chromatin pattern and increased mitotic activity (arrow). (B) Ki-67 indicates a high degree of proliferation.

84

blastoid B-cell DSCL with small cleaved lymphoid cells but with a high mitotic rate. The cells phenotyped as a monoclonal B-~ell population with strong expression of kappa and mu immunoglobulins and a very high proli­ferative index as revealed by Ki-67 (see figure 38).

DSCLs that originate in the T-cell arm of the immune system may be either immature ('thymic') or mature ('peripheral'). The immature T-cell phenotypes are those now designated as lymphoblastic lymphomas, distinct from classical DSCL with expression of thymic antigens [16, 31]. Histori­cally, before immunotyping, these immature LBL were frequently called DSCL. After immunotyping indicated its separate lineage, its distinctive morphologic features were retrospectively appreciated: high mitotic rate, dusky chromatin, etc. [131, 132]. The mature, post-thymic, peripheral T-cell phenotypes of DSCL (see 'peripheral T-cell' section) show 'idiosyncratic,' 'novel' expression of T-antigens (figure 39) [14, 15]. Both of these T-cell cousins of DSCL show increased proliferative activity with anti-Ki-67 anti­body staining (see figure 39).

The DSCLs, in summary, may arise de novo or may have a documented follicular predecessor. They are of both B-Iymphoid and T-Iymphoid origin; if even partially follicular, their immunophenotype relates to the B-cell lineage. In contrast to the follicular lymphomas, however, they show loss of ORCs and a concomitant modest increase in cell proliferation. The T­lymphoid varieties of DSCL may be found in either the immature or mature compartments of the T-cell system. The immature T-cell forms are now called lymphoblastic lymphoma and are readily separated from classic DSCL by immunologic and morphologic means. Survival in DSCL cannot be re­lated to morphology alone; immunophenotyping identifies some patients with mature T-cell DSCL who may have foreshortened survival not pre­dicted by morphology [14].

Diffuse mixed cell lymphoma

Diffuse mixed cell malignant lymphomas (DMxL) , perhaps more than any category of lymphoma, have very heterogeneous immunophenotypes [133-135]. The B-cell lesions are considered derived from follicular centers and the T-cell lesions from the peripheral (nonthymic) T-cell compartment.

The diffuse mixed lymphoma with a high content of epithelioid histio­cytes, is called 'Lennert's' lymphoma (LEL) [124, 125, 136-138]. Immuno­phenotypic analysis has shown most cases of Lennert's lymphoma are of a mature T-cell type, usually of T-helper/inducer (Leu 3) type [139, 140]. Many have novel T-cell phenotypes with pan-T antigen loss [14]. In spite of

Figure 39. Diffuse small cleaved cell lymphoma, T-cell phenotype. Left: Histologically, a blastoid variant of DSCL; mitotic activity is also readily identified (arrow). Right: A mature activated T-cell phenotype.

85

Figure 39.

86

extensive pan-T-antigen loss, recent genotypic studies indicate clonal T­antigen gene rearrangements, confirming Lennert's lymphoma as a true T­cell malignancy [141]. It may be lymphokines/cytokines produced by these functional T-cell malignancies which attract the histiocytes [140].

It is common for Lennert's lymphoma to occur throughout the reticu­loendothelial system, presenting as stage III or IV disease [124, 138]. Figure 40 illustrates a frequent site of involvement, the bone marrow, which shows the common Leu 3+ Leu 2- immunophenotype. Despite widespread disease and monotypic phenotype, the diagnosis is not straightforward, since it may easily be confused with a 'nonspecific' granulomatous reaction [142]. The diagnosis may be delayed for several months, and only when the patient has failed numerous courses of antibiotics and has obvious progression of disease is a malignant diagnosis suspected. Figure 41 shows such a case. The granu­lomatous appearance of these widespread pulmonary lesions (A and B) initially obscured the diagnosis. However, careful histologic examination of the lymphoid infiltrate revealed markedly atypical small lymphoid cells, and phenotyping showed an aberrant T-cell phenotype as associated with T-cell neoplasia and not T-cell inflammation [14, 22].

Lennert's lymphomas of B-ce\l origin are occasionally found as illustrated in figure 42. The histologic findings show features indistinguishable from the T-cell variety, complete with numerous epithelioid histiocytes. However, immunotyping reveals lambda +, IgM+ monoclonal staining. Interestingly, like many patients with Lennert's lymphoma, the diagnosis in this patient was also delayed, but for another reason. The earlier biopsies showed a pre­ponderance, but not a truly monoclonal, population of lambda+ cells. The diagnosis of lymphoid malignancy awaited several months evolution to monoclonality. The attractant for the epithelioid histiocytes in the B-ce\l variants of Lennert's lymphoma is not yet defined.

Diffuse large cell lymphoma including immunoblastic types

Diffuse large cell lymphomas (DLCL), including cleaved, noncleaved, immu­noblastic, polymorphous, plasmacytoid, clear cell, and epithelioid variants, represent a morphologically and immunologically diverse group [1, 4, 14, 15,38,87, 143-145]. Although they comprise a great variety of B-cell and T-cell phenotypes, they have in common full expression of activation anti­gens, consistent with their status as 'transformed' large cell malignancies. Some DLCL represent transformations from lower grade follicular and/or diffuse lymphomas [4, 122, 126, 146, 147]. These DLCL usually retain the Ig isotype of the parent indolent lymphoma, with the addition of more con-

Figure 40. Diffuse mixed cell lymphoma with a high content of epithelioid histiocytcs CLen­nert's' lymphoma), T-cell type, hone marrow. Immunophenotyping indicates a T-helper cell proliferation with Leu 3+ (left) and Leu Z- (right).

87

Figure 40.

88

Figure 41.

89

spicuous activation antigen expression (see figure 74) [110]. Nonetheless, some large cell lymphomas may present challenges to immunophenotypic characterization because loss of pan-B-antigens, surface Ig, and pan-T­antigens may accompany their transformed status. This antigen loss during transformation may account for the previous descriptions of a high percen­tage of 'null' phenotypes among DLCLs [38].

With an expanded battery of monoclonal antibodies, the DLCLs may now be more definitively characterized by immunotyping. As shown in figure 43, the null cell large cell lymphomas are decreasing: they now comprise 2% of LCL cases. The great majority (80%) are of B-cell origin, and the remaining (18%) are peripheral T-cell types. Of all DLCLs, more than half (57%) express monoclonal surface immunoglobulins and pan-B­markers, while a minority (20%) do not express immunoglobulins. A few (2%) contain cytoplasmic immunoglobulin, having lost surface Ig expression and having lost some pan-B-antigens (e.g., Leu 14) in the transition to more secretory B-cell status (see figure 3). The common ALL antigen (CALLA) is found in a minority (16%) of cases. The T-cell type of DLCLs shows tremendous diversity due to idiosyncratic pan-T-antigen loss, but all have at least one pan-T-antigen plus activation antigens, indicating a mature, acti­vated postthymic origin [14, 15]. These subgroup percentages of LCL are quite similar to those described by Freedman et at. [145].

The nuclear antibody Ki-67, which identifies cells in the growth phase of the cell cycle [90], is strongly positive in almost all DLCLs. This antigen is generally found in > 40% of DLCL cells (see figure 44C). The rare DLCL case with little Ki-67 activity indicates not all DLCL are highly proliferative, a finding perhaps relevant to prognosis.

An example of DLCL of B-cell lineage is found in figure 44. The mor­phologic features are consistent with a large transformed lymphoid cell malignancy, including the blastoid nuclear features and high mitotic activity (figures 44A and 44B), which correlate with the high proliferative activity defined by Ki-67 staining (figure 45C). Strong activity against pan-B-ceIl antigens is shown by the anti-L14 staining in figure 44D. This particular B­LCL tumor also reveals strong surface immunoglobulin expression, strong activation antigen expression (including Ia and the transferrin receptor [TRF]), and TAC (IL-2 receptor), indicating a mature, antigen-activated B­cell malignancy.

As these B-cells are driven toward more secretory immunoglobulin status, they take on the immunologic characteristics of plasmacytoid cells and are morphologically recognized as immunoblastic lymphomas (IBL) [5, 6, 148, 149]. The findings of a patient with B-IBL are shown in figure 45.

Figure 41. Diffuse mixed cell lymphoma with a high content of epithelioid histiocytes, T-cell type, lung. (A) and (B) Histologic details. (C) T-helpcr/inducer cell antigen expression. (D) Absence of T-cytotoxic/suppressor cell antigens.

90

. 4. . ~ 1

(1 • .~ ,'" ~ ~ " • " ~ ' . e ~ • ""' .. l~

. ..

Figure 42.

91

I. B Cell 80%

l. sIg+ pan B+ ~ Calla+ 7/49 14% 2. sIq+ pan R+ ~ Ca11a- 21/49 43% 3. sIg- pan B+ T- Calla+ 1/49 2% 4. sIg- pan B+ ~ Ca11a- 8/49 16% 5. sIg? pan B+ ~ 1/49 2% 6. sIg- CIg+ pan B+ Calla- 1/49 2%

II. T Cell: periphral T cell phenotype 18%

sIg- cIq- pan B- T+ 9/49 18%

III. Other 2%

Pan B- pan T- Ia+ LC+ ("null" cell?) 1/49 2% Figure 43. Tabulation of SO large cell lymphoma immunotypes at the University of Arizona.

The morphologic findings include an overt plasmacytic component admixed with the malignant cells which appear variably as immunoblasts and large plasmacytoid cells (figure 45A). The large cells show marked reactivity with anti-Ki-67 antibody (figure 45B), indicating their proliferative status. In keeping with the late secretory status of this lesion, surface Ig and some pan-B-antigens (Leu 14 and 16) were absent by frozen section assessment, but monoclonal Clg expression was found (figure 45C) using fixed tissue. The plasmacytoid nature of this lesion is further established by finding plasma cell associated antigen (PCA-1) on the neoplastic cells [150, 151]. The initial SIg -, pan-B - -phenotype could have led to a null cell designation [38]. However, the frozen section PCA-l determination coupled with the paraffin CIg assay ensured proper delineation.

B-cell IBL with preplasma cell features have been associated with a poor prognosis [152]. Additional studies have associated poor prognosis among LCL with surface Ig expression [38, 144]. In these studies SIg+LCL had a poor five-year survival (15%) relative to SIg- LCLs (63%) [144]. It appears surface markers are beginning to identify high-risk, poor-prognosis LCL patients who may benefit from different treatment strategies [153].

T-IBL lymphomas are quite distinctive from an immunologic point of view [14, 15]. As expected, they possess mature T-antigens, lack immature T-antigens (e.g., Leu 6, Tdt, CALLA), frequently have one subset antigen (either helper or suppressor) to the exclusion of the other, and commonly have abundant activation antigen expression (e.g., Ia+, TRF+, IL-2+ [TAC+]) [14,15]. Unexpectedly, they may have novel or idiosyncratic T-cell

Figure 42. Diffuse mixed cell lymphoma with a high context of epithelioid histiocytes CLen­nert's' lymphoma), B-cell type, lymph node. (A) Histologic appearance. (B) Histiocytic com­ponent staining with Leu MS. (C) Pan-B-cell marker LI2. (D) Pan-T-cell marker L4.

~,

f •

I -w

...'

. ....

. ~ ' •

.. • ~,~

"",

, .'

\l

• ~.

,.."

... ,'

....

~)." ~

I ~C

. -.

" '.~

~ : '"..

• •

• ...

. .

--.

' j,

, }

,.".,

. t

.. '.

oJ>

~

" ...

.. '

.1...

... ..

'.'~

. "i~·. ~.'

_ ...

.'i'

·,"

.;. 'f

l," .

... -.,..

,.,.,

""

, .. _

•• ,

~

• ~ '.

.

<i '

..

. ,

.....

~

4 , ,

..

..

.'

~..

f..

.. ...

. , "'

!I?

c'

,_. ~ .

.~

. . I

.l..:

.,~

. ~'.

n.

.., .. ~

.,' ,

-,. ..

.. -

. -'1

~ ~

• , ..

...

"4 ""

'..,,

&1 ,

. . 4'

' . '

'.' .

...

.,

.;'

~-

CO'

, .

,

••

";"

,'

\ ••• '

_I

I .• ,

. •

'Ii:

. ,

."

q;'1"

,. 011

...

•

• 'i ".,

_..

..

{]o

,

• ,

..,

. '"

,,"

.-. ....

' •

'IIi.

..

.t.

• .. ~ . .,.

.,."

..

. '-.' .. -

.....

~ -. "

t . Y

• ~.: ~

'"

'' ,

..

.~

... -...

-.. •• ".

r t

1...

•

• .

""'.......

......

...

....

• tt ..

...

....

]1

" ..

,.'

..

Fig

ure

44.

Dif

fuse

lar

ge c

ell

lym

phom

a, 8

-cel

l ty

pe.

(A)

His

tolo

gic

appe

aran

ce.

(8)

Tou

ch p

repa

rati

on.

(C)

K-6

7 st

aini

ng.

(D)

Leu

14,

pan

-8-c

ell

anti

body

.

'D

tv

Fig

ure

45.

Dif

fuse

lar

ge c

ell

lym

phom

a, i

mm

unob

last

ic t

ype,

B-c

ell

subt

ype.

(A

) H

isto

logy

. (B

) K

i-67

act

ivit

y, n

ote

pref

eren

tial

st

aini

ng o

f nu

cleo

li (

arro

w).

(C

) P

araf

fin-

embe

dded

tis

sue

stai

ned

for

lam

bda

ligh

t ch

ain

imm

unog

lobu

lin.

Kap

pa l

ight

cha

in n

egat

ive

(not

sho

wn)

.

\Q

~

94

phenotypes, with loss of some, but not all, pan-T-antigens. This aberrancy characterizes neoplastic mature T-cells and usually not inflammatory cells, so it serves as an indicator of likely T-cell malignancy [14, 15, 22].

An example of such a mature, activated or peripheral T-cell (PTL) IBL is illustrated in figure 46. This is an example of a 'plasmacytoid' T-cell IBL [1]. The name is paradoxical, since the plasmacytoid features lead to the ex­pectation of a B-cell immunotype. However, the large clear neoplastic cells in this case were of T-helper phenotype, and the surrounding plasma cells were of polyclonal Ig-bearing type, representing a reactive B-cell component. This case is a reminder that the correspondence between morphology and immunology is incomplete. This PTL appears to have retained true T-helper function, resulting in polyclonal recruitment of B-cells, perhaps due to Iymphokine effects [154].

Other well recognized subtypes of DLCL exist, including the sclerosing variant illustrated in figure 47. Initially described within the mediastinum [155-157] and retroperitoneum [158], the presence of sclerosis has been indicated as conferring either a worse [156, 157] or better [155, 158-160] prognosis. This variant of DLCL has been mistaken for carcinoma due to the nesting appearance of the cells (,compartmentalizing fibrosis'); however, lymphoid lineage is readily indicated with immunophenotyping (figure 47B). The sclerosis is considered a part of host response or a Iymphokine-related finding.

The propensity for DLCL to present in extranodal sites has been men­tioned [122]. It is emphasized that DLCL may involve any tissue site in the body. Some of these extranodal sites represent a major challenge to diag­nostic immunophenotyping. One such difficult circumstance is illustrated in figure 48. This represents a small endoscopic gastric biopsy performed for suspected gastric carcinoma. In spite of the scant amount of biopsy material immunophenotyped, a clear-cut indication of lymphoid neoplasia was found. As illustrated, the antikeratin antibody decorated the gastric glands and not the surrounding infiltrate, while the lymphoid markers identified pan-B­markers and a monoclonal isotype of Ig expression. By this means, gastrec­tomy was obviated and combined chemotherapy initiated.

Another extranodal diagnostic challenge is illustrated in figure 49. This DLCL involved the maxilla, requiring that undeca1cified snap-frozen bone be phenotyped. Nonetheless, a B-cell lineage was established as shown. In this instance, the cells also show a polylobated or multi lobated appearance, which has been described previously as a T-cell variant of DLCL [161]'

Figure 46. Diffuse large cell lymphoma, immunoblastic type T-cell subtype, with functional belper T-ccll activity. A pseudonodular appearance due to nests of pale malignant eells sur­rounded by polyclonal B-cells. Upper right: Cytoplasmic staining for lambda light chain in reactivc infiltrate. Both kappa light chain and mu heavy chain staining show a similar pattern of reactivity, indicating the polyclonality of the B-ccll infiltrate. The bottom three panels indicate the malignant T-cell phenotype: Leu 4+. Leu 2 .. , Leu 3+.

95

Figure 46.

Fig

ure

47.

Dif

fuse

lar

ge c

ell

lym

phom

a w

ith

scle

rosi

s. (

A)

His

tolo

gic

appe

aran

ce.

(B)

Leu

kocy

te c

omm

on (

LC

) an

tige

n ac

tivi

ty o

n ne

opla

stic

cel

ls;

anti

body

aga

inst

epi

thel

ial

anti

gens

(e.

g.,

cyto

kera

tin)

neg

ativ

e.

\0

0\

Fig

ure

48.

Ext

rano

dal

larg

e ce

ll ly

mph

oma,

sto

mac

h. (

A)

His

tolo

gic

appe

aran

ce i

n th

e m

ucos

a. (

B)

Pan

-B-c

ell

anti

body

Leu

14.

(C

) A

ntic

ytok

erat

in a

ntib

ody

(CK

).

\0

-...I

98

Figure 49. Extranodal large cell lymphoma , multilobulated /polylobated type , maxillary sinus. The polylobated nuclei are seen in a touch preparation. Upper right: Pan-B-cell antibody, B I , demonstrates activity.

although B-cell varieties are known [162, 163]. Interestingly , the association of B-cell polylobated DLCL and maxillary origin has been reported pre­viously [164].

True histiocytic malignancies are not included as part of the Working formulation classification [1]; previous descriptions of lymphomas as 'histio­cytic' referred to morphologic, not immunologic, features [1, 3]. Nonethe­less, true histiocytic malignancies do exist, and some of these may be morphologically indistinguishable from DLCL [1, 3]. In these cases, marker studies, not morphology, lead to proper lineage identification [165-167]' The salient features of a patient with malignant histiocytes, as found in malignant histiocytosis, are shown in figure 50. There is often widespread reticuloendothelial organ involvement. A lymph node biopsy showed the characteristic sinusoidal location of the malignant cells, which also possessed hemophagocytic activity [3]. The strong reactivity of the antimonocytic/ macrophage antibodies LM3 and LM5, coupled with lack of staining for B-

Figure 50. Malignant histiocytosis (MH). (A) Enlarged spleen in MH without conspicuous tumor. (B) Histiocytic marker, Leu MS, reactive with the neoplastic cells. (C) Sinusoidal in­volveme nt of a lymph node by MH. (0) Higher power detail shows the infiltrate admixed with residual lymphocytes. Erythrophagocytosis is demonstrated (arrows) in the atypical histiocytes.

99

Figure 50.

100

cell or T-cell markers in these large cells, places them in the histiocytic cate­gory [168] . Since some lymphoid malignancies manifest erythrophagocytosis and also recruit histiocytes, strongly simulating malignant histiocytosis, phenotypic discrimination among these entities is imperative [169].

Differential diagnosis of diffuse lymphomas

Benign disorders. Lennert's lymphoma may be misinterpreted as a reactive granulomatous process, suggesting an infectious etiology [142]. Persistent, usually progressive disease in the face of adequate antibiotic treatment will usually lead to diagnostic reconsideration, although often not until the lymphoma is far advanced [142]. Immunologic assessment may require re­peated biopsies, especially since-odd, novel phenotypes are the rule [14, 15].

Malignant disorders. 1. Diffuse small cleaved cell lymphoma may be misinterpreted as a lym­phoma of intermediate type (IDL) (see discussion of SLL for distinction). 2. Among the diffuse mixed cell lymphomas, the Lennert's lymphoma variant may be mistaken for Hodgkin's disease because of the epithelioid histiocytes and Reed-Sternberg-like cells [125]. This differential was dis­cussed previously (also see figures 40 and 41). 3. Diffuse large cell lymphomas have been confused with granulocytic sar­comas (see figure 61), metastatic carcinoma (see figure 47), plasmacytomas (see figure 70), and the lymphocyte-depleted form of Hodgkin 's disease. Granulocytic sarcoma and large cell peripheral T-cell lymphoma with aber­rant myelocytopoiesis [170] may be delineated using immunophenotyping (see figure 61). The lymphocyte-depleted form of Hodgkin's disease (LDHD) has been mistaken for DLCL because both may possess large pleomorphic cells [171-173). Immunologically, the DLCL lymphomas show either B-cell or T-cell lineage in the majority (98%) of cases, while neoplastic Reed­Sternberg cells of LDHD do not express markers of either lineage, although phagocytosis of 19 by Reed-Sternberg cells may confound the interpretation [174] .

Lymphoblastic lymphoma

General comments

Lymphoblastic lymphoma (LBL) is defined as a clinicopathologic entity characterized by frequent occurrence in young males, commonly with medi­astinal presentation and subsequent leukemic and CNS involvement [1, 4, 175 , 176] . This high-grade lymphoid neoplasm comprises> 30% of all childhood and 5% of adult non-Hodgkin's lymphomas [1, 176].

101

The maJonty of LBL are of immature T-cell phenotype [16, 31, 46, 177 -179]. The similarity of cytologic, immunologic, and clinical features of LBL to ALL has led to the view that LBL is a lymphomatous variant of T­ALL [180], although most LBL are of slightly more mature (cortical thymo­cyte) derivation than T-ALL [16, 46, 177, 181]. The similarity to T-ALL prompts treatment of LBL based on ALL principles (e.g., CNS prophylaxis) [25].

Histologic features

As revealed in figure 51, most LBL are comprised of small lymphoid cells with dusky chromatin, inconspicuous nucleoli, and convoluted nuclei [1, 3, 4]. The latter are distinguished from DSCL because of their high mitotic rate (5-7 mitoses/hpf) versus lower mitotic rate in DSCL (1-2 mitoses/hpf) [1]. LBL also reveal high Ki-67 poliferation relative to DSCL [90], in keeping with LBL high-grade status.

Occasional nonconvoluted or round cell LBL variants are described (figure 51), signifying that LBL and ALL may be morphologically indistin­guishable [176, 180]. Recently, an atypical pleomorphic or 'L-2' variant of LBL has been described [182]. Few atypical LBL have been phenotyped with immunologic completeness [31].

Figure 51. Histologic variants of lymphoblastic lymphoma. (A) Convoluted type . (8) Non­convoluted type.

102

Immunologic features

Most LBL (> 80%) have an immature T-cell phenotype, similar to cortical thymocytes at the middle stage of thymic differentiation (see figure 4) [46, 181, 183]. Unlike mature T-cell neoplasms (e.g., PTL and MF), LBL fre­quently coexpress Leu 2 and 3 in a pattern similar to cortical thymocytes [31,46,177,181]. LBL also frequently express the thymic antigen Leu 6 [16, 31, 46, 177-179] and the thymic-derived enzyme terminal deoxnucleotidyl transferase (Tdt) [184-188]' and may coexpress common ALL antigen (CALLA) [16, 31, 46, 189, 190]. Unlike mature activated T-cell lymphoma, LBL infrequently express la and infrequently demonstrate deficiency of pan­T-antigens (e.g., loss of Leu 114/9) [14, 15,31,46]. While immature T-LBL are usually positive for Tdt, almost all other lymphomas are negative [31, 184-188]. Tdt is therefore perceived as a highly useful marker for the diagnosis of LBL, adding to its definition as a unique immunologic, as well as a clinicopathologic, entity [191].

As illustrated in figure 52, the coexpression of Tdt and Leu 6 can be considered pathognomonic of immature T-LBL or T-ALL [9, 10]. As shown, Tdt occurs as a nuclear enzyme in a distinctive nuclear chromatin distribution, while Leu 6 occurs on the surface, giving the impression of a series of interconnecting rings in tissue section. Since Leu 6 is also expressed

Figure 52. Lymphohlastic lymphoma phenotype: (A) Nuclear Tdt expression. (8) Surface Leu 6 expression. (Grogan TM, Spier eM, Wirt DP, et al: The immunologic complexity of lympho­hlastic lymphoma. Diag Immun 4:81-88, 1986.)

103

on Langerhans cells, its presence is not T-cell specific [192]. However, Leu 6/Tdt coexpression outside of the thymus is almost exclusively associated with immature T-cell neoplasia (LBLIT-ALL) [31].

While most LBL have an immature T-cell phenotype, occasional LBL cases of more mature T-cell, as well as null, pre-pre-B-cell, pre-B-cell, and mixed lineage are described [16, 31, 46]. The range of T-antigen expression and early B-cell antigen expression among LBL patients is graphically illus­trated in figure 53. Using comparative serial section immunochemistry, we demonstrate four LBL immunologic subtypes: 1) immature T-cell type with coexpression of Leu 2/3/6/9/CALLA and Tdt; 2) less-immature T-cell type with co expression of Leu 3/6/9 and Tdt; 3) more mature T-cell type with Leu 9/Ia coexpression and absent Leu 6/Tdt expression; 4) pre-pre-B-cell pheno­type with simultaneous Tdt/CALLAlfaint B4 expression.

Note that among the most immature T-cell phenotypes CALLA expres­sion occurs with faint, simultaneous Leu 2/3 expression, as found in normal cortical thymocytes before subset differentiation [31, 46, 177]. Among less immature T-LBL, note that Leu 3 (T-helper) antigen expression by itself greatly exceeds the antigen density of joint Leu 2/3 expression. These Leu 3+ cells also lack associated CALLA expression, indicating they have gained more mature subset status [31].

As revealed in figure 53, occasional T-LBL cases without Tdt and Leu 6 expression occur [31, 46, 179, 185, 186, 188]. These rare T-LBL with more mature phenotypes demonstrate pan-T (Leu 9) and Ia expression, indicating activated T-cell status, as found in medullary thymocytes [31, 179]. The absence of Tdt in this form of LBL (see figure 53) is not an isolated antigenic change but part of a complete phenotypic profile difference (Leu 9+, Ia+, Leu 6-, TdC) from immature T-LBL [31, 179]. Some might question the inclusion of mature T-LBL among true LBL. However, these mature T-LBL patients otherwise fit the clinicopathologic profile of LBL with mediastinal presentation in young males, subsequent leukemic involve­ment, and classic convoluted LBL morphology [31, 179].

The illustrated pre-pre-B-cell LBL, along with previous descriptions of pre-B forms of LBL [16, 31, 46], affirm that LBL and ALL may be immuno­logically as well as morphologically indistinguishable. Both immature T-LBL and immature B-LBL are characterized by simultaneous Tdt and CALLA expression, signifying closely linked immature phenotypes in spite of separate lineage [190]. Indeed, Weiss et al. [46] describe LBL with mixed B-cell and T-cell lineage markers, indicating a biphenotypic origin. Thus, while the immunologic definition of LBL is stretched by finding early B-cell and mixed lineage forms, these variants remain more akin immunologically to T-LBL than other Tdt-/CALLA - lymphomas of the same B-lineage or non-T­lineage [31, 46]. Many non-T-LBL have occurred in an atypical clinical setting: among older patients with extramediastinal presentation [16, 31, 46]. Many non-T-LBL have presented as either cutaneous or isolated bone lesions, suggesting a distinct clinical profile for this subset of LBL. Different

104

Figure 53. Comparative immunotypes of lymphoblastic lymphoma. (Grogan TM, Spier CM, Wirt DP, et al : The immunolgic complexity of lymphoblastic lymphoma. Diag Immun 4:81-88, 1986.)

natural histories may yet be associated with pre-pre-B and mature T-subtypes of LBL [16, 31, 46].

Among LBL studies, there is a great deal of discrepancy in the reported incidence of certain T-cell antigens (e.g., Leu 2/4/CALLA) [16, 31, 46, 177-179). Some studies have reported a high percentage of cases with T-helper

105

status and no cases with coexpression of Leu 2/3 [16]. Some have found a low incidence of Leu 4/6 and CALLA expression [16]. It is very likely that many of these differences are methodologic [46].

As mentioned by Weiss et al. [46], Leu 4 may occur in a cytoplasmic distribution not detectable by cell suspension methods which only detect surface antigen. Cytoplasmic fixation of antigen as employed in tissue section methods would be necessary to detect this cytoplasmic antigen. Under these circumstances different methods would yield discrepant results. A further clue to potential methodologic differences is raised in figure 53. Note the faint coexpression of Leu 2/3 in case 1 and the faint B4 expression in case 11. Less than optimal immunoperoxidase or fluorochrome formulations would result in a failure to detect these antigens at the limits of detectability [60].

Whatever the discrepancies due to methodologic differences, these studies collectively document the heterogeneity and complexity of LBL and raise a challenge to the view that all LBL are the same [16, 31, 46]. These major T-cell and B-cell phenotypes of LBL might have therapeutic significance, and, as with ALL, LBL treatment may need to vary among immunologic subsets [31].

Variants

Lymphoblastic lymphomas usually result in diffuse effacement of the soft tissue. However, on occasion the neoplastic cells respecting the tissue planes, as in ALL, may have a lobular appearance which may be confused with follicular lymphoma. This entity has been called pseudofollicular lympho­blastic lymphoma [119]. Immunologic distinction from B-cell follicular lym­phoma is readily accomplished by detection of the following antigens: Tdt/ CALLA/Leu 1-6, 9 (see figure 35) [119].

Differential diagnosis

A serious error in diagnosis may result when lymphocytic thymoma is con­fused with LBL [193]. Since lymphocytic thymoma has a mediastinal pre­sentation and a predominance of immature T-cells with coexpression of TdtlCALLA/Leu 1-6, 9, this prospect is considerable on both a clinical and immunologic level. Since LBL requires elaborate multidrug chemotherapy with CNS prophylaxsis and thymoma requires simple surgery, the clinical stakes are high.

As illustrated in figure 54, the key to deciphering this difference is histologic and immunologic detection of the epithelial cell component [193]. Figure 54 demonstrates the simultaneous Leu 6 and cytokeratin expression in a lymphocytic thymoma. This 35-year-old male presented with a large anterior mediastinal mass. The biopsy obtained from mediastinoscopy was interpreted as a lymphoblastic lymphoma. A consulting hematopathologist concurred with this diagnosis, but only after confirming that the cytokeratin

106

Figure 54.

107

antibody on formalin fixed material was negative. Chemotherapy was in­itiated. However , since a thymoma was strongly suspected on radiologic grounds, an anti keratin antibody test on frozen tissue was performed, re­vealing the striking epithelial framework. Frozen tissue assessment indicated both an immature T-cell lymphoid component (Leu 1-6+/9+/CALLA +/ Tdt+) and epithelial components, mirroring the phenotype of normal thy­mus and supporting the diagnosis of a thymoma with a predominance of lymphocytes [193]. Note in figure 55 that the flow cytometry cell suspension analysis failed to detect keratin positive cells, reflecting the fact that tissue mincing may fail to liberate epithelial cells.

This case emphasizes several lessons in phenotyping: 1) immunotyping should not be performed in a clinical vacuum: The radiologist strongly suspected a thymoma; 2) avoid ' tunnel vision:' lymphoid phenotyping should be counterbalanced with other findings (e.g., cytokeratin, histologic evidence of lobation); 3) beware of false negatives, in general, and in formalin fixed material, in particular; freezing generally results in fewer antigenic false negatives [194, 195]; 4) cell suspension studies, without cytoplasmic fixatives and without methods to free epithelial cells from stroma, may be misleading [46].

Small non cleaved lymphoma (Burkitt'S and Burkitt's-like subtypes)

Histology

This high-grade lymphoma is a proliferation of intermediate-sized, round blastic cells with a high nuclear:cytoplasmic ratio , high mitotic rate and a 'starry-sky' pattern (figure 56) [1, 2, 196, 197] . The distinctive cytologic features include prominent cytoplasmic basophilia and vacuole formation (figure 56). Cytochemical studies indicate the vacuoles are lipid-laden and the cytoplasm is rich in RNA, which results in pyroninophilia with methyl green pyronine stains [1, 2, 196, 197].

Two histologic variants are described, including Burkitt's (BL) and non­Burkitt's types [1, 2, 197]. The Burkitt's type has a nearly monotonous appearance, comprised of regular blastic cells with round nuclei containing two to five basophilic nucleoli. The non-Burkitt's type, otherwise known as Burkitt's-like lymphoma (BLL) [197], shows greater morphologic hetero­geneity, with nuclear variability (pleomorphism) and nuclei containing one to two eosinophilic nucleoli [1 , 2, 196, 197].

Figu re 54. Lymphocytic thymoma phenotype. L6: Leu 6 staining of lymphoid thymocytes. CK: cytokeratin staining of epithelial cells.

108

CK 0%

leu 6 44%

leu 3 45%

L s • leu 2 55%

leu 1 94%

Control Figure 55. Lymphocytic thymoma phenotype by cytometry coplots. Lymphoid markers are readily detected in contrast with epithelial markers.

Figure 56. Histologic and cytologic features of Burkitt's (upper left) and Burkitt's-like (lower left, and right) lymphomas.

109

Clinical definition

Burkitt's lymphoma was initially described in Africa as a common childhood lymphoma involving the jaw and/or gonads of children under the age of three years [196]. Subsequently, rare cases outside of the endemic area were described [19S, 199]. These nonendemic BL were morphologically identical to endemic cases. They also occurred predominantly in children, but more commonly had an abdominal presentation and a median age at presentation of 12 years versus 3 years for endemic BL [S2, 200, 201].

The more pleomorphic Burkitt's-like cases also frequently have ab­dominal presentation but are more commonly encountered among adults [197, 202]. Interestingly, Burkitt's-Iike lymphoma, unlike Burkitt's lym­phoma, presents in three unusual circumstances: 1) it is the most common non-Hodgkin's lymphoma to develop in treated Hodgkin's disease patients [203]; 2) it is the most common non-Hodgkin's lymphoma complicating AIDS [204, 205]; 3) it occasionally arises from indolent follicular lymphoma [206, 207]. Some may consider the morphologic distinction between BL and BLL to be minor and academic. However, the distinctive epidemiologic, geographic, and clinical aspects of these two subsets warrant separate con­sideration [197].

Immunologic characteristics

Burkitt's and Burkitt's-like lymphomas are usually characterized as mono­clonal immunoglobulin-bearing B-cell neoplasms (see figure 57). Their con­stancy of pan-B-antigen expression and absence of T-cell antigens heralds this categorization [197,208]. Compared to other intermediate to high-grade B-cell tumors (e.g., LCL), Burkitt's and Burkitt's-like lymphomas manifest a much higher incidence of lambda light chain expression and a higher incidence of CALLA expression [20S]. Also, as revealed by anti-Ki67, there is a very high percentage of proliferating cells in BL and BLL (> 80%) compared to LCL (usually 40% -50%) [20S]. In contrast with LCL, which frequently express T AC or IL-2 receptors, BL and BLL usually lack IL-2 receptors [20S]. An archetypal BL phenotype (A +, CALLA +, pan-B+, high % of Ki67+) is revealed in figure 57. In BL and BLL, the CALLA expres­sion coupled with occasional Tdt coexpression may be taken as evidence of immature B-cell status (see figure 3) [20S].

In keeping with B-cell immaturity, occasional cases of BL and BLL may be of pre-pre-B-cell and pre-B-cell phenotype [37, 20S-210]. Although the pre-pre-B-cell Burkitt's lymphoma might appear indistinguishable from acute lymphocytic leukemia, it is separable on the basis of its 'L3' morphology with basophilic cytoplasm and vacuole formation [209, 210]. Cases of pre-B­BLL with cytoplasmic mu, CALLA, and Tdt expression have been described as shown in figure 5S [37]. Interestingly, this pre-B BLL case presented as a primary liver tumor. Since the liver is the primary fetal source of pre-B-cells,

110

111

Figure 58. Burkitt's-like lymphoma with pre-B-cell phenotype. (A) Histologic appearance. (B) Mu expression. (Verdi CJ, Grogan TM, Protell R, et al: Liver biopsy immunotyping to characterize lymphoid malignancies. Hepatology 6:6-13, 1986.)

this neoplasm appears to derive not only from a specific stage of early B-cell development, but also from an organ relevant to that stage [211].

In a detailed way, some differences between BL and BLL emerge (un­published data). BL demonstrate more consistent CALLA expression: 100% CALLA + BL versus 33% CALLA + BLL. BL are largely of mu, delta isotype without evidence of further heavy chain switching to gamma or alpha [208]; whereas, BLL tumors frequently demonstrate heavy chain switching like large cell lymphomas [208]. The CALLA and heavy chain findings together suggest BL are on the average less mature than BLL (see figure 3).

Extranodal presentation and spread is the rule with both BL and BLL [82, 200, 201]. A propensity to pleural and CNS involvement has been noted [200]. Due to this predilection, some patients may present with neurologic signs and/or pleural effusions, necessitating cytocentrifuge immunotyping, as shown in figure 59.

Occasionally, both BL and BLL may derive from lymph nodes in folli­cular form, implying a relationship to germinal centers (see figure 60) [206, 212]. The frequent finding in BUBLL of B2 (C3dr), an antigen most consis­tently expressed in germinal centers, corroborates this anatomic observation [208].

Figure 57. Burkitt's lymphoma phenotype.

112

Figure 59. CNS involvement by Burkitt's lymphoma. (A) Cytologic appearance ; (B) Lambda staining.(C) Absent kappa staining.

Rare immature and mature T-cell phenotypes of BUBLL may be found as illustrated in figure 61A. In this instance, a peripheral T-cell (PTL) phenotype is found in a tumor with BLL morphology. Surprisingly, this PTL-BLL appears to have aberrant admixed myelocytes probably due to lymphokine-related abnormal myelocytopoiesis [170]. Clearly, confusion with a granulocytic sarcoma is possible (see figure 61B) unless extensive phenotyping is employed [170].

Pathogenesis

Presently, there is an intense interest in Burkitt's lymphoma because it offers specific clues from epidemiology to molecular biology concerning the pathogenesis of lymphoma. Our understanding of the pathogenesis of B-cell neoplasia may well come from this rare tumor; from this rarity may come important generalities.

The African children with BL frequently have high antibody titers to Epstein-Barr (EB) virus and in some cases EB virus particles have been isolated from cultures of neoplastic cells [213]. Most Burkitt's lymphomas have demonstrated an 8, 14 chromosomal translocation [214]. Recently, the

Figure 60. Follicular variant of Burkitt's-like lymphoma. H & E: histologic appearance. Ki-67: positive cell population. K: absent kappa staining. A: abundant lambda staining.

113

" -. ,:~ . -"

..

... " .. -::.

, ~ < "

.. . ... ";,. ... :

." .) ... . . -. '. .. , ...

... " , ,

. .. , .. \-.

.. ~ c, . "

• "

Figure 60.

114

A. Burkitt's-like, T-cell, Abnormal My 81

•

8. Burkltt'a-Uke, a ..... ulocytlc .-~~~~~~~~~~

..

Figure 61. (A) Burkitt's-like lymphoma of mature T-cell type. L4: pan-T-antigen expression. MY7: this myeloid marker identifies scattered myelocytcs representing abnormal myelocyto­poiesis due to lymphokine effect. (B) Granulocytic sarcoma with Burkitt's morphology. Note near absence of pan-T-cells (L4) and abundance of myelocytes identified by chloroacetate esterase (see figure 2).

8, 14 translocation has been related to a distinctive cellular transforming gene or oncogene [215]. Specifically, in Burkitt's lymphoma the c-myc onco­gene is translocated from its usual position (chromosome 8) to a position adjacent to genes coding Ig heavy chains (chromosome 14) . It is believed that this translocation (8 to 14) results in activation of the oncogene (myc) near the site of immunoglobulin gene expression. The consequence is an unbridled, autonomous proliferation of a single clone of Ig-producing B-cells [215].

115

It is believed that the EB virus results initially in a polyclonal B-cell proliferation. The continuous B-cell divisions are then thought to facilitate the chromosomal translocation that activates the c-myc oncogenes and con­sequent monoclonal proliferation. The specific cause of the chromosomal translocation remains unknown. However, since EB virus is ubiquitous and Burkitt's lymphoma has a high incidence in restricted areas (e.g., Uganda's lowlands), an environmental cofactor (e.g., malaria) is suspected. Although malaria has long been suspected, it is only in 1984 that studies demonstrated that cytotoxic T-cells which normally control Epstein-Barr virus infected B­cells are functionally impaired in acute malarial infection [216]. EB virus infects us all as a lifelong infection. The initial primary infection may take the form of infectious mononucleosis or a nonspecific viral illness. Natural killer cells are crucial in stemming the initial onslaught. Subsequently, specific cytotoxic T-lymphocytes maintain lifelong surveillance of our viral carrier state. Alteration of these critical cells by malaria or cytotoxic/ immunosuppressive drugs (e.g., cyclosporin A) or by other non-EB viruses (see below) releases the EB virus to drive B-cell expansion [216]. Recently, EB virus receptors (C3dr) have been found in germinal centers, suggesting that germinal centers may be the immunologically-privileged site where latent EB virus lurks [57]. This is of special interest since BL is occasionally seen to arise in a follicular pattern from germinal centers [212] (see figure 60).

Since nonendemic cases of American Burkitt's lymphomas occur inde­pendent of malaria, other cofactors besides malaria are implicated. Clues come from a recent epidemic of EB virus-related Burkitt's-like lymphoma noted in homosexual men in San Francisco [204]. It is likely that this group of individuals with many hallmarks of acquired immune deficiency syndrome (AIDS) have T-cells infected and destroyed by the retrovirus known as HTL V -III [217]. The loss of T-cell surveillance of EB virus and B-cells seems to open the door to B-cell neoplasia. Many B-cell neoplasms are speculated to occur through a 'two-hit' mechanism of oncogenesis [218]. This appears to be an example of two hits by two viruses - one with T-cell tropism and the other with B-cell tropism. Destruction of dendritic cells may also be a factor. Recently, it is understood that the T -helper antigen (Leu 3), which is the receptor for HTLV-III on T-helper cells, is also found on histiocytes and dendritic cells [41, 42, 219]. In AIDS, HTLV-III appears to attach to dendritic cells in lymph node germinal centers, causing dendritic cell destruction and consequent aberrant B-cell proliferation, leading to neoplasia [43]. Interestingly, some of these patients in San Francisco had jaw tumors, as in Uganda. Of further interest, many of these patients used 'recreational drugs' known as 'poppers' or amyl nitrate ~ could this be yet another cofactor [204]?

The emerging relationship between oncogenes and lymphomas has wide­spread provocative implications. It now appears that the different lymphomas which derive from separate stages of B-cell and T-cell differentiation each have different, stage-specific transforming genes (eco-oncogene in pre-B-

116

neoplasms, myc in Burkitt's) [220]. This suggests that specific activated transforming genes may ordinarily regulate lymphoid cell proliferation during normal B-cell and T-cell differentiation. In short, understanding the im­munopathology of lymphomas (oncogenes and related specific subsets of lymphoma) is leading to an understanding of normal immunology (e.g., normal transforming genes).

Differential diagnosis

Granulocytic sarcoma. This myeloid neoplasm may have a morphologic appearance identical to BL or BLL, as shown in figure 61. Both cytochemical and full immunologic phenotype generally allow distinction of granulocytic from lymphoid neoplasms, although the recent descriptions of peripheral T-cell lymphomas (PTL) with aberrant myelocytes may closely simulate granulocytic sarcomas (see discussion of PTL) [170].

Poorly differentiated epthelial neoplasm. As revealed in figure 62, some poorly differentiated epithelial neoplasms may greatly simulate BL or BLL. In figure 62 note the touch prep highly suggestive of BL. Nonetheless, immunophenotyping revealed no lymphoid markers; histology showed cel­lular nesting more consistent with an epithelial neoplasm; ultrastructural analysis revealed dense core granules consistent with a neuroendocrine neo­plasm; uranaffin staining of these granules further corroborated their neuro­secretory status [221, 222]. Clearly, definitive diagnosis of immature blastic tumors is a difficult circumstance requiring an integrative approach using at once immunology, histology, ultrast1;uctural analysis, and cytochemical staining. Two notes of caution: 1) occasionally epithelial cells may cross­react with antilymphoid markers [48-53] and 2) lymphomas may possess neurosecretory-like granules [223, 224]. The confusion caused by cross­reactions may be obviated by use of a battery of MoAB, rather than isolated markers. The confusion caused by neurosecretory-like granules is assertively resolved through use of the uranaffin reaction, since lymphoma-related neurosecretory-like granules are not true neurosecretory granules and would be uranaffin negative [221, 222].

Peripheral T-cell lymphomas

General comments

The term peripheral T-cell lymphoma (PTL) is a broad immunologic cate­gorization of lymphomas, separate from the working formulation, which is characterized by mature, activated T-cell phenotypes [14, 15, 225]. These PTL are thought to derive from peripheral lymphoid sites outside the thy­mus, in contrast with central, thymic origin for immature T-cell lymphoblas­tic lymphomas [14, 15, 225].

117

Figure 62. Neuroendocrine tumor simulating Burkitt's lymphoma. (A) Touch preparation with Burkitt's cytologic appearance. (B) Nesting of cells in histologic section. (C) Neurosecretory granules at the ultrastructural level. L = Lysosome.

As neoplasms derived from mature T-cel\s, these cells expectedly express mature pan-T-antigens (e.g., Leu 1,4,5,9), lack immature T-antigens (Leu 6, Tdt, CALLA), and lack B-cell antigens [14,15,45,225-229]. In keeping with mature functional T-cell derivation, many express subset antigens (e.g., Leu 2 or Leu 3). Frequently, these subset antigens are expressed in a mutually exclusive fashion (Leu 2+, Leu r or Leu r, Leu 3+), signifying a monoclonal T-cell proliferation [14, 15, 45, 225-229]. Many PTL have 'activated' T-cell status with expression of activation antigens (la, TAC [IL-2 receptors], and TRF) [14, 15, 228, 229]. Many are also functionally active T-cell neoplasms with Iymphokine activity [230].

Although PTL phenotypes are broadly consistent with our expectations

118

(see figure 4) as neoplasms derived from mature, activated, functional T­cells, they nonetheless frequently show an unexpected aberration of pheno­type, which may prove most helpful in diagnosis [2, 4, 6]. In the majority (60%-80%) of PTL (excluding MF) there is loss of one or more pan-T­antigens (e.g., Leu 1, 4, 5, 9) [14, 15, 45]. This 'idiosyncratic' loss is not usual in inflammatory T-celI proliferations [22, 231, 232] but it is associated with clonal rearrangement of the T-antigen receptor genes in neoplastic T­cell proliferations [18] . These novel PTL phenotypes have no corresponding normal T-ontogeny phenotypes [14, 15]. This suggests, unlike B-celI neo­plasms, that PTL may not recapitulate normal lymphoid development. It is specificalIy this novel or idiosyncratic feature that helps characterize PTL as a distinct immunologic entity [14, 15].

MorphologicalIy, PTL comprise a broad spectrum of morphologic types (figure 63). Although PTL are most frequently classified in the WF as large celI (LCL) and/or immunoblastic (IBL) types, they may also be commonly found in the mixed category as well as any of the other WF diffuse cate­gories (see figure 63) [1, 72, 124, 154, 161, 233-245] . HistologicalIy, PTL are characterized by their pleomorphism and polymorphism. The pleomor­phism may at times be so startling that abundant polylobated Reed-Sternberg­like celIs beg distinction of PTL from Hodgkin's disease [228]. Indeed, this has led to the histologic designation of Hodgkin's-like PTL by some (figure 63) [228]. The polymorphism takes the form of frequent admixture of reactive elements within PTL: reactive B-celIs and T-cells, plasma cells, eosinophils, myeloid elements, fibroblasts, epithelioid histiocytes, and vas­cular elements [14, 15, 45, 225-229]. Commonly in PTL, these reactive elements outnumber neoplastic elements, leading to great difficulty in mor­phologic and immunologic diagnosis [14]. This difficulty is reflected in the length of the table in figure 63, indicating the great variety of names and eponymic designations given PTL. It is likely that this morphologic variation is largely due to the functional lymphokines produced by PTL cells [230].

Clinically, exclusive of MF, most patients are older (50-60 years of age), with frequent generalized lymphadenopathy, a high frequency of extranodal disease, B-symptoms, and a poor prognosis (median survival: 9-22 months) [14, 226, 227, 229]. Some have felt this poor prognosis is not unique to PTL, but simply reflects their frequent high-grade histology (e.g., LCLlIBL) [228, 246, 247]. Nonetheless, others have felt that clinical features (e.g., transient response to treatment, frequent relapse, fulminant course with localized presentation) might signify a more aggressive disease, requiring new thera­peutic strategies [14, 226, 227, 229, 230]. The issue of new therapeutic strategies remains unresolved, since unpredictably some PTL have indolent courses and they might do poorly with aggressive therapy [246].

The table in figure 63 indicates the great variety of variants under the designation of PTL [233-245]. Indeed, these rare PTL variants, considered to comprise only 5%-20% of alI lymphomas, outnumber the WF categoriza­tions of all other lymphomas combined. This listing reflects the morpho-

119

(Reference ~)

1. Wbrkinq formulation desiqnations: All diffuse: Larqe cell Burkitts-like Large cell, immunoblastic

• Mix~ • Small cell • Small cleaved cell

2. Historic and miscellaneous designations: T-cell LmmUnoblastlc sarcoma (clear cell) of

Lukes and Collins Multilobated T-cell lymphoma of Pinkus T-zone lymphoma (Lennert) Medium-sized cell lymphanas (Japanese) Node-based T-cell lymphcma Hodgkin's-like non-Hodgkin's lymphcma Plasmacytoid T-cell lymphomas

3. Lymphokine - related desiqnations: Mixed cell variants with abundant histiocytes (Lennert's

lymphana). T-cell lymphomas with plasmacytosis and hypergammopathy T-cell lymphoma with aberrant myelocytopoiesis

and eosinophilia Erythrophaqocytic T-cell lymnhoma simulatinq maliqnant

histiocytosis Larqe cell lymphcmas with compartmentalizinq fibrosis T-cell lymphomas with hypercalcemia

4. Grqan associated designations: Pulmonary: Waldron's 1 ymphana Splenic: 1° splenic PTL simulatinq hairy cell leukemia Cutaneous:

low qrade: Mycosis fungoides, small cerebriform cell. high grade: mostly dermal, mostly large cell

5. vessel-associated desiqnations:

(1)

(233)

(161) (234) ( 235) (236) (228) (237)

(124)

(154) (237,238) (227) (169,239)

(225) (227)

(225) (72)

(45) (14,45)

Angioimmunoblastic lymphadenopathy-like (AILD-like) (240) T-cell lymphana

Angiocentric, angiodestructive immunoproliferative lesions: (241) ? lymphanatoid granulomatosis (pulmonary) ? lethal midline granuloma (nasopharynx) (242) ? lymphanatoid papulosis (skin)

6. Virus-associated desiqnations: Human T-cell leukemia/lymphoma virus-associated (243)

(HTLV-I) lymphoid malignancies • Adult T-cell leukemia/lymphoma of Japan and the (244,245)

Caribbean

Figure 63. Tabulation of peripheral T-cell lymphoma variants.

logic, immunologic, functional, clinical, and etiologic diversity of PTL. The emergence of PTL as an entity, in spite of such diversity, reflects several factors: firstly, its immunologic specificity: mature, activated T-cell status coupled with novel pan-T expression. Secondly, as revealed in figure 63, some PTL have etiologic definition (HTLV-I-related PTL). Thirdly, revela­tion of distinctive clinical patterns of disease (e.g., primary splenic PTL simulating hairy cell leukemia, PTL with hypercalcemia, or eosinophilia). Fourthly, implications of poor prognosis status. Finally, the PTL categoriza­tion identifies a lymphoma with prominent activation antigen expression which might be subject to therapeutic control by use of lymphokines or antibodies to lymphokines and their receptors (e.g., TAC, IL-2, TRF).

120

Phenotypic identity of PTL may become mandatory to allow proper im­munotherapy [230].

Further immunologic detail

The salient immunologic features of PTL are illustrated in figure 64. This cutaneous PTL demonstrates a Leu 3+ Leu 2~ phenotype indicative of a T­helper cell proliferation. Recent studies demonstrate that > 75% of PTL have exclusive subset antigen expression (e.g., Leu 2~ Leu 3+ or Leu 2+ Leu r) [14, 15]. The cutaneous lesion illustrated has Ia+ expression, as found in > 70% of PTL, signifying activated status [14, 15]. Finally, this lesion unexpectedly reveals absent Leu 1 expression due to loss of pan-T­antigen. Absent Leu 1 expression has specifically been found in 46% of PTL [15]. Coupled with Leu 4, 5, and 9 loss, the incidence of pan-T loss in PTL exceeds 66% overall [15]. This aberrant pan-T expression pattern can be of diagnostic utility, particularly with Leu rr PTL. The latter pattern, lack of subset T-antigen expression, can be difficult to diagnose otherwise. Oc­casional Leu 3+ 2+ PTL are described [15]. This might suggest immature T-LBL lineage, except that these PTL lack Leu 6, Tdt, or CALLA coex­pression as expected in LBL [14, 15]. Because of the elaborate phenoty­pic differences between PTL and LBL, it should be appreciated that use of a panel of T-cell monoclonal antibodies, not an isolated marker, is critical [14, 15].

Figure 65 illustrates a common diagnostic dilemma encountered in PTL immunologic assessment. The lesion represents the most common form of PTL: a large cell lymphoma with numerous admixed reactive T-cells [14]. As illustrated, the reactive elements may greatly outnumber the neoplastic cells. Tissue section typing produces a 'hodge-podge' effect of both neoplas­tic and reactive cells. However, typing performed on cytocentrifuge prepara­tions allows simultaneous sizing and typing of cells, revealing that the large blastic cells, presumed malignant, are of T-helper (Leu 3) type and the smaller, presumed reactive, cells are of T-suppressor/cytotoxic type. Mere quantitation independent of morphologic considerations would have falsely suggested a normal ratio of T H to Ts/c cells.

Histologic variants

Two rare but distinctive histologic variants of PTL are illustrated. Figure 39 shows a PTL with intermediate-size 'blastic' cells, very similar to the Japan­ese description of medium-sized cell lymphomas of T-cell type [235]. This lesion has no established place in the WF scheme, although we may place it among the DSCL as an atypical blastoid form analogous to the description

Figure 64. Peripheral T-cell lymphoma of skin with a novel T-cell phenotype.

\ " , ; " . . '- .. _ J ' , :~~ ~ .. \ . >- ::~

.' ... \' . " .. :

.... " ,. , .~ . ... .

• '. •

.. '.

e' ...

• I , .. ,

.. .-" ..

Figure 64,

121

4 ' '. ."" •

,'J . ' '. \ ~I t ' . ......... e... ..~ .. ~

" I

. ... ... l_t ',' ~ I,.: •• •

) • .';. "t l . • ' , . ~. } I

I .... "'. / . \1 • . : l " .1... .,

." I .' I

t:'\ ' ,' ', -

. '~/: .. '/ :. .. , .. ~ -1'--.'"'1 .....

~ ',:. :. ... :'''''

•

122

'\

o c.

)

Figure 65. Peripheral T-eelllymphoma with abundant admixed reactive T-cells. The top panels show large blastic cells , presumed malignant, with numerous admixed small lymphoid cells, presumed reactive. Both were E-rosette positive (bottom left), but only the large cells stained with Leu 3 (bottom , center), while the small cells were positive for thc T-suppressor antibody Leu 2 (bottom , right). (Grogan TM, Fielder K, Rangel C, et al: Peripheral T-cell lymphoma: Aggressive disease with heterogeneous immunotypes. Am J Clin Pathol 83:279-288, 1985.)

of Come et al. [111]. Note the high proliferative index and the mature activated T-helper phenotype. This is a distinctive lesion morphologically; its absence in the WF probably reflects its rarity in the West. Its description in Japan reflects its common occurrence there [235].

While most PTL are of large celllimmunoblastic or mixed cell type , occa­sional small cell cases are observed, as indicated in figure 66. This lesion represents a primary splenic PTL, designated as SLL in the WF. While the immunologic distinction may seem academic, this lesion clinically simulated hairy cell leukemia [72]. Given the potential for mistreatment, phenotyping proved pivotal, ensuring that chemotherapy, usually not indicated in hairy cell leukemia, was given [72].

Fig

ure

66.

Pri

mar

y sp

leni

c pe

riph

eral

T-c

ell

lym

phom

a si

mul

atin

g ha

iry

cell

leuk

emia

. (A

) A

t lo

w p

ower

, re

d pu

lp i

nvol

vem

ent

by

smal

l ly

mph

ocyt

ic

lym

phom

a ce

lls

(SL

L).

(C

) T

ouch

pr

epar

atio

n of

sam

e (B

) P

an-T

-ant

igen

L

eu

4 sh

ows

stro

ng

reac

tivi

ty.

(Gre

enbe

rg B

, G

roga

n T

M,

Tak

asug

i B

, et

al:

A u

niqu

e m

alig

nant

T-c

ell

Iym

phop

roli

fera

tive

dis

orde

r w

ith n

eutr

open

ia s

imul

atin

g ha

iry

cell

leuk

emia

. C

ance

r 56

:282

3-28

30,

1985

.)

......

N

Vol

124

Lymphokine-related variants

Much of the diagnostic difficulty with PTL stems from the morphologic variation due to lymphokine-related proliferations of reactive cells. These major histologic variants may be due to proliferation of epithelioid histio­cytes (see figures 40-42) [124], myeloid elements (see figure 61) [227,237, 238), fibroblastic elements (figure 47) [225), and/or B-cells [154] (figure 46). Because of this factor, for example, Lennert's lymphomas with an abun­dance of epitheliod histiocytes are sometimes confused with widespread infectious granulomatous disease [124].

Etiologic variants

HTLV-I-associated T-cell leukemia/lymphoma may be thought of as a dis­tinct subset of PTL in which a specific etiologic agent (e.g., retrovirus, type C) is established [243, 245]. Like PTL, HTLV cases are of mature T­phenotype; they have pleomorphic and polymorphous histology; they are widespread at presentation, with frequent cutaneous, pulmonary, and bone marrow involvement; and they have a rapid course (median survival: nine months), even with aggressive combination chemotherapy [243]. Clearly, with this many similarities between PTL and HTLV, all patients with PTL in the future should be tested for possible evidence of the human T-cell leukemia/lymphoma virus. This includes American patients because, although HTL V -I -associated PTL is mostly endemic in Japan [30] and the Carribbean [245], nonendemic cases are known [243]. Although Southern blacks are most frequent among American cases, the case illustrated in figure 67 represents an HTLV-I-associated T-cell leukemia occurring in a white male Arizonian who spent 10 years in the export-import business out of Haiti. The estimated incidence of HTL V -associated PTL in the United States is between 10%-30% of PTL cases [227].

HTLV-II has been associated with a T-cell variant of hairy cell leukemia (HCL) [248]. The description in this chapter of primary splenic PTL which may simulate HCL suggests a need for viral assessment to exclude HTLV-II.

Functional aberrations

Occasional PTLs may manifest aberrant surface membrane recycling result­ing in signet-ring cell formation due to sequestration of T-antigens. These signet-ring cells can be confused with those found in signet-ring B-cell lymphomas or in certain epithelial neoplasms (figure 33) [107, 108].

Differential diagnosis

As emphasized by others [22, 241, 242, 249-251), there is an intriguing and as yet uncertain relationship between PTL and certain immunoproliferative

125

Figure 67. HTLV-I associated T-cell leukemia. (A) and (B) note striking nuclear irregularity. (C) Leu 1 reactivity. (D) E-rosette receptor activity.

disorders: lymphomatoid granulomatosis (LG), lymphomatoid papulosis (LP) , lethal midline granuloma (LMG), and angioimmunoblastic lympha­denopathy (AILD). laffe et al. [241] have listed these as angiocentric, angiodestructive forms of post-thymic T-ce\l neoplasia (see figure 63). The T-cell immunotypes and the frequent occurrence of malignant transforma­tion and metastasis in cases of LG argue for T-ce\l neoplasia [250]. With both LP [22] and LMG [242], novel idiosyncratic T-ce\l phenotypes not found in inflammatory conditions [18, 22, 232] suggest these entities are more akin to Iymphoproliferative than inflammatory disorders (see figure 68). Yet LP lesions occur predominantly in crops of self-healing, self-limited lesions. How could these be considered malignant? Recent studies revealing multiple clonal rearrangements of the T-antigen receptor genes in LP sug­gest attempted clonal expansion [249]. It is suspected that LP which eventuate as true metastatic lymphomas are immortalized clonal expansions which escape host response, while self-healing lesions are held in check by host­related response [249]. Antigen receptor gene analysis reveals similar clonal rearrangement in AILD [252]. AILD is well described to eventuate in lymphomatous transformation (AILD-Iike PTL) [252] . Since both novel T­antigen profiles and T-antigen receptor gene rearrangements, as found in PTL, occur in LP and AILD, and neither novel pan-T-antigen loss or clonal

126

Figure 68. Lethal midline granuloma, nose. (A) Histologic features. (8) Staining for Leu 9 (pan-T-cell) antigen.

gene rearrangements occur in inflammatory conditions, LP and AILD are now more closely associated with immunoproliferative T-cell disorders than with inflammation [249, 252].

In spite of all this knowledge of the relationship of LG, LP, LMG, and AILD to PTL, a practical uneasiness remains. As long as uncertainty re­mains about when these disorders constitute true malignancies, treatment will remain unsure.

Granulocytic sarcoma. Historically, the presence of scattered eosinophilic myelocytes among blastic cells has led to a histopathologic diagnosis of granulocytic sarcoma [170]. However, PTL with admixed eosinophilic mye­locytes which mimic granulocytic sarcoma are now described (figure 61). As iIIustrated, the primary lymphoid nature of the proliferation is revealed by immunohistochemical studies. The eosinophilic myelocytes are presumed due to lymphokine stimulation of abnormal myelocytopoiesis [230]. The aberrant myelocytopoiesis in PTL may also take the form of a chronic myeloproliferative disorder (CMD), especially among plasmacytoid PTL [237, 238]. CMD among PTL suggests the possibility of developing con­current myeloid and lymphoid neoplasia, making the distinction of PTL from granulocytic sarcoma academic [237, 238].

Lymphocyte-depleted Hodgkin's disease. It has been emphasized that the common occurrence of large polylobated Reed-Sternberg cells in PTL may

127

lead to confusion of PTL with Hodgkin's disease [172, 228]. In one study it was estimated retrospectively that 25% of LDHD cases were in fact non­Hodgkin's lymphomas, most likely of PTL type, although this study was simply a retrospective morphologic reconsideration, not a prospective im­munologic comparison [172]. Nonetheless, confusion between Hodgkin's­like PTL and LDHD is well acknowledged [14, 228].

Several studies have shown that the Leu-M1 antigen, a monocyte/granu­locyte-related marker, is consistently expressed by Reed-Sternberg cells in HD [253, 254]. This suggests that Leu Ml + Reed-Sternberg cells might be immunologically separable from PTL Reed-Sternberg-Iike cells. However, subsequent study has established that nearly 70% of PTL demonstrated Leu M1 + expression [53], precluding immunologic distinction.

More recently, assessment of the pan-leukocyte antigen (LC) indicates its presence in PTL and usual absence in HD, suggesting this marker may be of utility in the distinction of PTL from HD [174]. PTL would be expected to yield a LC+T+LM1 + phenotype, while HD would usually have a LC- T - LM1 + phenotype [174].

Mantle zone lymphoma/'intermediate' lymphocytic lymphoma

General comments

Mantle zone lymphoma (MZL) and 'intermediate' lymphocytic lymphoma (IDL) are considered different manifestations of the same malignant process [74, 75]. In MZL the cells grow around (often atrophic) germinal centers, appearing as expanded mantle zones of a follicle - hence the descriptive terminology [75]. In IDL, the cells are found in a diffuse proliferation [74]. In either situation, the cells bear striking similarity to each other and consist of small lymphoid cells with slightly irregular nuclear contours, 'interme­diate' in character between the cells of SLL and SCL [74, 75].

Widespread disease at the time of diagnosis is common in either disease [74-76]. The clinical course is prolonged, even without treatment [74]. Therefore, although they are not part of the WF classification, they are considered low-grade lymphomas.

Immunologic comments

The MZLlIDL immunophenotype is that of a monoclonal B-cell population [74-77]. Most also coexpress the pan-T-cell antigen Leu 1 (figure 69) [9]. Thus, in theory, these diseases could be classified as either SLL or SCL. However, Leu 1 coexpression is quite common in SLL (figure 11) and only occasionally present in SCL, hence the realignment of these cases as closer to SLL than SCL, as was the case previously [9].

128

• ...... , . . ~

, .. 1 " . . .. . ,".

(.. " .

r

Figure 69,

129

Differential diagnosis

Benign disorders. Either disorder may be misinterpreted as a reactive pro­cess, especially early MZL. Indeed, it is only by demonstration of mono­clonality that the true nature of the process may be discovered in some cases. The coexpression of Leu 1 by B-cells is found only rarely in non­malignant diseases, usually in the tonsil [68], and this unique combination would speak against a reactive process.

Malignant disorders. Depending upon experience and technical considera­tions, confusion may arise with either SLL or SCL, especially in IDL. This differential may be difficult to resolve, and even immunotyping may not prove definitive (as discussed previously). Since all are indolent diseases in the main, this issue may not have serious implications. There are some details beginning to emerge which suggest separation of these diseases will perhaps be made on a clinical basis (e.g., MZL may have a predilection for the gastrointestinal tract [78]).

Extramedullary plasmacytoma

The histologic and clinical details of extramedullary plasmacytomas are well described elsewhere and will be discussed herein only as they relate to lymphoma. As a neoplasm of mature plasma cells, plasmacytomas complete the scheme of B-cell tumors (see figure 3). True to B-cell ontogeny, these neoplasms demonstrate striking monoclonal cytoplasmic immunoglobulin expression [5, 255]. Both light and heavy chain restriction is the rule, with occasional cases of Bence-Jones type having only light chain expression [5, 255].

With such a well described cell of origin, the plasma cell, immunologic characterization would seem very straightforward. Nonetheless, a few com­plexities may foil the unwary. Firstly, polymorphism among light chains, especially lambda light chain, may result in a failure to react with a mono­clonal antilambda [27]. As revealed in figure 3, plasma cells have few surface markers and lack pan-B-antigens [5, 255]. Hence, when polymor­phism leads to an artifactual Clg- phenotype, it may also be coupled with a lack of pan-B markers to suggest an 'undifferentiated' malignancy. In this circumstance, antibodies directed to plasma cell (PC) and plasma cell asso­ciated (PCA) antigens may be the only means of lineage delineation [151, 152]. The PC+PCA + Clg- phenotype should then lead to repeat Clg deter-

Figure 69. Mantle zone lymphoma histology (A). Faint Leu 1 staining in mantic zone (B). Lambda light chain staining within atrophic germinal center; absent in mantle zone (C). Kappa staining of mantle zone and germinal center (D).

130

Figure 70. Plasmacytoma. (A) Histology. (B) Cytoplasmic kappa expression in paraffin­embedded tissue.

minations with polyclonal heterosera, allowing more certain Ig detection. A note of caution: Since PCA may also be found on myelomonocytic cells, a complete monocyte! granulocyte panel should also be performed [15], 152].

As revealed in figure 70, the most successful means of CIg demonstration in plasma cells employs heterosera on paraffin embedded, B5 fixed material.

A recent intriguing development has been the finding that some plasma­cytomas have an immature CALLA + component which portends a poor prognosis [32, 256, 257]. These immature plasma cell proliferations appear to be a morphologic and immunologic hybrid of plasma cell and lymphoid cell features [256, 258]. In some cases a striking pre-B-cell component may be found admixed with plasma cells (see figure 71). This pre-B-cell com­ponent which expresses mu, CALLA, and Tdt also expresses PC and PCA antigens, indicating these are aberrant pre-B-myeIOlJla cells, not normal counterparts [259]. This pre-B-component has been demonstrated to have a high proliferative index with Ki-67, indicating this may represent the stem cell component of myeloma. Immunotyping suggests that some cases of plasmacytoma and myeloma are more like immature lymphoma and!or leu­kemia than previously imagined [10, 259].

A

B

c F

igur

e 71

. P

lasm

acyt

oma

wit

h pr

e-B

-cel

l co

mpo

nent

. (A

) L

arge

pla

sma

cell

wit

h sm

all

adja

cent

lym

phob

last

ic c

ells

. (B

) Sm

all

pre-

B­

cells

sta

inin

g fo

r m

u. (

C)

Ki-6

7 re

acti

vity

am

ong

pre-

B-c

ells

ind

icat

ing

high

pro

life

rati

on r

ate.

......

w

......

132

Mycosis fungoides

Mycosis fungoides (MF) is an indolent cutaneous T-cell Iympholiferative disorder. These small cerebriform neoplastic lymphoid cells are usually of mature T-helper cell type (figure 72) [20,260,261]. They demonstrate epider­matropism with clustering around Langerhan's cells, resulting in a distinc­tive epidermal lesion referred to as a Pautier's microabscess (figure 72) [260, 262, 263]. The close association of Langerhan's cells and mature T-cells suggests Langerhan's cells may be part of the pathophysiologic condition leading to neoplasia [260]. Perhaps a carcinogenic agent in Langerhan's cells (e.g., virus) leads to aberrant T-cell homing and proliferation at this site [260].

It has been said that the early plaque stage of MF is both histologically and immunologically indistinguishable from chronic reactive cutaneous in­flammations since both conditions exhibit a predominance of T-helper cells [260]. Recently, it has been appreciated, to the contrary, that MF cells may have distinctive phenotypes [20, 261]. Leu 8 and Leu 9 antigens , normally present in the majority of mature T-cells in inflammation, are absent in the majority of MF cases (> 80%) [20,21,261,264]. This deficiency of Leu 8 and Leu 9 suggests that a major difference in cellular antigen expression may exist between benign and malignant cutaneous T-cell infiltrates [261]. As revealed in figure 72, the following phenotype may be considered most characteristic of MF: Leu 1+, r, 3+, 4+, 5+, 8-, 9- . Also described are occasional MF cases of T-suppressor-cytotoxic type (Leu 2+, r), T-subset negative type (Leu r, r), and those with T-subset coexpression of both T-helper and T-cytotoxic antigens (Leu 2+3+) [260]. Interestingly, these aberrant non-T-helper forms are most commonly associated with the late tumor stage of MF. These aberrant MF phenotypes have a propensity to rapid progression [260]. Interestingly, the large pleomorphic cells in advanced tumor phase MF express Leu Ml [1, 260], as found in large neoplastic cells in Hodgkin's disease and PTL [265], adding to the view that the large cells in the three disorders may be of related origin [251].

Composite lymphomas

Composite lymphomas represent the occurrence of two distinct architectural and cytologic subtypes of lymphoma at a single anatomic site [266]. Rarely this may represent the coincidence of Hodgkin's disease and non-Hodgkin's disease, as revealed in figure 73. In this example, there is a Burkitt's-like lymphoma in association with nodular sclerosis Hodgkin 's disease which occurred in a patient previously treated for Hodgkin's disease. The associ a-

Figure 72. Mycosis fungoides immunotype.

133

Figure 72.

134

Figure 73. Composite lymphoma: Hodgkin 's disease (left side) , Burkitfs-like lymphoma (right side). (Grogan TM: Histopathology of Hodgkin 's disease . Invited chapter in Surgical Pathology of the Lymph Node. E Jaffe, editor. Major problems in pathology series. WB Saunders, Philadelphia, PA, March, 1985.)

tion between Burkitt's like lymphoma and treated Hodgkin's disease may hardly be considered a coincidence given the high incidence of this associa­tion [203].

More commonly composite lymphomas represent two different subtypes of non-Hodgkin's lymphoma in which one component is related to the other. Figure 74 illustrates an instance of follicular small cleaved cell lym­phoma and surrounding diffuse large cell lymphoma. Note that the large cell component shows considerably greater transferrin receptor expression and proliferative activity with Ki-67, in keeping with its transformed status. Also note that the strong IgA, K-staining on the FSCL cells is substantially weaker on the LCL cells, also consistent with their transformed status. Although the kappa expression on the LCL cells is weak, it does suggest both lymphomas are of the same isotype and are likely to derive from the same clone [267]. Past experience indicates the clinical behavior of this lesion is likely to be dependent on the higher grade component [268, 269].

Figure 74. Immunotypes of a composite lymphoma. Illustrated are two nodules of a follicular small cleaved cell lymphoma (FSCL) surrounded by a large cell lymphoma (LCL).

135

Figure 74.

136

Conclusion

Two-dimensional electrophoretic gel analysis indicates 2013 distinct proteins in the average lymphocyte [270]. The extensive battery of monoclonals, described in this chapter picks up 40-50 characters in this 2013 word alpha­bet. Clearly, it is difficult to speak in sentences or paragraphs when so little of the immunologic alphabet is known. Given this degree of ignorance, arrogance or dogma is inappropriate. Until more is known, an openness to the inevitably increasing complexity of the lymphocytic language seems appropriate.

References

1. Rosenberg SA, Berard CW, Brown BW: National Cancer Institute sponsored study of classifications of non-Hodgkin'S lymphoma: Summary and description of a working for­mulation for clinical usage. Cancer 49:2112-2135, 1982.

2. Jaffe ES, Strauchen JA, Berard CW: Predictability of immunologic phenotype by mor­phologic criteria in diffuse aggressive non-Hodgkin'S lymphomas. Am J Clin Pathol 77:46-49, 1982.

3. Rappaport H: Tumors of the hematopoietic system. In: Atlas of Tumor Pathology. Fascicle 8, Washington, DC: Armed Forces Institute of Pathology, 1966.

4. Mann RB, Jaffe ES, Berard CW: Malignant lymphomas: A conceptual understanding of morphologic diversity. Am J Pathol 94:105-175,1979.

5. Nadler LM, Ritz J, Griffin JD, et al: Diagnosis and treatment of human leukemias and lymphomas utilizing monoclonal antibodies. Prog Hematol 12:187-225, 1981.

6. Foon KA, Todd RF: Immunologic classification of leuk,emia and lymphoma. Blood 68:1-31,1986.

7. Warnke RA, Rouse RV: Limitations encountered in the application of tissue section immunodiagnosis to the study of lymphomas and related disorders. Human Pathol 16: 326-331, 1985.

8. Bernard A, Bovmsell L, Dausset J, Milstein C, Schlossman SF (eds): Leucocyte typing: Human leukocyte differentiation antigens detected by monoclonal antibodies. Berlin, Springer-Verlag, 1984, pp 740-751.

9. Spier CM, Grogan TM, Fielder K, Richter L, Rangel C: Immunophenotypic profiles in 'well differentiated' lymphoproliferative disorders, with emphasis on small lymphocytic lymphoma. Human Pathol, 17:1126-1136, 1986.

10. Schwarting R, Stein H, Wang CY: The monoclonal antibodies anti-S-HCL-I and anti-S­HCL-3 allow the diagnosis of hairy cell leukemia. Blood 65:974, 1985.

11. Stein H, Bonk A, Tolksdorf G, Lennert K, Rodt H, Gerdes R: Immunohistologic analysis of the organization of normal lymphoid tissue and non-Hodgkin'S lymphomas. J Histo­chem Cytochem 28:746-760, 1980.

12. Grogan TM, Rangel CS, Richter LC, Wirt DP, Villar HV: Further delineation of the immunoarchitecture of the human spleen. Lymphology 17:61-68, 1984.

13. Levy R, Warnke R, Dorfman RF, et al: The monoclonality of human B-celllymphomas. J Exp Med 145: 1014-1028, 1977.

14. Grogan TM, Fielder K, Rangel C, et al: Peripheral T-cell lymphoma: Aggressive disease with heterogeneous immunotypes. Am J Clin Pathol 83:279-288, 1985.

15. Weiss LM, Crabtree GS, Rouse RV, Warnke RA: Morphologic and immunologic charac­terization of 50 peripheral T-cell neoplasms. Am J Pathol 118:316-324, 1985.

16. Cossman J, Chused TM, Fisher RI, et al: Diversity of immunological phenotypes of

137

lymphoblastic lymphoma. Cancer Res 43:4486-4490, 1983. 17. Grogan TM, Payne CM, Payne TB, Spier C, Cromey OW, Rangel C, Richter L: Cutane­

ous myiasis: Immunohistologic and ultrastructural morphometric features of a human botfly lesion. Am J Dermatopathol, 9:232-239, 1987.

18. Bertness V, Kirsch I, Hollis G, Johnson B , Bunn PA: T-cell receptor gene rearrange­ments as clinical markers of human T-cell lymphomas. N Engl 1 Med 313:534-538, 1985.

19. Kansas GS, Wood GS, Fishwild OM, Engleman EE: Functional characterization of human T lymphocyte subsets distinguished by monoclonal anti-Leu 8. J Immunol 134: 2995-3002, 1985.

20. Abel EA , Wood GS, Hoppe RT, Warnke RA : Expression of Leu 8 antigen, a majority T­cell marker, is uncommon in mycosis fungoides. 1 Invest Derm 85:199-202, 1985.

21. Gatenby PA, Kansas GS, Xi an CY, Evans RL, Engleman EG: Dissection of immuno­regulatory subpopulations of T lymphocytes within the helper and suppressor sublineages in man. J of Immunol 129:1997-2000, 1982.

22. Wood GS, Strickler JG, Deneau DG, Egbert B, Warnke RA: Lymphomatoid papulosis expresses immunophenotypes associated with T cell lymphoma but not inflammation. J Am Aead Dermatol, 15:444-458, 1986.

23. Warnke R, Levy R: Immunopathology of follicular lymphomas: A model of B-Iymphocyte homing. N Engl J Med 298:481-486, 1978.

24. Warnke R, Levy R: Detection of T and B cell antigens with hybridoma monoclonal antibodies: A biotin-avidin-horseradish peroxidas~ method. 1 Histochem Cytochem 28: 771 - 776, 1980.

25. Grogan TM, Jones SE: Atlas of the immunologic, pathologic and clinical aspects of the malignant lymphomas. Adria monograph series, Adria Laboratories, Columbus, Ohio, 1984.

26. Lokin MR, Herzenberg LA : Analysis of cell populations with a fluorescence-activated cell sorter. Ann NY Acad Sci 254:163-171,1975.

27. Cossman 1, Neckers LM, laffe ES: Monoclonality of B-ce\1 lymphomas: Demonstration by a simple assay for immunoglobulin in RNA. Lab Invest 50: 13A , 1984.

28. Casey IT, Posey DH, McCurley TL: OKT4 epitope deficiency in significant proportions of the black population. A cause for underestimation of helper/suppressor lymphocyte ratios. Arch Pathol Lab Med 1 \0:702- 704, 1986.

29. Grogan TM , Rangel CS, Wirt DP, Richter LC, Bever FN , Jolley CS, Villar HV, Jones SE: Immunotopography of splenic lymphoma of small cleaved B-cell type. Diagn Im­munol 3:126-132,1985.

30. Bindl 1M, Warnke RA: Advantages of detecting monoclonal antibody binding in tissue sections with biotin and avidin reagents in Coplin® jars. Am 1 Clin Pat hoi 85:490-493, 1986.

31. Grogan T, Spier C, Wirt DP, Hicks MJ, Paguin M, Hutter 1, Miller T, Rangel C, Richter L, 10nes S: Immunologic complexity of lymphoblastic lymphoma. Diagn Immunol 4:81-88, 1986.

32 . Durie BGM, Grogan TM: CALLA-positive myeloma: An aggressive subtype with poor survival. Blood 66:229-232, 1985.

33. Straus W: Peroxidase procedures, technical problems encountered during their applica­tion. J Histochem Cytochem 27: 1349, 1979.

34. Kronland R, Grogan T, Spier C, Wirt D, Rangel C, Richter L, Durie B, Greenberg B, Miller T, 10nes S: Immunotopographic assessment of lymphoid and plasma cell malignan­cies in the bone marrow. Hum Pathol 16: 1247-1254, 1985.

35. Warnke RA, Bindl J, Doggett R: Methods for decreasing interstitial immunoglobulin in tissue slices and cryostat sections. Histochem 1 15:637, 1983.

36. Wood GS , Warnke R: Suppression of endogenous avidin-binding activity in tissues and its relevence to biotin-avidin detection systems. 1 Histochem Cytochem 29: 1\ 96-1204, 1981.

37. Verdi CJ, Grogan TM, Protell R, Richter L, Rangel C: Liver biopsy immunotyping to characterize lymphoid malignancies. Hepatology 6:6- 13, 1986.

138

38. Warnke R, Miller R, Grogan T, et al: Immunologic phenotype in 30 patients with diffuse large-cell lymphoma. N Engl 1 Med 303:293-300, 1980.

39. Barnett Foster DE, Dorrington Kl, Painter RH: Structure and function of immunoglo­bulin domains. VIII. An analysis of the structural requiremcnts in human IgGI for binding to the Fc receptor of human monocytes. 1 Immunol 124:2186-2190, 1980.

40. Wang CY, Good RA, Ammirati P, et al: Identification of a p69,71 complex expressed on human T cells sharing determinants with B-type chronic lymphatic leukemic cells. J Exp Med 151:1539-1544, 1980.

41. Wood GS, Warner NL, Warnke RA: Anti-Leu 3/T4 antibodies react with cells of mono­cyte/macrophage and Langerhans lineage. 1 Immunol 131:212, 1983.

42. Dalgleish AG, Beverley PCL, Clapham PR, Crawford DH, Greaves MF, Weiss RA: The CD4(T4) antigen is an essential component of the receptor for the AIDS retrovirus. Nature 312:763, 1984.

43. Wood GS, Garcia CF, Dorfman RF, Warnke RA: The immunohistology of follicle lysis in lymph node biopsies from homosexual men. Blood 66:1092-1097, 1985.

44. Salahuddin SZ, Rose RM, Groopman lE, Markham PD, Gallo RC: Human T Iympho­trophic virus type III infection of human alveolar macrophages. Blood 68:281-284, 1986.

45. Wood GS, Burke lS, Horning S, et al: The immunologic and clinicopathologic hetero­geneity of cutaneous lymphomas other than mycosis fungoides. Blood 62:464-472, 1983.

46. Weiss LM, Bindl 1M, Pieozzi Vl, Link MP, Warnke RA: Lymphoblastic lymphoma: An immunophenotype study of 26 cases with comparison to T ccll acute lymphoblastic leukemia. Blood 67:474-478, 1986.

47. Roberts GT, Babawi SB, Sackey K, Splence D, Sheth KV, Aur RJ: Lineage amhiguity in acute leukemia. Cancer 58:1473-1478, 1986.

48. Garson lA, Beverly PC, Coakham HB, ct al: Monoclonal antibodies against human T lymphocytes label Purkinjc neuroncs of many species. Nature 298:375-377, 1982.

49. Rusthoven 11, Rohinson lB, Kolin A, Pinkerton PH: The natural killer cell-associated HNK-l (Leu 7) antibody reacts with hypertrophic and malignant prostatic epithelium. Cancer, in press.

50. Patterson Bl, Robinson lB, Pinkerton PH: Hematopoietic cell surface markers on meta­static small cell carcinoma with monoclonal antihodies. Am 1 Clin Pathol 84: 159-165, 1985.

51. Kawaguchi K, Koike M: Neuron-specific enolase and Leu-7 immunoreactive small round­cell neoplasm. The relationship to Ewing's sarcoma in bone and soft tissue. Am J Clin Pathol 86:79-83, 1986.

52. Gazdar AF, Bunn PA, Minna lD, Baylin SB: Origin of human small cell lung cancer. Science 229:679, 1985.

53. Sheibani K, Battifora H, Burke lS, Rappaport H: Leu-MI antigen in human neoplasms: An immunohistologic study of 400 cases. Am 1 Surg Pathol 10:227-236, 1986.

54. Metzgar RS, Borowitz Ml, lones NH, Dowell BL: Distribution of common acute lym­phoblastic leukemia antigen in nonhematopoietic tissues. 1 Exp Med 154: 1249-1254, 1981.

55. Grogan TM, Verdi C: CALLA expression on hi Ie canaliculi. N Engl 1 Med 312:993-994, 1985.

56. Young LS, Clark D, Sixhey JW, Riekinson AB: Epstein-Barr virus receptors on human pharyngeal epithelia. Lancet 1:240-242, 1986.

57. Fingeroth lD, Weiss 11, Tedder TF, Strominger 11., Bird PA, Fearon DT: Epstein-Barr virus receptor of human B lymphocytes in the C3d receptor CR2. Proc Nat! Acad Sci USA 81:4510-4514, 1984.

58. Greenspan lS, Greenspan D, Lennette ET, Ahrams DI, Corrant MA, Peterson V, Freese UK: Replication of Epstein-Barr virus within the epithelial cells of oral 'hairy' leukopla­kia, an AIDS-associated lesion. N Eng 1 Med 313:1564-70,1985.

59. Shaw GM, Harper ME, Hahn BH, Epstein LG, Gajdusek DC, Price RW, Navia BA, Petito CK, O'Hara Cl, Groopman lE, Cho E-S, Oleske 1M, Wong-Staal F, Gallo RC:

139

HTLV-IIl infection in brains of children and adults with AIDS encephalopathy . Science 227:177- 181,1985.

60. Grogan TM, Tubbs RR: A double blind comparative immunotypic study between two institutions phenotyping non-Hodgkins lymphomas. Am] Clin Pathol, 87:478-484, 1987.

61. Borowitz M], Newby S, Brynes RK, Cousar JB, Whitcomb CC, Crissman JD, Byrne GE, Collins RD: Multi-institution study of non-Hodgkin's lymphomas using frozen section immunoperoxidase: The Southeastern Cancer study group experience. Blood 63:1147-1152, 1984.

62. Palutke M, Schnitzer B, Narang S, Hyder DM, Parks S, Tabaczka P, Patch J: Storage and transportation of lymphoid tissue for immunophenotyping. Am J Clin Pat hoi 85:494-497, 1986.

63. Sheibani K, Winberg CD, Van De Velde S, Williams J, Rappaport H: Evaluation of expression of lymphocyte antigens in tissues placed in transport media. Comparison with cryostat fresh frozen section technique. An immunologic study of 56 cases. Am J Clin Pathol 85:297-304, 1986.

64. Jones SE, Grozea PN, Miller TP, Van Slyck EJ, Balcerzak SP, Costanzi JJ, Morrison FS, Eyre HJ, Fabian CJ, Dabich L, Dixon DO, Hartsock RJ, Grogan TM, Kjeldsberg C, Schnitzer B: Chemotherapy with cyclophosphamide, doxorubicin, vincristine, and pred­nisone alone or with Levamisole or with Levamisole plus BCG for malignant lymphoma: A Southwest Oncology group study. J Clin Oncol 3:1318-1324,1985.

65. Metter GE, Nathwani BN , Burke ]S , Winberg CD, Mann RB, Barcos M, Kjeldsberg CR, Whitcomb CC, Dixon DO, Miller TP, Jones SE: Morphological subclassification of folli­cular lymphoma: Variability of diagnoses among hematopathologists , a collaborative study between the repository center and pathology panel for lymphoma clinical studies. J Clin Oncol 3:25-38, 1985.

66. Pangalis GA, Nathwani BN, Rappaport H: Malignant lymphoma, well differentiated lymphocytic: Its relationship with chronic lymphocytic leukemia and macroglobulinemia of Waldenstrom. Cancer 39:999-1010, 1977.

67. Aisenberg AC, Long JC: Lymphocyte surface characteristics in malignant lymphoma. Am J Med 58:300, 1976.

68. Burns BF, Warnke RA, Doggett RS, Rouse RV: Expression of a T-cell antigen (Leu I) by B-cell lymphomas. Am J Pathol 113: 165-171, 1983.

69. Knowles DM II, Halper JP , Azzo W, Wang CY: Reactivity of monoclonal antibodies Leu 1 and OKT1 with malignant human lymphoid cells. Correlation with conventional cell markers. Cancer 52: 1369- 1377, 1983.

70. Cossman .I, Neckcrs LM , Hsu S-M, Longo D, Jaffe ES: Low grade lymphomas. Expres­sion of developmentally regulated B-cell antigens. Am J Pathol 115: 117-124. 1984.

71. Harris NL, Bhan AK: Distribution of T-cell subsets of follicular and diffuse lymphomas of B-cell type. Am J Pathol 113:172-180,1983.

72. Greenberg BR, Grogan TM , Takasug GJ , Trent JM , Hicks MJ, Durie BE: A unique malignant T-cell Iymphoproliferative disorder with neutropcnia simulating hairy cell leukemia. Cancer 56:2823-2830, 1985.

73. Colby TV, Warnke RA, Burke JS, Dorfman RF: Differentiation of chronic lymphocytic leukemia from Hodgkin's disease using immunologic marker studies. Am J Surg Pathol 5:707-710,1981.

74. Weisenburger DD. Nathwani BN, Diamond LW, Winberg CD, Rappaport H: Malignant lymphoma, intermediate lymphocytic type: A clinicopathologic study of 42 cases. Cancer 48:1415 - 1425, 1981.

75. Weisenburger DD, Kim H, Rappaport H: Mantic-zone lymphoma: A follicular variant of intermediatc lymphocytic lymphoma. Cancer 49: 1429- 1438, 1982.

76. Palutke M, Eisenberg L, Mirchandani I, Tabaczka P, Husain M: Malignant lymphoma of small cleaved lymphocytes of the follicular mantle zone. Blood 59:317- 322, 1982.

77. Weisenburger DD: Mantle-zone lymphoma. An immunohistologic study. Cancer 53:1073-1080, 1984.

140

78. Triozzi PL, Borowitz MJ , Gockerman JP: Gastrointestinal involvement and multiple lymphomatous polyposis in mantle-zone lymphoma. J Clin Oncol 4:866-873 , 1986.

79. Harousseau JL, Flandrin G, Tricot G , Brouet JC, Seligmann M, Bernard J: Malignant lymphoma supervening in chronic lymphocytic leukemia and related disorders. Richter's Syndrome: A study of 25 cases. Cancer 48 :1302-1308, 1981.

80. Wintrobe MM, Lee GR, Boggs DR, Bithell TC, Foerster J, Athens JW, Lukens IN: Clinical Hematology, 8th edition. Lea and Febiger, Philadelphia, 1981 , Chapters 67 and 69.

8!. Hicks MJ , Grogan TM, Fielder K, Spier CM: Differentiation of chronic lymphocytic leukemia from other 'well to intermediate differentiated ' lymphoproliferative disorders by the mouse rosette assay . Diag Immunol 4:31-36, 1986.

82. Jones SE, Fuks Z, Bull M, Kadin ME, Dorfman RF, Kaplan HS, Rosenberg SA, Kim H: Non-Hodgkin's lymphomas. IV. Clinicopathologic correlation in 405 cases . Cancer 31 : 806-823 , 1973.

83. Berard CW, Dorfman RF: Histopathology of malignant lymphomas. Clin Haematol 3:39-76, 1974.

84. Nathwani BH, Winberg CD, Diamond LW, Bearman RM, Kim H: Morphologic criteria for the differentiation of follicular lymphoma from florid reactive follicular hyperplasia: A study of 80 cases. Cancer 48:1794-1806, 1981.

85 . Jaffe ES, Shevach EM, Frank MM, Berard CW, Green I: Nodular lymphoma - evidence for origin from follicular B lymphocytes. N Engl J Med 290:813-819 , 1974.

86. Stein H, Berdes J, Mason DY: The normal and malignant germinal centre. Clin Haematol 11 :531-559, 1982.

87. Tubbs RR, Fishleder A, Weiss RA, Savage RA, Sebek BA, Weick JK: Immunohistologic cellular phenotypes of Iymphoproliferative disorders: Comprehensive evaluation of 564 cases including 257 non-Hodgkin's lymphomas classified by the International Working Formulation. Am J Pathol 113:207-221, 1983.

88. Grogan TM , Hicks MJ, Jolley CS , Rangel CS , Jones SE: Identification of two major B cell forms of nodular mixed lymphoma. Lab Tnvest 51 :504-514, 1984.

89. Harris NL, Nadler LM, Bhan AK: Immunohistologic characterization of two malignant lymphomas of germinal center type (centroblastic/centrocytic and centrocytic) with mono­clonal antibodies. Follicular and diffuse lymphomas of small-cleaved-cell type are related but distinct entities. Am J Pat hoi 117:262-272, 1984.

90. Gerdes J , Dallenbach F , Lennert K: Growth fractions in malignant non-Hodgkin 's lym­phomas (NHL) as determined in situ with the monoclonal antibody Ki-67. Hematol Oncology 2:365-371,1984 .

91. Rudders RA, Ahl ET, Delellis RA, Bernstein S, Begg CB: Surface marker identification of small cleaved follicular center cell lymphoma with a highly favorable prognosis . Cancer Res 42:349-353 , 1982.

92 . Harris NL, Bahn A: Immunoglobulin in lymphomas and reactive lymph nodes. Human Pathol 13:6JO-617, 1982.

93. Garcia CF, Weiss LM, Warnke RA, Wood GA: Cutaneous follicular lymphoma. Am J Surg Pathol JO:454-463, 1986.

94. Naiem M, Gerdes J, Abdulaziz Z, Stein H, Mason DY: Production of a monoclonal antibody reactive with human dendritic reticulum cells and its use in the immunohisto­logical analysis of lymphoid tissue. J Clin Pathol 36:167-175,1983.

95. Dvoretsky P, Wood GS, Levy R , Warnke RA: T-Iymphocyte subsets in follicular lym­phomas compared with those in non-neoplastic lymph nodes and tonsils. Human Pathol 13:618-625, 1982.

96. Braziel RM, Sussman E, Jaffe ES , Neckers LM, Cossman J: Induction of immunoglobulin secretion in follicular non-Hodgkins lymphomas: Role of immunoregulatory T cells. Blood 66 :128-134,1985.

97. Ford RJ, Kouttab NM , Shasrabuddhe CG , Davis FM, Mehta SR: Growth factor mediated proliferation in B cell non-Hodgkin'S lymphomas. Blood 65: 1335-1341 , 1985.

141

98. Gerdes J, Lemke H, Baisch H, Wacker HH, Schwab U, Stein H: Cell cycle analysis of a cell proliferation-associatcd human nuclear antigen defined by the monoclonal antibody Ki-67, J of Immunol 133:1710-1715, 1984.

99. Aisenberg AC, Wilkes PM, Harris NL: Monoclonal antibody studies in non-Hodgkin's lymphoma. Blood 61:469, 1983.

100. Aisenbcrg AC, Wilkes BM, Long JC, Harris NL: Cell surface phenotype in lymphopro­liferative disease. Am J Mcd 68:206, 1980.

101. Gregg EO, Al-Saffar N, Jones BB, Wright DH, Stevenson FU, Smith JL: Immunoglo­bulin negative follicle centre cell lymphoma. Br J Cancer 50:735-744, 1984.

102. Gerdes J, Stein H: Complement (C,) receptors on dendritic reticulum cells of normal and malignant lymphoid tissue. Clin Exp Immunol 48:348, 1982.

103. Stashenko P, Nadler LM, Hardy R, Schlossman SF: Expression of cell surface markers after human B lymphocyte activation. Proc Nat! Acad Sci USA 78:3848, 1981.

104. Iida K, Nadler LM, Nussenzweig V: Identification of the membrane receptor for the complement fragment C,d by means of a monoclonal antibody. J Exp Med 158:1021-1033, 1983.

105. Kim H, Dorfman RF, Rappaport H: Signet ring cell lymphoma: A rare morphological and functional expression of nodular (follicular) lymphoma. Am J Surg Pat hoi 2: 119-132, 1978.

106. Frizzera G, Anaya JS, Banks PM: Neoplastic plasma cells in follicular lymphomas. Clini­cal and pathologic findings in six cases. Virchows Arch (Pathol Anat) 409:149-162, 1986.

107. Grogan TM, Payne CM, Richter LC, Rangel CS: Signet-ring cell lymphoma of T-cell origin. An immunocytochemical and ultrastructural study relating giant vacuole formation to cytoplasmic sequestration of surface membrane. Am J Surg Path 9:684-692, 1985.

108. Weiss L, Wood G, Dorfman R: T-cell signet-ring cell lymphoma. Am J Surg Pathol 9:273-280, 1985.

109. Rappaport H, Winter WJ, Hicks EB: Follicular lymphoma: A re-evaluation of its position in the schem of malignant lymphoma, based on a survey of 253 cases. Cancer 9:792-821, 1956.

110. Erickson OJ, Cousar JB, Flexner JM, Collins RD: Transformation of follicular center cell (FCC) lymphomas (Lukes-Collins classification): Progression of small cleaved cell (SCC) type to transformed cell type. Lab Invest 44: 16A, 1981.

111. Come SE, Jaffe ES, Anderson JC: Non-Hodgkins lymphomas in leukemic phase: Clinico­pathologic correlations. Am J Med 69:667-674, 1980.

112. Weiss LM, Warnke RA: Follicular lymphoma with blastic conversion: A report of two cases with confirmation by immunoperoxidase studies on bone marrow sections. Am J Clin Pathol 83:681-686, 1985.

113. Mann RB, Jaffe ES, Braylon RC, Nanba K, Frank MM, Ziegler JL, Berard CW: Non­endemic Burkitt's lymphoma: A B cell tumor related to germinal centers. N Engl J Med 295:685-691, 1976.

114. Mintzer OM, Andreeff M, Filippa DA, Jhanwar SC, Chaganti RSK, Koziner B: Progres­sion of nodular poorly differentiated lymphocytic lymphoma to Burkitt's-like lymphoma. Blood 64:415-421, 1984.

115. Gay R, Fielder K, Grogan T: Quinidine-induced reactive lymph adenopathy. Am J Med 82:143-145, 1987.

116. Wirt DP, Grogan TM Jolley CS, Rangel CS, Payne CM, Hansen RC, Lynch PJ, Schuchardt M: The immunoarchitecture of cutaneous pseudolymphoma. Human Pathol 6:492-510, 1985.

117. Levy N; Nelson J, Meyer P, Lukes RJ, Parker JW: Reactive lymphoid hyperplasia with single class (monoclonal) surface immunoglobulin. Am J Clin Pathol 80:300-308, 1983.

118. Sklar J, Cleary ML, Thiclemans K, Gralow J, Warnke R, Levy R: Biclonal B-ccll lymphoma. N Engl J Med 311:20-27,1984.

119. Schwartz JE, Grogan TM, Hicks MJ, Weinrach RS, Miller TP: Pseudonodular T-cell lymphoblastic lymphoma. Am J Med 77:947-949, 1984.

142

120. Timens W, Visser L, Poppema S: Nodular lymphocyte predominance type of Hodgkin's disease is a germinal center lymphoma. Lab Invest 54:457-461, 1986.

121. Hansmann M-L, Wacker H-H, Radzun Hl: Paragranuloma is a variant of Hodgkin's disease with a predominance of B-cells. Virchows Arch (Pathol Anat) 409: 171-181, 1986.

122. Anderson T, Chabner BA, Young RC, Berard CW, Garvin Al, Simon RM, DeVita VT lr: Malignant lymphoma. I. The histology and staging of 473 patients at the National Cancer Institute. Cancer 50:2699-2707, 1982.

123. Foucar K, Armitage JO, Dick FR: Malignant lymphoma, diffuse mixed small and large cell. A clinicopathologic study of 47 cases. Cancer 51:2090-2099, 1983.

124. Kim H, Jacobs C, Warnke RA, Dorfman RF: Malignant lymphoma with a high content of epithelioid histiocytes. A distinct clinicopathologic entity and a form of so-called "Lennert's lymphoma." Cancer 41:620-635, 1978.

125. Lenncrt K, Mestdagh 1: Lymphogranulomatosen mit konstant hohem epitheloid­zellgehalt. Virchows Arch (Pathol Anat) 344:1-20, 1968.

126. Garvin Al, Simon RM, Osborne CK, Merrill J, Young RC, Berard CW: An autopsy study of histologic progression in non-Hodgkin's lymphomas. 192 cases from the National Cancer Institute. Cancer 52:393-398, 1983.

127. Rappaport H, Pangalis GA, Nathwani BN: The evolution of immunoblastic lymphomas in morphologically non-neoplastic immunoproliferative diseases. Rev Cancer 29: 1502-1512, 1972.

128. Leonhard SA, Muhleman AF, Hurtubise PE, Martelo 01: Emergence of Immunoblastic sarcoma in Waldenstrom's macroglomulinemia. Cancer 45:3102-3107, 1980.

129. Kradin RL, Young RH, Kradin LA, Mark El: Immunoblastic lymphoma arising in chronic lymphoid hyperplasia of the pulmonary interstitium. Cancer 50:1339-1343,1982.

130. Bauer TW, Mendelsohn G, Humphrey RL, Mann RB: Angioimmunoblastic lymph­adenopathy progressing to immunoblastic lymphoma with prominent gastric involvement. An immunohistochemical study of two cases. Cancer 50:2089-2098, 1982.

131. Barcos MP, Lukes RJ: Malignant lymphoma of convoluted lymphocytes - a new entity of possible T-cell type. In: Sinks LF, Godden J, eds: Conflicts in Childhood Cancer. An Evaluation of Current Management, Vol. 4, New York, Alan R Liss, 1975, pp 147-148.

132. Nathwani BN, Kim H, Rappaport H: Malignant lymphoma, lymphoblastic. Cancer 38: 964-983, 1976.

133. Cossman 1, laffe ES, Fisher RI: Immunologic phenotypes of diffuse, aggressive, non­Hodgkin's lymphomas. Correlation with clinical features. Cancer 54: 1310-1317, 1984.

134. Tubbs RR, Fishleder A, Weiss RA, Savage RA, Sebek BA, Weick lK: Immunohistologic cellular phenotypes of Iymphoproliferative disorders. Comprehensive evaluation of 564 cases including 257 non-Hodgkin's lymphomas classified by the International Working Formulation. Am 1 Pathol 113:207-221, 1983.

135. Nathwani BN, Metter GE, Gams RA, Bartolucci AA, Hartsock Rl, Neiman RS, Byrne GE lr, Barcos M, Kim H, Rappaport H: Malignant lymphoma, mixed cell type, diffuse. Blood 62:200-208, 1983.

136. Burke IS, Butler JJ: Malignant lymphoma with a high content of epithelioid histiocytes (Lennert's lymphoma). Am 1 Clin Pathol 66:1-9, 1976.

137. Hayes D, Robertson IH: Malignant lymphoma with a high content of epithelioid histio­cytes. J Clin Pathol 32:675-680, 1979.

138. Kim H, Nathwani BN, Rappaport H: So-called "Lennert's lymphoma." Is it a clinico­pathologic entity? Cancer 45: 1379-1399, 1980.

139. Palutke M, Varadachai C, Weise RW, Husain M, Tabacza P: Lennert's lymphoma, a T­cell neoplasm. Am 1 Clin Pathol 69:643-646, 1978.

140. Han T, Barcos M, Yoon JM, Rakowski I, Minowada 1: Malignant lymphoma with a high content of epithelioid histiocytes: Report of a T-cell variant of so-called Lennert lym­phomas and review of the literature. Med Pediat Oncol 8:227-236, 1980.

141. Feller AC, Griesser GH, Mak TW, Lennert K: Lymphoepithelioid lymphoma (Lennert's lymphoma) is a monoclonal proliferation of helper/inducer T cells. Blood 68:663-667, 1986.

143

142. Spier CM, Lippman SM, Miller TP, Grogan TM: Lennert's lymphoma. A clinicopathologic study with emphasis on phenotype and its relationship to survival. Cancer, in press.

143. Doggett RS, Wood GS, Horning S, Levy R , Dorfman RF, Bindl J, Warnke RA: The immunological characterizations of 95 nodal and extranodal diffuse large cell lymphomas in 89 patients. Am J Pathol 115:245-252, 1984.

144. Horning SJ , Doggett RS , Warnke RA, Dorfman RF, Cox RS, Levy R: Clinical relevance of immunologic phenotype in diffuse large cell lymphoma. Blood 63:1209-1215, 1984.

145. Freedman AS, Boyd AW, Anderson KC, Fisher DC, Pinkus GS, Schlossman SF, Nadler LM: Immunologic heterogeneity of diffuse large cell lymphoma. Blood 65:630, 1985.

146. Acker B, Hoppe RT, Colby TV, Cox RS, Kaplan HS, Rosenberg SA: Histologic con­version in the non-Hodgkin 's lymphomas. J Clin Oncol 1:11-16,1983.

147. Risdall R, Hoppe RT, Warnke R: Non-Hodgkin's lymphoma. A study of the evolution of the disease based upon 92 autopsied cases. Cancer 44:529-542, 1979.

148. Levine AM, Taylor CR, Schneider DR, Koehler SC, Forman SJ, Lichtenstein A, Lukes RJ , Feinstein DI: Immunoblastic sarcoma of T-cell versus B-cell origin: I. Clinical fea­tures. Blood 58:52-61, 1981.

149. Maurer R, Taylor CR, Parker JW, Cramer A, Lukes RJ: Immunoblastic sarcoma: Mor­phologic criteria and the distinction of Band T cell types. Oncology 39:42-50, 1982.

150. Anderson KC, Bates MP , Siaghenhoupt B, Schlossman SF, Nadler LM: A monoclonal antibody with reactivity restricted to normal and neoplastic plasma cells. J Immunol 132:3172-3179, 1984.

151. Anderson KC , Park EK, Bates MP, et al: Antigens on human plasma cells identified by monoclonal antibodies. J Immunol 130: 1132-1138, 1983.

152. Linden MD , Tubbs RR, Weick JK: Large cell lymphoma with diffuse sclerosis: A B cell' neoplasm of late secretory pre-plasma cells. Cleve Clin Q.

153. Grogan TM, Spier CM, Miller T: Immunotyping malignant lymphomas: A boon to diagnosis. Cleve Clin Q 53 :315-317, 1986.

154. Tamaki T, Katagiri S, Kanayama Y , Konishi I, Yonezawa T , Tarui S, Kitani T: Helper T-cell lymphoma with marked plasmacytosis and polyclonal hypergammaglobulinemia. Cancer 53: 1590-1595, 1984.

155. Rosas-Uribe A, Rappaport H: Malignant lymphoma , histiocytic type with sclerosis (sclerosing reticulum cell sarcoma). Cancer 29:946-953, 1972.

156. Miller JB, Variakojis D, Bitran JD, Sweet DL, Kwizie JJ, Golomb HM, Ultmann JE: Diffuse histiocytic lymphoma with sclerosis: A clinicopathologic entity frequently causing superior venacaval obstruction. Cancer 47:748-756, 1981.

157. Trump DL, Mann RB: Diffuse large cell and undifferentiated lymphomas with prominent mediastinal involvement. A poor prognostic subset of patients with non-Hodgkin's lym­phoma. Cancer 50:277-282, 1982.

158. Waldron JA Jr, Newcomer LN, Katz ME, Cadman E: Sclerosing variants of follicular center cell lymphomas presenting in the retroperitoneum. Cancer 52:712-720, 1983.

159. Bennett MH: Sclerosis in non-Hodgkin's lymphomata. Br J Cancer 31 (Suppl II): 44-52, 1975.

160. Ree HJ, Leone LA, Crowley JP: Sclerosis in diffuse histiocytic lymphoma: A clinico­pathologic study of 25 cases. Cancer 49:1636-1648, 1982.

161. Pinkus GS, Said JW, Hargreaves H: Malignant lymphoma, T-cell type. A distinct mor­phologic variant with large multilobulated nuclei, with a report of four cases. Am J C1in Pathol 72:540-550, 1979.

162. Cerezo L: B-cell multi lobulated lymphoma. Cancer 52:2277-2280, 1983. 163. Weiss RL, Kjeldsberg CR, Colby TV, Marty J: Multilobulated B cell lymphomas: A study

of seven cases. J Hematol 3:79-86, 1985. 164. Kluin PM, Stootweg PJ, Schuurman JH, Go DM, Rademakers LH, van der Putte SC, van

Unnik JA: Primary B-cell malignant lymphoma of the maxilla with a sarcomatous pattern and multilobualted nuclei. Cancer 54: 1598- 1605 , 1984.

165. Jaffe ES: Malignant histiocytosis and true histiocytic lymphomas. In: Surgical Pathology of

144

the Lymph Nodes and Related Organs. Jaffe ES, (cd): Philadelphia, WB Saunders, 1985, Chapter 15.

166. Mirchandani I , Shah I , Palutke M, Varadachai C, Tabaczka P, Franklin R, Bishop C: True histiocytic lymphoma. A report of four cases. Cancer 52:1911-1918, 1983.

167. Isaacson P, Wright DH, Jones DB: Malignant lymphoma of true histiocytic (monocyte/ macrophage) origin. Cancer 51:80-91, 1983.

168. Turner RR, Wood GS, Beckstead JH, Colby TV, Horning SJ, Warnke RA: Histiocytic malignancies: Morphologic , immunologic and enzymatic heterogeneity. Am J Surg Path 8:485-500, 1984.

169. Jaffe ES, Costa JC, Fauci A, Cossman J, Tsokos M: Malignant lymphoma and erythro­phagocytosis simulating malignant histiocytosis. Am J Med 75:741-749, 1983.

170. Whitcomb CC, Sternheim WL, Borowitz MJ, Davila E, Byrne GE Jr: T-cell lymphoma mimicking granulocytic sarcoma. Am J Clin Pat hoi 84:760-763, 1985.

171. Greer JP, Kinney MC, Cousar JB, F1exner JM, Dupont WD, Graber SE, Greco FA, Collins RD , Stein RS: Lymphocyte-depleted Hodgkin 's disease. Clinicopathologic review of 25 patients. Am J Med 81:208-214, 1986.

172. Kant JA, Hubbard SM, Longo DL, Simon RM, DeVita VT Jr, Jaffe ES: The pathologic and clinical heterogeneity of lymphocyte depleted Hodgkin's disease. J Clin Oncol 4:284-294, 1986.

173. Miller TP , Bryne GE, Jones SE: Mistaken clinical and pathologic diagnosis of Hodgkin 's disease : A Southwest Oncology Group study. Cancer Treat Rep 66:645-652 , 1982.

174. Strickler JG, Michie SA, Warnke RA, Dorfman RF: The "syncytial variant" of nodular sclerosing Hodgkin's disease. Am J Surg Pathol 10:470-477, 1986.

175. Barcos MP, Lukes RJ: Malignant lymphomas of convoluted lymphocytes. A new entity of possible T cell type . In Sinks LF, Godden JO (eds): Conflicts in Childhood Cancer, vol 4, New York , Alan R Liss, 1975, p 147.

176. Nathwani BN, Kim B , Rappaport H: Malignant lymphoma, lymphoblastic . Cancer 38: 964-983 , 1976.

177. Bernard A, Boumsell L, Reinberg EL, Nadler LM, Ritz J , Coppin H, Richard Y, Valens F, Dausset J, Flandrin G, Lemerle J, Schlossman SF: Cell surface characterization of malignant T cells from lymphoblastic lymphoma using monoclonal antibodies: Evidence for phenotypic differences between malignant T cells from patients with acute lympho­blastic leukemia and lymphoblastic lymphoma. Blood 57:1105-1110,1981.

178. Roper M, Crist WM , Metzgar R , Ragab AH, Smith S, Storling K, Pullen J , Leventhal B, Bartolucci AA, Cooper MD: Monoclonal antibody characterization of surface antigens in childhood T cell lymphoid malignancies. Blood 61:830, 1983.

179. Link M, Warnke R, Finlay J, Amylon M, Miller R, Dilley J, Levy R: A single mono­clonal antibody identifies T-cell lineage of childhood lymphoid malignancies. Blood 62: 722-728, 1983.

180. Jaffe ES, Berard CW: Lymphoblastic lymphoma: A term rekindled with new precision . Ann Intern Med 89:415-417, 1978.

181. Reinherz EL, Kung P, Goldstein G, Levey RH, Schlossman SF: Discreet stages of human intrathymic differentiation . Analysis of normal thymocytes and leukemic lymphoblasts of T cell lineage. Proc Natl Acad Sci USA 77:1588, 1980.

182. Kjeldsberg CR, Wilson JF, Berard CW: Non-Hodgkin's lymphoma in children. Human Pathol 14:612-627, 1983.

183. Bhan AK, Reinherz EL, Poppema S, McClusky RT, Schlossman SF: Location of T-cell and major histocompatibility antigens in the human thymus. J Exp Med 152:771-782, 1980.

184. Donlon JA, Jaffe ES, Braylan RC: Terminal deoxynucleotidyl transferase activity in malignant lymphomas. N Engl J Med 297:461-464, 1977.

185. Habeshaw JA, Catley PF, Stansfield AG, Ganeshaguru K, Hoffbrand AV: Terminal deoxynucleotidyl transferase activity in lymphoma. Br J Cancer 39:566-569, 1979.

186. Kung PC, Long JC, McCaffrey RP, Ratliff RL, Harrison TA, Baltimore D: Terminal deoxynucleotidyl transferase in the diagnosis of leukemia and malignant lymphoma. Am J

145

Med 64:788-794, 1978. 187 . Kurek AS, Davey FR: Terminal deoxynucleotidyl transferase. Human PathoI12:867-869,

1981. 188. Long JC, McCaffrey RP, Aisenberg AC, Marks SM, Kung PC: Terminal deoxynucleo­

tidy I transferase positive lymphoblastic lymphoma. Cancer 44:2127-2139, 1979. 189. Borowitz MJ, Croker BP, Metzgar RS: Lymphoblastic lymphoma with the phenotype of

common acute lymphoblastic leukemia. Am J Clin Pathol 79:387-391, 1983. 190. Ritz J, Nadler LM, Bhan AK, Notis-McConarty J, Pesandro JM, Schlossman SF: Expres­

sion of common acute lymphoblastic leukemia antigen (CALLA) by lymphomas of B-cell and T-cell lineage. Blood 56:648-652, 1981.

191. Murphy S, Jaffe ES: Terminal transferase activity and lymphoblastic neoplasms (editorial). N Engl J Med 311:1373-1374,1984.

192. Beckstead JH, Wood GS, Turner RR: Histiocytosis X cells and Langerhans cells: Enzyme histochemical and immunologic similarities. Hum Pathol 15:826, 1984.

193. Mokhtar N, Hsu SM, Rajnikant PL, Haynes BF, Jaffe ES: Thymoma: Lymphoid and epithelial components mirror the phenotype of normal thymus. Human Pathol 15:378-384, 1984.

194. Nagle RB, Clark VA, McDaniel KM, Davis JR: Immunohistochemical demonstration of keratins in human ovarian neoplasms. J Histochem Cytochem 31: 1010-1014, 1983.

195. Nagle RB, McDaniel KM, Clark VA, Payne CM: The use of antikeratin antibodies in the diagnosis of human neoplasms. Am J Clin PathoI79:458-467, 1983.

196. Berard C, O'Conor GT, Thomas LB, Torloni H: Histopathological definition of Burkitt's tumor. Bull WHO 1969; 40:601-607, 1969.

197. Payne CM, Grogan TM, Croney DW, Ejore CG, Kerrigan DP: An ultrastructural morphometric and immunophenotypic evaluation of Burkitt's and Burkitt's-like lympho­mas. Lab Invest 57:200-217, 1987.

198. O'Conor GT, Rappaport H, Smith E: Childhood lymphoma resembling 'Burkitt's tumor' in the United States. Cancer 18:411-417, 1965.

199. Dorfman RF: Childhood lymphosarcoma in St. Louis, Missouri clinically and histological­ly resembling burkitt's tumor. Cancer 18:418-430, 1965.

200. Levine PH, Cho BR, Connelly RR, et al: The American Burkitt's Lymphoma Registry: A progress report. Ann Intern Med 83:31-36,1975.

201. Cohen MH, Bennett JM, Berard CW, et al: Burkitt's tumor in the United States. Cancer 23:1259-1272,1969.

202. Miliauskas MB, Berard CW, Young RC, ct al: Undifferentiated non-Hodgkin's lym­phomas (Burkitt's and non-Burkitt's types). The relevance of making this histologic dis­tinction. Cancer 50:2115-2121, 1982.

203. Krikorian JG, Burke JS, Rosenberg SA, Kaplan HS: Occurrence of non-Hodgkin's lymphoma after therapy for Hodgkin's disease. N Engl J Med 300:452-458, 1979.

204. Ziegler JL, Drew WL, Miner RC, et al: Outbreak of Burkitt's-like lymphoma in homo­sexual men. Lancet 2:631-633,1982.

205. Ziegler JL, Beckstead JA, Volberding PA, et al: Non-Hodgkin's lymphoma in 90 homo­sexual men. Relation to generalized lymphadenopathy and the acquired immunodeficiency syndrome. N Eng1 J Med 311:565-570, 1984.

206. Mintzer DM, Andreeff M, Filippa DA, Jhanwar SC, Chaganti RSK, Koziner B: Progres­sion of nodular poorly differentiated lymphocytic lymphoma to Burkitt's-like lymphoma. Blood 64:415-421,1984.

207. Weiss LM, Warnke RA: Follicular lymphoma with blastic conversion: A report of two cases with confirmation by immunoperoxidase studies on bone narrow sections. Am J Clin PathoI83:681-686,1985.

208. Garcia CF, Weiss LM, Warnke RA: Small noncleaved cell lymphoma: An immunopheno­typic study of 18 cases and comparison with large cell lymphoma. Hum Pathol 17:454-461,1986.

209. Mangan KF, Rauch AE, Bishop M, et al: Acute lymphoblastic leukemia of Burkitt's type (L-3 ALL) lacking surface immunoglobulin and the 8; 14 translocation. Am J Clin Pathol

146

83:121-126, 1985. 210. Gluck WL, Bigner SH, Borowitz Ml, Benchman WD: Acute lymphoblastic leukemia of

Burkitt's type (L3-ALL) with 8;22 and 14;18 translocations and absent surface immuno­globulins. Am 1 Clin Pathol 85:636-640, 1986.

211. Cooper MD: Pre-B cells: Normal and abnormal development. 1 Clin Immunol 1:81-89, 1981.

212. Mann RB, laffe ES, Braylan RC, Nanba K, Frank MM, Ziegler lL, Berard CW: Non­endemic Burkitt's lymphoma: A B-cell tumor related to germinal centers. N Engl 1 Med 295:685, 1976.

213. Henle W, Henle G, Lennette ET: The Epstein-Barr virus. Sci Am 241:48-60, 1979. 214. Douglass EC, Magrath IT, Lee EC, Whang-Peng 1: Cytogenetic studies in non-African

Burkitt lymphoma. Blood 55: 148- 155, 1980. 215. Krontiris TG: The emerging genetics of human cancer. N Engl 1 Med 309:404, 1983. 216. Whittle, et al: T-cell control of Epstein-Barr virus-infected B-cells is lost during P

falciparum malaria. Nature 312:449, 1984. 217. Popovic M, et al: Detection, isolation, and continuous production of cytopathic retro­

viruses (HTLV-rrr) from patients with AIDS and pre-AIDS. Science 224:497, 1984. 218. Salmon S, Seligman M: B-cell neoplasia in man. Lancet 2: 1230, 1974. 219. Wood GS, et al: Human dendritic cells and macrophages: In situ immunophenotypic

definition of subsets that exhibit specific morphologic and microenvironmental char­acteristics. Am 1 Pat hoi 119:73, 1985.

220. Lane, et al: Stage-specific transforming genes of human and mouse B- and T-lymphocyte neoplasms. Cell 28:873, 1982.

221. Payne CM, Nagle RB, Borduin VF, Kim A: An ultrastructural evaluation of the cell organelle specificity of the uranaffin reaction in two human endocrine neoplasms. 1 Submicrose Cytol 15:833-841, 1983.

222. Payne CM, Nagle RB, Borduin V: An ultrastructural cytochemical stain specific for neuroendocrine neoplasms. Lab Invest 51:350-365, 1984.

223. Herrera GA, Crofts lL, Roberts CR: Neurosecretory granule-like structures in lym­phomas. Hum Pathol 11 :449-457, 1980.

224. Pileri S, Martinelli G, Rivano MT, Biagini G, Govoni E: Neurosecretory-like granules in non-Hodgkin's malignant lymphomas. Human Pathol 15:588-589, 1984.

225. Waldron lA, Leech lH, Glick AD, Flexner 1M, Collins RD: Malignant lymphoma of peripheral T-lymphocyte origin. Cancer 40: 1604-1617, 1977.

226. Brisbane lU, Berman LD, Neiman RS: Peripheral T-cell lymphoma: A clincopathologic study of nine cases. Am 1 Clin Pathol 79:285-293, 1983.

227. Greer lP, York lC, Cousar JB, Mitchell RT, Flexner 1M, Collins RD, Stein RS: Peripheral T-cell lymphoma: A clinicopathologic study of 42 cases. 1 Clin Oncol 2:788-798, 1984.

228. Weis lW, Winter MW, Phyliky RL, Banks PM: Peripheral T-cell lymphomas: Histologic, immunohistologic, and clinical characterization. Mayo Clin Proc 6:411-426, 1986.

229. Van der Valk P, Willemze R, Meijer CJ: Peripheral T-cell lymphomas: A clinicopatho­logical and immunological study of 10 cases. Histopathol 10: 235-249, 1986.

230. Wright DH: T-cell lymphomas. Histopathol 10:321-326, 1986. 231. Haynes BF, Hensley LL, legasothy BU: Phenotypic characterization of skin infiltrating T

cells in cutaneous T-cell lymphomas: Comparison with benign cutaneous infiltrates. Blood 60:463-473, 1982.

232. Todes-Taylor N, Turner R, Wood GS, Startle PT, Morhenn VB: T cell subpopulations in alopecia areola. 1 Am Acad Dermatol 11:216-223, 1984.

233. Luke RJ, Collins RD: New approaches to the classification of the lymphomata. Br J Cancer 31(suppl 2):1-28, 1975.

234. Lennert K: Malignant lymphomas other than Hodgkin's Disease: Histology, cytology, ultrastructure, immunology. New York, Springer-Verlag 1978, pp 196-209.

235. Watanabe S, Nakajima T, Shimosato Y, Shimoyama M, Minato K: T-cell malignancies:

147

Subclassification and interrelationship. Jpn J Clin Oncol 9(Suppl):423-442 , 1979. 236. Collins RD, Waldron JA, Glock AD: Results of multiparameter studies of T-cell lym­

phoid neoplasms. Am J Clin Pathol 72(Suppl):699- 707, 1979. 237. Muller-Hermelink HK, Steinmann G, Stein H, Lennert K: Malignant lymphoma of

plasmatoid T-cells. Morphologic and immunologic studies characterizing a special type of T-cell. Am J Surg Pathol 7:R49-R62, 19R3.

23R. Prastohofer E F , Grizzie WE, Prchal JT, Grossi CE: Plasmacytoid T-cell lymphoma associated with chronic myeloproliferative disorder. Am J Clin Pat hoi 9:380-387, 1985.

239. Kadin ME, Kamoun M, Lamberg J: Erythrophagocytic T A lymphoma. N Engl J Med 304:64R-653, 19R1.

240. Shimoyama M, Mirato K, Saito H, Takenaka T , Watanabe S, Nagatani T, Naruto M: Immunoblastic lymphadenopathy (TBL)-likc T-cell lymphoma. Jpn J Clin Oncol 9(Suppl): 347-356, 1979.

241. Jaffe ES: Pathologic and clinical spectrum of post-thymic T-cell malignancies. Cancer Invest 2:413-426, 19R4.

242. Lippman SM, Grogan TM, Spier CM, Koopman CF, Gall EP, Schimm OS, Durie BG: Lethal midline granuloma with a novel T-cell phenotype as found in peripheral T-cell lymphoma. Cancer, 59:936- 939, 19R7.

243. Jaffe ES, Blattner EA , Blayncw OW, et al: The pathologic spectrum of adult T-cell leukemia/lymphoma in the United States. Human T-cell leukemia!lymphoma vlrus­associated lymphoid malignancies. Am J Surg Pat hoi R:263-275, 1984.

244. Uchiyama T, Yodoi J, Sagawa K, Tatasuki K, Uchino H: Adult T-cell leukemia: Clinical and hematologic features of 16 cases. Blood 50:4RI-492, 1977.

245. Blattner W A, Kalyanaraman VS, Robert-Guroff M, et al: The human type-C retrovirus, HTLV, in blacks from the Caribbean region and relationship to adult T-cell leukemia! lymphoma. Int J of Cancer 30:2457-264, 19R2.

246. Cossman J , Jaffe ES, Risher Rl: Diversity of immunologic phenotypes of T-cell lym­phoma. Am J Surg Pathol 6:71, 1982.

247. Horning SJ , Weiss LM, Crabtress GS, Warnke RA: The clinical and phenotypic hetero­geneity of T cell lymphoma. Blood 67 : 157R-15R2, 1986.

248. Kalyanaraman VS , Sarngodharan MG, Robert-Guroff M, Miyoshi I, Blayngy 0, Golde D, Gallo RC: A new subtype of human T-cell leukemia virus (HTLV-II) associated with a T-cell variant of hairy cell leukemia. Science 218:571, 1982.

249. Weiss LM, Wood GS, Trela M, Warnke RA, Sklar J: Clonal T-cell populations in lymphomatoid papulosis. Evidence of a lymphoproliferative origin for a clinically bcnign disease. N Engl J Med 315:475-479, 19R6.

250. Colby TV, Carrington CB: Pulmonary lymphomas: Currcnt concepts. Hum Pathol 14: 884-R87 , 19R3.

251. Kadin M: Common activated helper - T cell origin for lymphomatoid populosis, mycosis fungoides, and some types of Hodgkin's disease. Lancet 2:864-865, 1985.

252. Weiss LM, Strickler JG, Dorfman RF, Horning SJ, Warnke RA, Sklar J: Clonal T-cell populations in angioimmunoblastic lymphadenopathy and angioimmunoblastic lympha­denopathy-like lymphoma. Am J Pathol 122:392- 397, 1986.

253. Hsu SM, Jaffe ES: Leu Ml and peanut agglutinin stain the neoplastic cclls of Hodgkins diseasc. Am J Clin Pathol 82:29-32, 1984.

254. Pinkus GS, Thomas P, Said JW, Leu Ml: A marker for Reed-Sternberg cells in Hodgkins disease. An immunoperoxidase study of paraffin-embedded tissues. Am J Pathol 119: 244- 252, 1985.

255 . Foon KA, Schroof RW, Gale RP : Surface markers on leukemia and lymphoma cells: Recent advances. Blood 60:1-19,1982 .

256. Ruiz-Arquelles G, Katzman JA, Greipp PR, Gonchoroff NJ, Garton JP, Kyle RA: Multiple myeloma: Circulating lymphocytes that express plasma cell antigens. Blood 64:352, 1984.

257. Ca •• garis-Cappio F, Bergui L, Tesio L, et al: Identification of malignant plasma cell

148

precursors in the bone marrow of multiple myeloma. J Clin Invest 76: 1243-1251, 1985. 258. Durie BGM, Vela E, Baum V, et al: Establishment of two new myeloma cell lines

from bilateral pleural effusions: Evidence for sequential in vivo clonal change. Blood 66:548-555, 1985.

259. Grogan TM, Durie BGM, Lomen C, et al: Delineation of a novel pre-B cell component in plasma cell myeloma: Immunochemical, immunophenotypic, cytologic, cell culture and kinetic features. Blood 70:932-942, 1987.

260. Ralfkiacr E, Wantzin GL, Mason DY, Hou-Jensen K, Stein H, Thomsen K: Phenotypic characterization of lymphocyte subsets in mycosis fungoides. Comparison with large plaque parapsoriasis and benign chronic dermatoses. Am J Clin Pathol 84:610-619, 1985.

261. Wood GS, Abel EA, Hoppe RT, Warnke RA: Leu 8 and Leu 9 antigen phenotypes: Immunologic criteria for the distinction of mycosis fungoides from cutaneous inflamma­tion. J Amer Acad Dermatol 14:1006-1013, 1986.

262. Rowden G, Lewis MG: Langerhans cells: Involvement in the pathogene isis of mycosis fungoides. Br J Dermatol 95:665-672, 1976.

263. Patterson JAK, Edelson R: Interaction of T cells with epidermis. Br J Dermatol 107: 117-122, 1982.

264. Payne CM, Grogan TM, Lynch PJ: An ultrastructural, morphometric and immunohisto­chemical analysis of cutaneous lymphomas and benign lymphocytic infiltrates of skin: Useful criteria for diagnoses. Arch Derm 122: 1139-1154, 1986.

265. Wieczorek R, Suhrland M, Ramsay D, Reed ML, Knowles DM: Leu M1 expression in advanced (tumor) stage mycosis fungoides. Am J C1in Pathol 86:25-32, 1986.

266. Kim H, et a1: Composite lymphoma. Cancer 40:949-976, 1977. 267. Jaffe ES, et al: Functional markers: A new perspective on malignant lymphomas. Cancer

Treat Rep 61:953-962, 1977. 268. Hoppe RT: Histologic variation in non-Hodgkins lymphomas: Commentary. Cancer Treat

Rep 65:935-939, 1981. 269. Fisher RI: Natural history of malignant lymphomas with divergent histologies at staging

evaluation. Cancer 47:2022-2025, 1981. 270. Anderson NG, Anderson NL: The human protein index. JAMA 246:2620-21, 1981.

3. Detection of central nervous system involvement in patients with leukemia or non-Hodgkin's lymphoma by immunological marker analysis of cerebrospinal fluid cells

Herbert Hooijkaas, Henk J. Adriaansen, and Jacques J.M. van Dongen

Introduction

Central nervous system (CNS) involvement is observed at diagnosis in up to 20% of all patients with acute lymphoblastic leukemia (ALL) , acute myeloid leukemia (AML), or non-Hodgkin 's lymphoma (NHL) [1-12]. Follow-up of these patients revealed that without prophylactic CNS therapy , up to 70% of the ALL, up to 25% of the AML, and, depending on the histological sub­type, up to 40% of the NHL patients will develop the serious complication of CNS involvement [1, 5, 7,12-26]. This increase in CNS involvement has been explained by the application of improved systemic therapy leading to longer survival of patients, which allows the growth of tumor cells present in relatively therapy-resistent sites such as the CNS. In addition , primary NHL can be located in the CNS, and they constitute about 1 % of all brain tumors and about 2% of all NHL [27-30]. These primary NHL located in the CNS are increasingly diagnosed, especially in immunodeficient individuals , prob­ably as a result of the increased use of radiotherapy and immunosuppressive drugs; furthermore, improved diagnostic procedures might be responsible for this increase [31-36].

Prevention and - if not possible - treatment of preferably early diag­nosed CNS involvement are important goals of therapy. Prophylactic treat­ments, such as cranial irradiation and intrathecal chemotherapy, have greatly reduced the occurrence of CNS involvement in ALL, AML, and some types of NHL; still, in 2%-15% of patients , primary meningeal relapse occurs [1, 6, 8-10, 13, 15, 19, 37-44]. This CNS involvement is especially a major concern in childhood ALL, where the majority of pa­tients now have become long-term survivors [45]. Although the detection of malignant cells in the CSF is often the first sign of CNS involvement, in most cases the disease process has started in an earlier phase by leukemic infiltra­tion of the superficial leptomeninges. Such a leukemic infiltration will even­tually lead to destruction of the trabeculae with release of cells in the CSF [46] . These infiltration processes may differ depending on the type of malig­nancy [47]. It has been pointed out that the malignant cells appear less often in the CSF when they are locally seeded and almost never when the tumor is

Bellllel/, J.M. alld 1"00". K.A ., (eds. ), lmmu"ologic Approaches 10 the Classificatio" a"d Mallagemellt of Lymphomas alld Leukemias_ © 1988 Kluwer Academic Publishers. tSBN978- t-4612-8965-4. All rights reser>ed.

150

limited to the brain and the pial surface has not been breached [48] . Thus, failure to find malignant cells in the CSF does not exclude the possibility of CNS involvement when clinically suspected. Therefore, multiple CSF sampling is preferred , especially in cases with little leptomeningeal involve­ment [49, 50]. In cases of NHL in the CNS, it has been recommended to perform at least three consecutive examinations; then a positive cytology is obtained in 60% - 70% of patients [12, 51]. Cytological examination of the CSF at regular intervals has been integrated in many protocols for the management of patients with leukemia or NHL. It is obvious that reliable methods for the detection of low numbers of malignant cells in the CSF will allow an earlier diagnosis of CNS involvement, which might be beneficial for the patient, as therapy can be started earlier , before massive infiltration and destruction of the meninges occur.

Methods for the detection of malignant cells in cerebrospinal fluid

CSF samples often contain few (tumor) cells, and this determines to a large extent which methods are suitable for the detection of malignant cells in the CSF [52]. The standard methods for the (early) detection of CNS leukemia or NHL are cell counting and cytomorphological examination of the CSF [50, 52]. This detection can be improved by immunological marker analysis of the CSF cells with monoclonal antibodies or conventional antisera, speci­fic for cell surface membrane antigens [52-73] or for the nuclear enzyme terminal deoxynucleotidyl transferase (TdT) [65, 72, 74-79]. In addition, cytogenetic analysis of CSF cells [80-84], electron microscopy [56, 85, 86], flow cytometric analysis of DNA and RNA content of cells in the CSF [87, 88], or measurements of CSF ~-2-microglobulin (~2m) [89, 90] have been proposed for the (early) detection of meningeal involvement in acute leuke­mia or NHL. Recently, gene rearrangement studies have been added as a diagnostic tool for proving monoclonality of CNS lymphomas [91]. The application of the aforementioned techniques, and especially the immuno­logical marker analysis , in the diagnosis of CNS involvement will be re­viewed and discussed.

Cytomorphology

For the morphological analysis of CSF cells, cytocentrifugation of the CSF is widely performed [92-98], although some prefer membrane filter prepara­tion [99-101] or gravity sedimentation of CSF on slides [102, 103]. Which method is best is still controversial [104-107]. Cell loss, recovery rate, and preservation of cell morphology are matters of debate, although it has been stated by Bigner and Johnston in their extensive review [52] that 'excellent cellular recovery and preservation can be obtained using any of these techniques, if careful attention to detail in cytopreservation is observed.' However, diagnostic problems can arise when the CSF cell count is low,

151

while cytomorphology is suggestive but not conclusive for the presence of malignant cells. Cases of both false-positive and false-negative findings have been reported [48, 53, 108, 109]. For example, viral infections or chemo­therapy can change normal mononuclear cells into atypical blast-like cells [93, 110]. On the other hand, lymphoblasts of ALL may have the appear­ance of relatively well differentiated lymphocytes [76]. Therefore, the sensi­tivity and specificity of cytological techniques for the detection of malignant hemopoietic cells is insufficient and may be enhanced by use of additional techniques.

Immunological marker analysis

Immunological marker analysis of normal and malignant hemopoietic cells has led to improved insight into leukocyte differentiation [111-115]. The expression of a particular set of markers designates a hemopoietic cell to a particular cell lineage and differentiation stage. It is now widely accepted that leukemia and NHL can be regarded as malignant counterparts of cells in the various hemopoietic differentiation stages. Hemopoietic differentia­tion schemes have been proposed which provide an indication of the cell lineage and differentiation stage of the various leukemias and NHL based on their immunological phenotype [111, 114, 115]. Some nonmalignant cells with a particular phenotype are only found in a number of specific ana­tomical sites. The presence of such cells in locations where they are not found normally, such as the CSF, provides a strong indication for the presence of malignant cells. Taken together, immunological marker analysis is a powerful approach to determine the nature of cells in the CSF if the appropriate markers are used, as will be discussed in subsequent sections.

Cytogenetic analysis

Cytogenetic analysis plays an important role in the characterization of hema­tological disorders [116-118]. In contrast to immunological markers, which are differentiation markers that appear also on normal cells, chromosomal aberrations are unique and specific features of malignant hemopoietic cells. In many cases nonrandom cytogenetic changes occur [119-121]. In a patient with meningeal involvement of leukemia or NHL, malignant cells may be found in the CSF by cytogenetic analysis if sufficient metaphases can be obtained. Since the cell count is often low in routine CSF examinations, cytogenetic analysis is less applicable in such situations. However, in ap­propriate cases, cytogenetic analysis can establish the presence of a neo­plastic clone [80-84].

Electron microscopy

Elaborate methods are available when ultrastructural studies are needed for the classification of CSF cells [86]. However, most studies have been per-

152

formed on biopsies [56], such as described in a case of a primary intracranial histiocytic lymphoma with Langerhans ' granules [85]. In practice, electron microscopy is too laborious a technique to be used for the routine analysis of CSF cells.

fJ-2-microglobulin and other biochemical parameters

It has been shown that the levels of ~-2-microglobulin (~2m), a cell mem­brane component of HLA class I antigens, are elevated in the serum of pa­tients with myeloproliferative as well as Iymphoproliferative disorders [122, 123]. This elevation has been assumed to reflect rapid cellular turn-over, resulting in excessive shedding of ~2m in the circulation. Similar results have been reported by Mavligit et al. for the CSF [89]. They and others [90] therefore suggested that ~2m levels in the CSF would be useful in early diagnosis of CNS involvement and in the monitoring of intrathecal therapy in patients with acute leukemia or NHL. However, this has not been con­firmed in a series of other papers [124-127], since the postirradiation syn­drome , intrathecal chemotherapy, as well as viral infections are accompanied with significantly higher ~2m levels than an incipient CNS relapse. This indi­cates that ~2m determination is not an appropriate test for diagnosing meningeal involvement in leukemia and NHL.

Other biochemical markers like ~-glucuronidase and carcinoembryonic antigen also failed to be diagnostic for the detection of CNS involvement of leukemia and NHL [128]. In addition, there appears to be no value in measuring protein and glucose levels when monitoring the CSF in acute leukemia for evidence of leukemic involvement of the CNS [94]. Some studies suggest that measurement of oligoclonal IgG bands in the CSF of patients with Burkitt's lymphoma might be useful for determining the pres­ence of the tumor in the CNS [129, 130].

Flow cytometric measurement of DNA and RNA content

Flow cytometric methods can be used to study the DNA and RNA content of leukemia and NHL cells. In acute leukemias in children, aneuploid DNA content can be identified in up to 40% of cases [131, 132]. In those cases, the abnormal DNA content serves as a marker for the presence of leukemic cells in BM and peripheral blood (PB). Since increased RNA content is characteristic for AML [133], Redner et al. [87] studied the CSF from leukemic patients by two parameter DNA-RNA flow cytometry. The small number of cells usually present in CSF samples was rarely a problem for this type of analysis. In 6 out of 15 patients with CNS relapse, the CSF contained cells with abnormal DNA content. In two additional cases, leukemic cells were identified by an abnormally high RNA content only. In 5 out of 6 other patients with a uniform or reactive cytomorphology, leukemic cells could be detected by DNA-RNA flow cytometry and, in addition, Iymphoblasts could be differentiated from non lymphoblastic cell types. It is obvious that diploid

153

leukemic cells cannot be detected by the measurement of DNA content alone.

Immunoglobulin gene and T-cell receptor gene rearrangements

Detection of Ig gene or T-cell receptor gene rearrangements is an accurate and highly specific method to identify clonal proliferations in hemopoietic cell populations [134, 135]. It does not strictly prove the delineation of these cells but does prove their monoclonal character. However, for current methods, 5-10 million cells per sample are needed, which does not make it a suitable technique for CSF analysis as yet. However, in a suspected cell population lacking surface membrane or cytoplasmic immunoglobulins and derived from a brain biopsy, a monoclonal Ig gene rearrangement was demonstrated [91]. The patient involved was a cardiac transplant recipient who apparently had developed a B-NHL in the brain [91].

Immunological marker analysis of cells in the cerebrospinal fluid

The immunological phenotypes of the acute leukemias (ALL and AML) correspond with immature differentiation stages, whereas the chronic leu­kemias such as chronic lymphocytic leukemia (CLL), hairy cell leukemia (HCL), and chronic myeloid leukemia (CML) are the malignant counter­parts of cells in more mature differentiation stages. Generally, NHL have also a more mature immunological phenotype . In ALL and AML, as well as in some types of NHL, CNS involvement is a matter of concern since, in spite of CNS prophylaxis, in up to 15% of cases CNS relapses still occur [1 , 6, 8--10, 13, 15, 19, 37-44]. CNS involvement in chronic leukemias is extremely rare [20, 52, 53, 98, 110, 136]. Only a few cases have been described such as in CLL [137-139] and HCL [140].

ALL can be divided into at least five different immunological phenotypes: null ALL, common ALL, pre-B-ALL, B-ALL, and T-ALL [111, 114 , 115] . AML, however, often consists of several subpopulations, which hampers its immunological characterization. Only double stainings for the various myeloid markers allow a more precise immunological marker analysis and hence a better classification of these leukemias [115].

Immunological characterization of NHL can reveal whether B-cells or T­cells are involved or can prove c10nality of the tumor by the establishment of the kappa-lambda ratio in the case of a B-cell malignancy. Cell surface antigens diagnostic of clonal proliferations of T-cells have not yet been recognized. The helper and suppressor phenotypes are not suitable to prove c1onality. In addition, the expression of an aberrant cellular phenotype not found among normal or reactive T-cells is helpful in diagnosing T-cell malignancy. For instance, TdT+ lymphoblastic lymphomas can be identified by the TdT-immunofluorescence (IF) assay [74-79] .

Using (monoclonal) antibodies, immunological marker analyses have

154

been performed on cells in the CSF from patients with leukemia or NHL at risk, or suspected of, meningeal involvement. One of the major obstacles in characterizing CSF cell populations by use of immunological markers is the low number of cells (0-5 cells/!-!!) usually present [52]. Nevertheless, it has been shown that in such CSF samples malignant cells can be found [141]. Several techniques are used: labeling of cells in suspension followed by cyto­centrifugation or preparation of cytocentrifuge slides followed by labeling with the appropriate antisera. In order to visualize reactivity of the anti­bodies, fluorochromes (fluorescein or rhodamine) or enzymes (peroxidase or alkaline phosphatase) are generally used as labels [61, 65, 70, 73, 76]. Re­cently, an immunorosetting technique using antibody labeled latex spheres has been described [68]. The simultaneous analysis of morphology and markers plus a permanent record of slides is an advantage of the enzyme or latex sphere techniques. Cell morphology, however, can also be studied in immunofluorescence studies using microscopes equipped with phase-contrast facilities [115].

Immunophenotyping of CSF cells has also been successfully applied in disease states other than leukemia or NHL, such as other malignancies [62, 142], infections [143-146], or neurological diseases [147 -150] . Normal values for marker expression were obtained, indicating that in normal situations 62% (range 53-80) of T-cells (figure 1), some B-cells, and 10%-30% of monocytes/macrophages (figure 2) can be present [68, 110, 151].

Kappa-lambda ratio

Since B-cell malignancies are clonal expansions of one B-cell, they will express only one type of Ig light chain: kappa or lambda [152]. Therefore , Ig light chain expression can be used as a marker of clonality, and the kappa­lambda ratio can give an impression about the number of malignant cells. This type of analysis may be useful for the detection of CNS involvement in surface membrane Ig or cytoplasmic Ig positive B-cell malignancies such as B-ALL and B-NHL, but not for other ALL, since only B-ALL express surface membrane Ig. Both primary [56, 57, 69, 71 , 73] and disseminated secondary CNS lymphomas [54, 55, 63, 66, 73] have been studied.

Despite multiple diagnostic procedures , in several cases the diagnosis of CNS lymphoma could not be made until immunological marker studies, including kappa-lambda ratio determination, were performed [63, 73]. Most lymphomas found in the CNS are of B-cell type but , using immunological marker analysis primary histiocytic lymphoma [85] or T-cell lymphoma [59, 153] have also been reported. In general, the kappa-lambda ratio determina­tion of CSF cells from patients suspected of a B-NHL in the CNS is useful in differentiating benign from malignant proliferations [60, 64, 73]. In such cases it might be effective to use a panel of additional B-cell and T-cell markers for the precise phenotyping of the cell population under study.

155

Figure I. CSF cells from an ALL patient in remission. Phase-contrast morphology (left) . Surface membrane immunofluorescence staining of CD3 (Leu-4) positive T-cells (right).

Surface membrane markers for acute lymphoblastic leukemia

In ALL, surface membrane marker analysis has been widely performed to detect malignant cells. Antibodies that recognize the common ALL antigen (CALLA or CDlO), which was initially regarded as leukemia specific [154], are frequently used. However, CALLA + cells have been demonstrated in normal BM and PB as well as in fetal hemopoietic tissues [155 , 156]. This marker may still be used for the detection of malignant cells in the CSF, although it will be only valuable in CALLA + ALL, which constitute up to 60%-80% of all ALL. In addition , loss or diminution of CALLA expression, as well as its acquisition, have been described [157]. Nevertheless , Veerman et aI., using an immunoperoxidase technique, investigated the presence or absence of CALLA + cells in 208 CSF samples from 62 children with ALL or other diseases [70]. It was shown that background levels of up to 3% CALLA + cells could be observed in conditions other than CALLA + leuke­mia (presence of solid tumors, convulsions, viral meningitis). Therefore, more than 5% CALLA + cells were thought to be indicative for leukemic involvement. All 21 CSF samples containing more than 5% CALLA + cells were derived from patients with an initial diagnosis of common ALL; 8 of these samples were considered to be normal and 3 uncertain (blood con-

156

Figure 2. CSF cells from an ALL patient in remiSSIOn. Phase-contrast morphology (left). Surface membrane immunofluorescence staining of CDI4 (My4) positive monocytes/macro­phages (right).

tamination or treatment for meningeal relapse) by standard cytomorpho­logical criteria. Six of the 8 cytologically normal samples were from patients who had clear meningeal involvement at diagnosis or developed a relapse later on. In the remaining 10 samples from 10 patients, a meningeal relapse was diagnosed based on cytological criteria. In one other patient, a morpho­logical diagnosis of meningeal relapse was not confirmed by the presence of cells positive for CALLA. This patient was still disease-free two years later without cytostatic treatment, suggesting that the morphological blasts were reactive lymphocytes. Veerman et a\. conclude that detection of surface markers with immunoperoxidase methods is possible even in normocellular CSF samples, as was also shown by Moir et a\. [61].

A similar study has been performed by Homans et a\. [68]. They used monoclonal antibody-coated spheres in a rosetting technique, which allowed the simultaneous detection of cell surface markers and cell morphology. In 199 CSF samples from 34 ALL patients and 14 control subjects, cells were studied for the expression of CD10 (15), CD2 (OKT-ll), and HLA-DR. In patients without leukemic meningitis, 69% of the CSF lymphocytes were mature CD2+ (OKT-ll +) T-cells, 8% were "B-cells" (HLA-DR +), and 3% of the CSF lymphoid cells expressed CDlO (15). Similar results were found in the control subjects. In 28 CSF samples from nine children with varying numbers of CSF Iymphoblasts (0% -100%) greater proportions of CALLA

157

and HLA-OR positive CSF cells (24% -96%) were detected. In five of these patients CNS leukemia was present or developed later on. However, three patients with small numbers of lymphoblasts and less than 5% CALLA + cells subsequently had normal CSF examinations. One patient died from pro­gressive BM involvement without further investigation of CNS malignancy.

Both groups claim the usefulness of their method for CALLA determina­tion, especially since cell morphology is also well preserved. However, since only a part of all ALL are CALLA +, we consider this method less suitable for routine evaluation of CSF samples from ALL patients. In our opinion, the TdT-IF assay combined with phase contrast morphology is superior (details to follow). When the cell count is high , it is , of course, informative to further analyze the malignant population with monoclonal antibodies for cell surface antigens, using the aforementioned methods or conventional IF microscopy [65] . The latter method for marker analysis can be easily com­bined with the TdT-IF assay for double staining of the suspected cell popula­tion in order to characterize the ALL population more precisely [76, 115, 158, 159, 160].

Surface membrane markers for acute myeloid leukemia

The presently available myeloid markers may be used for the detection of CNS involvement in AML. Myeloid-monocytic markers that are absent on normal or reactive myeloid cells, monocytes or macrophages but are present on AML cells are not yet available. However, using double IF stainings for the various myeloid markers, it might be possible to discriminate between reactive and leukemic cells. For example, cells positive for both C014 (My4) and C01S (VIM-OS) or both C01S (VIM-OS) and HLA-OR are generally only found in the BM, as they represent myeloid progenitor cells. The presence of these cells in the CSF of AML patients would therefore be indicative for CNS involvement. Indeed, such cells were found in CSF samples with low cell count and uncertain morphology from an AML patient with CNS leukemia (unpublished observations). More data, especially con­trol values, are needed to confirm these results and prove the value of double staining with these marker combinations for the establishment of CNS involvement in AML.

Terminal deoxynucleotidyl transferase

The enzyme TdT, a ONA polymerase lacking template requirements, is present on the nuclear membrane of immature lymphoid cells of both B-cell and T-cell lineages as well as their malignant counterparts (all ALL except the mature B-ALL) [160]. Occasionally, TdT+ AML have been observed [161]. Normally, a substantial proportion of the BM cells (up to 11%) and thymocytes can be TdT+, but low percentages of TdT+ cells have been detected at other sites as well , such as PB and lymph nodes « 0.4%) [160].

158

For the detection of small numbers of TdT+ cells, the biochemical TdT assay is rather insensitive, and therefore the TdT immunofluorescence (TdT­IF) assay is preferred [159, 162]. The TdT-IF assay has been employed to improve the accuracy of the diagnosis of extramedullary ALL, e.g., by the analysis of CSF. This will be reviewed in the next section.

Terminal deoxynucleotidyl transferase immunofluorescence assay on cells in the cerebrospinal fluid

The use of the TdT-IF assay for the detection of leukemic cells in the CSF was introduced by Bradstock et al. [74] and Peiper et al. [75]. Using cytocentrifuge slides of CSF cells from patients with ALL, they could detect TdT+ cells in CSF containing atypical cells or morphologically proven blast cells. In CSF from patients with infectious meningitis and from patients with TdT- malignancies, no TdT+ cells were detected. Nuclear TdT staining of cells in CSF can thus reveal small percentages of leukemic cells, allowing early diagnosis of CNS involvement [74-76]. The sensitivity of the TdT-IF assay in detecting the presence of leukemic blasts in CSF samples with a low cell count and/or inconclusive morphology was further tested by Casper et al. [77]. Agreement between TdT reactivity and cytomorphology was found in 55 of 60 samples (92%) from 28 children with TdT+ ALL. Two consecu­tive samples from one patient contained TdT+ cells but appeared to be mor­phologically normal. A subsequent sample showed definitive lymphoblasts and TdT+ cells. In three patients with positive morphology, but no evidence of TdT reactivity, TdT+ cells were found in subsequent CSF samples. In their series, Casper et al. considered the presence of more than 10% TdT+ cells in the CSF as strong evidence for CNS leukemia, even when the CSF cell count was less than 1O/!!I. In cases with less than 10% TdT+ cells, as well as a normal cell count, it was advised to repeat the lumbar puncture within three to four weeks. If TdT+ cells were found to increase in number, a diagnosis of CNS leukemia was made.

To determine whether the combination of cytomorphology and nuclear TdT analysis would be the most effective method for detecting or excluding minimal CNS leukemic infiltrates in patients with a TdT+ malignancy, a large longitudinal study was begun. In this study both the TdT-IF assay and conventional cytomorphology were employed to determine the presence or absence of leukemic cells in CSF samples from children with TdT+ ALL or TdT+ NHL at diagnosis as well as during follow-up. Preliminary data have been published [65, 78, 79] and will be reviewed with some of the unpub­lished results.

Two or more consecutive samples (n = 666) of 2.5 ml were obtained from 70 children with a TdT+ ALL or TdT+ NHL during maintenance therapy and/or after withdrawal of therapy. Cells were counted, and two cytocentri­fuge preparations were made (0.5 ml each) and stained with May-Grunwald-

159

Giemsa. Cytomorphology was considered posItive when more than 5% leukemic blasts were counted. CSF that contained less than 5% leukemic cells or contained atypical lymphocytes or lymphoblastoid cells was regarded as suspect.

For the TdT-IF assay CSF samples (1-4 ml) were centrifuged (300 g; 5 min). The cell pellets were resuspended to 100 [il and two cytocentrifuge preparations were made per CSF sample, air dried, fixed in methanol (30 min; 4°C), and washed in phosphate buffered saline (PBS). One preparation was incubated with 15 [il of optimally titrated rabbit anti-TdT antiserum, while the second one, as a control, was incubated with 15 [il of normal rabbit serum (moist chamber; 30 minutes; room temperature). After wash­ing in PBS buffer, the preparations were incubated with 15 [il of a fluore­scein conjugated goat anti-rabbit-Ig antiserum and washed again [158]. The anti-TdT and the second step antibody were obtained from Bethesda Re­search Laboratories, Bethesda, MD or from Supertechs, Bethesda, MD. Afterwards, the preparations were mounted in glycerol PBS (9:1) containing 1 mg p-phenylenediamine per ml (BDH Chemicals, Poole, UK) to prevent fading of fluorochrome [163], covered with a coverslip, and sealed with paraffin wax with ceresin (BDH Chemicals) [115].

Cells expressing nuclear TdT were enumerated using Zeiss fluorescence microscopes equipped with phase-contrast facilities. If possible, 100-2000 cells were screened. In 18% of cases, low cell content of the CSF allowed screening of only 10-100 cells. The percentage of TdT+ cells was calculated as the fraction of the number of nucleated cells.

For the evaluation of the nuclear staining of TdT-positive cells, phase­contrast morphology is a prerequisite. Figure 3 shows a typical example of TdT+ cells in the CSF.

The correlation between cytomorphology and the TdT-IF assay in 666 samples from 70 patients with a TdT+ ALL or NHL is summarized in Table 1. In 82.5% of the samples, both cytomorphology and the TdT-IF assay were negative; in 4.7% of the samples, they were both positive. The cyto­morphology was suspect in 44 samples (6.7%), but in 33 of these samples (5%) no TdT+ cells were observed. In the latter group, no meningeal relapse has occurred during a median follow-up period of 66 weeks (range: 12-156 weeks). In the other 11 cytomorphological suspect samples (1.7%), TdT+ cells were found. In 40 of the CSF samples (6.1%), no blasts were seen, although TdT+ cells were present. Figure 3 shows the presence of TdT+ cells in such a cytomorphological negative CSF sample from an ALL patient at diagnosis. The follow-up of 17 evaluable patients (23 samples) with TdT+ cells in the CSF but with a negative or suspect CSF cytomorphol­ogy is given in Table 2. Although some patients had several episodes of CNS involvement, only data of the first episode are shown. In eight patients more than 4% TdT+ cells were found in the CSF, and they all subsequently de­veloped CNS leukemia on classical cytomorphological and clinical grounds (follow-up: 0-86 weeks). In the CSF of two NHL patients, 2% TdT+ cells

160

Figure 3. CSF cells from an ALL patient at diagnosis. Phase-contrast morphology (left). Nuclear immunofluorescence staining of TdT positive cells (right).

were detected. One of these two patients developed a CNS leukemia (after 12 weeks); the other one did not (follow-up: 148 weeks). None of the remaining seven patients developed a CNS leukemia, although 0.2%-2.6% TdT+ cells were observed (follow-up: 50-106 weeks). Erythrocytes were seen in the CSF from three of these seven patients. Finally, we could not detect TdT+ cells in 69 CSF samples from 55 patients who did not suffer from ALL but suffered from other malignancies or infections.

In contrast with the results of Casper et aI., who found in three patients a

Table 1. Correlation between cytomorphology and the TdT-IF assay in 666 CSF samples from 70 patients with a TdT positive ALL (n = 67) or NHL (n = 3)*

Cytomorphology

Negative Suspect Positive Total

Negative

541 (82.5%) 33 (5%) 0(0%)

574 (87.5%)

* Ten samples were non-evaluable

TdT-IF assay

Positive

40(6.1%) 11 (1.7%) 31(4.7%) 82 (12.5%)

Total

581 (88.6%) 44(6.7%) 31 (4.7%)

656 (100%)

Tab

le 2

. F

ollo

w-u

p of

15

AL

L a

nd 2

NH

L p

atie

nts

wit

h T

dT

pos

itiv

e ce

lls

in t

he C

SF

but

wit

h a

nega

tive

or

susp

ect

cyto

mor

phol

ogy

Td

T I

F a

ssay

C

lini

cal

cour

se

Cli

nica

l ph

ase

Dia

gnos

is

Pat

ient

s C

ytom

orph

olog

y T

dT"

cell

s In

terp

reta

tion

C

NS

leuk

emia

' F

ollo

w-u

p (n

) ('Y

o)

(wee

ks)

Pri

mar

y di

agno

sis

AL

L

4 ne

gati

ve o

r su

spec

t 0

.2-2

.6

susp

ect*

' no

6

9-1

06

' A

LL

ne

gati

ve

11

posi

tive

ye

s 10

A

LL

ne

gati

ve

IS

posi

tive

ye

s 26

A

LL

ne

gati

ve

55

posi

tive

ye

s 0

AL

L

nega

tive

70

po

siti

ve

yes

S6

NH

L

nega

tive

2

posi

tive

no

14

S+

NH

L

nega

tive

2

posi

tive

ye

s 12

Fol

low

-up

unde

r m

aint

enan

ce t

hera

py

AL

L

3 ne

gati

ve o

r su

spec

t 0

.2-0

.S

susp

ect

no

50

-10

4+

AL

L

3 ne

gati

ve

4-S

po

siti

ve

yes

4-S

A

LL

ne

gati

ve

10

posi

tive

ye

s 20

, C

NS

leu

kem

ia o

n cy

tom

orph

olog

ical

and

cli

nica

l gr

ound

s *

* In

the

CS

F f

rom

thr

ee o

f th

ese

four

pat

ient

s er

ythr

ocyt

es w

ere

pres

ent

......

0'1 ......

162

positive cytomorphology but no evidence of TdT reactivity [77], in our series a clear cytomorphological and clinical relapse was always confirmed by a positive TdT-IF assay. Moreover, the finding of more than 4% TdT+ cells in the CSF, independent of whether it occurred together with a negative or a suspect cytomorphology, was always followed by overt CNS leukemia. The presence of a lower percentage of TdT+ cells « 2.6% in this series) did not result in CNS leukemia, although, in our opinion, every single TdT+ cell in the CSF, in principle , should be regarded as malignant [65, 76,77,159] and, in fact, as a representative of a much larger leukemic process [46]. The finding of low percentages of TdT+ cells in a CSF sample from a patient without CNS leukemia might be explained by contamination of the CSF with (sometimes hardly detectable) amounts of PB [93, 96, 164] or even BM [96, 165-167] during CSF sampling. Contamination during CSF sampling at the primary diagnosis of ALL, when high numbers of TdT+ cells are present in PB and BM, may especially cause false positive IF assays. In such situa­tions, simultaneous analysis of CSF as well as PB for TdT+ cells would be desirable. Several reports, however, clearly demonstrated that during trau­matic lumbar puncture only about 20% of the predicted numbers of white blood cells are found in the CSF, suggesting that a direct extrapolation of PB data to CSF findings is not accurate [168-170]. In three of our patients, erythrocytes were present in the CSF at primary diagnosis, while the PB also contained TdT+ cells. Such punctions cannot be evaluated properly and should be repeated. On the other hand, it is quite conceivable that low num­bers of TdT+ cells in the CSF can disappear as a consequence of either the induction or maintenance therapy or the prophylactic CNS therapy [48, 106]. This would explain why the presence of low numbers of TdT+ cells in the CSF does not always result in overt CNS leukemia and would also sup­ply additional support for giving CNS prophylactic therapy. More and longer follow-up studies may clarify whether low percentages of TdT+ cells repre­sent a malignant, contaminating, or normal cell population.

Our results and those of others clearly indicate that the TdT-IF assay can be easily performed, even when low numbers of cells are present in the CSF. It deserves a place next to, or even prior to , the cytomorphological evaluation of CSF samples from patients with TdT+ malignancies. The TdT­IF assay contributes to early detection as well as exclusion of CNS leukemia, thereby offering an opportunity to avoid both undertreatment and overtreat­ment of patients.

Conclusion

Detection of meningeal involvement in leukemia and lymphoma has been largely dependent on cell count and cytomorphology for many years. As ad­ditional tools for the diagnosis of CNS involvement, immunological marker analysis, gene rearrangement studies, cytogenetics, electron microscopy,

163

DNA and RNA flow cytometry, and measurement of biochemical param­eters in the CSF have been proposed and partially introduced. Several of these techniques require large cell numbers, are time-consuming or expen­sive, or are unreliable in terms of specificity and sensitivity and therefore are not suitable for routine examination of CSF in patients with leukemia or NHL. The aforementioned disadvantages are less applicable to immuno­logical marker analysis of CSF cells.

The use of immunological markers such as E-rosettes, surface Ig, and, later on, (monoclonal) antibodies to cell surface markers and to the nuclear enzyme TdT have greatly contributed to a more precise diagnosis in the CSF. We feel that in TdT+ malignancies, especially ALL and TdT+ NHL, the TdT-iF method is superior to all other techniques employed in detecting CNS leukemia. In several TdT- malignancies a well chosen panel of mono­clonal antibodies - tailored to the individual patient's situation - will provide important information. In cases of doubt between malignant and benign pleiocytosis, T-cell and B-cell subsets, as well as the number of monocytes/macrophages and granulocytes, should be enumerated and moni­tored in time.

The combination of cytomorphological characterization and immuno­logical marker analysis of CSF cells at diagnosis, as well as during follow-up, contributes to a more accurate diagnosis and an earlier detection or exclu­sion of CNS involvement in leukemia and NHL. It provides a powerful tool which enables more accurate diagnoses and, as a consequence, adjustment of staging and remission criteria. This can contribute to avoidance of both undertreatment and overtreatment of patients.

References

I. Evans AE, Gilbert ES, Zandstra R: The increasing incidence of central nervous system leukemia in children (Children's cancer study group A). Cancer 26:404-409, 1970.

2. Dahl GV, Simone lV, Hustu HO, Mason e: Preventive central nervous system irradiation in children with acute nonlymphocytic leukemia. Cancer 42:2187-2192, 1978.

3. Levitt LJ, Dawson DM, Rosenthal DS, Moloney We: CNS involvement in the non­Hodgkin's lymphomas. Cancer 45:545-552, 1980.

4. Meyer Rl, Ferreira PP, Cuttner l, Greenberg ML, Goldberg l, Holland JF: Central nervous system involvement at presentation in acute granulocytic leukemia. A prospective cytocentrifuge study. Am J Med 68:691-694,1980.

5. Venables GS, Proctor SJ, Bates D, Cartlidge NE, Shaw DA: Intracranial disease in non­Hodgkin's lymphoma. Quart J Med, new series XLIX 194:111-131, 1980.

6. Sullivan MP, Chen T, Dyment PG, Hvizdala E, Steuber CP: Equivalence of intrathecal chemotherapy and radiotherapy as central nervous system prophylaxis in children with acute lymphatic leukemia: A pediatric oncology group study. Blood 60:948-958, 1982.

7. Freund M, Ostendorf P, Gartner VH, Waller HD: CNS manifestations in non-Hodgkin lymphomas (NHL). Klin Wochenschr 61:903-909, 1983.

8. Sackmann MF, Svarch E, Pavlovsky S, Eppinger-Helft M, Braier J, Vergara B, Garay G, Kvicala R, Divito JM, Failaee R, Dibar E, Jimenez E: Comparison of central nervous system prophylaxis with cranial radiation and intrathecal methotrexate versus intrathecal

164

methotrexate alone in acute lymphoblastic leukemia. Blood 62:241-250, 1983. 9. Inati A, Sallan SE, Cassady JR, Hitchcock-Bryan S, Clavell LA, Belli JA, Sollee N:

Efficacy and morbidity of central nervous system 'prophylaxis' in childhood acute lympho­blastic leukemia: Eight years' experience with cranial irradiation and intrathecal metho­trexate. Blood 61:297-303, 1983.

10. Lampert F, Henze G, Langermann H-J, Schellong G, Gadner H, Riehm H-J: Acute lymphoblastic leukemia: Current status of therapy in children. In: E Thiel, S Thierfelder (eds): Recent Results in Cancer research, vol 93, Berlin, Springer-Verlag, 1984.

11. Pui C-H, Dahl GV, Kalwinsky DK, Look AT, Mirro J, Dodge RK, Simone JV: Central nervous system leukemia in children with acute nonlymphoblastic leukemia. Blood 66: 1062-1067, 1985.

12. Wolf MM, Cooper lA, Olver IN, Liew KH, Ding JC, Madigan JP: Non-Hodgkin's lymphoma involving the central nervous system. Aust NZ J Med 15:16-21, 1985.

13. Aur RJ, Simone JV, Hustu HO, Verzosa MS: A comparative study of central nervous system irradiation and intensive chemotherapy early in remission of childhood acute lymphocytic leukemia. Cancer 29:381-391, 1972.

14. West J, Graham-Pole J, Hardisty RM, Pike MC: Factors in pathogenesis of central­nervous-system leukaemia. Br Med J 3:311-314, 1972.

15. Hustu HO, Aur RJ, Verzosa MS, Simone JV, Pinkel D: Prevention of central nervous system leukemia by irradiation. Cancer 32:585-597, 1973.

16. Wolk RW, Masse SR, Conklin R, Freireich EJ: The incidence of central nervous system leukemia in adults with acute leukemia. Cancer 33:863-869, 1974.

17. Bunn PA Jr, Schein PS, Banks PM, DeVita VT Jr: Central nervous system complications in patients with diffuse histiocytic and undifferentiated lymphoma: Leukemia revisited. Blood 47:3-10, 1976.

18. Choi S-I, Simone JV: Acute nonlymphocytic leukemia in 171 children. Med Pediatr Oncol 2:119-146, 1976.

19. Kay HE: Development of CNS leukaemia in acute myeloid leukaemia in childhood. Arch Dis Child 51:73-74,1976.

20. Herman TS, Hammond N, Jones SE, Butler 11, Byrne GE, McKelvey EM: Involvement of the central nervous system by non-Hodgkin's lymphoma. The Southwest oncology group experience. Cancer 43:390-397, 1979.

21. Litam JP, Cabanillas F, Smith TL, Bodey GP, Freireich EJ: Central nervous system relapse in malignant lymphomas: Risk factors and implications for prophylaxis. Blood 54:1249-1257, 1979.

22. Green DM, Freeman AI, Sather HN, Sallan SE, Nesbit ME Jr, Cassady JR, Sinks LF, Hammond D, Frei E, III: Comparison of three methods of central-nervous-system prophylaxis in childhood acute lymphoblastic leukaemia. Lancet 1:1398-1402, 1980.

23. Johnson GJ, Oken MM, Anderson JR, O'Connell MJ, Glick JH: Central nervous system relapse in unfavourable-histology non-Hodgkin's lymphoma: Is prophylaxis indicated? Lancet II:685-687, 1984.

24. Raz D, Siegal T, Siegal T Polliack A: CNS involvement by non-Hodgkin's lymphoma. Arch Neurol 41:1167-1171, 1984.

25. Anaissie E, Geha S, Allam C, Jabbour J, Khalyl M, Salem P: Burkitt's lymphoma in the Middle East. A study of 34 cases. Cancer 56:2539-2543, 1985.

26. Ersboll J, Schultz HB, Thomsen BL, Keiding N, Nissen NI: Meningeal involvement in non-Hodgkin's lymphoma: Symptoms, incidence, risk factors and treatment. Scand J Haematol 35:487-496, 1985.

27. Freeman C, Berg JW, Cutler SJ: Occurrence and prognosis of extra nodal lymphomas. Cancer 29:252-260, 1972.

28. Henry JM, Heffner RR Jr, Dillard SH, Earle KM, Davis RL: Primary malignant lym­phomas of the central nervous system. Cancer 34: 1293-1302, 1974.

29. Jellinger K, Radaskiewicz IT, Slowik F: Primary malignant lymphomas of the central nervous system in man. Acta Neuropath (Berlin), Suppl VI:95-102, 1975.

165

30. Berry MP, Simpson WJ: Radiation therapy in the management of primary malignant lymphoma of the brain. Int J Radiation Oncology Bioi Phys 7:55-59, 1981.

31. Zimmerman HM: Malignant lymphomas of the nervous system. Acta Neuropath (Berlin), Suppl VI:69-74, 1975.

32. Varadachari C, Palutke M, Climie AR, Weise RW, Chason JL: Immunoblastic sarcoma (histiocytic lymphoma) of the brain with B cell markers. J Neurosurg 49:887-892, 1978.

33. Helle TL, Britt RH, Colby TV: Primary lymphoma of the central nervous system. Clinicopathological study of experience at Stanford. J Neurosurg 60:94-103, 1984.

34. Gill PS, Levine AM, Meyer PR, Boswell WD, Burkes RL, Parker JW, Hofman FM, Dworsky RL, Lukes RJ: Primary central nervous system lymphoma in homosexual men. Clinical, immunologic, and pathologic features. Am J Med 78:742-748, 1985.

35. Woodman R, Shin K, Pineo G: Primary non-Hodgkin's lymphoma of the brain. Medicine 64:425-430, 1985.

36. Hobson DE, Anderson BA, Carr I, West M: Primary lymphoma of the central nervous system: Manitoba experience and literature review. Can J Neurol Sci 13:55-61, 1986.

37. Aur RJ, Simone J, Hustu HO, Walters T, Borella L, Pratt C, Pinkel D: Central nervous system therapy and combination chemotherapy of childhood lymphocytic leukemia. Blood 37:272-281, 1971.

38. Nesbit ME, D' Angio GJ, Sather HN, Robison LL, Ortega J, Donaldson M, Hammond GD: Effect of isolated central nervous system leukaemia on bone marrow remission and survival in childhood acute lymphoblastic leukaemia. A report for children's cancer study group. Lancet 1:1386-1389, 1981.

39. Stewart DJ, Keating MJ, McCredie KB, Smith TL, Youness E, Murphy SG, Bodey GP, Freireich EJ: Natural history of central nervous system acute leukemia in adults. Cancer 47:184-196, 1981.

40. Komp DM, Fernandez CH, Falletta JM, Ragab AH, Bennet Humphrey G, Pullen J, Moon T, Shuster J: CNS prophylaxis in acute lymphoblastic leukemia. Comparison of two methods. A Southwest oncology group study. Cancer 50:1031-1036, 1982.

41. Anderson JR, Wilson JF, Jenkin RD, Meadows AT, Kersey J, Chilcote RR, Coccia P, Exelby P, Kushner J, Siegel S, Hammond D: Childhood non-Hodgkin's lymphoma. The results of a randomized therapeutic trial comparing a 4-drug regimen (COMP) with a 10-drug regimen (LSA2-L2 ). N Engl J. Med 308:559-565, 1983.

42. Gasparini M, Lattuada A, Lombardi F, Rilke F, Gianni C, Fossati-Bellani F: Childhood non-Hodgkin's lymphoma: Prognostic relevance of clinical stages and histologic sub­groups. Am J Pediatr HematollOncol 5:161-171, 1983.

43. Pinkerton CR, Chessells JM: Failed central nervous system prophylaxis in children with acute lymphoblastic leukaemia: Treatment and outcome. Br J Haematol 57:553-561, 1984.

44. Thompson CB , Sanders JE, Flournoy N, Buckner CD, Thomas ED: The risks of central nervous system relapse and leukoencephalopathy in patients receiving marrow transplants for acute leukemia. Blood 67:195-199,1986.

45. Editorial: Leukaemia and the central nervous system. Lancet 1:1196-1198, 1985. 46. Price RA, Johnson WW: The central nervous system in childhood leukemia: I. The

arachnoid. Cancer 31:520-533, 1973. 47. Hoogerbrugge PM, Hagenbeek A: Leptomeningeal infiltration in rat models for human

acute myelocytic and lymphocytic Icukemia. Leuk Res 9:1397-1404, 1985. 48. Glass JP, Melamed M, Chernik NL, Posner JB: Malignant cells in cerebrospinal fluid

(CSF): The meaning of a positive CSF cytology. Neurology 29:1369-1375,1979. 49. Olson ME, Chernik NL, Posner JB: Infiltration of the leptomeninges by systemic cancer.

A clinical and pathologic study. Arch Neurol 30:122-137, 1974. 50. An-Foraker SH: Cytodiagnosis of malignant lesions in cerebrospinal fluid. Revicw and

cytohistologic correlation. Acta Cytol 29:286-290, 1985. 51. Mackintosh FR, Colby TV, Podolsky WJ, Burke JS, Hoppe RT, Rosenfelt FP, Rosenberg

SA, Kaplan HS: Central nervous system involvement in non-Hodgkin's lymphoma: An

166

analysis of 105 cases. Cancer 49:586-595, 1982. 52. Bigner SH, lohnston WW: The cytopathology of cerebrospinal fluid. l. Nonneoplastic

conditions, lymphoma and leukemia. Acta Cytol 25:335-353, 1981. 53. Borowitz M, Bigner SH, lohnston WW: Diagnostic problems in the cytologic evaluation

of cerebrospinal fluid for lymphoma and leukemia. Acta Cytol 25:665-674, 1981. 54. Oehmichen M, Huber H: Supplementary cytodiagnostic analyses of mononuclear cells of

the cerebrospinal fluid using cytological markers. 1 Neurol 218: 187 -196, 1978. 55. Goodson lD, Strauss GM: Diagnosis of lymphomatous leptomeningitis by cerebrospinal

fluid lymphocyte cell surface markers. Am 1 Med 66:1057-1059, 1979. 56. lellinger K, Slowik F, Sluga E: Primary intracranial malignant lymphomas. A fine struc­

tural cytochemical and CSF immunological study. Clin Neurol Neurosurg 81:173-184, 1979.

57. Matsuda M, McMurria H, VanHale P, Miller CA: CSF findings in primary lymphoma of the CNS. Arch Neurol 38:397, 1981.

58. Ganick Dl, Sondel PM, Gilbert EF, Borcherding W: Leukemia presenting as central nervous disease without bone marrow involvement. Med Ped Oncol 11:229-232, 1983.

59. Marsh WL, Stevenson DR, Long III Hl: Primary leptomeningeal presentation of T-cell lymphoma. Report of a patient and review of the literature. Cancer 51: 1125-1131, 1983.

60. Magennis H, Markey G, Alexander HD, Morris TC: Leukaemia/lymphoma cells in cerebrospinal fluid. 1 Clin Pathol 36:606, 1983.

61. Moir Dl, Ghosh AK, Abdulaziz Z, Knight PM, Mason DY: Immunoenzymatic staining of haematological samples with monoclonal antibodies. Br 1 Haematol 55:395-410, 1983.

62. Coakham HB, Garson lA, Brownell B, Allan PM, Harper EI, Lane EB, Kemshead IT: Use of monoclonal antibody panel to identify malignant cells in cerebrospinal fluid. Lancet 1:1095-1098, 1984.

63. Ezrin-Waters C, Klein M, Deck 1, Lang AE: Diagnostic importance of immunological markers in lymphoma involving the central nervous system. Ann Neurol 16:668-672, 1984.

64. Hohlfe\d R, Brilskc-Hohlfeld I, Schwartz A, Brocke U, Toyka KV: Analyse von ober­fliichenmarkern auf liquorzellen. Beitrag zur differentialdiagnostik intrathekaler Iymphome. Dtsch Med Wschr 109: 1760-1762, 1984.

65. Hooijkaas H, van Dongen JJ, Hiihlen K, van Zanen GE: Immunological characterisation of cells in cerebrospinal fluid from patients with lymphoid malignancies. Lancet 1:518-519, 1984.

66. Martin SE, Zhang H-Z, Magyarosy E, Jaffe ES, Hsu S-M, Chu EW: Immunologic methods in cytology: Definitive diagnosis of non-Hodgkin's lymphomas using im­munologic markers of T- and B- cells. Am 1 Clin Pathol 82:666-673 , 1984.

67. Veerman AJ, Huismans DR, van Zantwijk CH, de Waal FC: Immunologische bepaling van oppervlaktedeterminanten op cellen in de liquor cerebrospinalis van kinderen met acute lymfatische leukemic. Ned Tijdschr Geneeskd 128:100-103, 1984.

68. Homans AC, Forman, EN, Barker BE: Use of monoclonal antibodies to identify cerebro­spinal fluid lymphoblasts in children with acute lymphoblastic leukemia. Blood 66: 1321-1325, 1985.

69. Jones GR, Mason WH, Fishman LS, DeClerck YA: Primary central nervous system lymphoma without intracranial mass in a child. Diagnosis by documentation of mono­clonality. Cancer 56:2804-2808, 1985.

70. Veerman AJ, Huismans LD, van Zantwijk IC: Diagnosis of meningeal leukemia using immunoperoxidase methods to demonstrate common acute lymphoblastic leukemia cells in cerebrospinal fluid. Leuk Res 9:1195-1200, 1985.

71. Hoefnagels W A, Speelman JD, Hische EA: Diagnosis of a primary non-Hodgkin's lym­phoma of the central nervous system by examination of the cerebrospinal fluid. Clin Vol Neurol Neurosurg 88:45-48, 1986.

72. Oscier DG, Mufti GJ, Hamblin n, Jones DB, Smith JL: Evolution of a terminal deoxynucleotidyl transferase-positive lymphoma from a chronic T cell lymphocytosis.

167

Scand .J Haematol 36:221-228, 1986. 73. Li C-Y, Witzig TE, Phyliky RL, Ziesmer SC, Yam LT: Diagnosis of B-ccll non-Hodgkin's

lymphoma of the central ncrvous systcm by immunocytochemical analysis of cerebrospinal fluid lymphocytes. Cancer 57:737-744, 1986.

74. Bradstock KF, Papageorgiou ES, Janossy G, Hoffbrand AV, Willoughby ML, Roberts PD, Bollum FJ: Dctection of leukaemic Iymphoblasts in CSF by immunofluorescence for terminal transferase. Lancet 1:1144, 1980.

75. Peipcr S, Stass SA, Bollum FJ: Detection of leukaemic Iymphoblasts in CSF. Lancet 2:39, 1980.

76. Bradstock KF, Papageorgiou ES, Janossy G: Diagnosis of meningeal involvement in patients with acutc lymphoblastic leukemia: Immunofluorcscencc for terminal transferasc. Cancer 47:2478-2481, 1981.

77. Casper JT, Lauer S.J, Kirchner PA, Gottschall JL, Camitta BM: Evaluation of cerebro­spinal fluid mononuclear cclls obtaincd from children with acute lymphocytic leukemia: Advantages of combining cytomorphology and terminal deoxynucleotidyl transfcrase. Am J Clin Pathol 80:666-670, 1983.

78. Hooijkaas H, van Dongen JJ, Hahlen K, van Zanen GE: Identification and monitoring of terminal deoxynucleotidyl transferase positive cells in the cerebrospinal fluid of children with lymphoid malignancies. In: Protides of the Biological Fluids. H. Peeters (ed): Oxford, Pergamon Press, 1984, pp 875-878.

79. Hooijkaas H, Adriaansen HJ, Hiihlen K, Dekker I, van Zanen GE, van Dongen JJ: The presence or absence of terminal deoxynucieotidyl transferase (TdT) positive cells in the cerehrospinal fluid (CSF) has implications for the diagnosis of meningeal involvement in patients with acute lymphoblastic leukemia. In: Pruzanski W, Seligman M (eds): Clinical Immunology. Amsterdam, Elsevier Science Publishers, 1986, pp. 431-434.

80. Kwaan HC, Pierrc RV, Long DL: Meningeal involvement as first manifestation of acute myeloblastic transformation in chronic granulocytic leukemia. Blood 33:348-352, 1969.

81. Mastrangelo R, Zuelzer WW, Ecklund PS, Thompson RI: Chromosomes in the spinal fluid: Evidence for metastatic origin of meningeal leukemia. Blood 35:227-235, 1970.

82. Zuelzcr WW, Inoue S, Thompson RI, Ottenbreit MJ: Long-term cytogenetic studies in acute leukemia of children: the nature of relapse. Am J Hematol 1:143-190, 1976.

83. Pearson J, Iigren EB, Spriggs AI: Lymphoma cells in cerebrospinal fluid confirmed by chromosome analysis. J Clin Pathol 35:1307-1311, 1982.

84. Dosik H, Rodriguez J, Verma RS: Cytogenetic studies of blast cells in the cerebrospinal fluid in meningeal leukemia. Cancer Genet Cytogenet 12:89-91, 1984.

85. Daniel SE, Scaravilli F, Hayward R, Henry K: Primary intracranial histiocytic lymphoma with Langerhans ' granules. A light microscopic and ultrastructural study. Cancer 56:2816-2822, 1985.

86. Kaps M, Burkhardt E: An improved method for electron microscopic observation of cerebrospinal fluid cells. Acta Cytol 29:484-486, 1985.

87. Redner A, Andreeff M, Miller DR, Steinherz P, Melamed M: Recognition of central nervous system leukemia by flow cytometry. Cytometry 5:614-618, 1984.

88. Avent J: Flow cytometry in body fluid analysis. Clin Lab Med 5:389-403, 1985. 89. Mavligit GM, Stuckey SE, Cabanillas FF, Keating MJ, Tourtellotte WW , Schold SC,

Freireich EJ: Diagnosis of leukemia or lymphoma in the central nervous system by betar microglobulin determination. N Engl J Med 303:718-722, 1980.

90. Koch TR, Lichtenfeld KM, Wiernik PH: Detection of central nervous system metastasis with cerebrospinal fluid betarmicroglobulin. Cancer 52:101-104,1983.

91. Cleary ML, Warnke R, Sklar J: Monocionality of Iymphoproliferative lesions in cardiac­transplant recipients. Clonal analysis based on immunoglobulin-gene rearrangements. N Engl J Med 310:477-482, 1984.

92. Drewinko B, Sullivan MP, Martin T: Usc of the cytocentrifuge in the diagnosis of meningeal leukemia. Cancer 31: 1331-1336, 1973.

93. Woodruff KH: Cerebrospinal fluid cytomorphology using cytocentrifugation. Am J Clin

168

Pathol 60:621-627, 1973. 94. Evans DI, O'Rourke C, Morris Jones P: The cerebrospinal fluid in acute leukaemia of

childhood: Studies with the cytocentrifuge . J C1in Pathol 27:226-230, 1974. 95. Hansen HH, Bender RA, Shelton BJ: The cyto-centrifuge and cerebrospinal fluid cytol­

ogy. Acta Cytol 18:259-262, 1974. 96. Komp DM: Diagnosis of CNS leukemia . Am J Ped Hemat/Oncol 1:31-35, 1979. 97. McIntosh S, Ritchey AK: Diagnostic problems in cerebrospinal fluid of children with

lymphoid malignancies. Am J Ped Hematol/Oncol 8:28-31 , 1986. 98. Ricevuti G , Savoldi F, Piccolo G Marchioni E, Rizzo SC: Meningeal leukemia diagnosed

by cytocentrifuge study of cerebrospinal fluid. A study of 631 cerebrospinal fluid samples from 87 patients. Arch Neurol 43:466-470, 1986.

99. Shankar SK, Verma K: Cerebrospinal fluid cytology in patients with leukemias and lymphomas. Indian J Med Res 71:90-95, 1980.

100. Barrett DL, King EB: Comparison of cellular recovery rates and morphologic detail obtained using membrane filter and cytocentrifuge techniques. Acta Cytol 20: 174-180, 1976.

101. Gondos B, King EB: Cerebrospinal fluid cytology: Diagnostic accuracy and comparison of different techniques. Acta Cytol 20:542-547, 1976.

102. Dyken PR: Cerebrospinal fluid cytology: Practical clinical usefulness. Neurology 25:210-217 , 1975.

103. Ducos R, Donoso J , Weickhardt U , Vietti TJ: Sedimentation versus cytocentrifugation in the cytologic study of craniospinal fluid . Cancer 43:1479-1482, 1979.

104. Pochedly C: Methods for cytological examination of the cerebrospinal fluid. In: C Poched­ley (ed): Leukemia and Lymphoma in the Nervous System, Springfield, Illinois, Charles C Thomas, 1977, pp 88-104.

105. Verma K, Tiwari MC: Millipore filter versus cytocentrifuge technique in cytopathological diagnosis . Indian J Med Res 74: l35-140, 1981.

106. Davey DD, Foucar K, Giller R: Millipore filter vs. cytocentrifuge for detection of child­hood central nervous system leukemia. Arch Pathol Lab Med 110:705-708, 1986.

107. Gondos B: Millipore filter vs. cytocentrifuge for evaluation of cerebrospinal fluid. Arch Pathol Lab Med 110:687-688, 1986.

108. Rawlinson DG, Billingham ME, Berry PF, Kempson RL: Cytology of the cerebrospinal fluid in patients with Hodgkin 's disease or malignant lymphoma. Acta Neuropath (Berl) Suppl VI :187-191 , 1975.

109. Davies SF, Gormus BJ, Yarchoan R, Kaplan ME: Cryptococcal meningitis with false­positive cytology in the CSF. Use of T-cell rosetting to exclude meningeal lymphoma. JAMA 239:2369-2370, 1978.

110. Choi H-S, Anderson PJ: Diagnostic cytology of cerebrospinal fluid by the cytocentrifuge method. Am J Clin Pathol 72:931-943, 1979.

111. Foon KA, Schroff RW, Gale RP: Surface markers on leukemia and lymphoma cells: Recent advances. Blood 60:1-19, 1982.

112. Bernard A, Boumsell L, Dausset J, Milstein C, Schloss mann SF (eds): Leucocyte typing. Human leucocyte differentiation antigens detected by monoclonal antibodies. Berlin, Springer-Verlag, 1984.

113. Reinherz EL, Haynes BF, Nadler LM, Bernstein ID (eds): Leucocyte typing II. Volume 1: Human T Lymphocytes; Volume 2: Human B Lymphocytes; Volume 3: Human Myeloid and Hematopoietic Cells. Berlin , Springer-Verlag, 1986.

114. Foon KA, Todd RF III: Immunologic classification of leukemia and lymphoma. Blood 68:1-31, 1986.

1l5. Van Dongen JJ, Adriaansen HJ, Hooijkaas H: Immunological marker analysis of cells in the various hematopoietic differentiation stages and their malignant counterparts. In: Ruiter DJ, Fleuren GJ, Warnaar SO (eds): Application of Monoclonal Antibodies in Tumor Pathology . Boston, Martinus Nijhoff Publishers, 1986, pp. 87-116.

1l6. Rowley JD: Chromosome abnormalities in human leukemia. Ann Rev Genet 14:17-39, 1980.

169

117. Chaganti RS: Significance of chromosome change to hematopoietic neoplasms. Blood 62:515-524, 1983.

118. Hagemeijer A, Adriaansen HJ, Bartram CR: New possibilities for cytogenetic analysis of leukemic cells. In: Hagenbeek A , Lowenberg B (eds): Minimal Residual Disease in Acute Leukemia 1986. Boston, Martinus Nijhoff Publishers, 1986, pp 1-11.

119. Sandberg AA: The chromosomes in human cancer and leukemia. New York, Elsevier North Holland, 1980.

120. Yunis JJ: The chromosomal basis of human neoplasia. Science 221:227-236, 1983. 121. Williams DL, Look AT, Melvin SL, Robertson PK, Dahl G, Flake T, Stass S: New

chromosomal translocations with specific immunophenotypes of childhood acute lympho­blastic leukemia. Cell 36:101-109,1984.

122. Shuster J, Gold P, Poulik MD: ~z-microglobulin levels in cancerous and other disease states. Clin Chim Acta 67:307-313, 1976.

123. Amiot PL, Adinolfi M: Serum ~z-microglobulin and its prognostic value in lymphomas. Eur J Cancer 15:791-796, 1978.

124. Clausen N, Ibsen, KK: Central nervous system relapse surveillance by serial betaz­microglobulin measurements in childhood acute lymphoblastic leukemia. Acta Paediatr Scand 73:848-854, 1984.

125. Imagawa S, Aoki Y, Ninomiya H, Yoda Y, Abe T: Beta-2-microglobulin concentration in cerebrospinal fluid. Monitoring of malignant lymphoma in the central nervous system. Acta Haematol Jpn 48:1059-1064,1985.

126. Pudek MR, Chan KW, Rogers PC, Teasdale JM: Betaz-microglobulin levels in cerebro­spinal fluid of children with leukemia and lymphoma. Clin Biochem 18:180-183, 1985.

127. Nagelkerke AF, van Kamp GJ, Veerman AJ, de Waal FC: Unreliability of ~-2-microglo­bulin in early detection of central nervous system relapse in acute lymphoblastic leukemia. Eur J Cancer Clin Oncol 21:659-663, 1985.

128. Schold SC, Wasserstrom WR, Fleisher M, Schwartz MK, Posner JB: Cerebrospinal fluid biochemical markers of central nervous system metastases. Ann Neurol 8:597-604, 1980.

129. Wallen WC, Biggar RJ, Levine PH, Neequaye J, Nkrumah FK: Cerebrospinal fluid markers in African Burkitt 's lymphoma with central nervous system involvement. J Natl Cancer Inst 69:787-792, 1982.

130. Wallen WC, Biggar RJ, Levine PH, Iivanainen MV: Oligoclonal IgG in CSF of patients with African Burkitt's lymphoma. Arch Neurol 40:11-13, 1983.

131. Look AT, Melvin SL, Williams DL, Brodeur GM, Dahl GV, Kalwinsky DK, Murphy SB, Mauer AM: Aneuploidy and percentage S-phase cell determined by flow cytometry correlated with cell phenotype in childhood acute leukemia. Blood 60:959-967, 1982.

132. Suarez C, Miller DR, Steinherz PG, Melamed MM, Andreeff M: DNA and RNA determination in 111 cases of childhood acute lymphoblastic leukaemia (ALL) by flow cytometry: Correlation of FAB classification with DNA stemline and proliferation. Br J Haematol 60:677-686, 1985.

133. Andreeff M, Darzynkiewicz Z, Sharpless T, Clarkson B, Melamed R: Discrimination of human leukemia subtypes by the flow cytometric analysis of cellular DNA and RNA. Blood 56:282-293, 1980.

134. Korsmeyer SJ, Arnold A, Bakhshi A, Ravetch lV, Sieben list U, Hieter PA, Sharrow SO, LeBien TW, Kersey JH, Poplack DG, Leder P, Waldmann TA: Immunoglobulin gene rearrangement and cell surface antigen expression in acute lymphocytic leukemias of T cell and B cell precursor origins. J Clin Invest 71:301-313,1983.

135. Van Dongen JJ' Quertermous T, Bartram CR, Gold DP, Wolvers-Tettero IL, Comans­Bitter WM, Hooijkaas H, Adriaansen HJ, de Klein A, Raghavachar A, Ganser A, Duby AD, Seidman JG, van den Elsen P, Terhorst C: The T cell receptor-CD3 complex during early T cell differentiation: Analysis of immature T cell acutc lymphoblastic leukemias (T­ALL) at DNA, RNA and cell membrane level. J Immunol, 138:1260-1269, 1987.

136. Young RC, Howser DM, Anderson T, Fisher RI, Jaffe E, DeVita VT Jr: Central nervous system complications of non-Hodgkin's lymphoma. The potential role for prophylactic therapy. Am J Med 66:435-443, 1979.

170

137. Reske-Nielsen E, Hoffmann-Petersen 1, Sylgaard H, Jensen KB: Leukaemia of the central nervous system. Lancet 1:211-212, 1974.

138. Getaz EP, Miller Gl: Spinal cord involvement in chronic lymphocytic leukemia. Cancer 43: 1858-1861, 1979.

139. P6ltorak M, Czlonkowska A, Nowicka K: Chronic lymphocytic leukemia: Study of cell subsets in cerebrospinal fluid and peripheral blood. Eur Neurol 22:289-292, 1983.

140. Wolfe DW, Scopelliti lA, Boselli BD: Leukemic meningitis in a patient with hairy cell leukemia. A case report. Cancer 54: 1085-1087, 1984.

141. Aaronson AG, Hajdu ST, Melamed MR: Spinal fluid cytology during chemotherapy of leukemia of the central nervous system in children. Am 1 Clin Pathol 63:528-537, 1975.

142. Hancock WW, Medley G: Monoclonal antibodies to identify tumour cells in CSF. Lancet 2:739-740, 1983.

143. Oehmichen M, Huber H: Supplementary cytodiagnostic analyses of mononuclear cells of the cerebrospinal fluid using cytological markers. 1 Neurol 218: 187 -196, 1978.

144. Harrison PB, Cripps A W: Exclusion of leukaemic meningitis by quantitation of T and B lymphocytes in cerebrospinal fluid. Aust NZ 1 Med 12:286-287, 1982.

145. Hohlfeld R, Schwartz A, Brocke U, Toyka KV: A practicable method for the analysis of T lymphocyte subsets in CSF lymphopleocytosis. Klin Wochenschr 61 :933-934, 1983.

146. Ambler MW, Homans AC, O'Shea PA: An unusual central nervous system infection in a young immunocompromised host. Arch Pathol Lab Med 110:497-501, 1986.

147. Cashman N, Martin C, Eizenbaum 1-F, Degos 1-D, Bach M-A: Monoclonal antibody­defined immunoregulatory cells in multiple sclerosis cerebrospinal fluid. 1 Clin Tnvest 70:387-392, 1982.

148. Hauser SL, Reinherz EL, Hoban Cl, Schlossman SF, Weiner HL: CSF cells in multiple sclerosis: Monoclonal antibody analysis and relationship to peripheral blood T-cell sub­sets. Neurology 33:575-579, 1983.

149. Richert lR, Rose lW, Xu X-H, Greenstein JI, McFarland HF, McFarlin DE: Characteri­zation of cerebrospinal fluid lymphocytes in multiple sclerosis. Annals NY Acad Sci 436:151-162, 1984.

150. Noronha A, Richman DP, Amason BG: Multiple sclerosis: Activated cells in cerebro­spinal fluid in acute exacerbations. Ann Neurol 18:722-725, 1985.

151. Moser RP, Robinson lA, Prostko ER: Lymphocyte subpopulations in human cerebro­spinal fluid. Neurology 26:726-728, 1976.

152. Levy R, Warnke R, Dorfman RF, Haimovich J: The monoclonality of human B-cell lymphomas. J Exp Med 145: 1014-1028, 1977.

153. Mann RB, laffe ES, Braylan RC, Eggleston JC, Ransom L, Kaizer H, Berard CW: Immunologic and morphologic studies of T cell lymphoma. Am J Med 58:307-313, 1975.

154. Ritz J, Pesando 1M, Notis-McConarty J, Lazarus H, Schlossman SF: A monoclonal antibody to human acute lymphoblastic leukaemia antigen. Nature 283:583-585, 1980.

155. Janossy G, Bollum Fl, Bradstock KF, Ashley J: Cellular phenotypes of normal and leukemic hemopoietic cells determined by analysis with selected antibody combinations. Blood 56:430-441, 1980.

156. Greaves MF, Hariri G, Newman RA, Sutherland DR, Ritter MA, Ritz 1: Selective expression of the common acute lymphoblastic leukemia (gplOO) antigen on immature lymphoid cells and their malignant counterparts. Blood 61:628-639, 1983.

157. Greaves M, Delia D, Janossy G, Rapson N, Chcssells J, Woods M, Prentice G: Acute lymphoblastic leukaemia associated antigen. TV. Expression on non-leukaemic 'lymphoid' cells. Leuk Res 4:15-32,1980.

158. Van Dongen JJ, Hooijkaas H, Comans-Bitter M, Hahlen K, de Klein A, van Zanen GE, van't Veer MB, Abels 1, Benner R: Human bone marrow cells positive for terminal deoxynucleotidyl transferase (TdT) , HLA-DR, and a T cell marker may represent prothymocytes. J Tmmunol 135:3144-3150, 1985.

159. Van Dongen JJ, Hooijkaas H, Hahlen K, Benne K, Bitter WM, van der Linde-Prccsman AA, Tettero IL, van de Rijn M, Hilgers J, van Zanen GE, Hagemeijer A: Detection of

171

minimal residual disease in TdT positive T cell malignancies by double immuno­fluorescence staining. In: Lowenberg B, Hagenbcek A (cds): Minimal Residual Disease in Acute Leukemia. Boston , Martinus Nijhoff Publishers , 1984, pp 67-81.

160. Van Dongen JJ , Hooijkaas H , Adriaansen HJ , Hahlen K, van Zanen GE: Detection of minimal residual acute lymphoblastic leukemia by immunological marke r analysis: Possi­bilities and limitations. In: Hagenbeek A, Lowenberg B (eds): Minimal Rcsidual Disease in Acute Leukemia 1986. Boston , Martinus Nijhoff Publishers , 1986, pp 113-133.

161. Rovigatti U, Mirro J, Kitchingman G, Dahl G, Ochs, J, Murphy S, Stass S: Heavy chain immunoglobulin gene rearrangement in acute non lymphocytic leukemia. Blood 63: 1023-1027, 1984.

162. Bollum FJ: Terminal deoxynucleotidyl transfe rase as a hematopoietic cell marker. Blood 54: 1203-1215, 1979.

163. Johnson GD , de C Nogueira Araujo GM: A simple method of reducing the fading of immunofluorescence during microscopy. J Immunol Methods 43:349-350, 1981.

164. Rohlfing MB, Barton TK, Bigner SH, Johnston WW: Contamination of cerebrospinal fluid specimens with hematogenous blasts in patients with leukemia. Acta Cytol 25:611-615, 1981.

165. Kruskall MS, Carter SR, Ritz LP: Contamination of cerebrospinal fluid by vertcbral bone­marrow cells during lumbar puncture. N Engl J Med 308:697-700, 1983.

166. Lane PA, Githens JH: Contamination of cerebrospinal fluid with bone-marrow cells during lumbar puncture. N Engl J Med 309:434-435, 1983.

167. Luban NL , Alessi RM , Gold BG, Kapur S: Cerebral spinal fluid pleocytosis with bone marrow contamination. J Pediatr 104:254-256, 1984.

168. Chow G, Schmidley JW: Lysis of erythrocytes and leukocytes in traumatic lumbar punc­tures. Arch Neurol 41:1084-1085,1984.

169. Novak RW: Lack of validity of standard corrections for white blood cell counts of blood­contaminated cerebrospinal fluid in infants. Am J Clin Pat hoi 82:95-97, 1984.

170. Rubenstein JS, Yogev R: What represents pleocytosis in blood-contaminated (,traumatic tap ') cerebrospinal fluid in children? J Pediatr 107:249- 251 , 1985.

4. Detection of residual disease in acute leukemia using immunological markers

Daniel H. Ryan and Jacques J.M. van Dongen

Rationale for residual disease monitoring

Central to our current strategies of cancer chemotherapy is the concept that all or nearly all malignant cells must be destroyed by therapy in order for treatment to be successful. Even in neoplasms that respond to chemother­apy, residual tumor cells too few in number to be detectable are capable of proliferating and causing relapse. The consequence of this is that cancer chemotherapy must be continued in patients who have attained a clinical remission for a period of time adequate to eliminate all residual clonogenic tumor cells in as many patients as possible. The optimum duration and intensity of maintenance chemotherapy is determined by clinical trials of patients stratified according to prognostic parameters.

Detection of small numbers of residual tumor cells has several potential advantages in this context. First, relapse may be detectable at an early stage by identification of rare tumor cells. Second, patients with a higher than average likelihood of relapse may be identified by detection of persistence of tumor cells after initiation of therapy or by detection of occult micrometas­tases in clinically localized tumor. Finally, the presence of neoplastic cells in bone marrow obtained during clinical remission for autologous transfusion may be detected by these methods.

Early detection of relapse

At diagnosis, the tumor burden in acute leukemia is approximately 1013 cells. Clinical remission is defined by the presence of fewer than 5% blast cells in the bone marrow, since that is the limit of sensitivity of morpholog­ical techniques for detection of residual leukemic cells. With a total body bone marrow cellularity of 2 x 1012 cells, this represents roughly 1011 cells. Below this level, the leukemic tumor burden can no longer be measured, and the patient continues to be treated according to protocol. If at the conclusion of therapy leukemia cells remain, they are likely to be at un­detectable levels. Assuming that the leukemia cells proliferate (i.e., that clonogenic stem cells remain), there will be a period of time until the tumor

BellI/eli, I .M. and Foon. K.A., (eds.), Immunologic Approaches 10 rhe Classificarion and Managemem of Lymphomas and Leukemias. © 1988 Kluwer Academic Publishers. /SBN978·/·4612·li965-4. All righrs reserved.

174

burden is once again roughly 1011 cells, which is the point of earliest detection of clinical relapse. Would it make a difference if the relapsing tumor cells could be detected at a tumor burden of 109 cells, rather than 10" cells?

The major clinical problem in treatment of relapsed malignant tumors is the emergence of chemotherapeutic drug resistance in the relapsed tumor. Often it is possible to obtain a partial or complete remission, but long-term survival is poor due to the presence of a small population of drug-resistant cells. Therefore, one may rephrase the question at the end of the preceding paragraph in a different way: What is the evidence that drug resistance is less likely to be present in 109 cells than 10" cells? There is considerable evidence that stable , drug-resistant, malignant cells arise spontaneously by mutation with a certain frequency, independently of the presence of a selecting agent [1]. Therefore, assuming a fixed mutation rate, the number of drug-resistant mutant clones in a tumor is related to the number of cell divisions that have occurred and therefore to the total tumor burden. This is the theoretical basis for the Goldie-Coldman model [2] of mutation to drug resistance in malignant tumors. Direct experimental proof of this model in man is limited by our inability to study rare drug-resistant tumor cells, but the model is consistent with animal models and indirect clinical evidence [3]. Mice injected with 105 L1210 leukemic cells are cured by two courses of ara­c, while animals injected with 107 L12l0 leukemic cells cannot be cured even by multiple courses of ara-c, due to the emergence of resistant cells [4]. The success of adjuvant chemotherapy in tumors such as osteogenic sarcoma [5] and breast cancer [6] supports the relevance of this model to human malig­nancy. A genetic mechanism (gene amplification) for multidrug (pleiotropic) resistance has recently been identified. Drug resistant cells, identified by monoclonal antibody to the P glycoprotein [7], were absent at diagnosis and remission but increased steadily prior to relapse in a patient with myeloid leukemia [8]. The Goldie-Coldman model has been adapted to include con­siderations of tumor cell death or differentiation and multiple drug-resistant phenotypes [9]. This analysis indicates that the probability of emergence of multiple drug-resistant clones rises as tumor burden increases, particularly when the rate of tumor cell death or differentiation is substantial. Since each cell division is associated with a risk of development of a drug-resistant mutation, drug resistance is more likely at any given tumor bulk when the rate of tumor cell death is high, because in this circumstance more tumor cell divisions are required to achieve a given tumor bulk.

It is important to note that the emergence of these drug-resistant clones is not dependent on the presence of the drug as a selecting agent. Thus the emergence of cells resistant to drug C becomes more likely with each round of cell division during (unsuccessful) chemotherapy with drugs A and B, even if A, B, and C are totally unrelated drugs . At relapse, one then discovers that the tumor now contains a small (but lethal) number of cells resistant to C as well as A and B. A goal of early detection of relapse in this

175

case would be to detect the relapsing cells prior to emergence of the clone resistant to drug C, allowing effective chemotherapy of the relapsed tumor using drug C. This analysis can be applied to bone marrow transplantation (BMT) in acute lymphoblastic leukemia, substituting BMT for drug C. The primary reason for failure of allogeneic BMT in patients who have relapsed with ALL is resistance of some leukemic cells to the induction chemotherapy for BMT, even when given at maximum tolerated doses (limited by nonmar­row toxicity in the case of BMT).

Figure 1 is a graphic illustration of the potential benefit of early detection of relapse in the context of emergence of multiply drug-resistant clonogenic leukemic cells. Assume that the proportion of tumor stem cells is approxi­mately 10% and that only two effective drugs are available for the treatment of relapsed disease. If the spontaneous rates of mutation to resistance to the two chemotherapeutic agents are independent and no resistant cells are originally present, the proportion of cells resistant to only one or to both of the drugs can be calculated [9]. If one postulates an average mutation rate in tumor stem cells (2.8 x 10-6 per cell division [10]) for resistance to each of two drugs, then at a tumor burden of 1011 cells, 10 doubly drug-resistant leukemic stem cells are likely to be present, while at a tumor burden of 109

cells, no doubly drug-resistant leukemic stem cell is likely to be present. For simplicity, this model does not include consideration of leukemic stem cell differentiation and death rates, but inclusion of these considerations, as calculated by Goldie and Coldman [9], will not significantly change the probability of a stable doubly drug-resistant stem cell phenotype. The like­lihood of there being any doubly resistant stem cell at any point in time (i.e., "curability" in this simplified model) is indicated in the lower panel of figure 1. Note the rapidity with which this probability drops from 97.5% to 4 %. Thus, in this model it makes little difference whether the relapse is treated at a tumor burden of 1012 vs. 1011 cells, but it does make a difference whether treatment is begun at 1011 vs. 109 cells.

There are some additional benefits to earlier detection of relapse, when the percent of leukemic cells in S phase (and susceptible to certain chemo­therapeutic agents) is higher and organ dysfunction secondary to tumor infiltration has not occured. A potential problem relates to sampling interval. The model in figure 1 does not take into account the slowing of tumor growth with increasing tumor bulk (Gompertzian kinetics) for the sake of simplicity. It is possible that the progression from 109 to lOlO cells may take place over days, while that from lOlO to 1011 or 1012 cells may take weeks or months to occur.

Assessment of response to induction chemotherapy

The rate at which tumor cells decrease during induction chemotherapy, as measured by bone marrow tumor burden at day 14 of induction therapy, has been correlated with ultimate clinical outcome in childhood acute lympho-

176

Relapse 100% ,...--------------==~~----''-----. 50% Probability of No Doubly Drug

Resistant Leukemia Stem Cells

0%

Time •

Leukemia Stem Cells

Stem Cells Resistant to One Drug

Stem Cells

Figure 1. Hypothetical model of tumor burden at various time points in the treatment of a patient with acute leukemia who relapses after an initial complete clinical remission. The standard detection limit for leukemia cells is taken to be 5% of the average bone marrow cellularity. The effect of detecting leukemic cells with a 100-fold greater sensitivity is considered. The leukemic stem cell population is estimated to be 10% of the total leukemic cells (which may be an overestimation [118]), and mutation to resistance to each or two chemotherapeutic agents is postulated to occur independently at a rate of 2.8 x 10- 6 per cell division, which is the mean of that reported for mutation at the hypoxanthine-guanine phosphoribosyltransferase locus by three Iymphoblastoid cell lines (10). The tumor burden of singly and doubly resistant leukemic stem cells and the probability of there being no doubly resistant leukemic stem cell at a particular point in time are taken from Coldman and Goldie [119). Note that early detection of relapse in this model occurs when 2240 singly drug-resistant leukemia stem cells are present, and there is a 97.5% probability that no doubly resistant cells are present. At the time of standard diagnosis of relapse, 280,000 singly drug-resistant leukemia stem cells are present, and there is less than 4% probability that no doubly resistant cells are present.

blastic leukemia (ALL) in the CCSG 160 series. Detection of residual leukemic cells by more sensitive methods may allow more accurate estimation of effectiveness of tumor cell killing by induction chemotherapy at diagnosis or in the initial treatment of relapse. In tumors with subpopulations of drug­resistant c\onogenic cells, the rate of killing of the drug-sensitive majority population of tumor cells may not reflect the treatability of the drug-resistant subpopulation. In tumors with a more homogeneous population of sensitive or partially resistant cells, the rate of tumor cell kill during induction may be of prognostic value.

177

Detection of occult metastatic disease

Immunologic detection of occult disseminated tumor in clinically localized disease has been described in breast cancer [11] and nodular lymphoma [12]. The finding of micrometastases in these tumors has not yet been shown to have prognostic significance. This application is relevant to acute leukemia only insofar as leukemic cells may be identified in relatively protected locations such as the CSF.

Autologous bone marrow transplantation

Autologous bone marrow transplantation is being studied in neuroblastoma and ALL, among other tumors. A potential problem with this therapy is the possibility of transplanting residual tumor cells along with the "remission" bone marrow. Detection of residual rare tumor cells in the transplanted remission bone marrow may help to determine whether a relapse arises from the transplanted autologous marrow or from bone marrow cells resistant to the preparative regimen.

Detection of residual tumor cells is potentially of greatest value in tumors for which curative therapy is available, but relapse occurs in a proportion of patients. Examples of such tumors include breast cancer, colon cancer, and childhood ALL. The leukemias in general are more amenable to many techniques of rare cell detection due to the ease with which these cells can be obtained in single cell suspension. Detection of residual acute leukemia using monoclonal antibodies is the subject of this review, but other methods of residual cell detection will be discussed for comparison. ALL has been more intensively studied than nonlymphoid leukemias because lineage and differentiation markers specific for early differentiation stages have been more clearly identified in the lymphoid lineage.

Methods for residual disease detection in acute leukemia

A successful clinical application for detection of residual disease in acute leukemia must satisfy a number of criteria, summarized in table 1. Not only does an assay have to be able to reliably distinguish leukemia cells from normal cells in a given tissue, but enough cells must be evaluated to be able

Table 1. Requirements for Successful Detection of Residual Leukemia

Conveniencc and speed of assay Ability of assay system to analyze large numbers of cells High frequency of leukemia marker in leukemia cells from different patients Rare expression of leukemia marker in normal tissue High likelihood of leukemic involvement in tissue sampled prior to clinical relapse Effective treatmcnt for patients with residual disease.

178

to detect rarely occuring events. Furthermore, given the ability to detect rare leukemia cells, there has to be a clinical benefit to the early identification of these cells relative to conventional means of residual disease detection. These considerations are important because the detection systems to be discussed below are limited with respect to different criteria.

Morphology

The (g) old standard for diagnosis of leukemia is the morphologic appearance of cells stained with the Wright/Giemsa stain. The major advantage of this technique is the long clinical experience with its use. Significant limitations include subjectivity, lack of standardization of dyes, and lack of sensitivity. Approximately 10% of children with ALL present with no morphologically identifiable leukemic cells in the peripheral blood but with substantial num­bers of cells positive for the common ALL antigen (CALLA) [13]. Auto­mated analysis of cell size and granule content by commercial leukocyte differential counters probably has a sensitivity similar to that of manual examination [14]. In childhood ALL, leukemic relapse is not suspected until over 5% blast cells are identifiable in the bone marrow and is not confidently diagnosable until over 25% blast cells are present. The limited sensitivity is due to the presence of small numbers of immature or reactive cells with morphologic features indistinguishable from ALL cells. These cells are more common in pediatric than adult marrows and may be particularly numerous in the bone marrow of patients after discontinuation of maintenance che­motherapy [15].

Measurement of secreted proteins

The search for abnormal serum proteins as 'tumor markers' has identified several proteins useful in the follow-up of patients with certain types of solid tumors [16]. Detection of minimal residual tumor depends on the quantity of protein shed or secreted by the tumor cells as well as the sensitivity of the assay method. Multiple myeloma is a hematopoietic tumor with a secreted product (monoclonal immunoglobulin [IgD that is roughly correlated with tumor burden [17]. Monoclonal Ig can also be detected by sensitive methods in the majority of patients with chronic lymphocytic leukemia [18]. Several proteins associated with ALL, including lactate dehydrogenase [19], HLA­DR antigen [20], terminal deoxynucleotidyl transferase (TdT), and CALLA [21], have been identified in serum, but only during overt clinical disease. A more recent study using a solid-phase immunoassay failed to find consistent TdT elevation in serum or plasma at the time of diagnosis of ALL [22]. Detection of residual leukemia using serum or urine levels of lysozyme, a protein associated with monocytic leukemia [23], suffers from the same problem of lack of sensitivity [24]. In addition, lysozyme is not specific for leukemic monocytes, but is characteristic of the monocytic lineage in general, and may be elevated in nonmalignant disorders [25].

179

Detection of specific proteins in individual cells

Detection of cellular proteins has the theoretical advantage of high sensi­tivity, since one abnormal cell could potentially be identified among millions of normal cells and multiple parameters could be used to identify such cells. Detection of rare breast cancer [11] and neuroblastoma [26] cells in the bone marrow using immunohistochemical techniques has been reported. Immu­nofluorescence analysis by flow cytometry has the advantages of speed and objectivity but for greatest sensitivity is limited to surface antigens. Analysis of cytoplasmic or nuclear antigens in rare cells by flow cytometry is difficult due to increased nonspecific binding of reagents to intracellular structures. Two markers of immature lymphoid cells, CALLA and TdT, have been helpful in the identification of residual leukemic cells in ALL. The detection of residual leukemia using immunologic markers will be discussed in the next section.

Detection of aneuploid cells

Chromosomal aneuploidy is a more specific marker for malignancy [27] than surface markers, although dysplastic cells from the colon and uterine cervix can be aneuploid. Using fluorescent dyes that bind stoichiometrically to DNA, the DNA content of individual cells can be determined by flow cytometry [28]. The limit of detection of alteration in DNA content for this method is about 5% of the total DNA content, or one to two chromosomes. As a marker for rare cells, aneuploidy is limited for a number of reasons. Hyperdiploid cells usually fall in the region between DNA index of 1 (2n) and 2 (4n), hence they are difficult to distinguish from normal S phase cells. Hypodiploid or hypertetraploid cells might be more distinctive, but the sensitivity of detection is limited by the invariable presence of small amounts of debris or cellular aggregates, even in the best of preparations. A com­bination of surface or nuclear marker and DNA content measurement may be a useful approach to this problem in the future [29]. A major problem in acute leukemia is that aneuploidy can be detected by flow cytometry in only about 30% of patients. In ALL, the hyperdiploid patients, in fact, have a better prognosis than those with normal karyotype or a translocation [30].

Conventional cytogenetic analysis

In contrast to the relative infrequency of gross aneuploidy in acute leukemia is the high incidence of chromosomal abnormalities detectable in nonlympho­cytic leukemia by standard cytogenetic analysis l31 J if sensitive techniques are used. Specificity is conferred not only by the fact that an abnormality exists, but also by the ability to compare the abnormality with that found in the initial neoplasm. For detection of rare cells, however, this methodology is severely limited. Cytogenetic analysis is very labor intensive, making traditional analysis of thousands of karyotypes in each sample clearly im-

180

practical. Another limitation is that cell mitosis is required to produce a karyotype, and leukemia cells frequently demonstrate less proliferative ac­tivity in vitro than do normal cell types, thereby obscuring the quantitative relationship between normal and abnormal cells in the sample. Stimulation of marrow samples by specific growth factors may be useful in improving the threshold of detection [32], but sensitivity will still be poorer than other methods. Chromosome spreads in ALL are notoriously poorly defined and difficult to read, further hampering attempts to detect rare ALL cells by this method.

Cytogenetic analysis by flow cytometry

An area of current interest is the application fo flow cytometry technology to identification and isolation of individual human chromosomes [33]. A major hurdle in this effort has been the similarity in size and DNA content of certain groups of human chromosomes. Recent studies have utilized two­color analysis using pairs of dyes such as chromomycin A3 (specific for GC base pairs) and Hoechst 33258 (specific for AT base pairs) to improve resolution of the larger human chromosomes [34]. Flow cytometric detection of chromosome translocations has been described [35].

Chromosome specific repetitive sequences of DNA have been described for several chromosomes. Repetitive sequences are more easily detected than individual genes, since they are present in multiple copies on the chromo­some. Repetitive sequences making up 1 % -10% of the genome can be detected in individual murine thymocytes using fluorochrome labeled DNA probes [36]. A possible clinical application of this technique is detection of translocation by identification of chromosome specific sequences on more than one different chromosome.

Potential advantages of these techniques include speed and the ability to sort individual abnormal chromosomes for further analysis. Problems hindering application of these methods to rare cell detection include the large number of mitoses needed for analysis (especially a problem in human leukemias 'and lymphoproliferative disorders, where the tumor cells often are less likely to undergo mitosis in vitro than the residual normal cells), discrimination of similar-sized chromosomes (such as 9-12 and 17-18), and detection of cytogenetic changes in a small subpopulation of abnormal cells (since only separated chromosomes can be studied, as opposed to analysis on a cell by cell basis).

Gene hybridization techniques

Antigen receptor gene rearrangement is characteristic of differentiated normal and abnormal T-lymphocytes and B-lymphocytes. The presence of a clonal rearranged band after hybridization of the appropriate probe to DNA cleaved by restriction endonuclease can be used to demonstrate the mono-

181

clonal origin of a cell population. In addition, information suggesting the T­cell or B-cell origin of a lymphoid malignancy may be provided by analysis of Ig38 or T-cell antigen receptor [37] (TcR) gene rearrangements, although the lineage suggested by gene rearrangements is not always the same as that indicated by surface phenotype. Since the detection system involves analysis of total DNA from a clinical sample, sensitivity depends on the ability to detect a particular band among a diffuse background of numerous bands arising from normal clones. In general, a 1 %-5% popUlation of monoclonal cells can be reasonably clearly identified by this technique. Mixing experi­ments using DNA isolated from ALL cells diluted in placental or peripheral blood leukocyte DNA showed that a rearranged Ig band from a 0.3% mixture of ALL cells was detectable [39]. In detection of relapse, one has the benefit of knowing where the rearranged band is likely to be, since rearranged Ig or T-cell antigen receptor (TcR) bands are similar at diagnosis and relapse in about 85% of cases [40]. A prepurification step involving depletion by antimyelomonocytic and erythroid monoclonal antibodies followed by DNA of the remaining cells may permit detection of a 0.05% population of malignant T-cells or B-cells [41]. These techniques are prom­ising, but some limitations remain. DNA rearrangements similar to those which occur in lymphocytes have not been found in other hematopoietic cell types. Prepurification steps, although capable of greater sensitivity, are more laborious and time-consuming than desirable for a test to be used in large numbers of patients as a screening test for relapse. The length of time required to complete the analysis and the requirement for approxi­mately 107 cells per digest may be limiting factors in general application of these techniques for this purpose at present. The limit of sensitivity in patient marrow samples in addition to in vitro mixtures of cells or DNA needs to be studied further.

Another approach is to identify DNA or mRNA which is specific for a particular leukemia type. For instance, the 9;22 translocation in CML results in formation of an abnormal her-abl mRNA transcript [42]. The protein product of this hybrid oncogene is abnormally large and possesses aberrant tyrosine kinase activity [43]. Such a protein is an example of a cell marker which may enable specific detection on the protein, mRNA, or DNA level of individual abnormal cells [44]. Proto-oncogenes are intimately involved in control of normal cell proliferation and are therefore not specific for malig­nancy, but aberrant forms of these genes, such as identified for e-ras [45], may be useful for detection of malignant cells. In situ hybridization may improve the sensitivity of detection of abnormal mRNA by allowing mor­phologic identification of positive cells [44, 46].

Clonogenie assays

The greatest potential for sensitivity and biological information is a clonogen­ic assay for abnormal cells. Cell lines have been established from morpho-

182

logically uninvolved bone marrow of 17% of patients with undifferentiated or lymphoblastic lymphomas , both highly proliferative tumors [47, 48]. The greatest difficulty in this endeavor is to provide the correct in vitro conditions to promote leukemic cell growth and to inhibit the growth of (far more numerous) normal progenitor cells under the same conditions. Again the problem of specificity for leukemic cells arises, only this time in the context of growth factor requirements. Specificity can also be conferred by phenoty­pic or cytogenetic analysis of the resulting colonies, but this adds a degree of complexity to the assay. Another limiting factor for sensitivity is that colony assays only detect c1onogenic cells, while phenotypic assays are potentially capable of detecting both end-stage and c1onogenic leukemia cells. If the number of c1onogenic cells is 0.01 %-1 % of the total leukemic cells, a loss of sensitivity of the colony assay will result. Conversely , the colony assay detects cells that are probably biologically more relevant than the total leukemic population. Although myeloid, monocytic, erythroid, and mega­karyocytic precursor cells can be readily grown in colonies in vitro, the leukemic counterparts of these cells often grow poorly or show limited pro­liferative potential. In vitro growth of the normal counterparts to ALL cells has not been achieved as of yet. Nevertheless, some success has been achieved recently in growing colonies of ALL cells in vitro [49]. In this study, colonies of cells with a phenotype similar to the original leukemia were iso­lated after short-term culture from the marrow of ALL patients shortly after entering complete remission. In one case, cytogenetic abnormalities were documented in the colonies. Two patients have remained in remission after leukemic cells were apparently grown from their marrow. Two explanations for this occurrence are that either the residual leukemic cells were killed by subsequent maintenance therapy or that the in vitro colony forming cells were not capable of long-term c1onogenic growth in vivo. Consolini et al. were able to grow B-cell colonies with phenotypic and cytogenetic abnorma­lities from most patients with ALL in remission. Of the five patients with no abnormal colonies identified, all remained in remission, but 3 of 12 patients with abnormal colonies relapsed one to five months afterwards [50]. The presence of abnormal colony forming cells may not therefore be inconsistent with continued remission in acute leukemia, as shown in acute myeloid leukemia by the identification of clonal granulopoiesis in patients with long­term remission of AML [51].

Detection of residual disease acute leukemia using immunologic markers

Methodological considerations

Of the criteria necessary for successful and clinically meaningful detection of minimal residual disease in acute leukemia, two relate to the specific metho­dology used: practicality and detection limit of the assay. For detection of

183

rare leukemia cells by monoclonal antibodies, two detection techniques have been commonly used, fluorescence microscopy or immunohistochemistry and flow cytometry. The microscopical approach has practical advantages of requiring less expensive instrumentation and permitting morphologic analysis of the cells, while flow cytometry allows larger numbers of cells to be analyzed quantitatively. The biological factors that tend to limit sensitivity of rare cell detection apply to both methods of detection, but we will use flow cytometry as an example to illustrate these problems.

Nonspecific binding of antibody. Detection of a rare leukemic population is limited as much by nonspecific binding of the normal cells as by the charac­teristics of the rare cells. Dantas et al. reported that detection of rare melanoma cells in bone marrow was limited to a sensitivity of 5%, not by the lack of reactivity of melanoma cells to the antibodies used, but by the high level of nonspecific binding of normal marrow cells [52]. The type of cells and antibody reagents studied significantly affect the degree of non­specific binding observed. Monocytes express large numbers of a particular type of IgG Fc receptor (FcRI) which avidly binds irrelevant monomeric murine IgG2a or IgG3 antibodies [53]. Since IgG2a is a common isotype of murine monoclonal antibodies, it is apparent that monocytes are a potential source of nonspecific binding of these reagents. Rare cell detection in bone marrow is also hampered by the large numbers of mature and immature myeloid cells that remain after ficoll density centrifugation (lower density bone marrow neutrophils appear to be less functionally mature than periph­eral blood neutrophils [54]). These cells have a higher autofluorescence in general than lymphoid cells and possess Fc receptors that can bind small amounts of aggregated antibody reagent.

Two general approaches to solve this problem are to deplete monocytes and myeloid cells before staining or identify and exclude them from evalua­tion after staining. Depletion techniques using physical properties of these cells abound, but all share some common problems, most notably a non­specific (and often nonrandom) loss of other cell types, including cells that might be of interest [55]. Depletion techniques (immunoadsorption on columns [56], panning on petri dishes [57]) using monoclonal antibodies have the advantage of specificity but introduce a degree of complexity to the assay and increase the length of time required.

In immunocytochemical techniques, monocytes and myeloid cells may be morphologically evaluable after staining and hence can be excluded from the count of positive cells. Flow cytometry techniques are well suited to multi­parameter analysis using light scatter and additional fluorescence parameters. Light scattered at 90° from the incident laser beam (90LS) is roughly pro­portional to the internal complexity of a cell, while light scattered at smaller angles (2° to 18° as usually measured) is more dependent on cell size. Mono­cytes and myeloid cells display high 90LS intensity [58] (due to cytoplasmic granules), while lymphoid cells and immature cells of all hematopoietic

184

lineages have low 90LS intensity. This parameter is important in clinical enumeration of T-cell subsets and has been shown to reduce the number of mononuclear cells showing bright nonspecific staining by 97% in a rare cell assay system [59]. This technique, of course, cannot be used if the rare cells may also have a high 90LS signal (i.e., myeloid leukemia cells of the FAB morphologic types M2 or M3).

A more flexible method for identifying cell types that obscure a rare population by nonspecific binding of antibody is to label these cells with a different colored fluorochrome from that used to identify the rare cells. This technique is only limited by the availability of monoclonal antibodies which stain normal cell types but not the leukemia cells . Double immunofluore­scence staining is useful not only for flow cytometric analysis, but also for fluorescence microscopy. Van Dongen and coworkers showed that T­lymphoma cells could not be detected in peripheral blood by single color flow cytometry at < 1% concentration but were detectable at a 0.01 % level in a double fluorescence microscopic assay [60].

Antigen density of rare cells. Related to the problem of nonspecific binding is the requirement for intensity of antigen expression of rare cells if they are to be detected. If one stains a 'suspension of peripheral blood mononuclear cells with an irrelevant monoclonal antibody directly conjugated to fluore­scein, the mean fluorescence intensity will be low, with a roughly log-normal distribution. If a large number of cells are counted, the probability of en­countering a cell with a fluorescence intensity higher than a given value can be estimated. Based on our data using peripheral blood lymphocytes stained with irrelevant murine monoclonal IgG2a, we have constructed computer simulations representing mixtures of negative cells and different proportions of positive cells at two different fluorescence intensities (figure 2). This type of modeling, as used by Leary and coworkers in detection of Rh positive fetal red cells in the maternal circulation [61], is important to the investigator attempting to detect a rare cell population. Knowing the distribution of negative (i.e., normal) cells stained with irrelevant control reagent and leukemic cells stained for the leukemia associated antigen, one can estimate the likely sensitivity of the assay system with respect to nonspecific binding of antibody reagent. The fluorescence intensity of the stained rare cell is of great importance. What seems to be a well separated population of positive cells at 50% frequency may become undetectable when only 0.1 % positive cells are present. Note in figure 2 that the apparent lower limit of detection is 100-1000 times lower for the bright cells than the weak cells, even though the bright cells are only 4.3 times as bright as the weak cells. This illustrates the critical difference that fluorescence intensity can make in the detection of rare cells. Moving a few channels out on the log fluorescence histogram places the positive population in a region where nonspecific flu­orescence of the negative cells is several fold lower. It does not necessarily follow from this that indirect staining procedures, which typically result in

185

50%

I fCA"O~LI ~.'"

en 10% I~ I) l---~ ::al u ~ ~

I j ~ I j L ',l:l 'Q;

1% 0 p..

'E ------...... <l)

II ~ u

II ~ en <l) ...

0.1% 0 ;:l

li:: ~ 'E

000% II LIJ ~ <l) u ~

p..

M01% II hll ~ Fluorescence Intensity Fluorescence Intensity

Figure 2. Computer modeling of a rare population of fluorescent cells constructed on a Mac­intosh computer. A log normal histogram with mode at channel 15 showed a striking resem­blance to logarithmically displayed fluorescence histograms of lymphoid cells stained with irrelevant monoclonal antibody. The "positive" population was represented by a normal distribution with mean at channel 85 (panels on left) or channel 62 (panels on right). The " negative" histogram was multiplied by a factor and added to the "positive" histogram to represent varying proportions of rare "positive" cells. The histograms were scaled up propor­tionately to show the positive population clearly on scale. Note that both the weakly and intensely positive cells are clearly defined and easy to recognize at 50'Yo and 10% concentration, but the weakly positive population becomes lost in the background fluorescence of the negative cells at a concentration of 0.1 % -0.01 % , while the strongly positive cell population is recogniz­able at a 0.001 % conccntration. To relate this simulation to logarithmic displays of fluorescence intensity, as commonly used in flow cytometry, the strongly positive cells would be 4.3 times as bright as the weakly positive population on such a scale. The fluorescence intensity shown by the brightly positive histogram is similar to the fluorescence intensity of strongly CD \0 positive leukemia cells, while that shown by the weakly positive cell population is similar to the intensity of CD2 on mature T-cells .

higher fluorescence, will always improve performance of a rare cell assay. Indirect immunofluorescence exposes cells to two sources of nonspecific binding and can result in higher nonspecific binding with resulting decrease in sensitivity.

Numbers of cells required. The number of cells required in a rare cell assay is related to the frequency of rare cells to be detected. To measure one cell

186

in 100,000, it is necessary to count 1,000,000 cells to detect 10 events, and 10,000,000 cells to detect 100 events. With modern commercial flow cyto­meters, count rates of 5000-10,000 per second are feasible, which permits analysis of 1,000,000 cells in two to four minutes. Therefore, clinical assays for the detection of rare cells in the order of 1/100,000 are quite feasible in terms of flow cytometer time. Detection of events of greater rarity (1/ 1,000,000 or 1/10,000,000) would require faster analysis, which is possible on modified instruments [62]. Ultimately, a practical limit on the number of cells attainable from routine bone marrow aspiration (about 100,000,000 total cells after ficoll density centrifugation) or peripheral blood (100,000,000 mononuclear cells requires at least 100 ml of blood) is reached. Account­ing for losses during centrifugations (at least 50%) and the need to stain a separate aliquot with control antibody, the practical limit of detection of residual leukemia in the bone marrow or peripheral blood by mono­clonal antibodies is about 1/500,000 to 1/1,000,000 (assuming detection of at least 10 abnormal cells is required to diagnose relapse). In the bone marrow, this means that detection of a leukemia burden of less than 106 cells is impractical by these methods (since the marrow cellularity is about 10 12

cells), even if a specific antibody were available. This fact in itself does not defeat the purpose of residual leukemia detection, since adequate treatment of a tumor burden of 109 cells may abort the development of a rare multidrug resistant clone that would otherwise result in relapse refractory to combina­tion reinduction chemotherapy, as illustrated in figure 1. At present the major limitations of residual disease detection in acute leukemia using monoclonal antibodies are the combined problems of nonspecific binding, weak staining intensity of the leukemic cells, and (especially) nonspecificity of anitbodies for the leukemic cells, not insufficient numbers of cells for study.

Biological factors. Certain biological factors can impair the sensitivity of an assay for residual leukemia. One of the most important is phenotypic changes in the leukemia from diagnosis to relapse. This appears to be a greater problem the more detailed the leukemia phenotype must be characterized to identify the leukemic cell correctly. Phenotypic changes from lymphoid to myeloid phenotype occur, but very infrequently (4% of cases in a recent study [63]), while shifts from CALLA positive to CALLA negative (about 15% [64, 65]) or TdT positive to TdT negative (24% [65]) within the ALL category are more common. In our experience , shifts from TdT positive to TdT negative occur infrequently at relapse in ALL.

Variability in leukemic infiltration measured in bone marrow from differ­ent anatomical sites has been observed in patients with relapsing ALL [66] and has been documented to occur in the transplantable Brown Norway rat myelocytic leukemia (BNML) model. In the BNML model, leukemic infil­tration seven to ten days after injection of 107 BNML cells is remarkably uniform in large and small bones. However, after chemotherapy , the varia-

187

tion between different animals in leukemic infiltration in small bones (rib, sternum, etc.) is 100 times greater the variation observed in large bones (femur, tibia) [67]. This may be related to statistical variations in repopula­tion of small bones from small numbers of residual leukemia cells in the larger bones. It is possible that such a variability might also apply to detection of leukemic cells in extramedullary sites seeded from the bone marrow. This may be a source of variability, even in technically accurate and sensitive detection assays.

Tissue localization of leukemic cells is a critical factor in detection of residual leukemia. The pattern of spread of leukemia cells through the body depends on the type of leukemia. The tissues commonly involved by a particular leukemia are often those which are normal homing areas for its normal counterpart, for example thymus in T-cell ALL.

Animal models

Animal models suitable for the study of leukemic growth in vivo can be used to assess feasability and possible clinical significance of residual disease detection in acute leukemia. An advantage of such systems is the ability to detect as few as 1-10 residual leukemia cells by in vivo adoptive transfer assays [68]. The BNML model is useful because the leukemia cells can be specifically identified by a monoclonal antibody [69] (Rm 124) at a sensitivity of 1 cell in 1000-10 ,000 [70]. The presence of clonogenic leukemia cells can be assayed by an in vivo assay that discriminates between leukemic and normal stem cells. Growth kinetics (doubling time) of leukemia cells in the liver after inoculation can be assayed not only at large tumor burdens (109

cells) but also after chemotherapy when only 103 _104 leukemia cells are present. Minimal residual leukemia cells grow more slowly than those at a tumor burden of 105 _108 cells [71]. There are significant differences between these animal models and human leukemia. Human leukemia populations have a much lower fraction of clonogenic cells than transplantable animal cell lines, and there appears to be considerable variation in growth rates of relapsing human leukemia, at least during the time when detectable leuke­mia cells are present (lOlO_1012 cells). In addition, there is at present neither a specific monoclonal antibody nor a leukemia specific clonogenic assay for human leukemia.

Immunologic subtypes of acute lymphoblastic leukemia

With the discovery of CALLA [72] and the production of a monoclonal antibody to this antigen [73], the majority of ALL cells could be immunolog­ically recognized by a marker that seemed to be relatively leukemia specific. With increasing knowledge of lymphoid differentiation, it has become ap­parent that the phenotype of ALL cells recapitulates the phenotype of normal differentiating lymphoid cells with some degree of fidelity. ALL

188

cells can be generally classified into common ALL (CALLA positive, T-cell antigen negative, surface Ig negative), T-ALL (T-cell antigen positive), pre­B-ALL (cytoplasmic Ig positive), B-ALL (surface Ig positive), and undiffer­entiated or null ALL (negative for all antigens except TdT and/or HLA­DR). The majority of ALL patients fall into the common ALL group (about 75% of childhood ALL and 50% of adult ALL). Figure 3 shows the pheno­typic resemblance of the subtypes of ALL to the sequence of normal lym­phoid differentiation. Antibodies to surface antigens of hematopoietic cells have been grouped into cluster designations (CD) based on patterns of cross­reactivity and structural similarities of the antigens to which they bind [74]. For instance, CALLA has been designated CDlO. The pan-B-cell marker CD19 (B4) is expressed throughout the B-celllineage and in nearly all non-

~ III pre-B-ALL III rIITB-ALLIIIII

~~ ;:g,:-,Q,,---eaQe~---Q ~'I- ~v / TdT TdT HLA-DR B lymphocytes ~ 0 HLA-DR HLA-DR (CDlO)

CD10 CD10 CD19

~ CD19 CD19 CD20

pro-B cell (CD20) (CD20) Smlg TdT CD34 weak Cyl'

CD19 O HLA-DR

~~"~;:~ CO"" ~"6~~~~ 0 H::-6R 0 0"lvI"$ 1: 0mature mature CD34 ~"I< common thymocyte T-helper/ inducer

< thymocytic T-ALL (TdT) lymphocytes prothymocyte

C02

TdT 0 0 CD3 HLA-DR --- CD5 C~ COO

thymocyte thymocyte <;>I'v-" ~"""'-TTn TdT TdT $ I-ALL

2g;4 immature common~ COB

gg~ (gg~) 0----0 CD5 CD?

(CD4/CDB) mature

thymocyte

(TdT) C02 CD3 CD4 CD5 CD?

mature T·suppressor/

cytotoxic lymphocytes

Figure 3. Hypothetical lymphoid differentiation scheme of the lymphoid lineage and the malignant counterparts of the early lymphoid cells. CD nomenclature for the surface markers is used. The immunological markers expressed by cells in the various lymphoid differentiation stages are indicated. Markers in parentheses are not uniformly expressed. The bars represent malignant counterparts , i.e. , the various (sub )types of ALL.

189

T-ALL. Expression of C020 (B1) occurs somewhat later in normal B-cell maturation. C02 (Leu 5) and C07 (3A1) are expressed very early in the T­cell lineage and are nearly always found in T-ALL. C05 (Leu 1) is expressed on all but the most immature thymocytes and most T-ALL. C04 (T4) and C08 (T8) are variably present in T-ALL, while C03 (T3) is usually absent.

Ig gene rearrangement can be demonstrated in all cases of common ALL, pre-B-ALL, B-ALL, and most cases of null ALL [75]. TcR gene rearrange­ment can be demonstrated in the majority of cases of T-ALL [76]. TcR rearrangements have been identified in common ALL [77], and, conversely, Ig gene rearrangement identified in T-ALL [78, 79]. Recent studies of v-abl transformed murine lymphoid precursors suggests that early rearrangement events of both Ig and T-cell antigen receptor genes may occur in both T-cell and B-cell lineages, but later stages of rearrangement may be more lineage specific [80].

Detection of minimal residual disease in common acute lymphoblastic leukemia

Attempts to detect residual leukemia in common ALL have utilized two markers that are associated with normal and neoplastic early B-cell differen­tiation, TdT and COW. TdT is an enzyme present on the nuclear membrane of immature B-cells and T-cells. Approximately 95% of cells from patients with ALL are TdT positive (B-ALL is characteristically TdT negative). TdT is a DNA polymerase which lacks template requirements and has been postulated to play a role in generating antibody diversity by introducing somatic mutations into immunoglobulin genes [81, 82]. The potential muta­genic ability of TdT has been documented [83]. Since neither of these markers are leukemia specific, diagnosis of residual disease can be made only if higher than normal numbers of antigen positive cells are identified in a given tissue or antigen positive cells are found in tissue locations outside the homing areas of their normal counterparts.

Terminal deoxynucleotidyl transferase. TdT positive cells are found normally in the thymus, where they make up the majority of cells, and the bone marrow, where they constitute a small percentage of lymphoid cells. Thymic TdT positive cells are early T-cell precursors, while the majority of marrow TdT positive cells express B-cell markers [84]. The number of TdT positive cells in the bone marrow decreases with age, averaging 3.4% in children and 1.1 % in adults [85], but may reach 10% in regenerating bone marrow [86]. Therefore, detection of residual leukemic cells in the bone marrow using immunofluorescence detection of TdT as a single marker is not practical [87].

TdT positive cells are exceedingly rare in peripheral blood, however, ranging from 0.005%-0.11% in children and 0%-0.045% in adults [85]. Using immunofluorescence detection of TdT positive mononuclear cells

190

in peripheral blood, Froelich et al. reported that elevated percentages (> 0.11 %) of TdT positive cells occurred three to five weeks before relapse in ALL [88]. A follow-up study on a total of 72 patients (612 samples) sug­gested that elevated numbers of TdT positive cells occur too often in the absence of later relapse to be useful in guiding therapy [89]. Six of ten relapses were preceded by four days to eight months with elevated (0.12%-0.23%) TdT+ cells in peripheral blood. However, an additional 21 children had elevated TdT + cells without subsequent relapse. Similarly, Barr et al. found that relapse could not be predicted by enumerating TdT positive cells in peripheral blood [90]. The significant problems these authors encountered with nonspecific binding of the TdT antiserum may be related to the use of unfixed cells for the demonstration of TdT by immunofluorescence. It should be noted that a recent report has described nonspecific staining with an anti­TdT antiserum of peripheral blood lymphocytes following PHA stimulation [91].

A sensitive solid-phase immunoassay has been developed for the detection of TdT antigen [92]. This assay has advantages of convenience, sample stability, and quantitative expression of results and appears to be useful in the diagnosis of ALL. It has been proposed as a means of detection of early relapse by monitoring the level of TdT antigen in peripheral blood mono­nuclear cells. Calculations based on the average amount of TdT antigen expresed per leukemic cell in ALL (11 fg/ce1l92-22 fg/ce1l93 ) and the normal range of TdT antigen in peripheral blood mononuclear cells « 50 ng/108 cells) suggest that the detection limit for residual ALL cells is about 2%. This is several orders of magnitude less sensitive than immunofluorescence assays at a single cell level to be described below and is unlikely to be of use in detecting early relapse in ALL.

Common ALL antigen. COlO has a much wider tissue distribution that TdT. It has been identified on normal marrow B-cell precursors [94], germinal center cells [95], granulocytes [96], renal parenchymal cells [97], breast myoepithelial cells, intestinal epithelium [98], and a subpopulation of thy­mocytes [99]. In bone marrow, the number of CDIO positive cells normally exceeds TdT positive cells, even when CDIO positive neutrophils are exclud­ed by either morphology or 90° light scatter. The TdT positive CDIO positive cells are phenotypically more immature than the COlO positive TdT negative lymphoid cells [100]. Oetection of minimal residual common ALL in mar­row using the COlO antigen is hampered by the frequency of normal mar­row COlO positive cells, averaging 4.5% in adult marrow but higher in children [100]. After cessation of chemotherapy for ALL, a rebound lym­phocytosis is commonly observed [101], with a high proportion of COW positive cells [102], occasionally over 50% of ficoll density centrifuged mono­nuclear cells. Although the average intensity of CDlO expression in normal marrow COlO positive cells is lower than in common ALL, the TdT positive COW positive subpopulation (about 18% of all normal marrow COW posi-

191

tive cells) is strongly CD10 positive like most common ALL cells [100], thus preventing distinction of normal CDlO positive cells from ALL cells by intensity of CDlO expression. Since nearly all CDlO positive ALL cells are also TdT positive, leukemic relapse may be suspected if the proportion of CDlO positive cells in the bone marrow which are also TdT positive is significantly higher than normal [85], but more clinical data are needed to establish this criterion as a useful guide to diagnosis of relapse. The sensitivity of detection of such a technique, unfortunately, is not low enough to allow detection of less than 0.5% -1 % leukemic cells, but may be useful in differ­entiating the marrow lymphocytosis that occurs post discontinuation of chemotherapy from leukemic relapse.

The number of CDlO positive cells in peripheral blood is much lower than in bone marrow, and therefore we have designed protocols to test the hypothesis that residual leukemia cells could be identified in the peripheral blood before the occurrence of clinical relapse. Assays for rare CDlO positive cells in peripheral blood were set up independently in the laboratories of the coauthors (DR-Rochester, NY; 11MvD-Rotterdam, The Netherlands). In the Rotterdam assay, peripheral blood mononuclear cells were stained with anti-CDlO using a rhodamine conjugated second-step antiserum; sub­sequently, cytocentrifuge preparations were made, which were subjected to indirect TdT staining using FITC conjugated second-step reagent. The cyto­centrifuge preparations were evaluated on a fluorescence microscope equipped with phase-contrast facilities and filter combinations for the selective visualization of FITC and rhodamine. At least 50,000 to 100,000 cells (i.e., three cytocentrifuge preparations) were evaluated per determination. All TdT positive cells were evaluated for the expression of CDlO antigen. Only cells expressing both antigens were scored as positive. In the Rochester assay [59], peripheral blood mononuclear cells were stained with a combina­tion of FITC conjugated anti-CD 10 and anti-common leukocyte antibody con­jugated to phycoerythrin (GAP 8.3 PE). A control aliquot of cells is incubated in a mixture of irrelevant mouse IgG2 FITC control and GAP 8.3 PE. For flow cytometry analysis, 1,000,000 cells in both the anti-CDlO and control stained sample are analyzed. A two parameter histogram of log green vs. log red fluorescence gated on forward angle light scatter (FALS) and 90° light scatter is obtained, as shown in figure 4. The gating for FALS excludes red cells, debris, dead cells, and platelets from the fluorescence analysis, because these elements are associated with a low F ALS signal. The gating on 90° light scatter is important to exclude 80% -90% of monocytes and > 99% of residual neutrophils from the fluorescence analysis. Over 90% of peripheral blood mononuclear cells are positive for the common leukocyte antigen, while cells from nearly all patients with CDlO positive ALL are negative. Thus this antibody is a useful parameter for distinguishing mature peripheral blood elements from rare leukemic cells. Cells displaying nonspecific binding of the anti-CDlO or control antibodies (such as residual monocytes) which would be otherwise counted as CDlO positive, are identified by two-color

192

u

..

(~V a:Jkx>){l1a'} UOWUIO:»

a:>ua:>sa~onld paN

Figure 4.

-o b c:: 8

193

immunofluorescence analysis. To increase specificity for ALL cells, only cells with strong CD10 intensity are counted as positive. Cells from 70% of patients with CALL stain brightly enough with anti-CDlO to be detected at a concentration of 0.001 % in the peripheral blood. The sensitivity of detec­tion will be slightly lower for patients with less strongly CDlO positive cells.

In both types of assays, the absolute number of strongly CD10 positive cells in the peripheral blood is calculated by multiplying the percent positive cells by the absolute lymphocyte count, correcting for the percentage of non lymphoid cells in the mononuclear cell preparation.

Samples (n = 237) from 35 patients with ALL at diagnosis (n = 9), during induction therapy (n = 11), during clinical remission (n = 202), prior to clinical relapse (n = 6 from four patients), and at clinical diagnosis of relapse (n = 10) have been studied in Rochester. These data indicate that the absolute number of strongly CDlO positive cells in the peripheral blood during clinical remission is lower than 100 cells per milliliter in > 99% of samples, while nearly all patients at the first diagnosis of relapse showed over 1200 cells per milliliter. The mean number of cells per milliliter is 8.8 in the former group and 11 ,900 in the latter.

Samples (n = 212) from 45 patients in remission (n = 204), prior to clinical relapse (n = 6 from two patients), and at clinical diagnosis of relapse (n = 2) with common ALL were studied in Rotterdam using the double immunofluorescence assay. The absolute number of CDlO+/TdT+ cells in peripheral blood was lower than 400 cells/ml in 99% of samples from patients in clinical remission.

Figure 5 compares the results from common ALL patients studied in Rochester with those studied in Rotterdam. Patients who have remained in remission for at least six months after analysis of a blood sample are compared to patients in whom relapse was diagnosed after analysis of a

Figure 4. Two-color flow cytometry assay for detection of rare COlO positive cells in peripheral blood. All figures are two parameter 64 x 64 channel histograms of cells gated on FALS and 90° light scatter (i.e ., most monocytes have already been gated out using 90° light scatter) . (A) Thawcd aliquot of COW positive cells obtained at diagnosis from a patient with common ALL stained with anti-COlO FlTC and anti-common leukocyte antigen phycoerythrin (PE). Note that 80% of the cells demonstrate strong green fluorescence (anti-COW) and no red fluorescence (anti-common leukocyte). The lower right quadrant indicates the region where residual COW positive leukemic cells are likely to be found. (B) Peripheral blood mononuclear cells from a patient with common ALL in remission stained with anti-COW FITC and anti-common leuko­cyte antigen phycoerythrin (PE). Note that only 8 cells out of 1,000,000 counted are in the lower right quadrant, while 74 cells are in the upper right quadrant (i.e. , high COW fluorescence but also positive for anti-common leukocyte). These cells are residual monocytes which bind the IgG2a anti-COW via an Fc receptor but can be distinguished from ALL cells by binding to anti-common leukocyte antibody . (C) Peripheral blood mononuclear cells from a patient with common ALL in remission stained with FlTC conjugated irrelevant murine IgG2a and anti­common leukocyte antigen phycoerythrin (PE). Note that no cells out of 1,000,000 counted are in the lower right quadrant, indicating very low nonspecific background staining in the region associated with abnormal COlO positive cells.

194

a) 100,000

10,000

1,000

• 100

10

Remission (n = 202)

CDIO (CALLA) Positive Cells per ml (Peripheral Blood)

Rochester Patients b) Rotterdam Patients

•

... 0 •

0

• 0

Prior to Clinical Remission Prior to Relapse Relapse (n = 204) Relapse

•

Clinical Relapse

Figure 5. (A) Absolute number of peripheral blood strongly CDlO+ cells measured by flow cytometry (Rochester assay) in patients with common ALL during clinical remission (sustained for at least nine months after assay), zero to four months prior to clinical diagnosis of relapse , and at the time of clinical relapse. Asterisk indicates a patient with CNS relapse without bone marrow involvement. (B) Absolute CDlO+lTdT+ cells measured by double immunofluore­scence staining (Rotterdam assay) in patients with common ALL during clinical remission (sustained for at least nine months after assay), zero to nine months prior to clinical diagnosis of relapse , and at the time of clinical relapse . Double asterisk indicates a patient with testicular relapse without bone marrow involvement. Absolute number of cells per milliliter was calculated from the percent positive cells, white blood cell count, and the percentage of lymphocytes in the mononuclear cell sample (cells with low 90° light scatter by flow cytometry or CDI5 negative cells in the double immunofluorescence assay).

blood sample. It appears from this data that the major limitation of the flow cytometer assay is the variability from patient to patient in egress of detec­table leukemic cells from the bone marrow into the peripheral blood prior to relapse. The difference between the Rochester and Rotterdam assays in the patient samples prior to relapse suggests that a population of weakly COlO+ /TdT+ leukemic cells may have been detected by the double im­munofluorescence assay but missed by the flow cytometry assay, which detects only strongly COlO+ cells. If the percentage of COW positive cells is plotted rather than absolute number, the discrimination between remission and impending relapse is less.

Figure 6 shows serial studies in two Rochester patients. The patient in figure 6A suffered a clinical relapse five weeks after an elevated number of COW positive cells was observed in the peripheral blood. The second patient (figure 6B) demonstrated an increased number of marrow blasts in the presence of normal numbers of COW positive cells in the peripheral blood.

195

100,000,000 a. 80

10,000,000 70

i 1,000,000 60

100,000 50 I 6 10,000 40 :a 0 1,000 30 ....... 0 ~ S 100 20 0

I:: ] ctS (]J 10 10 ::E -a (]J

'J:: 1 0 r::: (]J 0

e:, -12 -8 -4 0 +4 j:I:l '-'

a 100,000,000 80 CIl .... h. CIl

......... ctS

~ 10,000,000 70 S .... (]J r::: U 1,000,000 60 (]J

+ ~

~ 100,000 50 (]J p...

...l 10,000 40 ~ U

1,000 30

100 20

10 ..... 0 ........ 0···· ........ 10

0 -64 -48 -32 -16 0

Weeks Relative to Clinical Relapse

Figure 6. Follow-up studies of two patients with common ALL by flow cytometric determination of strongly CDlO positive cells in peripheral blood (Rochester assay). Open circles indicate absolute number of CD 10+ cells per milliliter; solid diamonds indicate percent blast cells in bone marrow. Time relative to the clinical diagnosis of relapse (based on bone marrow morphology) is indicated in weeks. (A) Patient with bone marrow relapse diagnosed five weeks after elevated number of CD10+ cells (> 100 cells/m!) identified in peripeheral blood. (8) Patient with normal number of CD 10+ cells (30 cells/ml) in peripheral blood at the time of bone marrow relapse (14% blasts).

T-cell acute lymphoblastic leukemia

Several studies indicate that cells positive for both a T-cell marker and TdT are normally present in the thymus, but not bone marrow or peripheral blood [103, 104]. The T-cell marker +/TdT+ phenotype is expressed by T­ALL and some T-cell non-Hodgkin's lymphomas (T-NHL), especially lym­phoblastic lymphomas. The detection of T-cell marker +/TdT+ cells in

196

extrathymic locations would therefore be a potential means of detecting relapse of T-ALL. A two-color immunofluoresce assay was developed by one of the authors (JJMvD) to test this hypothesis. Mixing experiments with T-NHL showed that the detection limit for these cells in peripheral blood mononuclear cells was at least 1110,000 [lOS]. To investigate whether normal T-cell marker +/TdT + cells occurs in bone marrow or peripheral blood, over SO samples from children and adults were analyzed. CDS+/TdT+ or CD1 +lTdT + cells were not identified in either peripheral blood or bone marrow in any sample. The rarity of T-cell marker +lTdT + cells in the peripheral blood has been confirmed by others [106]. CD7+/TdT+ and CD2+/TdT+ cells were detected in small numbers « 0.3%, usually 0.002%-0.OS%) in pediatric marrow samples. In adult marrow or regenerat­ing childhood marrow, the frequency of these cells was about five- to ten fold lower. These cells probably represent prothymocytes [107]; certain T­ALL cells also possess a similar immature prothymocyte phenotype [79]. CD7+/TdT+ and CD2+/TdT+ cells were very rare « 0.02%, usually 0.001 %-O.OlS%) in peripheral blood. Therefore the detection limit for CD7+/TdT+ or CD2+/TdT+ T-ALL cells is about 0.4% in bone marrow and 0.03% in peripheral blood, while the detection of CDS+/TdT+ and CD1+lTdT+ T-ALL cells is limited only by the number of cells counted in the assay [8S].

To prevent underdetection of malignant cells due to phenotypic shift, which has been reported in T-ALL [108], several different T-cell markers (CDl, CDS, CD7, CD3, CD4, or CD8, depending on the immunological phenotype of the malignancy at diagnosis) are tested in parallel for the T­cell marker/TdT two-color immunofluorescence assay. The peripheral blood and bone marrow of 12 children with a TdT positive T-cell malignancy have been monitored, including nine children with T-ALL and three with T­NHL. Three T-ALL patients have had a total of four bone marrow relapses. Using the two-color immunofluorescence staining, it was possible to detect abnormal cells in the peripheral blood two to three months prior to morpho­logical leukemic relapse. In the remaining nine patients, T-cell marker/TdT positive cells were below the threshold of detection in both peripheral blood and bone marrow in serial determinations [8S].

Serial studies on a patient with T-ALL are summarized in figure 7. After six weeks of induction therapy this patient was in remission according to morphological criteria « S% blasts in the bone marrow and no blasts in the peripheral blood). However, at that time, CDS+/TdT+ cells could be de­tected in both bone marrow and peripheral blood. The percentage of CDS+lTdT+ peripheral blood cells remained elevated on four follow-up determinations, and relapse occurred three months later. The patient re­ceived reinduction therapy and went into remission, determined by both morphology and double immunofluorescence staining. After ten weeks, CDS+lTdT+ cells were again detectable in peripheral blood, followed by relapse two months later.

Figure 7. Follow-up of a T-ALL patient by morphological techniques and CDSITdT double immunofluorescence staining (Rotterdam assay). (A) Percent CDS+lTdT + cells (solid squares) compared with percent blast cells (open squares) in peripheral blood. (8) Percent CDS+ ITdT+ cells in peripheral blood (solid squares) compared with percent CDS+lTdT+ cells in bone marrow (solid circles). Arrow with asterisk indicates start of induction therapy; arrow with triangle indicates start of reinduction therapy at time of clinical relapse.

The data from this patient and others suggest that detection of a very low percentage of phenotypically abnormal cells (about 0.01 %) in peripheral blood occurs two to three months prior to leukemic relapse in T-ALL. The increase in percentage of CD5+/TdT+ cells in peripheral blood parallels that in the bone marrow, suggesting rather free traffic of T-ALL cells from the bone marrow into the peripheral blood. This is consistent with the high likelihood of clinical involvement of peripheral tissues in T-ALL [109]. In a similar fashion, Bradstock and Kerr found occult involvement of peripheral blood at diagnosis and during remission in two patients with T-ALL using a double immunofluorescence assay [110].

These results demonstrate that the T-cell markerlTdT double immunoflu­orescence staining can be used during follow-up of patients with a TdT

198

positive T-cell malignancy to: 1) determine whether remiSSIOn has been obtained, 2) detect early relapse, and 3) rule out relapse when an increased number of morphologically suspicious, but T-cell marker/TdT negative lym­phoid cells are present.

The T-cell marker/TdT double immunofluorescence staining assay can also be used for staging of TdT positive T-NHL at diagnosis. We analyzed bone marrow and peripheral blood samples at diagnosis of three T-NHL patients who had stage lor II disease . Abnormal cells were identified at low frequencies in peripheral blood and bone marrow from all three patients. The data are consistent with the tendency of lymphoblastic lymphoma to dis­seminate and are similar to data collected in nodular B-cell lymphoma using comparison of kappa and lambda fluorescence intensity of peripheral blood cells [12].

Another potential use for T-cell markerlTdT double immunofluorescence staining is to detect residual leukemic T-cells in bone marrow obtained during remission for autologous marrow transplantation.

Nonlymphoid leukemia

The major limitation in application of techniques similar to those described above to nonlymphoid leukemia is the lack of leukemia specific antibody reagents. Early detection of relapse in AML using an antiserum to leukemia cells has been reported [111] but not confirmed. Our understanding of hematopoietic differentiation has advanced significantly with the recent identification of differentiation and lineage related antigens [112], but leuke­mia specificity has not been demonstrated for any of these markers. Multiple parameter analysis may be helpful in this regard to establish subtle differ­ences between leukemia cells and normal progenitors. Delwel et a1. have demonstrated differences in surface phenotype between 14117 cases of AML and normal CFU-GM using four immunologic reagents, CD34 (MYlO), UEA lectin, Vim-2, and RFB-1 [113]. Two-color flow cytometry assays using these antibodies may be useful in selectively obtaining AML clonogenic cells for in vitro culture [114] .

Present applications and future directions for detection of residual disease in acute leukemia

Current status of detection of residual leukemia by monoclonal antibodies

Sensitivity of assays. In both common ALL and T-ALL, no absolutely leukemia specific monoclonal antibodies exist. The strategy for detection of residual leukemia must therefore be either to identify abnormal cells outside the usual homing area of their normal counterparts (bone marrow T-cell antigen + /TdT + cells in T-ALL) or to identify increased numbers of cells

199

of a specific rare phenotype (peripheral blood COlO+/TdT + cells in com­mon ALL). Its access ability for frequent sampling and availability of cells in a single cell suspension make peripheral blood the logical location for study. Currently available assays show the most promise in T-ALL. The results so far suggest that T-ALL cells are more likely than common ALL cells to gain access to the peripheral blood from the bone marrow, so much so that the percentage of phenotypically abnormal T-ALL cells in the peripheral blood prior to overt relapse is very similar to that in the bone marrow itself.

The major problem in detection of residual disease in common ALL is the variable egress of leukemic cells from the bone marrow into peripheral blood. In one patient (figure 6B), less than 0.001 % strongly COlO positive cells were identifiable in the peripheral blood at a time when the bone marrow contained 14% lymphoblasts. In another patient (figure 6A), the peripheral blood contained 0.2% strongly COW positive cells, while 7% blasts were observed in the bone marrow. These differences were not due to COW negativity of the blasts, since at relapse nearly all marrow cells were COW positive. Smith et al. describe a similar problem in detection of periph­eral blood TdT + cells in common ALL. In three relapses occurring in two patients, samples of peripheral blood obtained six, five, and one week prior to relapse contained less than 0.11 % TdT + cells. At relapse 41 %, 95%, and 18% peripheral blood TdT + cells, respectively, were present [89].

An additional problem is the occurrence of leukemic relapse at sites other than bone marrow, such as the central nervous system (CNS) or testis. In one common ALL patient studied with the flow cytometry assay, CNS relapse with CD 10 positive cells occurred without the appearance of pheno­typically abnormal cells in the peripheral blood or morphologic bone mar­row involvement.

Clinical usefulness of assays. The final criterion for a successful assay for detection of residual disease in acute leukemia is the availability of effective treatment for patients in whom residual disease is detected. Treatment of the patient with ALL who relapses is a difficult problem, especially for the patient who relapses while on maintenance chemotherapy. Long-term sur­vival of the latter is less than 15% with present therapy. The degree to which reinduction for relapsed ALL would be successful if relapse could be diagnosed earlier is unknown and will remain undetermined until an effective assay for residual disease can be implemented in a clinical trial. There is no evidence for improved long-term survival of patients in whom relapse was detected (presumably earlier) by routine marrow aspiration in the presence of normal peripheral blood counts and those in whom relapse was clinically suspected (presumably more advanced) [115], although the latter group had shorter median duration of second remission in one study [116]. There are likely to be biological differences between these two groups of patients that make this data rather difficult to interpret in terms of detection of residual

200

disease. In addition, relapse detected either clinically or from morphologic examination of bone marrow is rather late in comparison to possibilites for early detection of relapse using monoclonal antibodies. It is important to keep in mind that the benefit of early relapse detection is related to the critical number of drug resistant clonogenic cells present at the time of detection, not the time difference between early detection of relapse and clinical manifestation of relapse. As noted earlier, tumor cell growth rates tend to slow with increasing tumor bulk. Monoclonal antibody detection of relapse three weeks before morphologic bone marrow relapse may be effec­tive if these three weeks represent the difference between 109 and 1011 total cells, as in figure 1. In contrast, whether one begins therapy at morphologic marrow relapse (10" cells) or symptomatic clinical relapse (10 12 cells) five weeks later may mean little, since many drug resistant cells may already be present when the leukemia burden is lO" total cells.

Future directions for detection of residual disease in acute leukemia using immunological markers

Detection of residual disease in acute leukemia using immunological markers is limited chiefly by the biological nonspecificity of currently available probes for malignancy, rather than technical limitations such as analysis speed or availability of sufficient cells for analysis. It is already possible to modify commercial flow cytometers to function at speeds (> 100,000 cells per second [62]) high enough to analyze all the mononuclear cells that can be obtained in one liter of blood or a very cellular bone marrow aspirate. Detection of abnormal cells outside their normal homing area (i.e., T-ALL cells in peripheral blood) does not require absolute leukemia specificity of the reagents, only that they be specific in that tissue for the leukemia­associated phenotype. The major limitation of this approach is that detection of leukemia cells outside their tissue of origin may not be sensitive enough to permit early diagnosis of relapse in certain leukemias, which do not as readily disseminate outside the bone marrow.

How can the specificity of monoclonal antibody reagents by improved? One approach is to identify aberrant proteins, such as those associated with bcr-abl hybrid mRNA transcripts in Phi positive leukemias. Further study of translocation sites in acute leukemia may identify further examples of ab­normal proteins. Another approach is a finer delineation of normal hemato­poietic cell maturation to identify markers or combinations of markers common in leukemia but rare in normal cells. For example, the cytoplasmic immunoglobulin (CIg)+/TdT+ phenotype is very rare in normal marrow [100], occurring in only one of every 400 CDlO+ cells, but present in pre-B­ALL. Aberrant differentiation can be observed in ALL as compared to normal marrow CDIO positive cells. TdT expression is tightly controlled in normal B-cell progenitors, so that CD20+/TdT+, CIg+/TdT+, or common leukocyte antigen + ITdT + cells are infrequent. TdT expression in common

201

ALL is unrelated at expression of these other antigens, resulting in aberrant phenotypes that may be useful in differentiating leukemic from normal early B-cell phenotypes. [120]

It is possible to perform three- and four-color immunofluorescence mea­surements on individual 'cells using new fluorochromes and two-laser flow cytometers [117], so that subtle differences in patterns of antigen expression between normal and leukemic cells can be used to detect aberrant cells. Several potential problems are associated with this multiparameter approach to residual leukemia detection. The first of these is that the more specific and detailed the description of the leukemia cell, the fewer patients any particular description will fit. The ultimate example of this is the tailor-made antibodies required for anti-idiotype therapy of surface Ig positive lym­phomas. An associated risk is that phenotypic shifts, which occur at relapse in a significant proportion of patients with ALL, will be more commonly observed the finer and more detailed the leukemic phenotypic description becomes. If the combination of markers used for detection has a functional significance for cell survival or proliferation, it is possible that phenotypic shifts will be less likely to occur.

Note added in proof: Using the polymerase chain reaction, cells carrying a t (14; 18) hybrid DNA sequence in a patient with follicular lymphoma could be detected in a 100,000 dilution. [121] This technique is very prom­ising, although it requires detailed information regarding the exact location of the translocation .

References

1. Goldie J , Coldman A: Genetic instability in the development of drug resistance. Semin Oneol 12:222-230, 1985.

2. Goldie JH, Coldman AJ: A mathematical model for relating the drug sensitivity of tumors to their spontaneous mutation rate. Cancer Treat Rep 63: 1727 -1733, 1979.

3. Dembo A: Spontaneous mutation to chemotherapy resistance: Implications of the Goldie­Coldman model for the management of ovarian cancer. J. Clin Oncol 2: 1311-1315, 1984.

4. Skipper H Schabel F: Quantitative and cytokinetic studies in experimental tumor systems. In : Holland J , Frei E (eds) : Cancer Medicine, Philadelphia, Lea and Febiger, 1982, pp 663-685.

5. Jaffe N, Frei E, Traggis D, Bishop Y: Adjuvant methotrexate and citrovorum-factor treatment of osteogenic sarcoma. N Engl J Med 291:994, 1974.

6. Bonadonna G, Brusamolino E, Balagussa P, Rossi A, Brugnatelli L, Brambilla C, De Luna M, Tancini G , Bajetta E , Museumeci R, Veronesi U: Combination chemotherapy as an adjuvant treatment in operable breast cancer. N Engl J Med 294:405 , 1976.

7. Kartner N, Riordan J , Ling V: Cell surface P-glycoprotein associated with multi drug resistance in mammalian cell lines. Science 221:1285-1288, 1983.

8. Ma D, Scurr R, Davey R, Mackertich S, Harman D, Dowden G, Isbister J, Bell D: Detection of a multidrug resistant phenotype in acute non-lymphoblastic leukemia. Lancet: 1:135-137, 1987.

9. Goldie JH, Coldman AJ: Quantitative model for multiple levels of drug resistance in clinical tumors. Cancer Treat Rep 67:923-931, 1983.

202

10. Seshadri R, Kutlaca R, Trainor K, Matthews C, Morley A: Mutation rate of normal and malignant human lymphocytes. Cancer Res 47:407-409, 1987.

11. Redding W, Coombes R, Monaghan P, Clink H, Imrie S, Dearnaley D, Ormerod M, Sloane J, Gazet J, Powles T, Neville A: Detection of micrometastases in patients with primary breast cancer. Lancet 2:1271-1273,1983.

12. Ault K: Detection of small numbers of monoclonal B lymphocytes in the blood of patients with lymphoma. N Engl 1 Med 300:1401-5, 1979.

13. Morgan E, Hsu C: Relationship of a leukemia-associated-antigen to the presence of lymphoblasts in the peripheral blood in children with acute lymphocytic leukemia. Blood 57:879-882, 1981.

14. Cranendonk E, Abeling N, Bakker A, de long M, van Gennip A, Behrendt H: Evalua­tion of the use of the hemalog D in acute lymphoblastic leukemia and disseminated non­Hodgkin's lymphoma in children. Acta Haematol 71:18-24, 1984.

15. Borella L, Green A Webster R: Immunolgic rebound after cessation of long-term che­motherapy in acute leukemia. Blood 40:42-51, 1972.

16. Cohn S, Lincoln S, Rosen S: Present status of serum tumor markers in diagnosis, prognosis, and evaluation of therapy. Cancer Invest 4:305-327, 1986.

17. Sullivan P, Salmon S: Kinetics of tumor growth of regression in IgG multiple myeloma. J Clin Invest 51: 1697-1708, 1972.

18. Deegan M, Abraham l, Sawdyk M, van Slyck E: High incidence of monoclonal proteins in the serum and urine of chronic lymphocytic leukemia patients. Blood 64:1207-1211, 1984.

19. Kornberg A, Polliack A: Serum lactic dehydrogenase (LDH) levels in acute leukemia: Marked elevations in lymphoblastic leukemia. Blood 56:351-355, 1980.

20. Herlyn M, Lange B, Bennicelli l, Blaszczyk M, Guerry D, Koprowski H: Increased levels of circulating HLA-DR antigen in sera of patients with acute lymphoblastoid leukemia. Leuk Res 8:323-334, 1984.

21. Carrel S, Buchegger F, Heumann D, Girardet C, Barras C, Losa G, Mach J, von Fliedner V: Detection of the common acute lymphoblastic leukemia antigen in the serum of leukemia patients. J Clin Invest 74:1882-1885, 1984.

22. Coleman M, Cibull M, Manderino G. A new solid-phase immunoassay for terminal deoxynucleotidyl transferase: Analysis of TdT antigen in Cells, plasma and serum. Blood 65:41, 1985.

23. Osserman E, Lawlor D: Serum and urinary lysozyme (muramidase) in monocytic and monomyelogenous leukemia. 1 Exp Med 124:921, 1966.

24. Catovsky D, Galton D, Griffin C: The significance of lysozyme estimations in acute myeloid and chronic monocytic leukaemia. Br J Haematol 21:565, 1971.

25. Prockop D, Davidson W: A study of urinary and serum lysozyme in patients with renal disease. N Engl J Med 240:269, 1964.

26. Moss T, Kindler A, Marangos P, Rajewsky M, Reynolds C, Seeger R: Immunohistologic detection and phenotyping of neuroblastoma cells in bone marrow using cytoplasmic neuron specific enolase (NSE) and cell surface antigens (CSA). In: Evans A, D'Angio G, Seeger R, (eds):' Advances in Neuroblastoma Research, New York, Alan Liss, 1985.

27. Barlogie B, Drewinko B, Schumann J, Bohde W, Dosik G, Johnston D, Frcirich: Cellular DNA content as a marker of neoplasia in man. Am J Med 69:195-203, 1980.

28. Braylan R: Attributes and applications of flow cytometry. Ann Clin Lab Sci 13:379-384, 1983.

29. Clevenger C, Bauer K, Epstein A: A method for simultaneous nuclear immunofluore­scence and DNA content quantitation using monoclonal antibodies and flow cytometry. Cytometry 6:208-214, 1985.

30. Look A, Roberson P, Williams D, Rivera G, Bowman W, Pui C, Abromowitch M, Kalwinsky D, Dahl G, George S, Murphy S: Prognostic importance of blast cell DNA content in childhood acute lymphoblastic leukemia. Blood 56:1079, 1985.

31. Yunis J, Bloomfield C, Ensurd K: All patients with acute non lymphocytic leukemia may

203

have a chromosomal defect. N Engl J Med. 305:135, 1981. 32. Michaeli J, Lerer I, Rachmilewitz E, Fibach E: Stimulation of proliferation of human

myeloid leukemia cells by a specilic growth factor: Applications for cytogenetic analysis (abstract). Leuk Rcs \0: 106, 1986.

33. Gray J, Langlois R: Chromosome classification and purification using flow cytometry and sorting. Ann Rev Biophys Chern 15: 195-235, 1986.

34. Langlois R, Yu L-C, Gray J, ct al: Quantitative karyotyping of human chromosomes by duel beam flow cytometry. Science 220:620, 1983.

35. Hagenbeek A, Martens A, Arkesteijn G. Dual laser bcam flow karyotyping as a new diagnostic tool in Leukemia (abstract). Blood 68:257a, 1986.

36. Track B, van dcn Engh G, Landegent J, in de Wal N, van der Ploeg M: Detection of DNA sequences in nuclei in suspension by in situ hybridization and dual beam flow cytometry. Science 230: 140 I - 1403, 1985.

37. Weiss L, Wood G, Trela M, Warnke R, Sklar J: Clonal T-cell populations in lymphomatoid papulosis. N Engl J Med 315:475-479, 1986.

38. Arnold A, Cossman, J, Bakhshi A, Jaffe E, Waldmann 1', Korsmeyer S: Immunoglo­bulin-gene rearrangements as unique clonal markers in human lymphoid neoplasms. N Engl J Med 309: 1593- 1599, 1983.

39. Zehnbauer BA, Pardoll DM, Burke PJ, Graham ML, Vogelstein B: Immunoglobulin Gene rearrangements in remission bone marrow specimens from patients with acute lymphoblastic leukemia (abstract). Blood 67:835-838, 1986.

40. Bunin N, Mirro J, Raimondi S, Behm F, Murphy SB, Kitchingman G: Gene rearrange­ments as clonal markers at diagnosis and relapse in chilhood acute leukemia (abstract). Blood 68: 196a, 1986.

41. Bregni M, Bonadonna G, et al: Minimal residual disease in remission bone marrow from acute lymphoblastic leukemia (ALL) patients detected by a highly sensitive and specific assay (abstract). Blood 68: 195a, 1986.

42. Collins S, KUDonishi I, Miyoshi I, Groudine M: Altered transcription of thc c-abl oncogene in K-562 and other chronic myelogenous leukemia cells. Science 225:72-74, 1984.

43. Konopka J, Watanabe S, Singcr J, Collins, S, Witte 0: Cell lines and clinical isolates derived from Ph I-positive chronic myelogenous leukemia patients express c-abl proteins with a common structural alteration. Proc Nat! Acad Sci USA 82: 1810-1814, 1985.

44. Grosveld G, Bootsma D, de Klein A, Heisterkamp N, Starn K, Groffen J: Specific Breakpoint analysis: Potentials for detection of minimal residual disease (abstract). Leuk Rcs 10:80, 1986.

45. Reddy F, Reynolds R, Santos R, Barbacid M: A point mutation is responsible for the acquisition of transforming properties by the 1'24 human Bladder Carcinoma oncogene. Nature 300: 149- 152, 1982.

46. Busch M, Beckstead J, Gantz D, Vyas G: Detection of human immunodeficiency virus infection of myeloid precursors in bone marrow samples from AIDs patients (abstract). Blood 68 (Supp 1):122a, 1986.

47. Benjamin D, Magrath I, Douglass E, Corash L: Derivation of lymphoma cell lines from Microscopically normal bone marrow in patients with undifferentiated lymphomas: Evi­dence of occult bone marrow involvement. Blood 61:I017-JOI9, 1983.

48. Smith S, Wood G, Fried P, Lowman J: In vitro growth of lymphoma colonies from children with non-Hodgkin's lymphoma. Cancer 48:2612-2623, 1981.

49. Estrov Z, Grunberger 1', Dube I, Wang Y, Freedman M: Detection of residual acute lymphoblastic leukemia cells in cultures of bone mmrow obtained during remission. N Engl J Med 315:538, 1986.

50. Consolini R, Breard J, Bourinbaiar A, Goutner A, Georgoulias V, Canon C, Brugerie E, Mathe G: Abnormal-In vitro differentiation of peripheral blood clonogenic B cells in common acute lymphoblastic leukemia during complete remission. Blood 67:796-8()1, 1986.

51. Fearon E, Burke P, Schiffer C, Zehnbauer B, Vogelstein B: Differentiation of leukemia

204

cells to polymorphonuclear leukocytes in patients with acute non lymphocytic leukemia. N Engl J Med 315:15-24, 1986.

52. Dantas M, Brown J, Thomas M, Robinson W, Glode L: Detection of melanoma cells in bone marrow using monoclonal antibodies. Cancer 52:949-953, 1983.

53. Anderson C, Looney J: Human leukocyte IgG Fc receptors. Tmmunol Today 7:264-266, 1986.

54. Berkow R, Dodson R: Purification and functional evaluation of mature neutrophils from human bone marrow. Blood 68:853-860, 1986.

55. Iwatani Y, Amino N, Mori J, Asari S, Ina K, Ennyu K, Miyai K: Effects of various isolation methods for human peripheral lymphocytes on T cell subsets determined in a fluorescence activated cell sorter (FACS) and demonstration of a sex difference of sup­pressor/cytotoxic T cells. J Immunol Methods 54:31, 1982.

56. Berenson RJ, Levitt LJ, Levy R, Miller RA: Cellular immunoabsorption using monoclonal antibodies. Transplantation 38: 136, 1984.

57. Greenberg P, Baker S, Link M, Minowada J: Immunologic selection of hematopoietic precursor cells utilizing antibody-mediated plate binding ("panning"). Blood 65: 190-197, 1985.

58. Ritchie A, Gray R, Micklem H: Right angle light scatter: A necessary parameter in flow cytofluorometric analysis of human peripheral blood mononuclear cells. J Immunol Meth 64:109-117,1983.

59. Ryan D, Mitchell S, Hennessy L, Bauer K, Horan P, Cohen H: Improved detection of rare CALLA-Positive cells in peripheral blood using multiparameter flow cytometry. J Immunol Meth 74:115, 1984.

60. van Dongen JJ, Hooijkaas H, Hahlen K, Benne K, Bitter W, van der Linde-Preesman A, Tetter I, van de Rijn M, Hilgers J, van Zanen G, Hagemeijer A: Detection of minimal residual disease in TdT positive T cell malignancies by double immunofluorescence stain­ing. In: Hagenbeek A, Lowenberg B (eds): Minimal Residual Disease in Acute Leukemia, The Hague, Martinus Nijhoff Publishers, 1984, pp 67-81.

61. Cupp JE, Leary JF, Cernichiari E, Wood JC, Doherty RA: Rare-event analysis methods for detection of fetal red blood cells in maternal blood. Cytometry 5: 138-144, 1984.

62. Peters D, Branscomb E, Dean P, Merrill T, Pinkel D, van Dilla M, Gray J: The LLNL high-speed sorter: Design features, operational characteristics, and biological utility. Cyto­metry 6:290-301, 1985.

63. Pui C, Raimondi S, Behm F, Ochs J, Furman W, Bunin N, Ribeiro R, Tinsley P, Mirro J: Shifts in blast cell phenotype and karyotype at relapse of childhood lymphoblastic leuke­mia. Blood 68:1306-1310,1986.

64. Greaves M, Paxton A, Janossy G, Pain C, Jobnson S, Lister T: Acute lymphoblastic leukemia associated antigen. III. Alterations in expression during treatment and in relapse. Leuk Res 4:1-14,1980.

65. Pui C, Raimondi S, Behm F, Ochs J, Furman W, Bunin N, Ribeiro R, Tinsley P, Mirro J: Shifts in blast cell phenotype and karyotype at relapse of childhood lymphoblastic leuke­mia. Blood 68: 1306-1310, 1986.

66. Hann I, Jones P, Evans D: Discrepancy of bone marrow aspirations in acute lymphoblas­tic leukemia in relapse (letter). Lancet 1: 1215-1216, 1977.

67. Hagenbeek A, Martens A: An immunological approach to analyze the kinetics of minimal residual disease in acute leukemia. In: Hagenbeek A, Lowenberg B (eds): Minimal Residual Disease in Acute Leukemia 1986, Dordrecht, Martinus Nijhoff Publishers, 1986, pp 76-85.

68. Krolick K, Uhr J, Vitetta E: Selective killing of leukaemia cells by antibody-toxin con­jugates: Implications for autologous bone marrow transplantation. Nature 295:604-605, 1982.

69. Martens A, Johnson R, Kaizer H, Hagenbeek A: Characteristics of a monoclonal anti­body (RMI24) against acute myelocytic leukemia cells. Exp Hematol 12:667-671, 1984.

70. Hagenbeek A, Martens A: Detection of minimal residual disease in acute leukemia:

205

Possibilities and limitations. Eur J Cancer Clin Oncol 21:389-395, 1985. 7l. Schulz F , Martens A, Hagenbeek A: Growth kinetics of minimal residual disease in the

brown Norway rate acute myelocytic leukemia. In: Hagenbeek A and Lowenberg B (eds). Minimal Residual Disease in Acute Leukemia 1986, Dordrecht, Martinus Nijhoff Publishers, 1986, pp 97-111.

72. Greaves M, Brown G, Rapson N, Lister T: Antisera to acute lymphoblastic leukemia cells. Clin Immunol Immunopathol 4:67-84, 1975.

73. Ritz J, Pesando J, Notis-McConarty J, Lazarus H, Schlossman S: A monoclonal antibody to human acute lymphoblastic leukemia antigen. Nature 283:583-585, 1980.

74. Reinherz E, Haynes B, Nadler L, et al (eds): Leucocyte Typing, Vols 1-3. New York, Springer-Verlag, 1986.

75. Korsmeyer S, Arnold A, Bakhshi A, Ravetch J, Sieben list U, Hieter P, Sharrow S, LeBien T, Kersey J, Poplack D, Leder P, Waldman T. Immunoglobulin gene rearrange­ment and cell surface antigen expression in acute lymphocytic leukemias of T cell and B cell precursor origins. J Clin Invest 71:301-313, 1983.

76. Minden M, Toyonaga B, Ha K, Yanagi Y, Chin B, Gelfand E, Mak T: Somatic re­arrangement of T-cell antigen receptor genc in human T-cell malignancies. Proc Natl Acad Sci USA 82:1224-1227, 1985.

77. Tawa A, Hozumi N, Minden M, Mak T, Gelfand E. Rearrangement of the T-cell receptor ~-chain gene in non-T-cell, non-B-cell acute lymphoblastic leukemia of childhood. N Engl J Med 313:1033-1037,1985.

78. Kitchingman G, Rovigatti U Mauer A, Melvin S, Murphy S, Stass S: Rearrangement of immunoglobulin heavy chain genes in T cell acute lymphoblastic leukemia. Blood 65:725-729, 1985.

79. van Dongen JJ, Quertermous T, Bartram C, Gold D, Wolvers-Terrero I, Comans-Bitter W, Hooijkaas H, Adriaansen H, De Klein A, Raghavachar A, Ganser A, Duby A, Seidman J , van den Elsen P , Tcrhorst C: T cell receptor-CD3 complex during early T cell differentiation. J Immunol 138:1260-1269, 1987.

80. Cook W, Balaton A: T-cell receptor and immunoglobulin genes are rearranged together in abelson virus-transformed pre-B and pre-T cells. Mol Cell BioI 7:266-272, 1987.

81. Baltimore D: Is terminal deoxynucleotidyl transferase a somatic mutagen in lymphocytes? Nature 248:409, 1974.

82. Desiderio S, Yancopoulos G, Paskind M, Thomas E, Boss M, Landau N, Alt F, Balti­more D: Insertion of N regions into heavy-chains genes is correlated with expression of terminal deoxytransferase in B cells. Nature 3131:752-755, 1984.

83. Kunkel T, Gopinathan K, Dube D, Snow E, Loeb L: Rearrangements of DNA mediated by terminal transferase. Proc Natl Acad Sci USA 83:1867-1871, 1986.

84. Brashem C, Kersey J, Bollum F, LeBien T: Ontogenic studies of lymphoid progenitor cells in human bone marrow. Exp Hematol 10:886, 1982.

85. van Dongen JJ, Hooijkaas H , Adriaansen H, Hahlen K, van Zanen G: Detection of minimal residual acute lymphoblastic leukemia by immunological marker analysis: Possi­bilities and limitations. In: Hagenbeek A, Lowenberg B (eds): Minimal Residual Disease in Acute Leukemia 1986, Dordrecbt, Martinus Nijhoff Publishers, 1986, pp 113-133.

86. Muehleck S, McKenna R, Gale P, Brunning R: Terminal deoxynucleotidyl transferase (TdT)-positive cells in bone marrow in the absence of hematologic malignancy. Am J Clin Pathol 79:277-284, 1983.

87. Stass S, McGraw T, Folds J , OdIe B , Bollum F: Terminal transferase in acute lymphoblas­tic leukemia in remission. Am J Clin Pathol 75:838-840, 1981.

88. Froehlich T, Buchanan G, Cornet J, Sartain P, Smith R: Terminal deoxynucleotidyl transferase-containing cells in pcripheral blood: Implications for the surveillance of patients with lymphoblastic leukemia or lymphoma in remission. Blood 58:214-220, 1981.

89. Smith R, Hetherington M, Huntsman P, Buchanan G: Surveillance of terminal-deoxynu­c1eotidyl transferase-positive cells in peripheral blood of patients with acute lymphoblastic leukemia (ALL). In: Hagenbeek A and Lowenberg B (eds): Minimal Residual Disease in

206

Acute Leukemia 1986, Dordrecht, Martinus Nijhoff Publishers, 1986, pp 134-140. 90. Barr R, Koekebakker M: Dctcction of circulating "terminal transferase-positive" cells

does not predict relapse in acute lymphoblastic leukemia. Leuk Res 8:1051-1055,1984. 91. Fletcher J, Bell R, Koekebakker M, Dowers A, McCaffrey R, Barr R: Spurious detection

of terminal deoxynucleotidyl transferase in phytohemagglutinin-stimulated lymphocytes. Blood 68:310-312, 1986.

92. Coleman M, Cibull M, Manderino G: A new solid-phase immunoassay for terminal deoxynucleotidyl Transferase: Analysis of TdT antigen in cells, plasma, and serum. Blood 65:41-45, 1985.

93. Ryan D, Chapple C, Cohen H: Subtypes of CALLA positive acute lymphoblastic leukemia (cALL) compared with phenotypic subtypes of normal CALLA positive marrow early B cells (abstract). Blood 68:251a, 1986.

94. Greaves M, Delia D, Janossy G, Rapson, N, Chessells J, Woods M, Prentice G: Acute lymphoblastic leukaemia associated antigen. IV. Expression on non-leukaemic lymphoid cells. Leukemia Research 4: 15, 1980.

95. Hoffman-Fezer G, Knapp W, Thierfelder S: Anatomical distribution of CALL antigen expressing cells in normal lymphatic tissue and in lymphomas. Leukemia Research 6:761, 1982.

96. Cossman J, Neckers LM, Leonard WJ, Greene We: Polymorphonuclear neutrophils express the common acute lymphoblastic leukemia antigen. J Exper Med 157:1064, 1983.

97. Platt JL, LeBien TW, Michael AF: Stages of renal ontogenesis identified by monoclonal antibodies reactive with Iymphohemopoietic differentiation antigens. J Exp Med 157: 155, 1983.

98. Metzgar R, Borowitz M, Jones N, Dowell B: Distribution of common acute lymphoblastic leukemia antigen in non hematopoietic tissues. J Exp Med 154: 1249-1254, 1984.

99. Neudorf S, LeBien T, Kersey J: Characterization of thymocytes expressing the common acute lymphoblastic leukemia antigen. Leuk Res 8:173, 1984.

lOll. Ryan D, Kossover S, Mitchell S, Frantz C, Hennessy L, Cohen H: Subpopulations of common acute lymphoblastic leukemia antigen-positive lymphoid cells in normal bone marrow identified by hematopoietic differentiation antigens. Blood 68:417-425, 1986.

101. Borella L, Green A, Webster R: Immunologic rebound after cessation of long-term che­motherapy in acute leukemia. Blood 40:42-51, 1972.

102. Bakkeren J, de Vaan G: Persistant presence of common ALL-antigen (cALLa)-positivc cells in regenerating bone marrow after cessation of maintenance therapy in ALL of childhood. Eur Paediatr Haematol Oncol 2:217-219, 1985.

103. Bradstock K, Janossy G, Tidman N, Papageorgiou E, Prentice H, Willoughby M, Hoff­brand A: Immunological monitoring of residual disease in treated thymic acute lympho­blastic leukemia. Leuk Res 5:301-309, 1981.

104. Janossy G, Bollum F, Bradstock K, McMichael A, Rapson N, Greaves M: Terminal transferase positive human bonc marrow cells exhibit the antigen tic phenotype of non-T, non-B acutc lymphoblastic leukemia. J Immunol 123:1525, 1979.

105. van Dongen J, Hooijkaas H, Hahlen K, Benne K, Bitter W, van der Linde-Preesman A, Tettcro I, van de Rijn M, Hilgers J, van Zanen G, Hagemeijcr A: Detection of minimal rcsidual disease in TdT positive T cell malignancies by double immunofluorescence stain­ing. In: Lowenberg B, Hagenbeek A (cds): Minimal Residual Disease in Acute Leukemia, Martinus Nijhoff Publishers, Boston, 1984, p 67-81.

106. Bradstock K, Kerr A: Phenotype of TdT-positive cells in normal peripheral blood: Implications for monitoring of ALL (abstract). Leuk Res 10:100, 1986.

107. van Dongen .I, Hooijkaas H, Comans-Bitter M, Hahlen K, de Klein A, van Zanen G van't Veer M, Abels J, Benner R: Human bone marrow cells positive for terminal deoxynucleotidyl transferase (TdT), HLA-DR, and a T cell marker may represent pro thy­mocytes. J lmmunol 135:3144-3150, 1985.

108. Bernard A, Raynal B, Lemerle J, Boumsell L: Changes in surface antigens on malignant T cells from lymphoblastic lymphomas at relapse: An appraisal with monoclonal antibodies

207

and microfluorimetry. Blood 59:809-815, 1982. 109. Bloomfield C, Gajl-Pcczalska K: The clinical relevance of lymphocyte surface markers in

leukemia and lymphoma. Curr Top Hematol 3: 175-240, 1980. 110. Bradstock K, Kerr A: Immunolgical detection of covert leukaemic spread in mediastinal

T-ccll lymphoblastic lymphoma, Leuk Res 9:905-911, 1985. III. Baker M, Falk J, Carter W, Taub R, thc Toronto Leukemia Study Group: Early diagnosis

of relapse in acute myeloblastic leukemia. New Engl J Med 301:1353-1357, 1979. 112. Foon K, Gale R, Todd, R: Recent advances in the immunologic classification of leukemia.

Semin Hematol 4:257-283, 1986. 113. Delwel R, Touw 1, Lowerberg B: Towards detection of minimal disease: Discrimination

of AML Precursors from normal myeloid precursors using a combination of surface markers. In: Hagenbeek A, Lowenberg B (cds): Minimal Residual Disease in Acute Leukemia 1986, Dordrecht, Martinus Nijhoff Publishers, 1986, pp 68-75.

114. Delwcl R, Lowenberg B: Isolation of acute myeloid leukemia precursors from normal myeloid precursors (abstract). Exp Hematol 6:471, 1986.

115. Haworth C, Heppleston A, Jones P, Campbell R, Evans D, Palmer M: Routine bone marrow examination in the management of acute lymphoblastic leukaemia of childhood. J Clin Pathol 34:483-485, 1981.

116. Rogers P, Bleyer W, Coccia P, Lukens J, Siegel S, Sather H, Hammond D: Yield of unpredicted bone-marrow relapse diagnosed by routine marrow aspiration in children with acute lymphoblastic leukaemia. Lancet 1: 1320-1322, 1984.

117. Lanier L, Loken M: Human lymphocyte subpopulations identified by using three-color immunofluorescence and flow cytometry analysis. Correlation of Leu-2, Leu-3, Leu-7, Leu-S and Leu-II cell surface antigen expression. J Immunol 132: 151, 1984.

118. Touw, I, Hofhuis W, van Zanen G, Delwe R, Lowenberg B: In vitro colony forming cells of acute lymphoblastic leukemia: Analysis of 24 cases with recombinant interleukin 2 as growth stimulus. In: Hagenbeek A, Lowenberg B (cds): Minimal Residual Disease in Acute Leukemia 1986, Dordrecht, Martinus Nijhoff Publishers, 1986, pp 141-147.

119. Coldman A, Goldie J: Role of mathematical modeling in protocol formation in cancer chemotherapy. Cancer Treat Rep 69:1041-1045, 1985.

120. Ryan D, Chapple C, Kossover S, Sandberg A, Cohen M. Phenotypic similarities and differences between CALLA-positive acute lymphoblastic leukemia cells and normal marrow CALLA-positive B cell precursors. Blood 70:814-S21, 1987.

121. Lee M, Chang K, Cabanillas F, Freireich E, Trujillo J, Stass S. Detection of minimal residual cells carrying the t (14; IS) by DNA sequence amplification. Science 237:175-178, 1987.

5. Radioimmunoscintigraphy of lymphoma with monoclonal antibodies

Jorge A. Carrasquillo and Steven M. Larson

Introduction

Patients with histologically proven lymphoma must undergo an extensive pretreatment diagnostic evaluation. Such testing is important in the choice of therapy as well as in defining the prognosis. Staging examinations must be repeated at intervals in order to evaluate the response to treatment [1-4].

The staging involves noninvasive as well as invasive procedures (i.e., biopsies or laparotomy) including radiographic studies (plain radiography, computerized tomography, ultrasound, magnetic resonance imaging, gallium scanning, and lymphangiography). These procedures are dependent on dif­ferences in attenuation, mass effect , paramagnetic properties, metabolism , or ultrasound reflection and therefore provide indirect evidence for the pres­ence of tumor. The role of these procedures in the staging of patients with lymphoma as well as their limitations has been reviewed [5]. Radioimmuno­scintigraphy (RIS) of tumors is an in vivo technique where radioactive labeled antitumor antibodies are administered to patients with malignancies as diagnostic reagents for tumor detection [6]. The interest in RIS has been stimulated by hybridoma technology [7] and advances in tumor immunology which have identified a multitude of tumor associated antigens present on human tumors. The underlying principle is that these antigens have a much higher expression in tumor than in normal tissue and when radio labeled antibodies specific for the antigen are administered to a patient, they will be carried passively to all tissues - because the tumor has the antigen - the antibody will bind, and the antibody progressively accumulate, permitting detection utilizing conventional nuclear medicine gamma cameras. This spec­ificity offers theoretical advantages over the majority of diagnostic methods available today. Unlike the "indirect" diagnostic methodologies, the presence of radio labeled antibody in a tumor mass is direct evidence of the presence of antigen-bearing tumor tissue.

Although preclinical work in animal tumor models has shown favorable targeting of radiolabeled antibodies [8-13], much of the human work with polyclonal and monoclonal antibodies has not shown the expected sensitivity for tumor imaging [14-17]. There are several factors that may be important

Bennel/, 1.M. a"d Foo". K.A., (eds.), Immunologic Approaches to the Classification and Managemelll of Lymphomas arid Leukemias. © 1988 KJuwer Academic Publishers. tS8N978· /·4612·8965-4. All rights reser.ed.

210

in improving tumor targeting. These are related to: 1) the antibody (specific­ity, affinity, labeling method, fragment, or dose); 2) the antigen (concen­tration, heterogeneity, and circulating antigen); 3) and tumor (blood flow, capillary permeability, and accessibility of the antibody); 4) the route of administration; 5) the isotope; and 6) the gamma camera resolution. Human imaging trials addressing these issues are in progress. The clinical experience with imaging of lymphomas with radiolabeled antibodies will be reviewed.

Intravenous delivery of radio labeled antibodies

The principles and efficacy of RIS have been documented in vivo in numer­ous animal models bearing solid human tumors xenografts (colon, melano­ma, ovarian, etc.) in which excellent localization of radiolabeled antibodies has been observed [8-13]. A few investigators have studied the localization of radiolabeled antibodies in animal lymphoma or leukemia. Scheinberg et al. demonstrated excellent targeting in the Rauscher murine erythroleukemia model with an Illln antibody [18] and demonstrated the superiority of Illln over 1311 radiolabeled antibody [19]. Badger et al. demonstrated efficient, specific targeting of mouse AKR T-cell lymphoma with 131 1 labeled MoAb directed at the Thy 1.1 differentiation antigen; in a number of animals treated with high dose 131} MoAb they observed therapeutic responses of established tumors [20].

Several therapeutic antibody trials utilizing unconjugated unlabeled MoAb in patients with human lymphoma and leukemia have been performed [21-26]. Although short-term therapeutic effects have been observed, many of these trials have had a limited long-term responses. Nevertheless, a large amount of information that is directly applicable to the field of RIS has been obtained. Minimal or no side effects were seen, even when repeated injec­tions of large amounts of monoclonal antibodies were given. The antibody was able to efficiently target circulating and nodal sites of disease, as demon­strated by fluorescein activated cell sorter analysis and immunohistology [22, 24]. Antigen modulation was frequently encountered and may have been partially responsible for the lack of tumor response [23]. Human anti-mouse antibody developed in some but not all of the patients treated with single or multiple injections [27].

Utilizing the RIS technique in patients with advanced cutaneous T-cell lymphoma (CTCL), the NIH group has demonstrated the ability of Illln labeled TI0l monoclonal antibody to localize in sites of lymph node involve­ment as well as erythroderma and skin tumors [28, 29]. T101 is a murine MoAb IgG2a that recognizes a pan-T-cell antigen (T65) present in normal T-Iymphocytes, T-cell lymphoma, including cutaneous T-cell lymphoma (CTCL), and also paradoxically expressed on chronic lymphocytic leukemia (CLL) [30, 31]. Although the antigen is present on normal lymphocytes, the antigen expression is higher in patients with CTCL [32]. Both of these

211

patient groups frequently have high numbers of circulating malignant cells, which makes targeting more favorable.

The antibody was radiolabeled using a modification of the mixed anhy­dride method [33] whereby a bifunctional chelate, diethylenetriaminepenta­acetic acid (DTPA), is attached to the antibody. This preparation forms a chelate with 1IIIn metal ion, resulting in a product with approximately 88% immunoreactivity. Patients received an intravenous injection of varying mass amounts of antibody, between 1 mg and 50 mg, with 5 mCi of III In nOl. To determine the kinetics of biodistribution, serial whole body gamma camera images, blood clearance, and whole body retention measurements were performed.

In 11 patients with CTCL, the scan showed concentration in clinically involved lymph nodes as well as in many previously unsuspected lymph nodes regions. The scan was positive in 38 of 39 clinically involved nodal sites (figure 1). Forty-four of 136 clinically negative nodal sites were positive on scan. Of the four sites biopsied, all were positive for the presence of occult lymphoma. The other 40 sites were not biopsied but may also repre­sent subclinical disease. Areas of cutaneous involvement with erythroderma or tumors were positive (figures 1 and 2), while skin plaques showed no localization. The negative findings in skin plaques may have been related to infiltration with a smaller number of malignant cells than tumors or erythro­derma. Control studies with III In alone or III In antimelanoma (control monoclonal antibody) showed no localization in involved regions. Although no normal patients were studied, the targeting of skin lesion and the asym­metrical nodal uptake observed in several patients suggest that normal nodes were not being targeted after intravenous injection.

In addition to concentration in involved nodes, all patients had high concentrations in liver, spleen, and bone marrow. The mechanism of uptake in these normal organs is likely to be multifactorial. The small amount of 111ln that leaches off the antibody while in circulation will bind to transferrin and may subsequently be deposited in liver, spleen, and bone marrow. However, this translocation from antibody to transferrin was measured and is too small to explain all the observed uptake. In studying 111In oxine labeled or Cr-51 labeled malignant T-cells in vivo, Miller [22] and Dillman [23] demonstrated that after targeting the T65 antigen with Leu 1 or TI01 MoAb, there was rapid clearance of radio labeled cells from the blood pool, predominantly into the liver. this rapid clearance of T-cells from the cir­culation was also observed in our studies and was thought to be secondary to removal of antibody-coated cells by the RES. The catabolism of immunoglo­bulin in the liver is also likely to be a contributory factor to the observed liver uptake. Whatever the mechanism of liver uptake, this "nonspecific" localization precludes evaluation of this organ for tumor involvement. Thus RIS does not supplant the need for liver biopsy, which is the most reliable test for staging of hepatic involvement by lymphoma. Also, splenic uptake of the radio labeled an~ibody is prominent and probably related to direct

212

Figure 1. Anterior and posterior whole body scan obtained 72 hours after intravenous injection of 5 mCi and 10 mg of 1IIIn TlO! MoAb. The study shows uptake in bilateral cervical, axillary, inguinal-femoral, epitrochlear, and popliteal nodes as well as in left supraclavicular and left iliac nodes. Uptake in skin (erythroderma) is also visualized. In addition, liver, spleen, and bone marrow uptake is seen. (From Larson SM, et al in Nucl Med Bioi Int J Radiat Appl Instrum 13(B): 207-213, 1986. Used with permission.)

binding to antigen-bearing cells, many of which may be normal T-Iympho­cytes. Studies by Bunn et al. have suggested that Sezary cells exist in three compartments: the skin, the blood, and the lymph nodes (where the cells are produced), with migration occurring from one compartment to another [34]. Miller et al. demonstrated that III In oxine labeled Sezary cells rein-

213

Figure 2. Anterior view of the skull obtained 72 hours after administration of 5 mCi, 1.5 mg of 111In T101 from a 69-year-old male with cutaneous T-cell lymphoma. There is focal uptake in a small subcutaneous tumor in the right cheek (open arrowhead) as well as in bilateral cervical and supraclavicular lymph node uptake (arrows). (Reprinted with permission from [38].)

jected intravenously would migrate to sites of skin involvement [35]. The biodistribution data observed in this study suggested that we targeted antigen bearing cells which, once targeted, can modulate and internalize the MoAb with the radio label [36]. The available evidence suggests that these cells will then migrate, carrying their contained radioactivity to the involved sites [29] (figure 3) . Biopsies showed good concentration in lymph nodes with 10 to 100 times the levels previously reported for RIS of solid tumors. The cohsistently good localization in involved sites of lymphoma is in contrast to the general experience of targeting solid tumors and is due to the excellent characteristic of the antibody , the biologic behavior of malignant T-cells, arid the favorable characteristics of the T65 antigen.

,In vitro work assessing the therapeutic efficacy of 1251 T101 has shown antigen specific cell killing of T-cell lines with several magnitudes of cell killing [37]. Imaging trials with 1311 radiolabeled T101 have shown very rapid breakdown and excretion of the 1311 from the body [38]. The striking difference in biodistribution of IIlIn and I3lI (figures 4 and 5) are evidently related to differences in handling of the isotope once the antibody is degraded and not to damage of the MoAb during the labeling process . In vivo, both antibodies behave similarly, targeting antigen bearing cells, after which the radio labeled antibodies undergo internalization and catabolism [36, 39]. Once 1311 is released from the antibody it does not bind to intracellular proteins and is released from the cell to be excreted in the urine; in contrast,

214

Figure 3. Anterior image of the pelvis from a 24-year-old male with cutaneous T-cell lymphoma , obtained at two hours (left panel) and 24 hours (right panel) after administration of 5 mCi, 1 mg of IllIn TIOI. There is bone marrow uptake with no localization in the lymph node at two hours , when 95% of the tracer had been cleared from the blood. At 24 hours, there is intense localization in the inguinal nodes. (Reprinted with permission from [29].)

once the 1IIIn is released from the antibody, it binds to intracellular proteins and is not easily released from the target cell [40].

Therapy studies with 131 I T101 have shown partial and complete responses which have been short-lasting [41]. Although it has not been confirmed that these are a direct consequence of the targeting of malignant T-cells, our studies suggest that isotopes with therapeutic potential which are similar to 1IIIn in their handling, with prolonged retention by the target cell (i.e., 90y), would be preferable candidates for therapeutic trials than 1311.

In summary, our preliminary experience with intravenous injection of 1IIIn T101 antibody in CTCL and CLL suggest that this approach should permit diagnostic evaluation of all nodal regions and extra nodal sites.

Lymphatic delivery

One of the disadvantages of intravenous delivery is that the MoAb undergoes dilution in the plasma volume and only a small proportion of the total antibody injected is available to flow through a small tumor or involved lymph node. Cross-reacting sites in normal tissue compete for the MoAb, decreasing the amount available for tumor binding. In addition , high levels of circulating radiolabeled antibody in the blood pool creates a high back­ground which decreases the tumor to nontumor ratios.

215

111ln-T101 (1.5mg, 5mCi) 1311_T101 (1.5mg, 2mCi)

(48 hr.)

Figure 4. Left upper and lower panels show spot images 48 hours after intravenous injection of 1IIIn THil (5 mCi, 1.5 mg); localization is in multiple involved sites: axillary and inguinal lymph nodes as well as erythroderma are seen. The upper and lower right panel show spot images 48 hours after nil TlOl (2 mCi, 1.5 mg). Rapid clearance from liver, spleen, and whole body is seen, with no accumulation in clinically involved skin and lymph nodes. Uptake in partially blocked thyroid and stomach is seen (right upper panel).

The lymphatic route of delivery has the advantage of delivering a high concentration of MoAb directly to the lymph nodes before any systemic dilution of the antibody occurs. Delivery via the lymphatics may be by direct intralymphatic administration, as in contrast lymphangiography, or via sub­cutaneous injection. When delivered subcutaneously, the antibodies enter the terminal lymph capillaries through junctions in the endothelial wall. From there they flow into larger vessels, arrive at the lymph node through the subcapsular sinus , and flow through the sinusoids into the medulla. The antibodies may bind to the target antigen, or, if not removed by the lymph nodes, they may pass into the efferent lymphatic vessels and from there into the blood stream via the thoracic duct. Animal studies performed by Wein­stein et al. targeting normal cell components of lymph nodes in mice (major histocompatibility antigen and a T-cell antigen) with 125I labeled MoAb demonstrated excellent targeting, with more efficient delivery and higher

216

Spleen 4%

Uver 3%

Para-aortic nodes 2% {

Iliac nodes 7% {

Inguinal- { femoral nodes 15%

Injection sites 14%

J- <1%

J- 3%

} 32%

25%

Figure 5. Anterior whole body from two separate patients with cutaneous T-cell lymphoma after subcutaneous administration of 0.5 mCi, 100 Ilg of IllIn TJOI MoAb in the web spaces of the feet. Left panel images were obtained at two days and right panel at three days. The study shows excellent localization in all the regional lymphatics extending from the femoral to the para-aortic nodes. (Reprinted with permission from [48].)

concentration of MoAb than when administered intravenously [42]. Utilizing this approach, these investigators demonstrated specific targeting of regional lymph node metastasis from the LlO hepatocarcinoma guinea pig model [43].

The clinical studies reported with this approach are limited, and, as with the intravenous delivery of radiolabeled MoAb, most of the work has been evaluating solid tumors such as breast, colon, and melanoma [44, 45, 46, 47]. Deland et al. [45] reported a sensitivity of 100% in seven patients with breast carcinoma metastatic to the axilla, although the images had poor resolution and no control antibodies were used to document the specificity. More recently Thompson et aI., utilizing MoAb against a breast tumor associated antigen, demonstrated a high sensitivity of tumor detection in patients with metastatic breast carcinoma to the regional lymphatics [46]. The NIH group targeting patients with melanoma and regional lymph node involvement has had little success in targeting metastatic lymph node disease

217

via subcutaneous delivery [47]; whether this was related to the type of spread of melanoma in the lymph node with distortion of lymph flow and inaccessi­bility of antibody to tumor or to the antibodies utilized is yet uncertain.

IllIn T101 was administered subcutaneously in patients with CTCL. The administration technique is similar to that for conventional lymphoscintigra­phy with Tc-99m antimony sulfur colloid. The patients received 500 rtCi, 100 rtg of III In T101 in divided doses in the web spaces of the feet. Serial 'imaging was performed of the injection sites, regional lymphatics, and distal lymphatics in order to obtain biodistribution information. Utilizing this technique, Keenan et al. [48], in a preliminary communication, demon­strated rapid and efficient delivery of IllIn TlOl to the regional lymph nodes, with concentration much higher than previously reported for intra­venous administration (figure 5) of radiolabeled monoclonal antibody. In two patients 24% and 36%, respectively, of the injected dose was retained in the regional lymphatics. Because of cross-reactivity with normal T-cells and visualization of all regional lymph nodes with this MoAb, uptake (al­though antigen specific) may not be related solely to tumor targeting. After intralymphatic delivery, both normal and neoplastic antigen-bearing cells are exposed to a high concentration of radio labeled antibody. The study clearly shows the potential of antigen targeting; perhaps with the use of more specific antibodies, improved tumor specificity will be obtained.

In addition to the regional lymphatics, sites outside the drainage area, such as the cervical, axillary, and hilar lymphatics, and subcutaneous tumor sites were also visualized. This distant targeting represented either overflow of antibody through the lymphatics or trafficking of targeted cells from the regional nodes into distant sites.

There are several potential limitations to the technique. Since the ap­proach is regional in nature, multiple injections in sites draining the lymph node regions of interest are necessary. Bulky lymph node disease could result in alterations in flow through the lymphatics and preclude delivery of MoAb. Further lymphatic studies with T101 or other more specific anti­bodies must be performed before routine application of this technique can be considered.

Human anti murine antibody response

With the use of MoAb for imaging as well as for serotherapy trials, the presence or the development of a host anti murine antiglobulin response has been of great concern because of the potential allergic reactions as well as the alterations in biodistribution [49]. Although several investigators have detected low levels of human IgG and IgM reactive with MoAb in healthy individuals as well as in patients with malignancies who have not previously received murine MoAb, in this setting clinical side effects have rarely been observed. The occurrence of human antimurine antibody response (HAMA)

218

after antibody administration is relatively high. When whole immunoglobulin was injected intravenously, 32% of patients receiving single injections devel­oped HAMA [50, 27]. The incidence is likely to be even higher in patients receiving multiple injections for serial imaging or therapy. The administra­tion of MoAb after development of HAM A has been associated with lack of tumor response [23], alterations in biodistribution, lack of tumor imaging, and allergic side effects [49, 51]. Fab fragments are considerably less immunogenic, and only 6% of the patients who received single injections of Fab developed HAMA. After multiple injections this incidence increased sharply [49]. Although it has been postulated that a host immune response with an anti-idiotype network may be beneficial in obtaining tumor responses [52], it is clear that for the purpose of imaging and repeated administration of therapeutic radiolabeled MoAb HAM A is detrimental.

In patients with immune deficiency, as seen with some lymphomas and eLL, an immune response may not be elicited in spite of multiple admini­stration of whole MoAb [27].

If the RIS technique is to find routine serial use as a screening agent for recurrent disease, some approach that deals with the development of HAMA will be necessary. Strategies to prevent HAMA development include: de­livery of immunosuppressants, alteration of the immunoglobulin, and use of human monoclonal antibodies. Also, levels of HAMA may be reduced once they have formed; plasmapheresis, affinity columns or preloading with un­labeled antibody are a few approaches which may help.

Future developments

Most RIS studies have used whole immunoglobulin [14, 15, 17, 29]. Because Fab fragments achieve an optimal target to background ratio more rapidly and are less immunogenic, several groups are exploring the use of Fab or F(ab')2 fragments with interesting preliminary results [53, 54, 55, 56, 57]. Fab fragments also have a number of drawbacks as well, including a very rapid excretion and accumulation in the kidneys. This latter problem has limited the applicability of III In labeled Fab. Further studies are needed to define whether fragments or intact IgG are optimal for clinical applications.

The most frequently used isotope for radio labeling of antibodies has been 1311, by the chloramine T [58] or iodogen [59] methods because of the well known chemistry and their documented utility. Iodine isotopes suffer several limitations and are not ideal for imaging; 1311 has a high-energy gamma ray (364 kev), which requires thick lead septa collimation, reducing the sensitivity and resolution of the gamma camera; the associated beta radiation and the long TI/2 result in a relatively high absorbed dose. While 1231 has been utilized successfully [60, 61, 62], with the advantage that its energy charac­teristics are more suitable (159 kev and no beta radiation), its short TlJ2 (13 hours) is not ideal as a tracer for the relatively slow in vivo targeting of

219

radiolabeled antitumor IgG's, and the associated use of antibody fragments may be required.

The problem of in vivo dehalogenation has been discussed by several authors [9, 15, 38]. Halpern et al. demonstrated in vivo dehalogenation of radioiodinated antibodies and improved targeting and stability of III In label­ed MoAb, which distributed similarly to endogenously labeled 18Seleno­methionine MoAb [60]. In addition to higher concentration of III In MoAb in tumor, III In has more favorable characteristics than 131 I for imaging with an energy of 173 ke V and 247 ke V - which is better suited for the gamma camera - a high photon abundance, and a T1j2 of 2.83 days. New bifunc­tional chelates which are more stable have been developed and may lead to improved imaging.

99mTc is the ideal isotope for imaging with the gamma camera because of its energy and the resultant low absorbed does to the patient. Technetium chemistry is difficult, and methods for producing stable conjugates are still in development. The short TI/2 of six hours is a disadvantage when utilizing whole immunoglobulin, since most studies have found optimal images beyond the 24 hours postinfusion.

Advances in instrumentation and computer software have permitted three­dimensional imaging with single photon emission computed tomography (SPECT). SPECT offers improved contrast resolution [63] as compared to routine gamma camera imaging. Preliminary reports with RIS and SPECT suggest improved sensitivity [64].

Conclusions

The feasibility of targeting lymphoma with radiolabeled antibodies has been demonstrated. The current experience suggests that, in addition to targeting solid tumor (i.e., colon, ovarian, melanoma,) with radiolabeled MoAb, a larger emphasis should be placed on targeting lymphoma, since accessibility of antigen-bearing tumor cells appears to be more favorable. Because of the inherent specificity of radiolabeled MoAb, there is the great potential to target a variety of lymphomas. Questions that need to be answered in the experimental setting include: 1) Which antibodies give the best tumor locali­zation: pan-T-cell, pan-B-cell, more specific antibodies, or a cocktail of antibodies? 2) Are whole antibody or fragments superior for clinical studies? 3) What is the best dose?

References

1. Carbone PP, Kaplan HS, Musshoff K, et al: Report of the committee on Hodgkin's disease staging classification. Cancer Res 31: 1860-1861, 1971.

2. Chabner BA, Johnson RE, Young RC, et al: Sequential nonsurgical staging of non­Hodgkin's lymphoma. Ann Med 85:149-154, 1976.

220

3. Bunn PA Jr, Huberman MS, Whang-Peng J, et al: Prospective staging evaluation of patients with cutaneous T-cell lymphomas: Demonstration of high frequency of extracu­taneous dissemination. Ann Med 3:223-230, 1980.

4. Bunn PA, Lamberg SI: Report of the committee on staging and classification of cutaneous T-cell lymphomas. Cancer Treat Rep 63:725-728, 1979.

5. Hamlin J: Radiographic approach to the staging of lymphoma, including Hodgkins disease. In: JM Bennett (ed): Lymphoma, 1 Martinus Nihoff Publishers, 1981, pp 1, 177,233.

6. Goldenberg DM, DeLand FH, Kim E, et al: Use of radiolabeled antibodies to carcinoem­bryonic antigen for the detection and localization of diverse cancers by external photoscan­ning. N Engl J Med 298:1384-1386. 1978.

7. Kohler G, Milstein C: Continuous culture of fused cells secreting antibody of predefined specificity. Nature (London) 256:495-497. 1975.

8. Colcher D, Zalutsky M, Kaplan W, Kufe D, Austin F, Schlom J: Radio-localization of human mammary tumors in athymic mice by a monoclonal antibody. Cancer Res 43:736-742, 1983.

9. Halpern SE, Hagan PL, Garver PR, Koziol JA, Chen AWN, Frincke JM , Bartholomew RM, Davis GS, Adams TH: Stability, characterization, and kinetics of IIIIn-labeled mono­clonal antitumor antibodies in normal animals and nude mouse-human tumor models. Cancer Res 43:5347-5355, 1983.

10. Moshakis V, McIlhinney RA, Raghavan D, Neville AM: Localization of human tumor xenografts after I.V. administration of radiolabeled monoclonal antibodies. Br J Cancer 44:91-99, 1981.

11. Pimm MV, Embleton MJ, Perkins AC, Price MR, Robins RA, Robinson GR, Baldwin RW: In vivo localization of anti-osteogenic sarcoma 791T monoclonal antibody in osteogenic sarcoma xenografts. Int J Cancer 30:75-85, 1982.

12. Ghose T, Ferrone S, Imai K, Norvell ST, Luner SJ, Martin RH, Blair AH: Imaging of human melanoma xenografts in nude mice with a radiolabeled monoclonal antibody. J Natl Cancer Inst 69 (4):823-826, 1982.

13. Beaumier PL, Larson SM, Krohn KA, Ferens JM, Nelp WB, Brown JP: Effects of iodinatin of p97 antigen-specific binding and in vivo tumor localization using monoclonal antibody fragments. J Nucl Med 24:116, 1982.

14. Mach JP, Carrel S, Forni M, et al: Tumor localization of radiolabeled antibodies against carcinoembryonic antigen in patients with carcinoma. N Engl J Med 5-10, 1980.

15. Sullivan DC, Silva JS, Cox CE, et al: Localization of 131 1 labeled goat and primate anti­carcinoembryonic antigen (CEA) antibodies in patients with cancer. Invest Radiol 17:350-355, 1982.

16. Ott RJ, Grey LJ, Zivanovic MA, Flower MA, Trott NG, Moshakis V, Coombes RC, Neville AM, Ormerod MG, Westwood JH , McCready VR: The limitations of the dual radionuclide substraction technique for the external detection of tumours by radioiodine­labelled antibodies. Br J Radiol 56: 101-108, 1983.

17. Mach JP, Chatal JF, Lumbroso JD , et al: Tumor localization in patients by radiolabeled monoclonal antibodies against colon carcinoma. Cancer Res 43:5593-5600, 1983.

18. Scheinberg DA, Strand M, Gansow 0: Tumor imaging with radioactive metal chelates conjugated to monoclonal antibody. Science 215:1511-1513,1982.

19. Scheinberg DA, Strand M: Kinetic and catabolic considerations of monoclonal antibody targeting in erythroleukemic mice. Can Res 43:265-272, 1983.

20 . Badger CC, Krohn KA, Peterson A, et al : Radiotherapy of murine lymphoma with 131 1 labeled anti-Thy 1.1 monoclonal antibody. Can Res 45:1536-1544, 1985.

21. Miller RA, Oseroff AR, Stratte PT, Levy R: Monoclonal antibody therapeutic trials in seven patients with T-cell lymphoma. Blood 62:988-995, 1983.

22. Miller RA , Maloney DG, McKillop J, Levy R: In vivo effects of murine hybridoma monoclonal antibody in a paticnt with T-cell leukemia. Blood 58:78, 1981.

23. Dillman RO, Sobol RE, Collins H, Beauregard J , Royston I: TIOI monoclonal antibody therapy in chronic lymphocytic leukemia. In: Oettgen H, Mitchell M (eds): Hybridomas in

221

Cancer Diagnosis and Treatment, New York, Raven Press, 1982, pp 151-17l. 24. Foon KA, Schroff RW, Bunn RA, Mayer D, Abrams PG, Fer MF, Ochs JJ, Bottino GC,

Sherwin SA, Herberman RB, Oldham RK: Effects of monoclonal antibody therapy in patients with chronic lymphocytic leukemia. Blood 64:1085-1094, 1984.

25. Ritz 1, Pensando 1M, Sullan SE et al: Serotherapy of acute lymphoblastic leukemia with monoclonal antibody. Blood 58:141-152, 1981.

26. Nadler LM, Stashenko P, Hardy R et al: Serotherapy of a patient with monoconal antibody directed against a human lymphoma-associated antigen. Cancer Res 40:3147-3154, 1980.

27. Schroff RW, Foon KA, Beatty SM, Oldham RK, Morgan AC: Human anti-murine immu­noglobulin responses in patients receiving monoclonal antibody therapy. Cancer Res 45: 879-885, 1985.

28. Bunn PA, Carrasquillo JA, Keenan AM, et al: Imaging of T-cell lymphoma by radio­labeled monoclonal antibody. Lancet 2: 1219-1211, 1984.

29. Carrasquillo JA, Bunn PA, Keenan AM et al: Radioimmunodetection of cutaneous T-cell lymphoma with lllIn labeled nOI monoclonal antibody. N Engl 1 Med 315:673-680,1986.

30. Royston I, Majda lA, Baird SM, et al: Human T-cell antigens defined by monoclonal antibodies: The 65,000-dalton antigen of T cells (T65) is also found on chronic lymphocytic leukemia cells bearing surface immunoglobulin. 1 Immunol 125:725-731, 1980.

31. Dillman RO, Beauregard JC, Lea JW, et al: Chronic lymphocytic leukemia and other chronic lymphoid proliferation surface marker phenotypes and clinical correlations. J Clin Onco: 1:190-197, 1983.

32. Ledbetter JA, Frankel AE, Herzenberg LA, Herzenber, HA: Human Leu T-cell differen­tiation antigen: Quantitative expression on normal lymphoid cells and cell lines, In: Gl Hammeriing, U Hammerling, lF Kearney (eds): Research Monographs in Immunology, volume 3, Monoclonal antibody and T-cell hybridoma, p 16-22, 1981.

33. Krejcarek GE, Tucker KL: Covalent attachment of chelating groups to macromolecules. Biochem Biophys Res Commun 77:581-585, 1977.

34. Bunn PA Jr, Edelson RL, Ford SS, Shackney SE: Patterns of cell proliferation and cell migration in the Sezary syndrome. Blood 57:452-463, 1981.

35. Miller RA, Coleman CN, Fawcett HD, Hoppe RT, McDougall IR: Sezary syndrome: A model for migration of T-Iymphocytes to skin. N Engl J Med 303:89-92, 1980.

36. Schroff RW, Klein RA, Farrell MM, Stevenson, HC: Enhancing effects of monocytes on modulation of a lymphocyte membrane antigen. J Immunol 133(4):1641-1648, 19 .

37. Boven E, Lindmo T, Mitchell J, Bunn P: Selective cytotoxicity of 12sI-labeled monoclonal antibody nOI in human malignant T cell lines. Blood 67:429-435, 1986.

38. Carrasquillo JA, Mulshine lL, Bunn P, et al: Tumor imaging of lllIn T101 monoclonal antibody is superior to ulI T101 in cutaneous T-cell lymphoma. 1 Nucl Med. in press.

39. Shawler DL, Miceli MC, Wormsley SB, Royston I, Dillman RO: Interaction of in vitro and in vivo antigenic modulation by the anti-human T-cell monoclonal antibody TlOl. Cancer Res 44:5921-5927, 1984.

40. Thakur M, Segal AW, Louis WL et al: In-Ill labeled cellular blood components: Mecha­nism of labeling and intracellular location in human neutrophils. J Nucl Med 18: 1022-1026, 1977.

41. Spies SM, Zimmer AM, Spies WG, et al: Initial experience with radioimmunodetection and therapy of cutaneous T-cell lymphoma using murine rllonoclonal antibodies (abstract). J of Nucl Med 27:903, 1986.

42. Weintstein IN, Parker RJ, Keenan AM et al: Monoclonal antibodies in the lymphatics: Toward the diagnosis and therapy of tumor metastases. Science 218: 1334-1337, 1982.

43. Weinstein lN, Steller MA, Covell DG, et al: Use of monoclonal antibodies for diagnosis and therapy of tumor metastases in lymph node metastases of a solid tumor. Science 222: 423-426, 1986.

44. Order SE, Bloomer WD, Kaplan JA, et al: Radionuclide immunoglobulin lymphangiogra­phy: A case report. Cancer 35:1487-1489,1985.

45. Deland FH, Kimm EE, Goldenberg DM: Lymphoscintigraphy with radionuclide-labeled

222

antibodies to carcinoembryonic antigen. Cancer Res 40:2997-3000, 1980. 46. Thompson CH, Lichtenstein M, Stacker SA, Leyden MJ, Salehi N, Andrews JT, McKenzie

IF: Immunoscintigraphy for detection of lymph node metastases from breast cancer, Lancet 2(8414): 1245-7, 1984.

47. Lotze MT, Carrasquillo JA, Weinstein IN, et al: Monoclonal antibody imaging of human melanoma: Radioimmunodetection by subcutaneous or systemic injection. Ann Surg, in press.

48. Keenan AM, Weinstein IN, Mulshine lL, et al: lmmunolymphoscintigraphy in patients with lymphoma after subcutaneous injection of Indium-Ill TI01 monoclonal antibody. 1 Nucl Med 28:42-46, 1987.

49. Carrasquillo JA, Krohn KA, Beaumier P, et al: Diagnosis of and therapy for solid tumors with radiolabeled antibodies and immune fragments. Cancer Treat Rep 68:317-318, 1984.

50. Reynolds JC, Carrasquillo lA, Larson SM: Human anti-murine response following admini­stration of radiolabeled monoclonal antibodies, 1 Nucl Med 26(227):55.

51. Pimm MY, Perkins AC, Armitage NC, Baldwin RW: The characteristics of blood-borne radiolabels and the effect of anti-mouse IgG antibodies on localization of radiolabeled monoclonal antibody in cancer patients. 1 Nucl Med 26: 1011-1023, 1985.

52. Koprowski H, Herlyn D, Lubeck M, et al: Human anti-idiotype antibodies in cancer patients: Is the modulation of immune response benefitial for the patient? Proc Natl Acad Sci USA 81(1):216-219, 1984.

53. Epenetos AA, Snook D, Hooker G, Begent R, Durbin H, et al: Indium-Ill labelled monoclonal antibody to placental alkaline phosphatase in the detection of neoplasms of testis, ovary, and cervix. Lancet 2(8451):350-3, 1985.

54. Larson SM, Carrasquillo JA, McGuffin RW, et al: Use of 1311 labeled, murine Fab against a high molecular weight antigen of human melanoma: Preliminary experience. Radiology 155:487-492, 1985.

55. Larson SM, Carrasquillo JA, Krohn KA, Brown JP, McGuffin RW, Fereus 1M, Graham MM, Hill LD, Beaumier PL, Hellstrom KE, Hellstrom I: Localization of 131 1 labeled p97 specific F AB fragments in human melanoma as a basis for radiotherapy. J Clin Invest 72:2101-2114, 1983.

56. Buraggi GL, Callegaro L, Mariani G, et al: Imaging with 13l1-labeled monoclonal anti­bodies to a high-molecular-weight melanoma-associated antigen in patients with melanoma: Efficacy of whole immunoglobulin and its F(ab')2 fragments. Cancer Res: 45(7):3378-87, 1985.

57. Moldofsky Pl, Powe l, Mulhern CB, Hammond N, Sears HF, Gatenby RA, Steplewski Z, Koprowski H. Metastatic colon carcinoma detected with radiolabeled F(ab')2 monoclonal antibody fragments. Radiology 149:549-555, 1984.

58. McConahey Pl, Dixon Fl: A method of trace iodination of proteins for immunologic studies. Int Arc Aller 29: 185-189, 1966.

59. Fraker Pl, Speck JC lr: Protein and cell membrane iodinations with a sparingly soluble chloroamide, 1,3,4, 6-tetrachloro-3a, 6a-diphenylgly-coluril. Biochem Biophys Res Com­mun 80:849-857, 1978.

60. Epenetos AA, Shepherd l, Britton KE, et al: 1231 radio iodinated antibody imaging of occult ovarian cancer. Cancer 55(5):984-7, 1985.

61. Pateisky N, Philippk, Burchel J, et al: Radioimmunodetection in patients with suspected ovarian cancer. J Nucl Med 26: 1369-1376, 1985.

62. Davies 10, Davies ER, Howe K, Jackson PC, et al: Radionuclide imaging ovarian tumors with 123I-labelled monolonal antibody (NDOG2) directed against placental alkaline pho­sphatase. Br J Obstel Gynaecol 1985. 92(3):277-86.

63. Ell Pl, Khan 0: Emission computerized tomography: Clinical applications. Semin Nucl Med 11:50-60, 1981.

64. Berche C, Mach JP, Lumbroso JD, et al: Tomoscintigraphy for detecting gastrointestinal and medullary thyroid cancers: First clinical results using radiolabeled monoclonal anti­bodies against carcinoembryonic antigen. Br Med 1 285:1447-1451,1982.

6. Radiolabeled Antibodies in Hodgkin's Disease

Stanley E. Order

The theoretical concept of a restrictive cytotoxic radio labeled antibody ad­ministered cyclically for the treatment of malignancy has now become a clinical reality both with polyclonal and monoclonal antibodies [1-3]. Due to their relative radiosensitivity compared to other solid tumors, lymphomas (i.e., Hodgkin's disease or others) make ideal tumors for clinical investigation of radiolabeled antibodies and for monitoring investigative clinical progress [4]. This chapter will summarize to date both laboratory and clinical data pertinent to the development and clinical use of radiolabeled antibodies in the treatment of Hodgkin's disease.

Ferritin as a tumor antigen

The finding that ferritin is a tumor-associated protein in Hodgkin's disease led to investigation of its cellular source and relevance to the disorder [5]. Assaying the least dense layers of bovine serum albumin gradients in cellular extracts from the tumor, our laboratory demonstrated that the lymphocytes contained ferritin both within the cell and on the cell surface [6]. Sarcione found that the T-lymphocytes from Hodgkin's patients synthesized and secreted ferritin preferentially compared to normal lymphocytes, and in the same lymphocytes puramycin inhibited ferritin synthesis [7]. In contrast, protein synthesis was greater in normal lymphocytes than in Hodgkin's lymphocytes, even though ferritin synthesis was greater in the Hodgkin's lymphocytes [7]. Pretlow noted that T-lymphocytes in Hodgkin's spleens showed surface ferritin and would form rosettes with Hodgkin's cells [8]. More recently, Strauchen published a new observation that pre-AIDS pa­tients with Hodgkin's disease had T-suppressor cell dominance and a T-cell helper deficit, unlike other Hodgkin's patients with both T-suppressor and T-helper cells infiltrates in the tumor [9]. Presently we are evaluating these observations in our study of ferritin distribution in Hodgkin's disease.

In addition, in patients with acquired immunodeficiency and in Hodgkin's patients after successful remission, the inexplicable observation remains that serum ferritin levels are elevated [10]. We observed increased ferritin content

BellI/eli, I .M. and Foon. K.A., (eds.), Immunologic Approaches 10 rhe Classificarion and Managemem of Lymphomas and Leukemias. © 1988 Kluwer Academic Publishers. ISBN978-1-4612-8965-4. All rig/liS reserved.

224

in circulating lymphocytes, as well, in patients with active Hodgkin's disease [11]. These initial limited observations have not been pursued further in our laboratory.

As an antigenic target, ferritin in Hodgkin's disease is present on both the lymphocyte surface and, importantly, in a halo-like distribution surrounding the tumor in the stroma. The stromal ferritin provides an excellent zone of high concentration antigen for antibody targeting (figure 1) [12].

The isotype of ferritin in Hodgkin's disease is similar to splenic ferritin [13]. One may ask, then, why the radiolabeled 1-131 antiferritin does not target normal tissues containing ferritin such as muscle, bone marrow, heart, and spleen?

Biologic Window

There are no animal models for Hodgkin's disease. The fundamental problem concerning the preferential targeting of ferritin tumors, as contrasted with normal tissue, has been explored in rodent hepatoma and in clinical hepato-

Figure 1. Gallium scan with positive Hodgkin's tumor infiltrates (arrows) due to increased transferrin and ferritin activity in the tumors.

225

cellular cancer [14-18]. The neovasculature of tumors provides the neces­sary access for radiolabeled antibody (lgG - 150,000/dL) penetration to allow the antibody to bind ferritin. Normal tissue lacking such neovascula­ture does not allow 1-131 antiferritin binding. A direct comparison of 1-131 normal IgG and 1-131 antiferritin IgG in experimental models demonstrated no difference in distribution in normal tissue [14-15]. The 1-131 antiferritin concentration in the tumor was three fold greater than the radio labeled normal IgG, thus demonstrating the role of specificity in the tumor [14].

Further evaluation of the tumor neovasculature has demonstrated that radiation enhances vascular permeability and increases radiolabeled anti­body deposition both experimentally and in the clinic [18]. Thus, "the biologic window" consists of the presence of a high concentration of ferritin (antigen), neovasculature of the tumor, and the vascular permeability [18-20]. Hopefully, future studies will find agents other than radiation to selec­tively increase tumor vasculature permeability.

1-131 antiferritin in advanced Hodgkin's disease

Our initial series of patients were those who failed two major courses of combined drug chemotherapy [21]. The patients were treated if they had no evidence of bone marrow invasion, adequate bone marrow function (defined as WBC > 3000/111, platelets> 100,000/111, liver function < 2 mg/dL), and adequate kidney functions (BUN < 25 mg/dL and creatinine < 1.5 mg/dL). An elevated alkaline phosphatase did not disqualify patients for the study [21 ].

All patients were assigned to a single treatment dose and schedule. Patients were treated with 30 mCi of 1-131 antiferritin by a bolus injection in an i.v. drip on day 0 and a second dose of 20 mCi on day 5. The scheduling was based on the known effective half-life of the radiolabeled antiferritin [22, 23]. Patients were discharged from the hospital when their total body irradiation fell below 1.8 mR/hr measured 1 meter from the patient. The average duration of time spent in the hospital to receive each two-dose cycle of treatment was eight days. The patients were then evaluated for response and toxicity on an outpatient basis. Responses were characterized as com­plete, partial, no change, or progression, using Eastern Cooperative Oncol­ogy Group criteria [21].

Two months after the injection of radioimmunoglobulin, a second cycle of 1-131 antiferritin was to be administered if the patient's blood counts were adequate. If the patient's platelet counts did not recover to 100,000/111 but had stablized between 50,000/I1L and 90,000/I1L, 1-131 was administered at a modified dose of 20 mCi on day 0 and 10 mCi on day 5. Each subsequent therapy was derived from a different animal species so that no patient received the same foreign protein from the same animal more than once in the initial phases of the study. The species used were rabbit, pig, and monkey.

226

Thirty-eight patients with advanced and progressive Hodgkin's disease who had relapsed from, or had not responded to, two potent curative combination chemotherapy regimens had a 40% partial remission to single agent 1-131 antiferritin (figures 2 and 3). Symptomatic response, with the relief of B-symptoms in particular, was recorded in 77% of the patients. Toxicity was restricted to bone marrow depression with thrombocytopenia being greater than leukopenia. Of particular note was the fact that only 25 of the 37 patients entered could achieve a second cycle [21]. Even then, the second cycle was delayed at least two weeks or more beyond the anticipated and desired time of administration [21].

90-yttrium antiferritin therapy in advanced Hodgkin's disease

Several conclusions evolved from the previous study. The hematologic damage from previous chemotherapy, i.e., MOPP, ABVD, etc., in com-

Figure 2. Resolution of gallium masses following 1-131 antiferritin treatment. This patient did not take his saturated solution of potassium iodide as directed, and therefore thyoid uptake was noted (arrow).

227

Figure 3. SPECT (single photon emission tomography) demonstrates lack of resolution of nodal disease on conventional scan. Double white line (arrow) represents tomographic section, and remaining scans demonstrate nodal localization (arrow) of 1-131 antiferritin .

bination with radiolabeled antibodies, would in the future require autologous marrow transplantation in order to reduce hematologic toxicity and to shorten cycle time if more powerful radiolabeled antibodies are to be used to achieve remission.

The known dose-response curve for daily fractionated radiation in Hodg­kin's disease, 4000 rad in four weeks, should require less dose with the continuous radiation of radiolabeled antibody [4]. 90-yttrium, a pure beta emitting isotope, when compared to 1-131, would yield up to three times the dose rate for an equal dose administered and at least another 50% total tumor dose if measurements were analogous to those measured in hepato­cellular cancer, where such evaluations are possible [18]. 111-indium antifer­ritin is always administered before 90-yttrium antiferritin, as it provides an estimate of distribution of the radiolabel as well as dosimetry to be deter­mined. The 111-indium is a gamma emitter, and this allows for external counting. The chelation for indium and yttrium are similar, and their biodis-

228

tributions are the same. In the present protocol, autologous marrow is stored in three aliquots and given 16 days after 90-yttrium antiferritin; rapid hematologic recovery is expected and treatment continues into the next cycle of therapy with the radiolabeled antibody. Shortening the cycle time of treatment due to autologous marrow infusion and increasing the dose rate and total tumor dose of radiation by use of 90-yttrium radiolabel should increase the response rate. In our first clinical experience this has been consistent.

The finding that external radiation increases radio labeled antibody depo­sition has been incorporated in this protocol, where 150 rad x 2 is given to lesions over 5 cm3 . The first patient treated with bilateral lymph nodes, 2 cm or less, and a mediastinal mass has had complete remission of the lymph node masses and most of the mediastinal mass after only two of the proposed three cycles of treatment. This has occurred without any significant hema­tologic toxicity, even though the third cycle of treatment has not been administered.

Potential new developments

Completion of a dose escalation trial with 90-yttrium antiferritin with autol­ogous marrow rescue will determine toxicity, response, and the value of the autologous marrow. In larger masses, lll-indium studies will allow dosimet­ric determination of tumor dose in Hodgkin's disease for the first time. With the chelated Ill-indium and 90-yttrium antiferritins, comparison of polyclonal and monoclonal antibody will also be accomplished, since the chelated isotopic antibody combinations are stable on monoclonal antibody [24].

Finally, it may be possible to consider drug integration and 90-yttrium antiferritin as a new second line of chemotherapy. Realizing that the radio­labeled antibodies cause no acute symptoms, the need for further investiga­tions is necessary for radio labeled antibodies to fulfill their possible potential in the treatment of Hodgkin's disease .

References

1. Order SE: The history and progress of serologic immunotherapy and radiodiagnosis. Radiology 118:219-223, 1976.

2. Ettinger OS, Order SE, Wharam MD, Parker MK , Klein JL, Leichner PK: Phase I-II study of isotopic immunoglobulin therapy for primary liver cancer. Cancer Treat Rep 66:289-297, 1982.

3. Ettinger OS, Dragon LH, Klein JL, Sgagias M, Order SE: Isotopic immunoglobulin in an integrated multi model treatment program for a primary liver cancer: A case report. Cancer Treat Rep 68 :131-134, 1979.

4. Kaplan HS: Hodgkin's Disease. Cambridge, Mass, Harvard University Press, 1980.

229

5. Katz DH, Order SE, Graves M, Benacerraf B: Purification of Hodgkin's disease tumor­associated antigens. Proc Natl Acad Sci USA 70:396-400, 1973.

6. Order SE, Colgan J, Hellman S: Distribution of fast and slow migrating Hodgkin's tumor associated antigen. Cancer Res 34:1182-1186, 1974.

7. Sarcione EJ, Smalley JR, Lenia MJ, et al: Increased ferritin synthesis and release by Hodgkin's disease peripheral blood lymphocytes. Int J Cancer 20:339-346, 1977.

8. McGuire RA, Pretlow TG, Wareing TH, et al: Hodgkin's cells and attached lymphocytes. Cancer 44:183-197, 1979.

9. Unger PD, Strauchen JA: Hodgkin's disease in AIDS complex patients. Cancer 58:821-825, 1986.

10. Blumberg B, Hann H, Mildvan D, et al: Iron and iron binding proteins in persistent generalized lymphadenopathy and AIDS. Lancet 347, 1984.

11. Order SE, personal communication. 12. Order SE, Porter M, Hellman S: Hodgkin's disease: Evidence for a tumor associated

antigen. N Engl J Med 285: 471-474, 1971. 13. Eshhar Z, Order SE, Katz D: Ferritin: A Hodgkin's disease associated antigen. Proc Natl

Acad Sci 71:3956-3960, 1974. 14. Rostock RA, Klein JL, Leichner PK, Kopher KA, Order SE: Selective tumor localization

in experimental hepatoma by radiolabeled antiferritin antibody. Inti J Rad Onc BioI Phys 9:1345-1350, 1983.

15. Rostock RA, Klein JL, Kopher KA, Order SE. 1984. Variables affecting the tumor localization of 1-131 antiferritin in experimental hepatoma. Am J Clin Oncol 6:9-18.

16. Rostock RA, Klein JL, Leichner PK, Order SE: Distribution of physiologic factors that affect 1-131 antiferritin tumor localization in experimental hepatoma. Int J Rad Oncol BioI Phys 10:1135-1141, 1984.

17. Rostock RA, Kopher KA, Bauer TW, Klein JL: Factors that affect antiferritin localization in four rat hepatoma models. Cancer Drug Deliv 2:139-145, 1985.

18. Order SE, Klein JL, Leichner PK, Frincke J, Lollo C, Carlo DJ: 90-yttrium antiferritin: A new therapeutic radiolabeled antibody. Int J Rad Oncol BioI Phys 12:277-281, 1986.

19. Order SE: Analysis, results and future prospective of the therapeutic use of radio labelled antibody in cancer therapy. In: Monoclonal Antibodies for Tumor Detection and Drug Targeting, London, Academic Press, 1986, pp 306-316.

20. Order SE: New radiotherapeutic approaches to hepatoma. In: Biliary and Hepatic Cancer, New York, Marcel Dekker, 1986.

21. Lenhard RE, Order SE, Spunberg JJ, Asbell SO, Leibel SA: Isotopic immunoglobulin: A new systemic therapy for advanced Hodgkin's disease. J Clin Oncol 3:1296-1300, 1985.

22. Leichner PK, Klein JL, Garrison JB, Jenkins RE, Nickoloff EL, Ettinger DS, Order SE: Dosimetry of 1-131 labeled antiferritin in hepatoma. A model for radio immunoglobulin dosimetry. Int J Rad Oncol BioI Phys 7:323-333, 1981.

23. Leichner PK, Klein JL, Siegelman SS, Ettinger OS, Order SE: Dosimetry of 1-131 labeled antiferritin in hepatoma: Specific activities in the tumor and liver. Cancer Treat Rep 6:647-657, 1983.

24. Order SE: Monoclonal antibodies potential in radiation therapy and oncology. Int J Rad Oncol. BioI Phys 8:1193-1201,1982.

7. Interferon therapy for lymphoproliferative disorders

Mark S. Roth, Paul A. Bunn, and Kenneth A. Foon

Introduction

Interferon was the term originally applied to a soluble factor that was recognized by its ability to induce interference against viral infection of chick chorio-allantoic membrane by influenza A virus [1]. It has subsequently been shown to be a family of closely related proteins and glycoproteins which, in addition to antiviral activity, are potent regulators of cellular function and structure and possess direct antiproliferative activities. These latter properties underlie the current interest in interferon as an anticancer agent.

Three major species of human interferon are recognized and designated interferon-a, interferon-~, and interferon-y [2] (table 1). Interferon-a is produced by leukocytes (B-cells, T-cells, null cells, and macrophages) upon exposure to B-cell mitogens, viruses, foreign cells, or tumor cells. Interferon-~ is produced by fibroblasts upon exposure to viruses or foreign nucleic acids. Interferon-y is produced by T-Iymphocytes upon stimulation with T-cell mitogens, specific antigens, or interleukin-2 [3]. By use of recombinant DNA techniques, complete nucleotide sequences for interferons a, ~, and y have been defined and amino acid sequences derived.

The genes recognized to code for interferon-a have been assigned to chromosome nine [4]. Sixteen distinct sequences for interferon-a have been described [4]. Each is approximately 166 amino acids in length, with an additional 20 amino acid secretory peptides present on the amino-terminal end. The human genes differ by approximately 10% in nucleotide sequence and 15%-20% in amino acid sequence [5]. Two recombinant human inter­ferons, aA and aD, comprise over 60% of interferons present after buffy coat stimulation and have been extensively studied [6]. While they possess different antiviral and antiproliferative activity in vitro, similar in vivo effects on immune effector cells have been observed [6]. The interferon-a used in the first human clinical trials was obtained from Sendai virus stimulated

BellI/eli, I .M. and Foon. K.A., (eds.), Immunologic Approaches 10 rhe Classificarion and Managemem of Lymphomas and Leukemias. © 1988 Kluwer Academic Publishers. ISBN978-1-4612-8965-4. All righrs reserved.

Tab

le J

. In

terf

eron

s in

clin

ical

use

Typ

e S

ubty

pe"

Sou

rce

(new

nom

encl

atur

e)

a L

euko

cyte

(IF

N-a

[L

ED

L

euko

cyte

s fr

om n

orm

al b

lood

L

ymph

obla

stoi

d L

ymph

obla

stoi

d (N

amal

va)

cells

(I

FN

-alf

a-N

l)

in c

ultu

re

Wel

lfer

onR

(B

urro

ughs

Wel

lcom

e C

o.)

Rec

ombi

nant

a2

Tra

nsfo

rmed

E.

coli

(I

FN

-alf

a-2b

) In

tron

R-A

(S

cher

ing

Cor

p.)

Rec

ombi

nant

aA

T

rans

form

ed E

. co

li

(IF

N-a

lfa-

2a)

Rof

eron

R-A

(H

offm

an-L

a R

oche

, In

c.)

Rec

ombi

nant

aD

T

ansf

orm

ed E

. co

li

(IF

N-a

D)

Rec

ombi

nant

a-2

arg

Tra

nsfo

rmed

E.

coli

(I

FN-a

lfa-

2c)

~ F

ibro

blas

t F

etal

for

eski

n fi

brob

last

in

cult

ure

(IFN

-~)

Rec

ombi

nant

~cy

s T

rans

form

ed E

. co

li

(rIF

N-~c

ys)

Rec

ombi

nant

~se

r T

rans

form

ed E

. co

li

(rIF

N-B

ser)

y Im

mun

e T

-Iym

phoc

ytes

fro

m n

orm

al b

lood

(I

FN

-y)

Rec

ombi

nant

y

Tra

nsfo

rmed

E.

coli

(r

IFN

-y)

• T

hese

cru

de p

repa

rati

ons

can

be p

urif

ied

to n

ear

hom

ogen

ety

(see

tex

t) .

• N

ew n

omen

clat

ure

was

pro

pose

d at

a j

oint

mee

ting

of

the

WH

O a

nd U

SAN

cou

ncil

in M

ay,

1985

. Pur

ity

<1

%*

<

1%

*

>9

5%

>9

5%

>9

5%

>9

5%

< 1<

}lo*

>9

5%

>9

5%

<1

%'

>9

5%

Am

ino

acid

dif

fere

nces

Arg

inin

e at

pos

itio

n 23

; de

leti

on a

t po

siti

on 4

4 w

hen

com

pare

d to

oth

er a

SU

btyp

es.

Lys

ine

at p

ositi

on 2

3;

dele

tion

at

posi

tion

44.

29 v

aria

tion

s fr

om a

A

Arg

inin

e at

pos

itio

n 23

; ar

gini

ne a

t po

siti

on 3

4.

Cys

tein

e at

pos

ition

17.

Ser

ine

at p

osit

ion

17.

tv

W

tv

233

buffy coat leukocytes and represented 1 % purity (106 units/mg protein)l [7]. Refinement in purification methods by use of high performance liquid chro­matography, two-dimensional polyacrylamide gel electrophoresis, and im­munoaffinity chromatography has allowed purification to homogeneity (l08 units/mg protein) [8, 9, 10]. The use of recombinant DNA techniques with splicing of the interferon-a gene into E. coli has further allowed for pure single species interferon-a in larger quantities.

Unlike interferon-a, only a single protein species has been identified for both interferon-~ and interferon-y [S]. Interferon-~ consists of 166 amino acids with 4S% homology of nucleotides and 29% amino acids compared to interferon-a [S]. Interferon-y consists of 146 amino acids in length and has approximately 12% amino acid sequence homology with interferon-a [11]. Interferon-y may exist in biologic fluids in a dimeric form [12].

Industrial scale production of interferon-~ and y has only recently been accomplished, and clinical trials are limited in number. Interferon-a, how­ever, has been extensively studied for the past decade in both basic science and clinical research, and it is among the most potent biologic agents ever administered to man. While antitumor activity has been seen both in vitro and in vivo in some solid malignancies (breast cancer, renal cell cancer, Kaposi's sarcoma, bladder cancer, ovarian cancer, and melanoma) [13, 14], the most impressive responses have occurred in the hematologic malig­nancies. A review of these results and proposed mechanisms of action is presented.

Clinical experience

A summary of clinical trials for the hematologic malignancies using inter­feron-a is presented in table 2. Some reported trials have used highly purified interferon-a (108 units/mg protein), while others have used crude preparations of interferon-a (106 units/mg protein). Impurities in the latter preparations include albumin, transferrin, and additional lymphokines. Despite these contaminants, toxicities and antitumor responses seen with both preparations have been similar. The major side effects have been those of a flu-like illness (fever, chills, muscle aches, headache, gastrointestinal upset, and fatigue). The onset of fever is generally 4-8 hours after the dose, with a duration of approximately 12 hours. With repeated administration, tachy-phylaxis to fever usually occurs, while fatigue and anorexia increase with dosage and duration of treatment and remain the usual dose-limiting toxicities. Other reported side effects include dose-related myelo-suppres­sion, elevated transaminases, paresthesias, anosmia, somnolence, confusion, and impotence in males. One case of interstitial nephritis has been reported [IS], and elevation of hepatic transaminases was the dose-limiting toxicity in

'One unit of interferon is roughly the amount which reduces viral replication in tissue culture by half.

Tab

le 2

. C

linic

al t

rial

s w

ith i

nter

fero

n-a:

Hem

atol

ogic

Mal

igan

cies

Tum

or

Num

ber

of

. Res

pons

e ra

tes

% t

otal

ev

alua

ble

pati

ents

re

spon

se

CR

P

R

MR

Hai

ry c

eill

euke

mia

* 15

8 22

86

44

96

N

on-H

odgk

in's

lym

phom

a lo

w-g

rade

92

9

30

6 42

in

term

edia

te-

and

high

-gra

des

36

1 4

2 14

H

odgk

in's

dis

ease

8

0 0

2 0

Cut

aneo

us T

-cel

l ly

mph

oma

20

2 7

2 45

C

hron

ic ly

mph

ocyt

ic l

euke

mia

67

0

12

18

Mul

tiple

mye

lom

a 22

4 3

41**

20

C

hron

ic m

yelo

geno

us l

euke

mia

68

8

40

7 81

E

ssen

tial

thr

ombo

cyth

emia

4

3 0

75

Acu

te l

euke

mia

62

0

19#

31

CR

= c

ompl

ete

resp

onse

; P

R =

par

tial

res

pons

e; M

R =

min

or r

espo

nse;

% t

otal

res

pons

e =

CR

+ P

Rin

umbe

r ev

alua

ble

pati

ents

. Ref

eren

ces

18-2

7

30

-36

3

0,3

2,3

4,3

6,3

7

34

38,

39

36

,41

-45

3

6,4

7-5

4

57

-59

60

61

-65

* C

ompl

ete

resp

onse

mea

ns a

bsen

ce o

f ha

iry

cells

in

the

bone

mar

row

and

nor

mal

izat

ion

of p

erip

hera

l bl

ood

whi

te c

ells

, pl

atel

ets,

and

ery

thro

cyte

s.

Par

tial

res

pons

e m

eans

a n

orm

aliz

atio

n of

per

iphe

ral

bloo

d w

hte

cells

, pl

atel

ets,

and

ery

thro

cyte

cou

nts

an

d>

50%

red

ucti

on i

n ha

iry

cells

in

the

bone

m

arro

w.

Min

or r

espo

nse

gene

rally

mea

ns i

mpr

ovem

ent

in h

emog

lobi

n to

mor

e th

an 1

0 gl

dl,

impr

ovem

ent

in p

late

lets

to

mor

e th

an 1

00,0

00 c

ells

/rd,

or

impr

ovem

ent

in n

eutr

ophi

ls t

o m

ore

than

100

0 ce

llsht

!' %

tot

al r

espo

nse

for

hair

y ce

ll le

ukem

ia i

nclu

des

min

or r

espo

nses

. *"

Com

plet

e re

spon

se a

nd p

arti

al r

espo

nse

not

avai

labl

e fr

om a

ll tr

ials

; %

tot

al r

espo

nse

incl

udes

all

resp

onse

s.

# M

ost

resp

onse

s w

ere

of s

hort

dur

atio

n.

N

VJ """

235

another study [16]. All of the above toxicities are reversible on cessation of the drug.

Hairy cell leukemia

Hairy cell leukemia is a well characterized lymphoproliferative disorder in which cells with lymphoid morphology and villous cytoplasmic projections infiltrate the bone marrow, blood, and reticuloendothelial system. It is of B­cell origin and usually presents with cytopenias [17]. The disease is often indolent, with median age of onset 50 years and a 5: 1 male to female pre­dominance. Standard initial therapy has been splenectomy that often restores hematologic parameters to normal, however, most of these patients relapse weeks to years post splenectomy. Treatment of relapses has been generally poor with standard cytotoxic agents. Excellent responses were reported [18] in seven patients with hairy cell leukemia (three complete and four partial) treated with crude interferon-a. Similar data have been reported by a number of investigators using recombinant interferon-a. Response rates have been comparable with recombinant preparations following three times per week therapy or daily therapy with doses ranging from 3-6 x 106 units intramus­cularly or subcutaneously [18-27]. While the initial report suggested that complete responses were frequent, this has not been confirmed, with only 22 of 158 complete responses reported [18-27]. More important , however, is that virtually all of the responding patients normalize their peripheral blood counts and maintain this while on interferon therapy. Many of these patients had no prior therapy, including splenectomy. Responding patients have not been reported to become refractory to interferon-a; many patients have been followed for over three years. In addition, improvement in natural killer activity and immunologic surface markers parallels the hematologic recovery [24]. In a recent study [20] conducted to address the issue of duration of treatment, interferon was discontinued in 25 patients after 12 months of treatment. In 8 of the 25 patients, a progressive decline in the granulocyte count and reappearance of circulating hairy cells occurred at a median of six months after cessation of treatment. Reinitiation of treatment resulted in reinduction of remission in five of the eight patients who have completed three months of administration. In the remaining 17 patients, a sustained hematologic remission remains at a follow-up of 6 to 22+ months. Interestingly, an initial increase in mean granulocyte and platelet counts for the first month post interferon-a cessation has also been reported, suggesting interferon-a causes myelosuppression [28]. Therefore, ongoing studies to assess low (3-4 x 106 units) vs. ultra low dose (0.3-0.4 x 106 units) interferon-a are currently underway. Phase III trials randomizing newly diagnosed patients with hairy cell leukemia to splenectomy or interferon-a are also underway. Although hairy cell leukemia accounts for less than 2% of all leukemias, its response to interferon-a makes it an ideal disease to study the putative mechanisms of activity that are addressed below.

236

Non-Hodgkin's lymphoma and Hodgkin's disease

The histologic classification of non-Hodgkin's lymphoma has recently under­gone reformulation from the commonly used Rappaport system. Based on prognosis and morphology, the histologic types of malignant lymphoma have been grouped into low-grade, intermediate-grade, and high-grade malignancy under the working formulation [29]. Although many chemo­therapy agents produce responses, patients with low-grade non-Hodgkin's lymphoma are not curable with currently available treatment. This, in com­bination with the indolent nature of the disease, leads to multiple episodes of treatment and relapse, with the patient eventually dying of unrelated causes, toxicity of therapy, progressive disease, or emergence of a more aggressive histology. The low-grade non-Hodgkin's lymphoma have shown responses to interferon-a [30-36]. Early results with crude interferon-a preparations reported responses to interferon-a in four of seven patients [30, 31]. In the largest series reported to date [32], previously treated pa­tients received recombinant leukocyte interferon-a at a dose of 50 x 106

u/m2 of body surface area intramuscularly three times weekly. Thirteen re­sponses were obtained (four complete responses and nine partial responses) among 24 evaluable patients with a median duration of response of eight months. The role of interferon-a in combination with standard cytotoxic agents is currently under investigation as first-line therapy.

Interferon-a has shown less effectiveness in the intermediate-grade and high-grade lymphomas. Thirty-six cases have been treated with both crude and recombinant interferon-a and five responses reported [30, 32, 34, 36, 37]. Further study of interferon-a in unfavorable non-Hodgkin's lymphoma may be warranted to establish which histologic subgroups might benefit from treatment.

Eight patients with advanced refractory Hodgkin's disease have been treated with crude interferon-a [34]. Only two brief minor responses were reported. In a recent study, with recombinant interferon-a, however, approx­imately 30% of patients with advanced refractory Hodgkin's disease have responded (E. Bonnem, personal communication).

Cutaneous T-cell lymphoma

Cutaneous T-cell lymphoma (mycosis fungoides and the Sezary syndrome) is a non-Hodgkin's lymphoma, characterized by a malignant proliferation of mature helper T-Iymphocytes, that presents with skin infiltration and an indolent clinical course. Effective therapies include topical mechlorethamine, psoralen plus ultraviolet light (PUV A), total skin electron beam irradiation, and systemic chemotherapy. Unfortunately, prolonged disease-free survival has been reported only rarely with these modalities, and the best response rates for advanced disease are reported to be about 25%, with short duration of response [38]. Responses in 9 of 20 patients (two complete, seven partial)

237

with advanced stages of disease refractory to prior therapy were observed [39] using recombinant interferon-a at a dose of 50 x 106 u/m2 body surface area intramuscularly three times weekly. Responses defined as at least 50% reduction in the sum of perpendicular measurements of malignant lesions lasting at least one month, occurred within four weeks of therapy and lasted three months to more than 25 months. Extracutaneous responses also oc­curred. Reduction in the size of large lesions by more than 90% occurred in a number of patients, suggesting interferon-a is perhaps the best single agent for cutaneous T-cell lymphoma.

Chronic lymphocytic leukemia

Chronic lymphocytic leukemia is a hematologic malignancy characterized by proliferation and accumulation of relatively mature-appearing lymphocytes. Most cases involve a clonal proliferation of B-lymphocytes [40]. Chronic lymphocytic leukemia typically occurs in persons over 50 years (median age, 60 years) and affects males more than females at a ratio of 2:1 [40]. The disease is usually stable over months to years, but transformation to a more aggressive disease state does occur. Alkylating agents, radiation therapy, and corticosteroids are commonly used to treat patients, although few data show that survival is substantially improved. In a number of early studies, crude interferon-a preparations were reported to be active in patients with advanced chronic lymphocytic leukemia [36, 41, 42]. In a phase II trial of recombinant interferon-a, 18 patients were treated with both high dose (50 x 106 u/m2 intramuscularly) and low dose (5 x 106 u/m2 intramuscularly) recombinant interferon-a three times weekly [43], with only two brief re­sponses reported. Five patients appeared to have an acceleration of disease while receiving recombinant interferon-a. This low response rate was con­firmed by a number of investigators [33, 34, 42, 44, 45]. This finding is in marked contrast to responses in patients with chemotherapy-refractory low­grade non-Hodgkin's lymphoma and hairy cell leukemia, as described above. The possible mechanism for this will be addressed below.

Multiple myeloma

Multiple myeloma is a disease of uncontrolled proliferation of malignant plasma cells in the marrow manifested clinically by tumor formation, oste­olysis, hemopoiesis, hypogammaglobulinemia with a paraprotein monoclonal spike, and renal disease. The mean age at the time of diagnosis is 62 years. Patients with multiple myeloma respond initially to a variety of chemothera­peutic agents, however, once patients become refractory to first-line therapy, further responses become difficult [46]. A number of trials with crude and recombinant interferon-a for patients with multiple myeloma have been reported [47-54]. In a pilot study, four previously untreated patients were treated with crude interferon-a 3 X 106 units intramuscularly daily. All patients obtained durable responses (two complete responses, two partial

238

responses) lasting 3 to 19 months [47]. This study was extended into a prospective randomized trial comparing interferon-a (crude) 3 X 106 units intramuscularly daily with melphalan/prednisone on a six-week schedule. Fifty-three patients were alloted to melphalan/prednisone and 62 patients to interferon-a. Total response rate was higher in the melphalan/prednisone group (41%) than in the interferon-a group (14%) (p < O.OS; response defined as > SO% reduction in paraprotein) [48].

Recombinant interferon-a has been administered in a number of trials [SO-S4]. Doses ranged from 2 x 106 units/m2 - 100 X 106 units/m2 daily. Only 22 of 122 previously treated patients responded, while 7 of 19 untreated patients responded. Of note is a recent observation of synergy between interferon-a and high dose chlorambucil in refractory myeloma [SS]. Further trials of combination alkylating agent and interferon are ongoing.

Chronic myelogenous leukemia

Chronic myelogenous leukemia is a neoplastic disease characterized by clonal proliferation of a myeloid stem cell . A unique chromosomal trans­location, the Philadelphia (PhI) chromosome, is present in about 90% of patients. The peak age of onset is 40 years. The clinical manifestation of the disease relates to accumulation in the blood and abdominal viscera of large numbers of immature and mature granulocytic cells. In most cases that proliferation of the hematopoetic cells can be suppressed for one to four years with cytotoxic agents, but over 80% of patients transform to an acute leukemia or blast crisis [S6]. In the acute phase, therapeutic agents including those useful in the treatment of acute leukemia are ineffective. Fifty-one patients in the chronic phase of their disease were treated with 3-9 x 106

units intramuscularly daily of crude (106 units/mg protein) interferon-a [S7]. Forty-one of the patients responded to therapy, achieving complete (36 patients) or partial (S patients) response in their peripheral blood. Respond­ing patients showed a gradual decrease of spleen size to normal and decrease in bone marrow cellularity. Suppression of the Ph' chromosome occured in varying degrees in 20 of Sl patients and was complete in two patients. Successful lowering of platelet counts in nine patients (all previously treated) with severe symptomatic thrombocytosis has also been demonstrated [S8] with crude interferon-a. A recent study using 5 X 106 units/m2 daily of recombinant interferon-a demonstrated 13 hematologic remissions and one partial hematologic remission among 17 patients [S9]. In six of the patients with hematologic remission, there was complete suppression of the Phi clone on at least one examination. While these are very exciting data, they are preliminary and will require confirmation.

Essential thrombocythemia

Essential thrombocythemia is a myeloproliferative disease defined by a platelet count generally in excess of one million per microliter, megakaryo-

239

cyte hyperplasia in the bone marrow, and the absence of a predisposing cause (i.e., Ph I chromosome, increased red cell mass, infection, or iron deficiency). Essential thrombocythemia usually appears between the ages of 50 and 70 years. The major morbidity of the disease is bleeding and thrombosis, with a 50% five-year survival. A number of agents have been effective e2p, L-phenylalanine mustard, busulfan, uracil mustard, and hydroxyurea) in lowering platelet counts. Recombinant interferon-a has been administered to four previously untreated patients with essential thrombocythemia at a dose of 5-10 x 106 units/day intramuscularly for 30 days [60]. Platelet counts returned to normal in three of the four patients. Maintenance interferon-a twice weekly was given after 30 days, and patients were followed up to 80 days without relapse. As no known leukemogenic potential axists for interferon-a, it may become a useful initial treatment of essential thrombocythemia.

Acute leukemia

Acute leukemia is a malignant stem cell disorder characterized by uncon­trolled growth of poorly differentiated lymphoblasts. Early studies with crude interferon-a were reported to produce responses in six of seven pa­tients with acute lymphoblastic leukemia and two of three with acute non­lymphoblastic leukemia at doses of 0.5 - 5 million units/kg intravenously daily for two weeks to two months. In phase I and II trials [63, 64], 53 patients were treated with partially pure lymphoblastoid interferon-a (5-200 x 106 units/m2 daily x 10 days). Five of 33 patients with leukemia experi­enced significant (80% -99%) drops in circulating blast counts, but bone marrow pathology revealed only three patients with any degree of improve­ment in bone marrow infiltration (two transiently and one for three months). Recombinant interferon-a (25-100 x 106 units/day x 7 days) was admini­stered to 13 heavily pretreated patients with only two minimal responses [65]. Interferon-a in the high doses used above has had limited effectiveness for management of acute leukemia. The potential role of lower dose inter­feron-a has yet to be determined.

Mode of action

The effect of interferon at the cellular level is initiated by binding of the interferon molecule to a cell-surface membrane receptor [66]. Competitive binding studies indicate that interferon-a and interferon-~ interact with one cell surface receptor, while interferon-y may interact with another receptor [66]. Following binding to the cell surface membrane, human interferon is rapidly internalized and degraded [67]. Whether this internalization is re­quired for the biological responses to interferon has not been resolved. Analogous to several polypeptide hormones and their target cells, a down

240

regulation of interferon receptors after exposure of cells in vitro to inter­feron occurs [67, 68]. Similar results are seen in patients after daily inter­feron treatment, suggesting that an interval treatment program (e.g., every other day) might be better than a continuous one, as it would allow for recovery of cell surface interferon receptors [68].

Direct and indirect mechanisms of anticancer activity of interferon will likely result from a number of different mechanisms, including induction of several intracellular proteins, enhancement of immune effector celis, and changes in cellular surface structure (table 3). Two enzymes appear to playa major role in interferon activity. Treatment of cells in culture with inter­feron results in an increase in 2' -5' -oligoadenylate synthetase (2-SA syn­thetase) [69, 70]; studies suggest that this response represents the induction of a gene which is subject to control by interferon [71]. 2-SA synthetase is capable of synthesizing a novel series of oligonucleotides 2' -S' oligo­adenylates in the presence of double stranded RNA and ATP. These oligo­nucleotides range from 2 to IS in length and are collectively referred to as 2-SA. 2-SA in turn activates a latent endoribonuclease which is capable of cleaving both viral and host RNA (messenger RNA and ribosomal RNA), effectively inhibiting transcription and translation [66]. 2-SA introduced into normal and neoplastic cells appears to inhibit both protein and DNA synthesis [72]. The second enzyme activated by exposure of cells to inter­feron is a protein kinase capable of phosphorylating peptide eukaryotic initiation factor (eIF-2a) and ribosome associated protein PI [66, 73]. Recent observations are suggestive that the interferon induced protein kinase is protein PI [73]. The net result of the kinase activation is the inhibition of peptide chain initiation. The exact role of these observations in relation to anticancer activity remains undetermined. Preliminary data exists correlat-

Table 3. Cellular events after treatment with interferon-a

Intracellular protein changes Increased 2-SA synthetase Increased protein kinase activity

Direct antiproliferative Antiproliferative effect on tumor cell lines Antiproliferative effect in vivo in murine tumors Antiproliferative effects on transplanted tumors in nude mice

lrnrnunornodulatory activities Enhance (low dose) or suppress (high dose) natural killer activity Augment antibody-dependent cellular cytotoxicity (ADCC) Enhance tumoricidal activity of macrophages Regulation of antibody production in B-cells Enhanced cytotoxic phase of mixed lymphocyte culture (MLC) Depressed Iymphoproliferative phase of mixed lymphocyte culture Increase expression of cell surface antigens, HLA-A; B, C, and B2 microglobulin Decreased oncogene expression

241

ing the levels of induced 2-5 synthetase with a-interferon administration [74), however, correlation with antitumor activity has not been made [75].

Antiproliferative effect

Interferon-a has anti proliferative activity on some malignant tumor cells. Dose-dependent in vitro inhibition of hematologic cell lines using interferon­a has been shown in Burkitt's lymphoma, lymphocytic lymphoma, acute myelogenous leukemia, chronic myelogenous leukemia, chronic lymphocytic leukemia, and multiple myeloma [76-80]. Interestingly, in comparative anti­proliferative studies interferon-a has shown a greater inhibitory effect in cells of hematopoietic origin than either interferon-13 or interferon-y using both crude and recombinant interferons [77, 78, 81]. Of note, non cycling tumor cells (Go-Gd appear to be a more sensitive target for the antipro­liferative activity of human interferon [82, 83].

Crude murine interferon-a preparations have been shown to inhibit the growth of transplantable tumors of diverse origins (melanoma, friend leu­kemia, osteogenic sarcoma, Lewis lung, Ehrlich ascites) [84-87]. In support of a direct anti proliferative effect are studies of transplanted human tumors in immunodeficient nude mice in which immunomodulatory effects of ad­ministered human interferon-a are minimal [88,89]. Dose-dependent growth inhibition is observed in these models and persists only for the duration of treatment [88, 89]. Evidence for direct antiproliferative effect in human trials is suggested in cutaneous T-cell lymphoma. Four of ten patients who had had a relapse while receiving a 10% maintenance dose responded after reescalation to a 100% dose [38].

Immunomodulatory activity

Immunomodulatory activities of interferon are also of considerable interest and possibly playa role in the anticancer effect. The first evidence of this indirect effect of interferon was demonstrated when mice inoculated with L1210 cells derived from an interferon-resistant clone but were still pro­tected by daily interferon treatment [90]. Since the resistant cells did not revert to interferon sensitive ones in vivo, these experiments were inter­preted as suggesting an antitumor effect was mediated by the host, rather than a direct effect on cell multiplication. Subsequently it has been shown that interferon-a can enhance as well as suppress cell mediated and humoral immune responses, which are believed to play an active role in tumor surveillance.

Natural killer activity and antibody dependent cellular cytotoxicity

Natural killer cells are a heterogeneous population of lymphocytes which are cytotoxic against several cell types and tumors in vitro [91]. Morphologically,

242

human natural killer cells are identified as large granular lymphocytes. Direct evidence exists that natural killer cells inhibit tumor growth in vivo in animals, while inhibition in humans has thus far been indirect and less consistent [92]. In in vitro systems, human natural killer cell cytolytic activa­tion is consistently enhanced in the presence of both crude and pure inter­feron-a [93-95]. There is conflicting evidence regarding the natural killer cell activating effects of interferon used in human therapy. Using both crude and pure interferon-a, many clinical studies have documented interferon­induced increases in natural killer cell activity [96-97], whereas other in­vestigators reported a lack of effect on natural killer activity or occasional depression of natural killer cytolytic activity by interferon [98, 99]. In order to define parameters governing this effect, in one study purified lympho­blastoid interferon was given in six doses ranging from 105 to 3 X 107 units intramuscularly weekly to cancer patients [100]. A negative correlation be­tween the amount of interferon injected and the natural killer cell activity was found, with cytolysis peaking 24 hours post injection of 3 x 106 units (threefold increase).

The exact mechanism by which interferon stimulates natural killer cell activity and by which natural killer cells lyse their targets is not fully under­stood. There is some evidence that suggests that interferon is able to induce differentiation of precursor cells into mature natural killer cells and to directly activate pre-existing mature natural killer cells [101, 102]. More recently, release of a natural killer cytotoxic factor has been demonstrated in supernatants of natural killer cells exposed to appropriate tumor target cells, and it is believed to be involved in natural killer mediated cytolysis [103]. Addition of interferon-a to human lymphocytes results in augmenta­tion of natural killer cytotoxic factor production [103]. Furthermore, in vitro studies suggest that interferon-a may be required for both the production of natural killer cytotoxic factor and for the modulation of its lytic activity [104].

The lysis of specific antibody coated target cells, known generally as the antibody-dependent cellular cytotoxicity, has also been shown to be mediated by large granular lymphocytes [105]. Several studies have indicated that both crude and purified interferon-a preparations are able to augment antibody dependent cellular cytoxicity responses mediated by human lymphocytes in vitro [93, 106]. This increase also occurred against target cells resistant to natural killer activity [107]. More recently, it has been shown that pure interferon-a enhances the antibody-dependent celular cytotoxicity of human polymorphonuclear leukocytes against several hematologic cell lines in vitro [108]. Interestingly, the effect was most pronounced when the IgG anti­bodies in the antibody-dependent cellular cytotoxicity reaction were present in suboptimal amounts, suggesting that, in vivo, interferon may playa role in initial immune response when IgG levels are still low [108]. Interferon may augment this activity by increasing the expression of FcG receptors on the lymphocyte cell surface, enhancing the binding of immunoglobulin­coated target cells [109]. Like the natural killer cell response in human

243

trials, antibody-dependent cellular cytotoxicity response in patients receiving pure interferon-a is variable [110].

Monocyte function

Monocytes or macrophages are bone marrow derived cells that have the capacity of phagocytosis and pinocytosis and, more recently, have been shown to be tumoricidal in vitro and in animal models [111]. Like the natural killer cell activation, interferon-a in vitro enhances tumoricidal monocyte function [94, 112]. Unlike the natural killer cell activation, human clinical trials with both crude and recombinant interferon-a have shown consistent activation of monocyte tumoricidal function [98, 99]. The exact mechanisms by which interferon activates monocytes in man remains un­known. Recent studies using recombinant interferons suggest that interferon acts as an inducer of macrophage Fc receptor-mediated phagocytosis [113]. Recombinant y-interferon was significantly more potent than either inter­feron-a or interferon-~. Some studies suggest that interferon-y is the major natural human lymphokine (known as macrophage-activating factor) cap­able of inducing monocyte-macrophage tumoricidal activity [114] . The role of activated macrophages in tumor surveillance or tumoricidal activity in humans is currently under investigation.

B-lymphocytes

In vitro and in vivo studies on the effect of interferon-a on immunoglobulin synthesis by B-cells demonstrate the importance of dose and time of ex­posure . Pretreatment with both crude and pure interferon-a of human peripheral blood B-lymphocytes before addition of mitogen enhances im­munoglobulin production, but interferon treatment after exposure to mitogen supresses production [115, 116]. Lower doses of interferon-a enhance ma­turation of B-cells, while at higher doses suppression occurs [117, 118]. The enhanced immunoglobulin production occurs even when peripheral blood lymphocytes are separated into T-cell and B-cell subpopulations prior to interferon administration, suggesting a direct effect on B-cells [115, 116]. The first evidence suggesting an effect of interferon in vivo on antibody formation came from studies with mice [119]. Preliminary studies in human trials have demonstrated minimal increases of immunoglobulin secretion at 30 x 106 units , but not at other doses [110]. Similar to the interferon effect on macrophages, gamma interferon has been shown to be a more potent regulator of antibody response than alpha or beta on an antiviral unit basis [120] .

T-cells

T-lymphocytes are the effector cells of cell mediated immunity, and they perform a variety of functions, including cellular cytotoxicity, helper and

244

suppressor activity, and the production of lymphokines. The effect of inter­feron on T-cells is complex in that some phases of the T-cell responses may be enhanced and others depressed. In vitro, interferon-a enhances the cyto­toxicity of the mixed lymphocyte cultures, however, proliferation is inhib­ited [121]. Both inhibition and activation of T-suppressor cells from mixed lymphocyte cultures has been observed [122, 123]. In clinical trials, both crude and recombinant interferon have been shown to depress lymphoprolif­erative response to mitogens and mixed lymphocyte culture [98, 99]. The importance of the effects of interferon on T-cells as it relates to antitumor effect is not known.

Modulation of cell surface antigens

Interferon induces a variety of changes in the cell surface, including in­creases in the expression of Fc receptors on lymphocytes and macrophages, which enhance tumoricidal activity [109, 113]. Consistent increased expres­sion of HLA antigens A, B, and C and the HLA subunit Brmicroglobulin is observed with interferons-a, -~, and -y both in vivo and in vitro [124, 125]. Only interferon-y has consistently increased expression of HLA-DR [126] and, moreover, human interferon-y, unlike -a or -~, is able to increase expression of HLA-A, B, and C proteins on the cell surface at concentra­tions which are considerably lower than those necessary to induce an anti­viral effect [127]. Since the HLA-DR system in humans appears to playa major role in the presentation of antigen for immune response [128], gamma interferon may have a more important role in treatment directed at cell surface proteins than either a or ~.

Oncogene expression

Neoplastic transformation of normal cells to malignant cells is now believed to be regulated by expression of cellular oncogenes. Rat fibroblast cells when exposed to the Rous sarcoma virus undergo malignant transformation resulting from the expression of the viral src oncogene. The product of this gene has been shown to be a tyrosine phosphokinase (pp60src) [129] that is capable of inducing this transformation. Treatment of Rous sarcoma virus transformed rat cells with rat crude interferon-a resulted in a 50% reduction in intracellular pp60src associated protein kinase activity and a more normal growth pattern [130]. Moreover, eH] leucine pulse labeling experiments showed that interferon worked by selectively reducing the synthesis of the src gene product [130].

Recombinant human interferon-a has been shown to decrease accumula­tion of the cellular myc oncogene messenger RNA in the Daudi cell line (Burkitt's lymphoma) [131]. The effect is dose-dependent and occurs before any inhibition of cell growth can be detected. Interestingly, no effect was seen on c-myc transcription rates, but rather an accelerated degradation of

245

c-myc mRNA was noted (67%-80% reduction in c-myc mRNA half-life) [131].

The effect of interferon-a (crude) on oncogene expression of peripheral blood cells in two patients with chronic myelogenous leukemia has also been studied [132]. While the expression of several oncogenes (sis, ras-Harvey, ras-Kirsten, and myc) remained unchanged during interferon therapy, a significant reduction in abl oncogene expression was detected within a few days after initiating treatment in both patients. The results of these three studies suggest another mechanism by which interferon may inhibit tumor growth.

Mechanisms of interferon activity in specific diseases

Hairy cell leukemia is the model disease to study the effects of interferon-a. Patients with hairy cell leukemia have a severe deficiency in natural killer cell activity. Recovery of natural killer activity has been reported [24, 133] in most patients with hairy cell leukemia following interferon-a therapy. The recovery of natural killer cells paralleled hematologic recovery. It remains unclear whether the natural killer cells played a direct role in hematologic recovery or were simply a byproduct of interferon-induced hematologic recovery. However, it was of interest that the low natural killer activity in the untreated cells was not really attributable to a relative deficiency or dilution of the effector cells, since the percent of Leu-ll + cells which identify the natural killer cells was within the normal range, suggesting that interferon-a activated these cells into functional effector cells [24]. In addi­tion to natural killer cell recovery, improvement in the total numbers of T-Iymphocytes , including both helper and suppressor populations, and monocytes paralleled the improvement in the other hematologic parameters following interferon-a therapy.

Hairy cell leukemia and low-grade lymphomas are both indolent diseases and B-cell in origin. Interferon-a has a high degree of activity in both dis­eases [18-27 , 30-36]. The lack of responsiveness of another indolent B-cell malignancy, chronic lymphocytic leukemia, has as yet been an unexplained finding [36, 41-45]. A comparison of binding of iodinated recombinant interferon-a to normal peripheral blood mononuclear cells, hairy cell leu­kemia cells, and chronic lymphocytic leukemia cells demonstrated that hairy cells bound approximately twice as much iodinated interferon as chronic lymphocytic leukemia and normal cells, however, the hairy cells had twice the surface area, which may explain the greater number of receptors [134]. This suggests that the responsiveness of a particular lymphoproliferative disease cannot be predicted or explained solely by the degree of interaction between interferon and its cell surface receptor.

Interferon-a has been reported to induce cell surface and intracellular proteins induced by interferon-a in patients with hairy cell leukemia [135].

246

Autoradiographic analysis of one-dimensional polyacrylamide gels showed induction of at least six proteins in nine patients treated with recombinant interferon-a. Overall protein synthesis was not significantly altered. Some of these proteins were in the cell membrane, leading the authors to suggest that interferon induces a protein signal in the hairy cell enabling their destruction [135].

Most recently Baldini and coworkers from Milan [136] isolated hairy cells from the spleen from previously untreated patients and cultured them in the presence of recombinant human interferon-a. Monoclonal antibody surface­marker studies revealed a significant enhancement of class II HLA antigens (HLA-DR). Since HLA antigens have been shown to be involved in cell­mediated cytotoxicity [128], they speculated that selective enhancement of class II HLA antigen may be another in vivo therapeutic mechanism of interferon-a.

Conclusion

The relative importance of interferon as a direct antitumor agent or a biological response modifier remains an unanswered question in the treat­ment of malignant diseases. While it is clear that interferon will not be effective in the majority of cancers, we have reviewed herein the effective­ness it can have in managing some of the hematologic malignancies. Even in these diseases, we don't know the optimal dose of interferon to use. While high doses may have greater direct anti proliferative activity, they may in fact suppress the immune system, while low doses may be more effective in enhancing the immune system. The role of interferon as a first-line treat­ment, or in combination with standard cytotoxic drugs or other biological response modifiers, are areas of ongoing research. Regardless of the even­tual role of interferon-a in the treatment of cancer, it is an important first member of a family of biological response modifiers used in treating human malignancies.

References

1. Isaacs A, Lindenman J: Virus interference: I. The interferon. Proc R Soc Land Bioi 147:258-267, 1957.

2. Stewart WE II, Blalock JE, Burke DC, et al: Interferon nomenclature. J Immunol (letters to the editor) 125:2353, 1980.

3. Fleischmann WR, Klimpel GR, Tyring SK, Voss WR, Baron S: Interferon and cancer: Current use and novel approaches. In: Sunkara PS (ed): Novel Approaches to Cancer Chemotherapy, Orlando, Academic Press, 1984, pp 1-22.

4. Sehgal PB: The interferon gene. Biochim Biophys Acta 695:17-33, 1982. 5. Borden EC: Progress toward therapeutic application of interferons, 1979-1983. Cancer

54:2770-2776, 1984.

247

6. Hawkins MJ, Borden EC, Merritt JA, et al: Comparison of the biologic effects of two recombinant human interferons alpha (rA and rD) in humans. J Clin Oncol 2:221-226, 1984.

7. Cantell K, Hirvonen S: Preparation of human leukocyte interferon for clinical use. Texas Rep Bioi Med 35:138-144,1977.

8. Pestka S: The purification and manufacture of human interferons. Sci Am 249:37-43, 1983.

9. Lin LS, Stewart WE II: Purification of human leukocyte interferon by two-dimensional polyacrylamide gel electrophoresis. Methods Enzymol 78:481, 1981.

10. Berg K, Heron I: Antibody affinity chromatography of human leukocyte interferon. Methods Enzymol 78:487, 1981.

11. Epstein LB: Interferon-gamma: Success, structure and speculation. Nature 295-454, 1982.

12. Yip YK, Pang RH, Urban C, Vilcek J: Partial purification and characterization of human gamma (immune) interferon. Proc Nat! Acad Sci USA 78:1601-1605, 1981.

13. Kirkwood JM, Ernstoff MS: Interferons in the treatment of human cancer. J Clin Oncol 2:336-352, 1984.

14. Bonnem EM, Spiegel RJ: Interferon-a: Current status and future promise. J Bioi Resp Modif 3:580-598, 1984.

15. Auerbuch SD, Austin HA, Sherwin SA, et al: Acute interstitial nephritis with nephrotic syndrome following recombinant leukocyte A interferon therapy for mycosis fungoides. N Engl J Med 301 :32-5, 1984.

16. Sherwin SA, Knost JA, Fein S, et al: A multiple-dose phase I trial of recombinant leukocyte A interferon in cancer patients. JAMA 248:2461-6, 1982.

17. Golomb HM: Hairy cell leukemia: An unusuallymphoproliferative disease. A study of 24 patients. Cancer 42:946-956, 1978.

18. Quesada JR, Reuben J, Manning JT, Hirsch EM, Gutterman JY: Alpha interferon for induction of remission in hairy cell leukemia. N Engl J Med 310:15-18, 1984.

19. Quesada JR, Hersh EM, Gutterman JU: Treatment of hairy cell leukemia with alpha interferon (abstract). Proc Am Soc Clin Oncol 3:207, 1984.

20. Quesada JR, Hersh EM, Manning JT, et al: Treatment of hairy cell leukemia with recombinant a-interferon. Blood 68:493-497, 1986.

21. Ratain MJ, Golomb HM, Vardiman JW, Vokes EE, Jacobs RH, Daly K: Treatment of hairy cell leukemia with recombinant alpha2 interferon. Blood 65:644-8, 1985.

22. Jacobs AD, Champlin RE, Golde DW: Recombinant a-2-interferon for hairy cell leu­kemia. Blood 65:1017-1020,1985.

23. Thompson JA, Brady J, Kidd P, Fefer A: Recombinant a-2-interferon in the treatment of hairy cell leukemia. Cancer Treat Rep 69:791-3, 1985.

24. Foon KA, Maluish AE, Abrams PG, et al: Recombinant leukocyte A interferon therapy for advanced hairy cell leukemia: Therapeutic and immunologic results. Am J Med 80:351-356, 1986.

25. Worman CP, Catovsky D, Bevan PC, et al: Interferon is effective in hairy-cell leukemia. Br J Haem 60:759-763, 1985.

26. Habermann T, Hoagland H, Chang M, Phylidy R: A phase II trial of alpha recombinant interferon in hairy cell leukemia in patients with advanced disease (abstract). Blood (suppl 1) 66:2009, 1985.

27. Castaigne S, Sigaux F, Cantell K, et al: Interferon alpha in the treatment of hairy cell leukemia. Cancer 57:1681-1684, 1986.

28. Ratain M, Golomb H, Vardiman J, et al: Durability of reponses to a-2-interferon in advanced hairy cell leukemia (abstract). Proc Am Soc Clin Oncol 5:163, 1986.

29. The non-Hodgkin's lymphoma pathologic classification project: National Cancer Institute sponsored study of classifications of non-Hodgkin's lymphomas. Summary and description of a working formulation for clinical use. Cancer 49:2112-2135, 1982.

30. Merigan TC, Sikora K, Bredden JH, Levy R, Rosenberg SA: Preliminary observations on

248

the effect of human leukocyte interferon in non-Hodgkin's lymphoma. N Engl J Med 299:1449-1453, 1978.

31. Louie AC, Gallagher JG, Sikora K, Levy R, Rosenbery SA, Merigan TC: Follow-up observations on the effect of human leukocyte interferon in non-Hodgkin's lymphomas. Blood 58:712-718, 1981.

32. Foon KA, Sherwin SA, Abrams PG: Treatment of advanced non-Hodgkin's lymphoma with recombinant leukocyte A interferon. N Engl J Med 311:1148-1152, 1984.

33. O'Connell MJ, Colgan JP, Oken MM, Ritts RE, Kay NE, Itri LM: Clinical trial of recombinant leukocyte A interferon as initial therapy for favorable histology non­Hodgkin's lymphomas and chronic lymphocytic leukemia. J Clin Oncol 4: 128-136, 1986.

34. Horning SJ, Merigan TC, Krown SE, et al: Human interferon alpha in malignant lym­phoma and Hodgkin's disease. Cancer 56:1305-1310, 1985.

35. Leavitt RD, Kaplan R, Ozer H: Low dose alpha-3 interferon treatment for low grade non­Hodgkin's lymphoma (abstract). Blood (Supplement I) 64:182a, 1984.

36. Gutterman JV, Blumenschein GR, Alexanian R: Leukocyte interferon-induced tumor regression in human metastatic breast cancer, multiple myeloma, and malignant lym­phoma. Ann Intern Med 93:399-406, 1980.

37. Leavitt RD, Ratanatharathorn Y, Ozer H, Rudnick S, Ferraresi R: A phase II study of recombinant alpha 2 interferon in patients with malignant lymphoma with unfavorable histology (abstract). Proc Am Soc Clin Oncol 2:54, 1983.

38. Bunn PA, Foon KA, Ihde DC: Recombinant leukocyte A interferon: An active agent in advanced cutaneous T-cell lymphomas. An Intern Med 101:484-7, 1984.

39. Bunn PA, Idhe DC, Foon, KA: The role of recombinant interferon alpha-2a in the therapy of cutaneous T-cell lymphomas. Cancer, 57:1689-1695, 1986.

40. Gale RP, Foon KA: Chronic lymphocytic leukemia, recent advances in biology and treat­ment. Ann Intern Med 103:101-120, 1985.

41. Misset JL, My the G, Gastiaburu J, et al: Treatment of leukemias and lymphomas by interferons: Phase II trial of chronic lymphoid leukemia by human interferon alpha. Biomed Pharmacother 36:112-6, 1982.

42. Huang A, Laszlo J, Brenckman W: Lymphoblastoid interferon (wellferon) trial in chronic lymphocytic leukemia. Proc Am Assoc Cancer Res 23:113, 1982.

43. Foon KA, Bottino G, Abrams PG: Phase II trial ofrecombinant leukocyte A interferon in patients with advanced chronic lymphocytic leukemia. Am J Med 78:216-220, 1985.

44. Ozer H, Leavit R, Ratanatharathorn Y: Experience in the use of DNA alpha-2 interferon in the treatment of malignant lymphoma (abstract). Am Soc Hematol 62:211a, 1983.

45. Schulof RS, Lloyd MI, Stallings JJ, et al: Recombinant leukocyte A interferon in B-cell chronic lymphocytic leukemia: In vivo effects on autologous antitumor immunity. J Bioi Response Mod 4:310-323, 1985.

46. Bergsagel D: Plasma cell myeloma. In: Williams W, Beutler E, Ersleu A, Lichtman M (eds): Hematology, New York, McGraw Hill, 1983, pp 1078-1099.

47. Mellstedt H, Bjorkholm M, Johansson B, Ahre A, Holm G, StrandeI' H: Interferon therapy in myelomatosis. Lancet 1 :245-8, 1979.

48. Bjorkholm M: Daily low dose interferon versus intermittent high dose melphalan/ prednisone treatment in multiple myeloma (abstract). Proc Am Soc Clin Oncol 2:242, 1983.

49. Osserman EF, Sherman WH, Alexanian R, Gutterman J: Preliminary results of the American Cancer Society (ACS) sponsored trial of human leukocyte interferon (IF) in multiple myeloma (MM) (abstract). Am Assoc Can Res 21:161,1980.

50. Ohno R, Kimura K: Treatment of multiple myeloma with recombinant interferon alpha-2a. Cancer 57:1685-1688, 1986.

51. Costanzi J, Ozer H, Scarffe H: A Phase II trial of rDNAurinterferon in multiple mye­loma. Presented at the International Symposium on Progress and Controversies in Multi­ple Myeloma, Padua, Italy, April 12, 1984.

52. Quesada JR, Alexanian R, Gutterman JV: Induction of remission and restoration of

249

serum immunoglobulins in multiple myeloma with recombinant DNA-derived Interferon (rIFNaA) (abstract). Blood (supplement I) 64:183a, 1984.

53. Wagstaff 1, Loynds P, Scarffe JH: Phase II study of rDNA human alpha-2 interferon in multiple myeloma. Cancer Treat Rep 69:495-498, 1985.

54. Case DC, Sonneborn HL, Paul SO, et al: Phase II study of rDNA alpha-2 interferon (Intron A) in patients with multiple myeloma utilizing an escalating induction phase (abstract). Blood (Suppl 1), 66:213a, 1985.

55. Clark RH, Dimitrov NV, Axelson lA, Charamella LJ: Leukocyte interferon as a possible biological response modifier in Iymphoproliferative disorders resistant to standard therapy. J BioI Resp Modif 3:613-619, 1984.

56. Champlin RE, Golde DW: Chronic myelogenous leukemia: Recent advances. Blood 65:1034-1047,1985.

57. Talpaz M, Kantarjian H, McCredie Ml, et al: Clinical study of human alpha interferon in chronic myelogenous leukemia (abstract). Blood (Suppl 1) 66:209a, 1985.

58. Talpaz M, Mavligit G, Keating M, Walters RS, Gutterman lU: Human leukocyte inter­feron to control thrombocytosis in chronic myelogenous leukemia . Ann Intern Med 99:789-792,1983.

59. Talpaz M, Kantarjian HM, McCredie K, Trujillo 1M, Keating Ml, Gutterman lU: Hematologic remission and cytogenetic improvement induced by recombinant human interferon alpha A in chronic myelogenous leukemia. N Engl 1 Med 314:1065-1069, 1986.

60. Velu T , Delwiche F , Flument 1, Monsieur R, Stryckmans P, Wybran J: Therapy of essential thrombocythemia with human a2 recombinant interferon (abstract) . Blood (Sup­plement 1) 64:176a, 1984.

61. Hill NO, Pardue A, Khan A, Aleman C, Hilario R, Hill JM: Clinical trials of human leukocyte interferon in malignancy. Texas Rep Bioi Med 41:634-638, 1982.

62. Hill NO, Pardue A, Khan A, Aleman C, Dorn G, Hill JM: Phase I human leukocyte interferon trials in cancer and leukemia. J Clin Hem Onco 11:23-25, 1981.

63. Rohatiner AZ, Balkwill FR, Griffin DB, Malpas JS, Lister TA: A phase I study of human Iymphoblastoid interferon administered by continuous intravenous infusion. Cancer Chemother Pharmocol 9:97-102 , 1982.

64. Rohatiner AZ, Balkwill FR, Malpas JS, Lister TA: Experience with human Iympho­blastoid interferon in acute myelogenous leukemia. Cancer Chemother Pharmacol 11:56-58, 1983.

65. Leavitt RD, Duffey P, Wiernik PH: A phase 1111 study of recombinant leukocyte-A interferon in previously treated acute leukemia (abstract) . Am Soc Hematol 62:205a, 1983.

66. Williams BRG: Biochemical actions of interferon. In: Sikora K (ed): Interferon and Cancer, New York, Plenum Press, 1983, pp 33-52.

67. Feinstein S, Traub A, LaZar A, Mizrahi A, Teitz Y: Studies on cell binding and internali­zation of human Iymphoblastoid interferon. 1 IFN Res 5:65-7, 1985.

68. Lau AS, Hannigan GE, Freedman MH, Williams BR: Regulation of interferon receptor expression in human blood lymphocytes in vitro and during interferon therapy. J Clin Invest 77: 1632-1638, 1986.

69. Ball LA: Induction of 2' -5'-oligoadenylate synthetase activity and a new protein by Chick interferon. Virology 94:282-296, 1979.

70. Ball LA: 2'-5'-0Iigoadenylat synthetase, In: PD Boyer (ed): The Enzymes vol XV. New York Academic Press 1982, pp 281-313.

71. Merlin G, Chebath 1, Benech P, Metz R, Revel M: Molecular cloning and sequence of partial cDNA for interferon-induced (2'-5') oligo(A) synthetase mRNA from human cells. Proc Nat! Acad Sci USA 80:4904-4908, 1983.

72. Revel M, Kimchi A, Shulman L, et al: Role of interferon induced enzymes in the antiviral and anti mitogenic effects of interferon. Ann NY Acad Sci 350:349-472, 1980.

73. Bischoff lR, Samuel CE: Mechanism of interferon action . 1 BioI Chern 260:8237-8239, 1985 .

250

74. Schattner A, Merlin G , Wallach D , et al: Monitoring of interferon therapy by assay of 2' , 5' -oligoisoadenylate synthetase in human peripheral white blood cells. J IFN Res; 1:587-594, 1981.

75. Merritt JA, Borden EC, Ball LA: Measurement of 2' , 5'-oligadenylate synthetase in patients receiving interferon-alpha. J IFN Res 5:191-198, 1985.

76. Balkwill FR, Oliver RTD: Growth inhibitory effects of interferon on normal and malig­nant human haemopoietic cells. Int J Cancer 20:500-505, 1977.

77. Borden EC, Hogan TF, Voelkel JG: Comparative antiproliferative activity in vitro of natural interferons 0. and ~ for diploid and transformed human cells. Cancer Res 42:4948-4953, 1982.

78. Chadha KC, Srivastava BI: Comparison of the antiproliferative effects of human fibroblast and leukocyte interferons on various leukemic cell lines. J Clin Hem One 11:55-60, 1981.

79. Salmon SE, Durie BG, Young L, Liu RM, Trown P, Stebbing N: Effects of cloned human leukocyte interferons in the human tumor stem cell assay . J Clin Oncol 1:217-225, 1983.

80. Denz H , Lechleitner M, Marth CH, Daxenbichler G , Gast! G , Braunsteiner H: Effect of human recombinant alpha-2- and gamma-interferon on the growth of human cell lines from solid tumors and hematologic malignancies. J IFN Res 5:147-157, 1985.

81. Blalock J, Georgiades JE, Langford MP, Johnson HM: Purified human immune inter­feron has more potent anticellular activity than fibroblast or leukocyte interferon. Cell Immun 49:390-394, 1980.

82. Horoszewicz JS, Leong SS, Carter WS: Noncycling tumor cells are sensitive targets for the antiproliferative activity of human interferon. Science 206:1091-1093, 1979.

83. Creasey AA, Batholomew JC, Merigan TC: Role of GO-Gl arrest in the inhibition of tumor cell growth by interferon. Proc Nat! Acad Sci USA 77:1471-1475, 1980.

84. Bart RS, Porzio NR, Kopf AW, Vilcek JT, Cheng EH, Farcet Y: Inhibition of growth of B-16 murine malignant melanoma by exogenous interferon. Cancer Res 40:614, 1980.

85. Rossi GB, Marcheglani M, Matarese GP and Gresser I : Brief communication : Inhibitory effect of interferon on multiplication of friend leukemia cells in vivo . J Nat! Cancer Inst 54:993, 1975.

86. Crane JL, Glasgow LA, Kern ER, Youngner JS: Inhibition of murine osteogenic sar­comas by treatment with type I or type II interferon. J Nat! Cancer Inst 3:871, 1978.

87. Greseer I, Tovey M: Antitumor effects of interferon. Biochem Biophys Acta 516:231-247, 1978.

88. Yoshitake Y, Kishida T, Esaki K, Kawamata J: Antitumor effects of interferon on transplanted tumors in congenitally athymic nude mice . Giken J 19:125-7, 1976.

89. Balkwill FR, Moodie EM, Freedman V, Frantes KH: Human interferon inhibits the growth of established human breast tumors in the nude mouse. Int J Cancer 30:231-235, 1982.

90. Gresser I: Antitumor effects of interferon. In: Klein G, Weinhouse S (eds): Advances in Cancer Research, New York, Academic Press, 16:97-140, 1972.

91. Barlozzari T, Leonhardt J, Wiltrout RH , Herberman RB, Reynolds CW: Direct evidence for the role of LGL in the inhibition of experimental tumor metastases. J Immunol 134:2783-2789, 1985.

92. Herberman RB, Ortaldo JR: Natural killer cells: Their role in defenses against diseases. Science 214:24-30, 1981.

93. Herberman RB, Ortaldo JR, Bonnard G: Augmentation by interferon of human natural and antibody-dependent cell mediated cytotoxicity. Nature 277:221-223 , 1979.

94. Herberman RB, Ortaldo JR, Rubinstein M, Pestka S: Augmentation of natural and antibody-dependent cell mediated cytotoxicity by pure human leukocyte interferon. J Clin Immunol 1:149-153, 1981.

95. Herberman RB, Ortaldo JR, Mantovani A, Hobbs DS, Kung H-F, Pesta S: Effect of human recombinant interferon on cytotoxic activity of natural killer cells and monocytes. Cell Immunol 67:160-167, 1982.

96. Huddleston JR, Merigan TC, Oldstone MBA: Induction and kinetics of natural killer cells

251

in humans following interferon therapy. Nature 282:417-419, 1979. 97. Borden EC, Holland JF, Dao T, et al: Leukocyte-derived interferon (alpha) in human

breast carcinoma. The American Cancer Society phase II trial. Ann Intern Mcd 97: 1-6, 1982.

98. Maluish AE, Ortaldo JR, Sherwin SA, Oldham RK, Herberman RB: Function in patients receiving natural leukocyte interferon. J Bioi Resp Modif 2:418-427, 1983.

99. Maluish AE, Leavitt R, Sherwin SA, Oldham RK, Herberman RB: Effects of recom­binant interferon-a on immune function in cancer patients. J Bioi Resp Modif 2:470-81, 1983.

100. Edwards BS, Merritt JA, Fuhlbrigge RC, Bordon EC: Low doses of interferon alpha result in more effective clinical natural killer cell activation. J Clin Invest 75:1908-1913, 1985.

101. Targan S, Dorey F: Interferon activation of 'pre-spontaneous killer' cells (pre-sk) and alteration in kinetics of lysis of both 'pre-sk' and active sk cells. J Immunol 124:2157, 1980.

102. Timonen T, Ortaldo JR, Herberman RB: Analysis by a single cell cytotoxicity assay of natural killer (NK) cell frequencies among human large granular lymphocytes and of the effects of interferon on their activity. J Immunol 128:2514, 1982.

103. Wright SC, Bonauida B: Role of natural killer cytotoxic factors (NKCF) in the mechanism of NK cell mediated cytotoxicity. In: Herberman RB (ed): NK Cells and Other Natural Effector Cells, New York, Academic Press, pp 961-968.

104. Steinhauer EH, Doyle AT, Kadish AS: Human natural killer cytotoxic factor (NKCF): Role of IFN-a. J Immunol 135:294-299, 1985.

105. Timonen T, Ortaldo JR, Herberman RB: Characteristics of large granular lymphocytes and relationship to natural killer and killer cells. J Exp Med 153:569-582, 1981.

106. Masucci MG, Sziget R, Klein E, et al: Effect of interferon-a 1 from E. coli on some cell functions. Science 209: 1431-5, 1980.

107. Ortaldo JR, Pestka S, Slease RB, Rubinstein M, Herberman RB: Augmentation of human K-cell activity with interferon. Scand J Immunol 12:365-369, 1980.

108. Hokland P, Berg K: Interferon enhances the antibody-dependent cellular cytotoxicity of human polymorphonuclear leukocytes. J Immunol 127:1585-8, 1981.

109. Djeu JY: Regulation of cell functions by interferon. In: Zoon KC, Noguchi PO, Liu T-Y (eds): Interferon: Research, Clinical Application, and Regulatory Consideration, New York, Elsevier, 1984, pp 125-131.

110. Ozer H, Gavigan M, O'Malley J, et al: Immunomodulation by recombinant interferon-a2 in a phase I trial in patients with lymphoproliferative malignancies. J Bioi Resp Modif 2:499-515, 1983.

111. Rosenstreich DL: The macrophage. In: Oppenheim JJ, Rosenstreich DL, Potter M (eds): Cell Functions in Immunity and Inflammation. New York, Elsevier, North Holland, 1981, p 127.

112. Sone S, Utsugi T, Shirahama T, Ishii K, Mutsuura S, Mitsumasa 0: Induction by interferon-a of tumoridical activity of adherent mononuclear cells from human blood: Monocytes as responder and effector cells. J Bioi Resp Modif 4:134-140, 1985.

113. Fertsch D, Vogel SN: Recombinant interferons increase macrophage Fe receptor capacity. J Immunol 132:2436-2439, 1984.

114. Sadlik JR, Hoyer M, Leyko MA, et al: Lymphocyte supernatant-induced human mono­cyte tumoricidal activity: Dependence on the presence of y-interfcron. Cancer Research 45:1940-1945,1985.

115. Harfast B, Huddleston JR, Casali P, Merigan TC, Oldstone MB: Inteferon acts directly on human B lymphocytes to modulate immunoglobulin synthesis. J Immunol 127:2146-2150, 1981.

116. Rodriguez MA, Prinz WA, Sibbitt WL, Bankhurst AD, Williams RC: a-interferon in­creases immunoglobulin production in cultured human mononuclear leukocytes. J Im­munol 130:1215-1219, 1983.

252

117. Choi YS, Lim KH, Sanders FK: Effect of interferon-a on pokeweed mitogen-induced differentiation of human peripheral blood B lymphocytes. Cell Immunol 64:20-28, 1981.

118. Fleisher TA, Attallah AM, Tosato 0, Blaese RM, Greene WC: Inhibition of human polyclonal immunoglobulin synthesis. J Immunol 129:1099-1103, 1982.

119. Braun W, Levy HB: Interferon preparations as modifiers of immune responses. Proc Soc Exp Bioi Med 141:769-773, 1972.

120. Sonnefeld G: Effects of interferon on antibody formation. In: Vilcek J, DeMaeyer E (eds): Inteferon, Volume 2: Interferons and the Immune System, Amsterdam, 1984, Elsevier Science Publishers pp 85-99.

121. Heron I, Berg K, Cant ell K: Regulatory effect of interferon on T-cells in vitro. J Immunol 117:1370-1373,1976.

122. Fradelizi D, Gresser I: Interferon inhibits the generation of allospecific suppressor T lymphocytes. J Exp Med 155:1610-1622, 1982.

123. Schnaper HW, Aune T, Pierce C: Suppressor T cell activation by human leukocyte interferon. J Immunol 131:2301-2306, 1983.

124. Gresser I: The effect of interferon on the expression of surface antigens. In: Vilcek J, DeMaeyer E (eds): Interferons and The Immune System. Elsevier Amsterdam, Science Publishers, 1984, pp 113-132.

125. Heron I, Hokland M, Berg K: Enhanced expression of B2-microglobulin and HLA antigens on human lymphoid cells by interferon. Proc Nat! Acad Sci 75:6215, 1978.

126. Kelley VE, Fier W, Strom TB: Cloned human interferon-y, but not interferon-~ or a, induces expression of HLA-DR determinants by fetal monocytes and myeloid leukemic cell lines. J Immunol 132:240, 1984.

127. Wallach D: The HLA proteins and a related protein of 28 KdA are preferentially induced by interferon-y in human WISH cells; Eur J Immunol 13:794, 1983.

128. Meur SC, Schlossman SF, Reinherz EL: Clonal analysis of human cytotoxic T-lympho­cytes T-4+ and T-8+ effector T-cells recognize products of different major histocompat­ability complex regions. Proc Nat! Acad Sci USA 79:4395, 1982.

129. Erikson RL, Purchio AF, Erikson E, Collet MS, Brugge JS: Molecular events in cells transformed by Rous sacrcoma virus. J Cell Bioi 87:319-325, 1980.

130. Lin SL, Garber EA, Wang E, et al: Reduced synthesis of pp 60 and expression of the transformation-related phenotype in interferon-treated Rous sarcoma virus-transformed rat cells. Mol Cell Bioi 3:1656-1664,1983.

131. Dani CH, Mechti N, Piechaczyk M, Leblcu B, Jeanteur PH, Blanchard JM: Increased rate of degradation of c-myc mRNA in interferon-treated Daudi cells. Proc Natl Acad Sci, 82:4891-4899, 1985.

132. Strayer DR, Gillespie DH, Bressuer J, Brodsky I: Oncogene expression decreased in two patients treated with interferons (abstract). Blood (supplement I) 64:175a, 1984.

133. Semenzato G, Pizzolo G, Agostini C, et al: a-interferon activates the natural killer system in patients with hairy cell leukemia. Blood 68:293-296, 1986.

134. Faltynek CR, Princler GL, Rusetti FW, Maluish AE, Abrams PG, Foon KA: Relation­ship of the clinical response and binding of recombinant interferon alpha in patients with lymphoproliferative diseases. Blood 67:1077-1082, 1986.

135. Samuels BL, Brownstein BH, Golomb HM: Effect of interferons on patterns of protein synthesis in hairy cells (abstract). Proc Am Assoc Can Res 26:20, 1985.

136. Baldini L, Cortelezzi A, Polli N, et al: Human recombinant interferon a-2C enhances the expression of class II HLA antigens on hairy cells. Blood 67:458-464, 1986.

8. Monoclonal antibody therapy of lymphomas and leukemia

Mark S. Kaminski and Kenneth A. Foon

Passive immunotherapy using heteroantisera for the treatment of cancer in animals and humans has been studied for over 50 years. Attempts have been made to treat animal tumors with sera from immunized syngeneic, allogeneic, or xenogeneic animals. A number of studies of passive immuno­therapy using heterologous antisera in humans have also been performed [1]. These studies have generally been attempted in patients with large tumor burdens, and, as would be expected, responses have been transient at best. A wide variety of patients with leukemias and lymphomas have been treated with antisera raised in sheep, horses, rabbits, and goats. Problems such as anaphylaxis, serum sickness , and severe cytopenias have been en­countered with these antisera.

There are a number of potential mechanisms by which unconjugated antibodies might be cytotoxic to tumor cells. Antibodies bound to the cell surface membrane of tumor cells may lead to cell lysis by complement­dependent or antibody-dependent cellular cytotoxicity. Circulating tumor cells bound by antibody may be more susceptible to phagocytosis by the reticuloendothelial system. Antibody bound to the cell surface membrane of tumor cells may enhance immunogenicity of the tumor cell, leading to acti­vation of the host's immune system. In any of these cases, successful therapy with antibodies is dependent on the accessibility of the antibody to the tumor, the density and heterogeneity of antigen expression by the tumor, the natural immunity of the host, the degree of specificity of the antibodies used for targeting, and the class of antibody injected.

Due to the potential for targeting of cytotoxic agents, attempts have been made to link tumor-specific heteroantisera to drugs such as methotrexate, chlorambucil, and doxorubicin. Other agents such as radioisotopes , toxins, and enzymes have also been conjugated to antibody. One of the major problems encountered in these initial attempts at immunoconjugate prep­aration has been the inability to develop tumor-specific antibodies with sufficient specificity and in sufficient amounts suitable for in vivo therapy.

Monoclonal antibodies have created a new wave of enthusiasm for using antibodies for the treatment of cancer. Monoclonal antibodies are specific for a single target antigen, can be produced in large quantities with high

Bennel/, 1.M. a"d Foo". K.A., (eds.), Immunologic Approaches to the Classification and Managemelll of Lymphomas arid Leukemias. © 1988 KJuwer Academic Publishers. tS8N978· /·4612·8965-4. All rights reser.ed.

Tab

le 1

. C

linic

al t

rial

s w

ith u

nlab

eled

mon

oclo

nal

anti

bodi

es

Dis

ease

A

ntib

ody/

Cla

ss

Num

ber

of

Tox

icity

E

ffec

t R

efer

ence

s pa

tien

ts

B-l

ymph

oma

A68

9/Ig

G2

a R

enal

T

rans

ient

red

ucti

on i

n ci

rcul

atin

g 2

cells

B

-lym

phom

a A

nti-

idio

type

/IgG

I 1

\ F

ever

, ch

ills,

nau

sea,

vom

itin

g,

1 co

mpl

ete

and

3,4

o

r Ig

G2

a o

r Ig

G2

h

head

ache

, di

arrh

ea,

tran

sien

t 5

part

ial

resp

onse

s dy

spne

a B

-lym

phom

a IF

5/Ig

Gza

4

Fev

er,

mye

losu

ppre

ssio

n 1

part

ial

and

7 1

min

or r

espo

nse

B-C

LL

A

nti-

idio

type

/IgG

2b

F

ever

, ur

tica

ria

Tra

nsie

nt r

educ

tion

in

circ

ulat

ing

6 an

d Ig

G1

cells

B

-CL

L

TlO

llIg

G2

a 18

D

yspn

ea,

feve

r, m

alai

se,

Tra

nsie

nt r

educ

tion

in c

ircu

lati

ng

8-1

1

urti

cari

a, h

ypot

ensi

on

cells

C

TC

L

TlO

llIg

Gza

30

D

yspn

ea,

urti

cari

a, f

ever

, 10

min

or r

emis

sion

s 1

0-1

4

Ant

i-L

eu-l

IIgG

za

cuta

neou

s pa

in

cAL

L

anti

-J5/

IgG

2a

4 F

ever

T

rans

ient

red

ucti

on in

cir

cula

ting

15

ce

lls

AM

L

PM/8

11Ig

M

3 F

ever

, ba

ck p

ain,

art

hral

gia,

T

rans

ient

red

ucti

on i

n ci

rcul

atin

g 16

A

ML

-2-2

3/Ig

Gzh

m

yalg

ia

cells

PM

N 2

9/Ig

M

PM

N 6

/IgM

A

TL

A

nti-

Tac

/IgG

2a

2 N

one

1 pa

rtia

l re

spon

se

17

* B

-CL

L,

B c

hron

ic l

ymph

ocyt

ic l

euke

mia

; C

TC

L,

cuta

neou

s T

-cel

l ly

mph

oma;

cA

ll,

com

mon

acu

te l

ymph

obla

stic

leu

kem

ia;

AM

L,

acut

e m

yelo

geno

us

leuk

emia

; A

TL

, ad

ult

T-c

ell

leuk

emia

/lym

phom

a.

N

U1

-1'

0-

255

degrees of purity, and can be uniformly coupled to drugs, toxins, and radionuclides. The specificity of monoclonal antibodies should theoretically reduce toxicity to normal tissues that are nonreactive with the antibody conjugate. Unlike crude heteroantisera, the monoclonal antibodies require no adsorption and are of a single immunoglobulin subclass. Monoclonal antibodies can be produced in large quantities from ascites fluid or by tissue culture production techniques. Purity of such antibodies can range from 95%-99%.

Results of clinical trials with unlabeled antibodies

Several investigators [2-17] have attempted to treat lymphoid or myeloid leukemias with unlabeled monoclonal antibodies (table 1). In some studies, patients with advanced B-cell-derived chronic lymphocytic leukemia (CLL) received nOl monoclonal antibody (anti-CDS) [8-11]. nOl recognizes the CDS antigen, which is a 65 kilodalton glycoprotein antigen found on normal and malignant T-cells and B-cell chronic lymphocytic leukemia cells. T101 could be safely infused and led to transient reductions in circulating leu­kemia cells. There was no sustained effect on the bone marrow, involved lymph nodes, or other organs. This therapy resulted in some intravascular leukemia cell injury, but destruction in the spleen, liver, and lungs was probably more important. Similar results have been reported in patients with adult T-cell leukemia/lymphoma, acute lymphoblastic leukemia (ALL), and acute myelogenous leukemia (AML) treated with other monoclonal antibodies [12, 15, 16]. Patients with cutaneous T-cell lymphoma who re­ceived (anti-CDS) antibody have had only transient improvement in skin lesions and lymphadenopathy [10-14].

Results of clinical trials with anti-idiotype antibodies

One particular therapeutic approach with monoclonal antibodies which merits more extensive discussion involves the use of monoclonal anti-idiotype anti­bodies in B-cell malignancies. Unlike the antibodies used for therapy dis­cussed above, in which the target antigen is tumor-related or associated, anti-idiotype antibodies have as their target a tumor-specific antigen, the idiotype of the cell surface immunoglobulin present on B-cells. Indeed, this antigen is the closest we have come to identifying a tumor-specific antigen in man. This specificity is based on the fact that individual B-cells are com­mitted to the synthesis of only one immunoglobulin species with a unique variable region structure (idiotype). Moreover, since B-ce\l lymphomas and leukemias are clonal in nature, members of the malignant clone should express the same immunoglobulin molecule, and hence the same idiotype. This feature thus represents a marker by which these tumor cells can be

256

distinguished from normal cells of the host. These facts also imply that an individual patient's tumor cell idiotype will be different from that of other patients, hence anti-idiotype antibodies must be 'tailor-made' for the in­dividual patient. Because of the highly specific nature of these antibodies, treatment with these antibodies have yielded important results regarding the ultimate potential of monoclonal antibody therapy.

The largest experience reported with anti-idiotype therapy is the work of Levy and coworkers. Their first attempt at this therapy was in a patient originally diagnosed as having a malignant lymphoma of the nodular, poorly differentiated, lymphocytic type (follicular small cleaved cell lymphoma) [3]. At the time of treatment, the patient had evidence of rapidly progressive systemic disease symptoms which were resistant to chemotherapy and inter­feron. Following eight continuous six-hour intravenous infusions spaced over the period of one month, the patient entered a complete clinical remission that has been sustained for more than five years without further treatment (R. Levy, personal communication). The mechanisms accounting for this dramatic response are not clear. Because it was noted that the patient's antitumor response continued after the period of passive antibody adminis­tration, evidence of an anti-idiotype antibody response by the patient him­self was investigated, but none was detected. It is still possible that indirect mechanisms could have been involved. Since the immune system may be regulated in part by networks of interactions between idiotypes and anti­idiotypes [18], the administered anti-idiotype could have triggered these types of networks of interactions which led to an anti proliferative response against the patient's tumor.

Encouraged by the above result, Levy et al. have now treated an addi­tional 13 patients with individually tailored anti-idiotype antibodies of vary­ing antibody subclasses [4]. Some patients have been treated with more than one antibody (differing in isotype or epitope specificity) during the course of an individual treatment period. The dramatic result of the first patient treated has not been reproduced so far. Instead, significant tumor responses have been demonstrated in 50% of the patients, but these have not been complete responses and have not lasted for longer than a few months. Never­theless, several important lessons have been learned from these studies. It was found that up to 900 mg of monoclonal anti-idiotype antibody could be infused safely as a single dose, provided the level of circulating free antigen (idiotype) was low or nondetectable and if no immune response by the host against the infused mouse protein (human anti mouse antibodies) was pres­ent. The presence of both serum idiotype and human antimouse antibodies were correlated with acute toxicity during infusions consisting of fever, rigors, dyspnea, arthralgias, and headache, with thrombocytopenia occurring less commonly. This was presumably due to immune complex formation. The presence of significant levels of serum idiotype was found to clearly be a barrier to antibody penetration to tumor sites and thus to a clinical re­sponse. Plasmapheresis was shown to transiently reduce serum idiotype

257

levels but not to a degree sufficient to eliminate this barrier. The effect of the presence of an antimouse response by the host was similar in that tissue penetration and clinical response were prevented by these antibodies. About one third of patients developed this response within a two-week period after the initial infusion. This thus appears to be a less frequent phenomenon in B-cell lymphoma patients than in patients with solid tumors and T-cell lymphomas.

Another means by which patients' tumors could evade the therapeutic effects of anti-idiotype antibodies was by the emergence of idiotype variants within tumors during treatment [19]. This phenomenon was recognized when tumors of two patients lost reactivity with the anti-idiotype antibody generated against the respective original tumors during treatment. Subse­quent studies have shown that this loss of reactivity was not due to anti­genic modulation. Comparison of immunoglobulin gene rearrangements by Southern blot analysis in pretherapy and post-therapy tumors taken from each patient revealed identical rearrangements in each case. This strongly suggests that all cell populations studied were part of a single monoclonal lymphoma in each patient. In one of these cases, the anti-idiotype antibody was known to react with only the heavy chain variable region of the surface IgM protein of the pretherapy tumor and not with light chain regions. Eight independent heavy chain variable region isolates from tumors prior to and after treatment were subjected to nucleotide sequence analyses [20]. Ex­tensive point mutations were demonstrated in all isolates and no two se­quences were identical. A clustering of mutations encoding for amino acid changes was observed in the CDR2 region. Comparison of pretherapy and post-therapy sequences implicated a single amino acid in CDR2 at position 54 as being important in determining reactivity with the anti-idiotype anti­body. Three of the post-therapy sequences had a common substitution at that position, and a fourth post-therapy sequence had other substitutions in a neighboring position. Thus, clones with mutations in this region ap­parently escaped the antibody'S strong negative selection pressure in vivo. Further analysis indicated that there was a significant bias against mutations resulting in amino acid changes in portions of the V region gene other than CDR2, even in the absence of any selection by antibody treatment. Thus the nonrandom clustering in CDR2 may have been due to endogenous selective forces interacting with tumor cell surface immunoglobulin. The generality of these concepts is now being explored in other patients' tumor samples. It is now believed that somatic mutation accounted for tumor escape in more than these two patients and that somatic mutation prior to any therapy may be the rule rather than the exception [21, 22]. This poses an additional problem for anti-idiotype antibody therapy in that more than one antibody may need to be developed for each individual patient so that idiotypic variants within the tumor can be recognized.

It is still unclear why the excellent response in the first treated patient has not been reproduced. Various factors have been examined for their ability

258

to predict response to this therapy [23]. Included among these are the isotype of the anti-idiotype antibody used, the density of cell surface idio­type, the epitope recognized by the anti-idiotype antibody, the affinity of anti-idiotype antibody for antigen, the relative ability of the anti-idiotype antibody to modulate surface antigen, the direct effect of antibody on tumor cell proliferation in vitro, and the degree of T-cell infiltration present in pretherapy tumor specimens. None of these factors has been positively correlated with good clinical outcome, except the number of T-cells present in pretherapy tumor tissue [23, 24]. In the two best responding cases, the T­cells actually outnumbered the tumor cells. The majority of these T-cells were of the helper/inducer phenotype (CD4). Whether the anti-idiotype antibodies given to these patients augmented an ongoing cell-mediated cyto­toxic response by the host against the tumor is not clear. Certainly more observations on pretherapy T-cell infiltration must be made before the actual significance and function of this finding become apparent. Another factor which has yet to be fully explored is the nature of somatic mutation in the immunoglobulin genes of the various tumors of patients undergoing treatment, as these may more fully define an endogenous host response that may regulate tumor growth and eventual response to therapy.

While anti-idiotype therapy remains an interesting area of investigation, its general applicability to the treatment of B-cell lymphoma still remains to be defined. Certainly, the time-consuming nature of the isolation of individ­ually tailored antibodies limits the availability of these reagents for therapy. This latter problem is compounded by the finding of the emergence of idiotypic variants which might require the isolation of additional antibodies for the treatment of individual patients. Overcoming these shortcomings will go a long way in increasing the feasibility of this approach.

Problems with unlabeled monoclonal antibody therapy

Monoclonal antibody therapy has several shortcomings that must be ad­dressed. First, with few exceptions, unlabeled antibodies are clearly not very effective in destroying tumor cells. While they target quite well to tumor cells in vivo, most murine antibodies do not fix human complement and do not effectively mediate tumor lysis through human effector cells. By con­jugating toxins, drugs, and/or isotopes to antibodies, the limitations inherent to the antibody may be overcome. Treatment with antibodies such as anti­CD5 and anti-J5 (anti-CDlO) results in modulation of the antigen from the cell surface, which prevents further antibody from binding to the tumor cells. The antigen-antibody complex is pinocytosed into the cytoplasm [25], a phenomenon that might be advantageous when drugs or toxins are linked to the antibody to enhance its cytotoxicity. Antigen in the circulation poses another potential problem because it may prevent the antibody from reaching the tumor cells. This was clearly a major problem with anti-idiotype anti-

259

bodies. Plasmapheresis was not effective in reducing the circulating idiotype. Furthermore, murine antibodies can stimulate production of human anti­mouse antibodies which lead to antibody neutralization. This situation may be correctable by treatment with high initial doses of antibody (> 500 mg) to induce tolerance or by simultaneous treatment with immunosuppressive drugs. Another potential problem is that the heterogeneity of antigen ex­pression on tumor cells may necessitate therapy with more than one anti­body.

Imaging and therapy trials with labeled monoclonal antibodies

Antisera and monoclonal antibodies conjugated to radionuclides for tumor imaging have been extensively studied [26]. The n01 antibody conjugated to 11lindium has been used for imaging in 12 patients with cutaneous T-cell lymphoma [27, 28]. Tumors as small as 0.5 em have been localized; however, nonspecific uptake of the immunoconjugate in the liver and spleen have prevented critical evaluation of these organs. This problem has been partially circumvented by the administration of intracutaneous injections of the im­munoconjugate so that it is taken up by the lymphatics directly to the lymph node sites of disease [29]. This procedure does not, of course, facilitate visualization of extra lymphatic disease.

Recent results have demonstrated that when 100-300 mCi of 131iodine was linked to the T101 or the Lym-l antibodies and injected intravenously into patients with T-cell and B-cell non-Hodgkin's lymphoma, respectively, excellent antitumor responses resulted [30, 31] (table 2). Antitumor activity has also been reported using antiferritin heteroantisera labeled with 131iodine in patients with advanced Hodgkin's disease [32]. A number of centers are studying toxin and drug conjugates with murine antibodies; clinical trials have just begun, and, while favorable responses have not yet been reported, this remains an important avenue of investigation [33, 34].

Toxicity

Side effects of unlabeled monoclonal antibody therapy are usually minor. Respiratory distress following the rapid infusion of monoclonal antibody has been described [9, 10], and some patients have demonstrated transient elevation of creatinine and hepatic enzymes [2, 12]. Fever and urticaria are common but are rarely dose-limiting. Nausea and vomiting have also been reported [4].

The major dose-limiting toxicity with radiolabeled antibodies has been myelosuppression [28, 29]. This is secondary to both specific localization of antibody in the bone marrow as well as the nonspecific effects of total body radiation.

tv

S

Tab

le 2

. C

linic

al t

rial

s w

ith

labe

led

mon

oclo

nal

anti

bodi

es

Dis

ease

A

ntib

ody/

Cla

ss

Lab

el

Num

ber

of

Tox

icit

y E

ffec

t R

efer

ence

s pa

tien

ts

B-l

ymph

oma

Lym

-lII

gG2

a 13

1,

9 N

one

3 co

mpl

ete,

30

1

part

ial,

and

3

min

or r

espo

nses

C

TC

L1

TlO

lllg

G2

a 13

1,

6 D

yspn

ea,

feve

r ur

tica

ria,

2

part

ial

and

31

mye

losu

ppre

ssio

n 3

min

or r

espo

nses

H

odgk

in's

an

tife

rrit

in/

131

1 38

M

yelo

supp

ress

ion

40%

par

tial

32

di

seas

e he

tero

anti

sera

re

spon

ses

B-C

LL

' T

lOll

lgG

2a

rici

n-A

5

Non

e T

rans

ient

red

ucti

on in

cir

cula

ting

3

3,3

4

cells

1 crC

L,

cuta

neou

s T

-cel

l ly

mph

oma;

B-C

LL

, B

-chr

onic

lym

phoc

ytic

leu

kem

ia.

261

Conclusion

The use of monoclonal antibodies and antibody immunoconjugates in the treatment and radioimaging of cancer is in its infancy. Although much work remains to be done to clarify the issues surrounding the use of monoclonal antibodies, studies in animal tumor models and humans have clearly dem­onstrated that antibodies alone or antibody conjugates can be safely ad­ministered with minimal adverse effects; in selected cases these may have diagnostic and therapeutic value. Nonspecific localization of antibody in the reticuloendothelial system, host antibody response, and antigenic hetero­geneity are major obstacles to safe and effective treatment with monoclonal antibodies. These issues are under investigation in animal models and humans. Although anti-idiotype antibodies are highly specific and have produced excellent responses in a small number of patients, problems such as biclonality of some lymphomas [35, 36], instability of the idiotype, and the difficulty of tailoring antibodies to individual patients clearly limit the role of anti-idiotype therapy. Perhaps the most important future role for monoclonal antibody therapy will be in patients with minimal disease in the "adjuvant" setting, in whom antibody conjugates might eliminate micro­metastatic deposits of tumor cells. This remains to be addressed in controlled trials.

References

1. Rosenberg SA, Terry WD: Passive immunotherapy of cancer in animals and man. Adv Cancer Res 25:323-388, 1977.

2. Nadler LM, Stashenko P, Hardy R, et al. Serotherapy of a patient with monoclonal antibody directed against a human lymphoma-associated antigen. Cancer Res 40:3147, 1980.

3. Miller RA, Maloney DG, Warnke R, Levy R: Treatment of B-celllymphoma with mono­clonal anti-idiotype antibody. N Engl J Med 306:517, 1982.

4. Meeker TC, Lowder J, Maloney DG, et al: A clinical trial of anti-idiotype therapy of B cell malignancies. Blood 65:1349, 1985.

5. Rankin EM, Hekman A , Somers R, Huinink B: Treatment of two patients with B cell lymphoma with monoclonal anti-idiotype antibodies. Blood 65:1373, 1985 .

6. Giardina SL, Schroff RW, Woodhouse CS, et al: The generation of monoclonal anti­idiotype antibodies to human B cell-derived leukemias and lymphomas. J ImmunolI35:653, 1985.

7. Press OW, Appelbaum F, Ledbetter JA, et al: Monoclonal antibody IF5 (anti-CD20) serotherapy of human B-cell lymphomas. Blood 69:584, 1987.

8. Dillman RO, Shawler DL, Sobol RE, et al: Murine monoclonal antibody therapy in two patients with chronic lymphocytic leukemia . Blood 59:1036, 1982.

9. Foon KA, Schroff RW, Bunn RA, et al: Effects of monoclonal antibody therapy in patients with chronic lymphocytic leukemia. Blood 64:lO85, 1984.

lO. Dillman RO, Shawler DL, Dillman JB, Royston R: Therapy of chronic lymphocytic leukemia and cutaneous T-cell lymphoma with TI01 monoclonal antibody. J Clin Oncol 2:881,1984.

11. Bertram IH, Gill PS, Levine AM, et al : Monoclonal antibody TlOl in T cell malignancies:

262

A clinical, pharmacokinetic, and immunologic correlation. Blood 67:1680, 1986. 12. Foon KA, Schroff RW, Bunn RA: Monoclonal antibody therapy for patients with leukemia

and lymphoma. In: Foon KA, Morgan AC (eds): Monoclonal Antibody Therapy of Human Cancer, Boston, Martinus Nijhoff Publishing, 101:85, 1985.

13. Miller RA, Oseroff AR, Stratte PT, Levy R: Monoclonal antibody therapeutic trials in seven patients with T-cell lymphoma. Blood 62:988, 1983.

14. Miller RA, Levy R: Response of cutaneous T cell lymphoma to therapy of hybridoma monoclonal antibody in a patient with T cell leukemia. Lancet 2:226, 1981.

15. Ritz J, Pesando JM, Sallan SE, et al: Serotherapy of acute-lymphoblastic leukemia with monoclonal antibody. Blood 58:141, 1981.

16. Ball ED, Bernier GM, Cornwell GG, McIntyre OR, O'Donnell JF, Fanger MW: Mono­clonal antibodies to myeloid differentiation antigens: In vivo studies of three patients with acute myelogenous leukemia. Blood 62:1203-1210, 1983.

17. Waldmann TA, Longo DL, Leonard WJ, et al: Interleukin 2 receptor (Tac antigen) expression in HTLV-l-associated adult T-cell leukemia. Cancer Res 45:4559s-4562s, 1985.

18. Jerne NK: Towards a network theory of the immune system. Ann Immunol 125C:373, 1974.

19. Meeker T, Lowder J, Cleary ML, et al: Emergence of idiotype variants during treatment of B-cell lymphoma with anti-idiotype antibodies. N Engl J Med 312:1658, 1985.

20. Cleary ML, Meeker TC, Levy S, et al: Clustering of extensive somatic mutations in the variable region of an immunoglobulin heavy chain gene from a human B cell lymphoma. Cell 44:97, 1986.

21. Raffeld M, Neckers L, Longo DL, Cossman J: Spontaneous alteration of idiotype in a monoclonal B-cell lymphoma. Escape from detection by anti-idiotype. N Engl J Med 312:1653, 1985.

22. Carroll WL, Lowder J, Streifer R, Warnke R, Levy S, and Levy R: Idiotype variant cell populations in patients with B cell lymphoma. J Exp Med 164:1566, 1986.

23. Lowder IN, Meeker TC, Campbell M, et al: Studies on B lymphoid tumors treated with monoclonal anti-idiotype antibodies: Correlation with clinial responses. Blood 69:199, 1987.

24. Garcia CF, Lowder J, Meeker TC, Bindl J, Levy R, Warnke RA: Differences in "host infiltrates" among lymphoma patients treated with anti-idiotype antibodies: Correlation with treatment responses. J Immunol 135:4252, 1985.

25. Schroff RW, Farrell MM, Klein RA, Oldham RK, Foon KA: T65 antigen modulation in phase I monoclonal antibody trial with chronic lymphocytic leukemia patients. J Immunol 133:1641, 1984.

26. Goldenberg DM, DeLand FH: History and status of tumor imaging with radiolabelled antibodies. J BioI Resp Modif 1:121, 1982.

27. Bunn PA, Carrasquillo JA, Keenan AM, et al: Successful imaging of malignant non­Hodgkin's lymphoma using radio labeled monoclonal antibody. Lancet 2:1219, 1984.

28. Carrasquillo JA, Bunn PA, Keenan AM, et al: Radioimmunodetection of cutaneous T-cell lymphoma with I II In-labelled TIOI monoclonal antibody. N Engl J Med 315:673, 1986.

29. Mulshine J, Keenan A, Carrasquillo J, et al: Successful immunoscintigraphy after lymphatic delivery of 1IIIn_TI01. Proc Am Soc Clin Oncol 4:205, 1985.

30. DeNardo SJ, DeNardo GL, O'Grady LF, et al: Radioimmunotherapy of patients with B­cell lymphomas using 1131 Lym-l MAb (abstract). J Nuc Med. 27:903, 1986.

31. Rosen ST, Zimmer AM, Goldman-Leikin R, et al: Radioimmunodetection and radio­immunotherapy of cutaneous T-cell lymphomas using an 1311-labeled monoclonal antibody: An Illinois Cancer Council Study. J Clin Oncol 5:562-573, 1987.

32. Lenhard RE, Order SE Jr, Spunberg JJ, Asbell SO, Leibel SA: Isotopic immunoglobulin: A new systemic therapy for advanced Hodgkin's disease. J Clin Oncol 3:1296-1300, 1985.

33. Laurent G, Pris J, Farcet JP, et al: Effects of therapy with TIOI ricin A-chain immunotoxin in two leukemia patients. Blood 68:752, 1986.

34. Hertler AA, Schlossman DM, Borowitz MJ, Frankel AE: A phase I study of TIOI RTA

263

immunotoxin in refractory chronic lymphocytic leukemia. Biood 68(suppl 1):223a, 1986. 35. Sklar J , Cleary ML, Thielmans K, Gralow J, Warnke R, Levy R: Biclonal B-celllymphoma.

N Engl J Med 311 :20, 1984. 36. Giardina SA, Schroff RW, Woodhouse CS, et al: Detection of two malignant B-cell clones

in a single patient using anti-idiotype monoclonal antibodies and immunoglobulin gene rearrangement. Blood 66:1017, 1984.

9. Autologous bone marrow transplantation in acute leukemia and lymphoma following ex vivo treatment with monoclonal antibodies and complement

Arnold S. Freedman, Tak Takvorian, Lee M. Nadler, Kenneth C. Anderson, Stephen E. Sallan, and Jerome Ritz

Introduction

Recent experience with both allogeneic and syngeneic bone marrow trans­plantation (BMT) has demonstrated that high-dose chemoradiotherapy can lead to long-term disease-free survival in patients with various hematologic neoplasms, including acute myeloblastic leukemia (AML) , acute lympho­blastic leukemia (ALL), chronic myelogenous leukemia (CML) , and non­Hodgkin's lymphoma (NHL) [1, 2]. In most instances, bone marrow donors have been siblings who are identical with the patients at the major histocom­patibility complex. Therefore, BMT has been limited to the 30%-40% of patients with histocompatible siblings to serve as normal marrow donors. One of the major obstacles, even in patients with HLA-matched donors, is graft versus host disease (GVHD), which occurs in 30%-70% of patients who receive standard prophylaxis regimens. GVHD is thus a significant cause of morbidity and mortality, particularly in older patients. The use of autologous bone marrow for patients who lack histocompatible donors is an attractive alternative for two reasons. First, it eliminates the need for histo­compatible donors, and secondly, the complications associated with GVHD do not occur. Unfortunately, in both leukemias and lymphomas, tumor cells are often present in the bone marrow during complete hematologic remission [3]. In Burkitt's lymphoma, several groups have demonstrated that tumor cell lines can be derived from histologically uninvolved marrow [4]. More recently, Hu et al. have demonstrated clonal immunoglobulin gene rear­rangements in peripheral blood lymphocytes of patients with low-grade NHLs [5]. Approximately 60% of patients with no evidence of disease demonstrated circulating lymphoma cells by this technique. Therefore, prior to autologous BMT, the removal of malignant cells from bone marrow may

Supported by National Institutions of Health Grant CA 34183. ASF is supported by PHS grant number 5K08 CAOl105-01, awarded by the National Cancer Institute, DHHS. KCA is a recipient of a Junior Faculty Research Award from the American Cancer Society. JR is a Scholar of the Leukemia Society of America. 2 We would like to thank Ms. Marie Sweeney for her excellent preparation of the manuscript.

BellI/eli, I .M. and Foon. K.A., (eds.), immunologic Approaches 10 rhe Classijicarion and Managemem of Lymphomas and Leukemias. © 1988 Kluwer Academic Publishers. ISBN978-1-4612-8965-4. All rig/us reserved.

266

be a prerequisite. For these reasons, a number of techniques have been explored to deplete neoplastic cells from the bone marrow of patients with leukemias and lymphomas.

This chapter examines the preclinical and clinical experience of autologous BMT in acute leukemia and NHLs. It focuses on studies that have utilized monoclonal antibodies (MoAbs) directed against antigens expressed on leukemia and lymphoma cells to eliminate tumor cells ex vivo from autol­ogous donor marrow.

Lymphocyte differentiation antigens defined by monoclonal antibodies

B-cell antigens

Monoclonal antibodies have been useful in identifying cell surface molecules on normal and malignant hematopoietic and lymphoid cells. These reagents have been used to define cellular lineage and stages of differentiation of lymphoid and myeloid cells. All of the antigens found on neoplastic cells represent normal differentiation antigens, and true leukemia or lymphoma specific antigens have not been identified. By extensively screening MoAbs against a panel of B-cells isolated from normal peripheral blood, adult, and fetal lymphoid tissues, and patients with a variety of B-cell leukemias and lymphomas, a hypothetical model of normal and neoplastic B-cell differen­tiation has been constructed based on differences in antigen expression [6, 7].

In general, five major groups of B-cell restricted and associated antigens have been identified. The first group represents pan-B-cell antigens, which are expressed throughout ontogeny from the pre-B-cell stage, prior to the presence of cytoplasmic mu and surface Ig, and lost only at the terminal! plasma cell stage of differentiation. These include the B-cell restricted anti­gens, B4 (Cluster designation, CDI9) [8] and Bl (CD20) [9, 10], and the B-cell associated antigens, Ia (HLA-DR) [11] and CD24, defined by the MoAb BA-l [12]. Ia antigen is not B-cell specific, since it is also expressed on monocytes, activated T-cells, and myeloid progenitor cells. The second group of antigens are expressed only during limited stages of B-cell differen­tiation. Antigens in this group include the B-cell restricted antigens, surface immunoglobulin (sIg), and B2 (CD21) [13], which are present on small resting B-cells and are lost following activation in vivo and in vitro . The B2 antigen which contains the C3d/Epstein Barr virus (EBV) receptor [14-16] is not present on normal or neoplastic pre-B-cells, but is expressed on mature B-cells. B2 is no longer detected on resting B-cells by three days after stimulation with a variety of B-cell mitogens including pokeweed mito­gen, anti-immunoglobulin, protein A, and EBV [17-19]. Surface IgD follows a similar pattern of loss of expression. B2 is expressed on most chronic lymphocytic leukemias (CLLs), nodular and diffuse poorly differentiated

267

lymphocytic lymphomas (NPDLs/DPDLs), few diffuse large cell lymphomas (DLCLs), and not on Waldenstroms or myelomas. This further supports the idea that B2 is expressed in the mid-stages of B-cell differentiation.

Another series of antigens are not expressed on small resting B-cells but appear after in vitro activation with mitogens. These include the B-cell restricted antigens B5 (20), Blast-1 (21), Blast-2 (CD23)(22), and BB1 (23), and the B-cell associated antigens, the IL-2 receptor (CD25)(IL-2R) [24, 25], and the transferrin receptor (T9) [26]. Studies of in vitro activation of resting splenic B-cells with anti-Ig [18] have demonstrated that B5 and the IL-2R appear 24 hours after activation, followed by the appearance of T9 and BB1 at two days. Blast-1 and Blast-2 appear at three days, and all of these antigens are maximally expressed from three to four days, which coincides with the time of maximal proliferation. By six days following stimulation, when IgG production is detectable, the expression of these activation antigens begins to decrease to background levels. The expression of these antigens after activation of normal B-cells, which are then com­petent to respond to several growth factors including low and high molecular weight B-cell growth factor, IL-2, and interferon gamma, makes them excel­lent candidates for growth factor receptors and other important regulatory molecules. We have recently examined the expression of these activation antigens on a variety of B-cell malignancies [27] and have noted that B5, BB1, and T9 are expressed on NPDLs, Burkitt's lymphomas, and DLCLs. In contrast, the IL-2R and B5 are expressed on hairy cell leukemias and DPDLs. Except for T9, which is present on many non-T-cell ALLs, these activation antigens are not expressed on normal or neoplastic pre-B-cells or terminally differentiated B-cells.

A group of antigens that are -expressed at very limited stages of differen­tiation have also been described. This group includes the common ALL antigen (CALLA) (CDlO), a 100 kd glycoprotein that is expressed on a subset of normal adult and fetal pre-B-cells, and on cells in the germinal center of secondary lymphoid follicles [28, 29]. CALLA is also present on 80% of non-T-cell ALLs, most NPDLs, Burkitt's lymphomas, and T-cell lymphoblastic lymphomas [30]. Another B-cell associated antigen with limit­ed expression is gp26 (CD9) (defined by MoAbs 12 and BA-2) [31, 32]. This antigen is expressed on fetal hematopoetic cell in bone marrow and liver, less than 5% of adult bone marrow, thymocytes, activated T-cells, and mature platelets. It is also present on 83% of CALLA+ non-T-cell ALLs and 68% of the CALLA- non-T-cell ALLs. CD9 antigen is also infrequently expressed on B-cell and ALLs of T-cell origin. One additional antigen, Y 29/55, which is B-cell restricted, is expressed on B-cells from peripheral blood, lymph node, tonsil, and spleen both prior to and after pokeweed mitogen activation [33]. Y 29/55 is expressed on malignancies which corre­spond to the mid-stages of normal B-cell ontogeny, which include CLL, Burkitt's lymphoma, most NHLs, and hairy cell leukemias, but not non-T­cell ALLs, Waldenstroms, or myelomas.

268

The last group of antigens appear only at the terminal stages of B-cell differentiation and include PCA-l [34] and TlO [35]. PCA-l is expressed on granulocytes as well as normal and neoplastic terminally differentiated B­cells. In addition to normal and neoplastic plasma cells, TlO is also present on thymocytes and myeloid progenitor cells and activated T-cells.

T-cell antigens

Analogous to the expression of B-cell antigens, the stages of T-cell differen­tiation have also been characterized by the expression of a series of antigens. These have been extensively reviewed elsewhere [36, 37]. Early thymo­cytes (stage I) express TlO (also expressed on terminally differentiated B­cells), T9, CD7 (defined by the MoAbs WTl, 3Al, and Leu 9), and Tll (C02), which is the sheep red blood cell receptor. Tll is a pan-T-cell antigen, similar to B4, and is expressed on early thymocytes through to mature T-cells. Stage II or common thymocytes, express T4 (CD4), T6 (COl), T8 (CD8), TlO, CD7, and TIL The mature thymocytes acquire Tl (CDS), which is also expressed on B-cell CLLs, fetal B-cells, and a minor subpopulation of normal marginal zone B-cells. The mature thymocytes (stage III) branch off into cells which only express T4 or T8. Mature thymocytes also acquire the T3 (Cd3)/Ti T-cell antigen receptor complex on their surfaces, as well as Tl2 a 120 kd glycoprotein (CD6) [38]. Tl2 is also present on all mature T-cells of either the helper/inducer (Tl, T3, T4, Tll, CD7) or cytotoxic/suppressor (Tl, T3, T8, Tll, CD7) phenotype.

The majority of T-cell ALLs generally express the phenotype of early thymocytes (T9, TlO, Tll, CD7) [39, 40]. In contrast, patients with T-cell lymphoblastic lymphomas have the phenotype of common or mature thymo­cytes [41, 42]. The adult T-cell leukemias and lymphomas phenotypically correspond to mature T-cells [43, 44]. Most of these leukemias and lympho­mas and all cases of Sezary syndrome express the helper/inducer phenotype, with the suppressor/cytotoxic T-cell phenotype less commonly expressed by T-cell CLLs and NHLs. More recently, the HTLV-I associated adult T-cell leukemias have been characterized as having a T-helper cell phenotype, as well as expression of the T-cell growth factor receptor (IL-2R) [45].

Preclinical studies of autologous BMT

A variety of techniques have been used for ex vivo treatment of autologous bone marrow to remove residual tumor cell populations prior to autologous BMT [46]. These have included separation of normal marrow cells from tumor cells using density gradient separation, lectin agglutination, chemical separation such as in vitro incubation with 4-hydroperoxycylophosphamide (4-HC), as well as immunologic techniques using heteroantisera and comple­ment (C'). With the unique specificity and high titer of MoAbs reactive with leukemia and lymphoma cells, these reagents appear ideal for immunologic

269

manipulation of donor bone marrow prior to freezing and storage. Impor­tantly, the almost unlimited availability of a particular MoAb also permits a large number of patients to be treated uniformly with the same reagent.

Preclinical studies of the use of MoAbs and C' for in vitro elimination of tumor cells have focused on ALL and B-cell NHL. Requirements which are necessary prior to clinical studies include high affinity of the MoAbs for strongly expressed antigens on the neoplastic cells and lack of expression of the antigen on normal hematopoetic stem cells so that in vitro treatment will still permit normal hematopoietic reconstitution. For example, although 1% of normal BM cells express CALLA, treatment with J5 (anti-CALLA) MoAb and rabbit C' did not selectively deplete committed myeloid stem cells (BFU-E, CFU-E, CFU-C) or mixed colonies (CFU-G/M) [47]. Similar studies demonstrated that gp26 (defined by J2 MoAb) was not expressed on CFU-GM, BFU-E, CFU-E, and CFU-GEM [48]. These studies, coupled with the specificity and high reactivity of J5 and 12 for ALL cells, suggested that marrow treated with J5 and J2 followed by C' would be ideal for autologous BMT. Bast and coworkers [47] subsequently defined the optimum conditions for elimination of greater than 99% of S1Cr labelled CALLA + cells from a 100-fold excess of normal human bone marrow. It was observed that three treatments with C' for 30 minutes were more effective than two treatments for 45 minutes or one treatment for 90 minutes. This treatment did not deplete committed hematopoietic stem cells. Additional studies demonstrated that treatment with a combination of J5 with 12 and C' proved more effective than either single reagent for eliminating clonogenic Burkitt's lymphoma cells from an excess of normal bone marrow (4 logs J5/12, 2 logs J5, 3 logs J2). This combination did not inhibit growth of normal hemato­poietic precursors. The efficacy of multiple antibodies for the elimination of neoplastic cells has been similarly demonstrated by the group at University of Minnesota [49]. They observed in both S1Cr release and clonogenic assays that the combination of the BA-l (CD24), BA-2 (CD9/gp26), and BA-3 (CALLA) MoAbs and rabbit C' was as effective or more effective at lysing cell lines or fresh leukemia cells than a single antibody. In addition, this antibody combination was effective as a single treatment of 60 minutes in the presence of 100-fold excess of normal human bone marrow.

These preclinical studies examining the utility of MoAbs and C' in deplet­ing ALL cells, led to investigations of two antibodies anti-Bl [48] and anti-Y 29/55 [50], for their use in depleting NHL cells ex vivo prior to autologous BMT. In studies similar to those with J5, anti-Bl (murine IgG2a isotype) and rabbit C' eliminated greater than two logs of clonogenic tumor cells in 100-fold excess of normal bone marrow. Moreover, in these experiments, committed hematopoietic stem cells were unaffected by the treatment. The anti-Y 29/55 MoAb (murine IgG2a isotype) in the presence of rabbit C', lysed about 90% of B-cell CLL target cells in vitro. This treatment was without effect of normal CFU-C cells. A similar series of studies utilizing MoAbs to T-cell antigens (CDS and CD7) demonstrated effective depletion

270

of tumor cells in an excess of normal bone marrow, with no effect of committed hematopoietic stem cells, either with C' or as ricin immunotoxin conjugates [51, 52]. Therefore, these MoAbs were found to be useful rea­gents for purging marrow for autologous BMT in NHL due to their strong expression on B-cell and T-cell NHLs, the antibodies fix C' well, and lack toxicity for hematopoietic progenitor cells as measured in the available culture systems.

The vast majority of clinical studies of ex vivo marrow depletion with MoAbs have involved the use of C' mediated cytotoxicity. There are several advantages to this approach. These include highly specific and effective killing of antibody coated cells. In addition, C' mediated lysis requires shorter incubation periods than immunotoxin conjugates. Several problems with the use of C' include variability of different lots of C', loss of cells due to washing and nonspecific cytotoxicity, and inadequate lysis of tumor cells that may not express the target antigen or react only weakly with the MoAb. This has led to the examination of other techniques of depleting marrow of unwanted cells such as the use of immunotoxins, or MoAbs bound to particles, metal colloids, or microspheres. These have been recently reviewed [46] and will only be discussed briefly. The immunotoxin which has been most frequently used clinically is the anti-CDS (Tl, TlO1, Leu 1) bound to whole ricin or ricin A chain. Preclinical studies have demonstrated six logs of tumor cytoreduction, with no toxicity for CFU-C or BFU-E, with T101-ricin A chain. Immunophysical techniques to deplete cells from bone marrow have included directly conjugated MoAbs to gold particles or metal colloids (iron, cobalt) or indirectly via a second anti mouse antibody. The depletion using gold particles involves a density gradient centrifugation to effect sepa­ration, whereas the iron or cobalt spheres are removed using magnets. In vitro studies with magnetic beads have demonstrated three logs of neoplastic cell removal in the presence of a 100-fold excess of normal bone marrow.

Clinical studies of autologous BMT in ALL

The initial studies of in vitro depletion of malignant cells in ALL involved the use of rabbit heteroantisera and complement [53]. Hematologic engraft­ment was observed, but few long-term remissions were achieved. With the availability of a number of MoAbs directed against leukemic cells and the preclinical studies demonstrating effective depletion of malignant cells with preservation of hematopoietic progenitor cells, a number of clinical studies were initiated.

A protocol was started in 1980 at Dana-Farber Cancer Institute (DFCI) involving the in vitro treatment of bone marrow from patients with CALLA + non-T-cell ALL using the J5 and 12 MoAbs [54]. For patients to be eligible for this program, they required histologic evidence of relapse following standard therapy, reactivity of leukemic cells with J5 and/or J2 MoAbs, successful reinduction into a complete remission, and informed consent.

271

Any patient with a histocompatible donor available for allogeneic BMT was excluded.

To date, 39 patients with relapsed CALLA+ non-T-cell ALL have been treated on this protocol. It should be noted that the first 13 patients received 850 rad total body irradiation (TBI), with subsequent patients receiving 1200 rad fractionated TBI over three days. The bone marrow of 14 of the initial 15 patients was treated in vitro with J5 alone, one patient's marrow was treated with J2, and the remaining patients with the combination. This group of 39 patients, with a median age of 8.5 years (range 3-54), included 7 adults over age 18. Of the 39 patients treated (table 1), 12 died while in remission within four months of autologous BMT, the majority of infectious complications. Thirteen of the patients relapsed in marrow with CALLA + ALL, seven of these within the first four months post-autologous BMT, the others within 14 months. The remaining 14 patients are in unmaintained remission from 1 to 67 months, with a median follow-up of 26 months. A Kaplan Meier analysis of probability of survival at 60 months is approxi­mately 30%. The survival of these patients is similar to studies of allogeneic BMT of patients in second and subsequent remissions [55], with long-term disease-free survival of about 30%. Although it is unclear if relapse was due to inadequate ablative therapy or reinfusion of unpurged leukemic cells, the similar results obtained with both autologous and allogeneic BMT suggest that failure due to relapse is primarily due to inadequacy of the patient treatment regimen.

The initial question which was answered by this protocol was whether J5/J2 treated marrow could reconstitute hematological function following ablation with TBI. Hematologic engraftment was noted in all patients who survived more than 20 days. There was no correlation between numbers of treated marrow cells infused and rate of engraftment. T-cells, which were predominantly T3+T8+T10+, were observed between 9-19 days post BMT, but the return of normal numbers of B-cells was significantly longer (31-128 days).

The length of a patient's prior remission influenced disease-free survival after autologous BMT. Of the nine patients currently in unmaintained re­missions of greater than two years, eight had initial remissions of greater than two years (range 29-84 months). Nevertheless, six of these nine pa­tients now have remissions after autologous BMT which exceed their longest

Table 1. Current status of patients treated with J5/12 autologous BMT (n = 39)

Status

Remission Relapse Remission deaths

Number

14 13 12

Months post-BMT

67,63,53,50,38,33,27,26,25,12,4,3,2,1 14,14, 12, 8, 7, 5, 4, 4, 3, 2, 2, 2,2 4,3,2,2,1,1,1,1,1,1,1,1

272

previous remissions. In contrast, the patients who had short first remissions or relapsed on therapy tended to relapse after autologous BMT.

Other groups have utilized different MoAbs and C' to deplete ALL cells prior to autologous BMT. The University of Minnesota has treated bone marrow from 23 pediatric patients in second (9 patients), third (13 patients), and fourth remission (1 patient) with the BA-l, BA-2, and BA-3 MoAbs and rabbit C' [56]. Fifteen of the 23 patients (65%) relapsed, with a median time to relapse of about four months (1.4-7.4 months); one died at day 21 of sepsis. The remaining seven patients remained disease-free 6+-23+ months (median of 21+ months) post-transplant, with a probability of remaining disease-free at one year of 29%. This is similar to the DFCI experience; however, four of the seven long-term surviving patients in the Minnesota experience received some form of maintenance therapy after autologous BMT. The hematologic reconstitution of these patients was simi­lar to that seen in the J5/12 autologous BMT patients. Although the Min­nesota study demonstrated a similar disease-free survival and hematologic reconstitution to the DFCI protocol, the relapse rate was significantly higher in Minnesota (65% vs. 29%), while the number of remission deaths was lower (4% vs. 35%). Although the in vitro treatments were similar, the conditioning regimens differed between the two programs. The more inten­sive chemotherapy conditioning of the J5/12 protocol (VM-26, ARA-C, cyclophosphamide, and TBI) might account for the increased number of remission deaths and decreased relapse rate, in contrast to the conditioning regimen of the Minnesota group (cyclophosphamide and TBI).

Monoclonal antibodies which react with T-ALL cells have also been used in autologous BMT to deplete T-lymphoblasts. However, in contrast with the non-T-cell ALL autologous BMT programs, many of the studies to be reviewed have utilized immunotoxins, and few patients with T-cell ALL have undergone BMT with anti-T-cell MoAb and C' depleted autologous marrow. Six patients with T-cell ALL have been reported in studies using autologous BMT with TlO1-ricin (three patients) or T101-ricin/TAl-ricin (three patients) in vitro treated bone marrow [57, 58]. Four of these patients were in second remission, two in first remission. The follow-up of these patients is short, with three patients alive and disease-free at 45+, 79+, and 90+ days. Two patients relapsed at 53 and 64 days, and one died in remission of CMV penumonia at 120 days. Hematologic engraftment data (platelets/granulocytes) of these patients was similar to that reported for the previously described studies involving MoAb and C' treated bone marrow. Although a definite antileukemic effect has not been clearly demonstrated for those patients who received TI01-ricin treated bone marrow, toxicity to hematopoietic stem cells and toxicity to the patient from the purged marrow has not been seen.

In summary, several approaches have been taken to demonstrate that autologous BMT is an option for patients with ALL in second or subsequent remission who lack suitable HLA-matched donors. Hematologic engraftment

273

has not been a significant problem, however, relapsed leukemia as seen in allogeneic BMT is. The major problem therefore would appear to be inade­quate ablative regimens and resistant leukemia. The similarity in disease­free survival between allogeneic and autologous BMT indicates that changes in conditioning regimens may improve these results with both modalities of treatment.

Clinical studies of autologous BMT in non-Hodgkin's lymphomas

The vast majority of patients with relapsed NHL are incurable. However, many of these patients are still responsive to combination chemotherapy and, more recently, to intensive chemoradiotherapy with infusion of either syngeneic, allogeneic, or autologous bone marrow to protect against myelo­toxicity [59-63]. The long-term survival of these patients approaches 25%. A major obstacle to the use of autologous BMT in NHL is the high fre­quency of bone marrow involvement, 20%-50% at diagnosis and significant­ly higher at relapse. With the preclinical and clinical studies in autologous BMT for ALL, several programs were subsequently initiated utilizing mono­clonal antibodies directed against antigens on the surface of B-cell and T-cell NHLs for in vitro purging prior to reinfusion of autologous bone marrow.

A study initiated by Baumgartner et al. in 1983 utilized anti-Y29/55 and rabbit C' to treat the bone marrow of pediatric patients with Burkitt's lymphoma or other B-cell type NHLs [64]. To date, seven patients with advanced stage disease were treated after remission induction, with vincri­stine, adriamycin, cyclophosphamide, and 600 rads TBI followed by infusion of anti-Y29/55 treated marrow. Six of seven patients were in a complete remission at the time of transplant. Of the six patients in first remission, five were reported to be in continuous complete remission from 8+ to 34+ months (median 27+ months) post BMT. The other patient transplanted in first remission, who had marrow involvement at diagnosis, subsequently relapsed in the marrow at six weeks, and one additional patient in second partial remission at the time of transplant relapsed in the abdomen. Hema­tologic engraftment of anti-Y29/55 purged marrow was similar to unpurged marrow, the median time to achieving a white blood count of lOOO/mm3 was 13 days, and a platelet count greater than 50,000/mm3 was seen at a median of 27 days. Four additional patients with Burkitt's lymphoma who received anti-Y29/55 treated marrow have been reported by Philip et al. [65]. Of these four patients, three were in complete remission at the time of trans­plant and two of these initially had a positive bone marrow. After autologous BMT, two of the patients were reported alive at 30+ and 186+ days in continuous complete remission. Similar to previous reported studies, the hematologic recovery of these patients did not differ from patients who received unpurged marrow.

In 1982 a protocol was initiated at DFCI for in vitro treatment of bone marrow in patients with relapsed B-cell NHL using anti-BI MoAb [66]. The

274

criteria for eligibility for this protocol included patients with 1) age less than 65,2) tumor cells which express B1, and 3) disease which had relapsed after standard treatment. In addition, the patients had to attain a near complete remission prior to autologous BMT (no masses greater than 2 cm2 , less than 5% bone marrow involvement, both histologically and by flow cytometric analysis) with either chemotherapy or local radiotherapy. As of November

Table 2. Clinical characteristics of Bl autologous BMT patients

Patient Age Sex LN histology

At diagnosis At transplant

1 50 M NM DM 2 45 M DLCL DLCL 3 57 M DPDL DLCL 4 46 M DM DLCL 5 30 M DPDL DLCL 6 46 F DPDL DPDL 7 42 F DPDL DPDL 8 39 M NPDL DLCL 9 57 M DLCL DLCL

10 57 F NPDL DPDL 11 52 M DLCL DLCL 12 38 F DWDL NPDL 13 36 F NPDL NPDL 14 38 F NPDL DPDL 15 42 M NPDL DLCL 16 58 M NM NM 17 42 M DUL DUL 18 25 F DM DLCL 19 26 F DLCL DLCL 20 48 M NH DLCL 21 61 M DPDL DPDL 22 51 M DLCL DLCL 23 36 F DLCL DLCL 24 44 M NPDL DM 25 57 M DPDL DLCL 26 43 F NPDL DM 27 44 M NPDL DPDL 28 40 M DLCL DLCL 29 33 M NPDL DLCL 30 26 M NPDL NM 31 52 F DLCL DLCL 32 34 M DPDL DPDL 33 48 M NPDL DLCL 34 45 M NPDL NPDL 35 31 F DLCL DLCL

Lymphoma histologies include nodular mixed lymphocytic histiocytic (NM), diffuse large cell (DLCL), diffuse mixed lymphocytic-histiocytic (DM), diffuse poorly differentiated lymphocytic (DPDL) , nodular poorly differentiated lymphocytic (NPDL) nodular histiocytic (NH) subtypes.

275

1986, 35 patients with recurrent B-cell NHL, with a median age of 46, have been treated on this protocol (table 2) [67]. Essentially all of the patients presented with unfavorable histologic subtypes and advanced stage disease (III/IV). All patients were induced into either a disease-free (16 patients) or minimal disease state (12 patients). Residual disease at the time of transplant included nodes only in six patients and focal bone marrow involvement with less than 5% BI+ cells in eight patients. The bone marrow from an addi­tional seven patients with greater than 5% bone marrow involvement was treated with anti-Bl and one or more monoclonal antibodies (including anti­B5, 15, and 12). Following bone marrow harvest, patients were treated with 60 mg/kg of cyclophosphamide over two days, followed by 1200 rad TBl fractionated over a three-day period. A complete response was sustained in all 35 patients, however eleven patients relapseu, nine within the first five months (table 3). One patient relapsed at seven, and a second at eight, months post-transplant. Essentially all patients relapsed in sites of previous bulk disease, and eight of the patients who relapsed died within three months of the time of relapse. Twenty-three patients are alive in unmain­tained remissions with a median survival of 14+ months (range 6+ to 51 + months).

The conditioning regimen was well tolerated by all patients. Nausea, vomiting, mucositis, weight loss, and fever were frequently seen. One patient died of veno-occlusive disease of the liver on day 20 following bone marrow infusion. Hematologic reconstitution was similar to other previously discussed autologous BMT protocols with purged marrow [68]. The first evidence of WBe recovery was noted by days 10-12, with a granulocyte count greater than 500/mm3 from 10 to 45 days (median of 22 days). A stable platelet count above 20,000/mm3 was seen between 14 and 57 days (median of 28).

The reconstitution of B-cells has been of interest in view of the presence of the BI antigen on 50% of pre-B-cells and all mature normal B-cells. Bl+ cells in peripheral blood were first detected 37 - 57 days after autologous BMT, with normal numbers (5% BI+ cells) detected between two and three months (68). Immunoglobulin levels dropped to less than 25% of normal and returned to normal levels between three and six months. T-cells were detected at around 10-17 days after transplant. These cells expressed T3, Tll, T12, and la, suggesting that they were mature activated T-cells. The T4/T8 ratio was reversed in all patients for 6-12 months. These results

Table 3. Current status of patients treated with Bl autologous BMT (n = 35)

Status

Remission

Relapse Remission deaths

Number

23

11

Months post-BMT

51,50,39,38,36,23,22, 18, 18, 16, 13, 13, 12, 12, 11,9,9,8,8,7,7,6,6

8,7,5,4,4,3,2,2,2,2,1

276

closely resemble the immunologic reconstitution seen in the J5112 autologous BMT patients.

In summary, the preparative regimen used in these multiply relapsed patients with B-cell non-Hodgkin's lymphoma could induce a complete re­sponse in all patients. Anti-Bl treated bone marrow could lead to normal hematologic and immunologic reconstitution. The toxicity of this program was acceptable, with only one treatment related death. Currently, 23 of 35 (66%) patients remain in unmaintained remission with a median disease-free survival of 14+ months.

There have been two additional recent reports of the use of anti-B-cell MoAb purged bone marrow in relapsed NHL. Rohatiner et al. have reported 12 patients with relapsed NHL, ten with follicular lymphoma, and two with high grade histology who received anti-Bl purged marrow [69]. Eleven of the patients remain in continuous complete remission with similar toxicity and hematologic and immunologic reconstitution to that seen at DFCI. With the predominant histologic subtype of nodular lymphoma, this series of patients will require longer follow-up to determine the therapeutic efficacy of this treatment. The University of Minnesota has reported 12 patients with a variety of histologic subtypes of B-cell NHL who received marrow treated in vitro with BA-l, BA-2, and BA-3 plus complement [70]. The Kaplan­Meier projected survival at 32 months is 55 ± 30%, and none of the patients relapsed after achieving CR with ABMT.

Two groups have examined the use of anti-T-cell MoAbs defining pan-T­cell antigens and C', as well as T101-ricin A chain, for purging marrow of patients with T-cell lymphoblastic lymphoma prior to autologous BMT. Seven patients have been reported from Johns Hopkins (ages 5-39) who received marrow treated with either Leu-l (anti-CD5) alone or a combination of Leu-l and Leu-9 (anti-CD7) and C' [51]. Three of these patients were reported to be disease-free at 131 + to 1320+ days, but four patients relapsed. Gorin et al. have reported three patients with advanced stage, T-Iympho­blastic lymphoma whose marrow was purged with T101-ricin A chain as a means of depleting neoplastic T-Iymphoblasts [52]. All three patients en­grafted with a time to developing adequate granulocytes and platelets similar to that seen in previously reported autologous BMT with purged marrow programs. At the time of the report, only one patient remained disease-free at 11 + months. Although it is difficult to know if marrow purging had a significant effect in these studies, anti-T-cell MoAb with C' and immunotoxin T-I01 treated marrow was capable of reconstituting normal hematopoietic elements.

Future directions of in vitro use of monoclonal antibodies in autologous bone marrow transplantation

The studies which have been reviewed have demonstrated that bone marrow which has been treated in vitro with monoclonal antibodies and C' and/or

277

immunotoxins is not toxic to the patient and is capable of reconstituting hematopoietic and lymphoid function following ablative therapy. Leukemia or lymphoma relapse, however, remains a major problem in these studies in patients who have previously demonstrated resistance to standard therapy. Further changes in the ablative regimen and marrow cleanup may there­fore improve the disease-free survival. With the experience that has been accumulated to date, coupled with the development and characteriz­ation of additional MoAbs, several new directions will be taken in auto­logous BMT.

There is extensive in vitro evidence that multiple MoAbs are synergistic in the elimination of neoplastic cells from an excess of normal bone marrow. Studies from this laboratory have demonstrated the coexpression of a variety of B-cell antigens on NHLs. For example, NPDLs generally coexpress CALLA, B1 and B5, and B1 and B5 are present on most DLCLs [27, 30]. Previous studies have demonstrated the lack of toxicity of the combination of anti-B1 and 15 to hematopoietic progenitor cells, and more recently the combination of anti-B11B5 has also been evaluated. As previously discussed, we have used these antibody combinations for marrow treatment in seven patients with overt marrow involvement. Consistent with the in vitro studies, these patients have all engrafted with antibody treated marrow. A very attractive combination would be the use of antibodies to the clonogenic cell of NHLs. It has been suggested, at least in myeloma, that the neoplastic event may involve a very early cell in ontogeny [71, 72]. B4, which as previously discussed, is a pan-B-cell antigen, expressed from the pre-B-cell stage up to the plasma cell stage of differentiation. B4 is not expressed on bone marrow progenitor cells, therefore, the use of B4 in combination with other B-cell antibodies may provide a more effective means of neoplastic cell elimination.

The appropriate selection of patients for autologous BMT will influence future application for this treatment modality . Several series have observed that autologous BMT in relapsed NHL leads to prolonged disease-free survival of only 20% -25% , with lethal toxicity in 20% -40% of patients [73, 74]. In contrast to these reports, the results of the DFCI experience demon­strate relatively high success rates with low mortality. Moreover, the observ­ed disease-free survival in B1 autologous BMT for relapsed patients is comparable to conventional combination therapy for primary NHL. These finding are likely due to the clinical features of patients selected for this program, including good performance status, being in a minimal disease state, and continued responsiveness to conventional chemotherapy or radia­tion therapy. This suggests that autologous BMT for relapsed NHL, which is potentially toxic and costly, is highly effective in carefully selected patients. A potential future application of this procedure would be as a consolidative treatment for patients with particularly poor prognostic features for relapse after standard therapy. Thus, those subgroups of patients with a low prob­ability of achieving long-term survival after conventional chemotherapy

278

[75-80] may benefit from autologus BMT as part of their induction therapy. Similarly, in adult ALL where conventional therapy leads to long-term disease-free survival in only 30% of patients [81], the application of autolo­gous BMT in first remission may improve relapse-free survival.

A major limitation to the use of MoAbs for purging marrow in AML prior to autologous BMT has been the cross-reactivity between AML cells

.and normal hematopoietic stem cells. To effectively deplete AML cells, the antibody used would have to react with clonogenic AML cells but not with pluripotent normal stem cells. Recently, six patients with AML in remission have been treated by Ball et al. on a protocol using two MoAbs with reac­tivity with AML cells, committed myeloid progenitor cells, but not pluri­potent progenitor cells [82]. After conditioning with cyclophosphamide and TBI, patients received PM-81 and AML-2-23 treated marrow. Six patients were in second remission, two in third, and two in early first relapse. Seven of the ten patients are in unmaintained complete remission from 2+ to 21 + months, one relapsed at 67 days, and two died in remission at 6 and 7.7 months. Eight of the patients had full hematopoietic recovery, but two never achieved a platelet count above 20,000/mm3 . At DFCI a program has been recently initiated using the anti-My9 MoAb and C' to deplete clonogen­ic AML cells from marrow of patients with AML in second remission prior to autologous BMT. Anti-My9 identifies a 68 kd surface antigen expressed on leukemic cells and leukemic colony forming cells from more than 80% of patients with AML [83]. MY9 is present on peripheral blood monocytes, but not T-cells, B-cells, or granulocytes. In the bone marrow, 20%-30% of cells are My9 positive, including 90% of CFU-GM and approximately half the BFU-E and CFU-GEMM cells. Nevertheless, normal bone marrow cells treated with anti-My9 and C' were still able to proliferate and differentiate in a normal fashion in long-term marrow culture assays, and therefore anti­My9 may be an ideal reagent for autologous BMT in AML.

In summary, a series of MoAbs which define normal hematopoietic differ­entiation antigens have been used in a variety of techniques to purge bone marrow from patients with leukemias and lymphomas prior to autologous BMT. Bone marrow treated ex vivo with either MoAbs and C', or immuno­toxin conjugates can effectively deplete neoplastic cells in vitro without affecting normal hematopoietic stem cell function. The majority of studies reviewed involved patients with relapsed non-T-cell ALL and B-cell NHLs. After ablative therapy, patients had complete hematologic and immunologic engraftment, with approximately 60% of the NHL patients and 30% of the ALL patients having prolonged disease-free survivals. The development of alternative ablative therapies and purging methods may improve the impact of this approach of treatment on these diseases in relapsed patients. The use of autologous BMT as a means of consolidation, may improve the disease­free survival for subgroups of patients with high risk of relapse following standard therapy.

279

References

l. Thomas ED: Marrow transplantation for malignant diseases (Karnofsky lecture). J Clin Oncol 1:517-31, 1983.

2. Champlin RE, Gale RP: Role of bone marrow transplantation in the treatment of hema­tologic malignancies and solid tumors : Critical review of syngeneic, autologous, and allo­geneic transplants. Cancer Treat Rep 68: 145-61 , 1984.

3. Anderson T , Chabner BA, Young RC, Berard CW, Garvin AJ , Simon RM, DeVita VT Jr: Malignant lymphoma I. The histology and staging of 473 patients at the National Cancer Institute . Cancer 50:2699-2707, 1982.

4. Benjamin D, Magrath IT, Douglass EC, Corash LM: Derivation of lymphoma cell lines from microscopically normal bone marrow in patients with undifferentiated lymphomas: Evidence of occult bone marrow involvement. Blood 61:1017-9, 1983.

5. Hu E , Trela M, Thompson J, Lowder, J , Horning S, Levy R , Sklar J : Detection of B-cell lymphoma in peripheral blood by DNA hybridization. Lancet 2:1092-5, 1985.

6. Nadler LM: B cell/leukemia panel workshop: Summary and comments. In: Reinherz EL, Haynes BF, Nadler LM, Bernstein ID (eds): Leukocyte Typing II, vol. 2, New York, Springer-Verlag, 1986, pp 3-43.

7. Anderson KC, Bates MP, Slaughenhoupt BL, Pinkus GS , Schlossman SF, Nadler LM: Expression of human B cell associated antigens on leukemias and lymphomas: A model of human B cell differentiation. Blood 63 :1424-33, 1984.

8. Nadler LM, Anderson KC, Marti G, Bates M, Park E, Daley JF, Schlossman SF: B4, a human B cell associated antigen expressed on normal, mitogen activated, and malignant B lymphocytes. J Immunol 131:244-50, 1983.

9. Stashenko P, Nadler LM, Hardy R, Schlossman SF: Characterization of a new B lymphocyte specific antigen in man. J Immunol 125:1678-85, 1980.

10. Nadler LM, Stashenko P, Ritz J, Hardy R, Pesando JM, Schlossman SF: A unique cell surface antigen identifying lymphoid malignancies of B cell origin. J Clin Invest 67:134-40, 1981.

11. Nadler LM, Ritz J, Griffin JD Todd RF, Reinherz EL, Schlossman SF: Diagnosis and treatment of leukemias and lymphomas utilizing monoclonal antibodies. Progress in Hema­tology 12:187-225, 1981.

12. Abramson C, Kersey JH , LeBien TW: A monoclonal antibody (BA-I) primarily reactive with cells of human B lymphocyte lineage. J Immunol 126:83-7, 1981.

13. Nadler LM, Stashenko P, Hardy R, van Agthoven A , Terhorst C, Schlossman SF: Charac­terization of a human B cell specific antigen (B2) distinct from B1. J Immunol 126:1941-7, 1981.

14. Iida K, Nadler LM, Nussenzweig V: Identification of the membrane receptor for the complement fragment C3d by means of a monoclonal antibody. J Exp Med 158:1021-33 , 1983.

15. Fingeroth JD , Weis J , Tedder TF, Strominger JL, Biro PA, Fearon DT: Epstein-Barr virus receptor of human B lymphocytes is the C3d receptor CR2. Proc Nat! Acad Sci USA 81 :4510-4, 1984.

16. Nadler LM, Boyd AW, Park E, Anderson KC, Fisher D, Slaughenhoupt B, Thorley­Lawson DA, Schlossman SF: The B cell-restricted glycoprotein (B2) is the receptor for Epstein-Barr virus. In: Reinherz EL, Haynes BF, Nadler LM, Bernstein ID (eds): Leuko­cyte Typing II, vol. 2, New York, Springer Verlag, 1986, pp 509-518.

17. Stashenko P, Nadler LM, Hardy R, Schlossman SF: Expression of cell surface markers following human B lymphocyte activation . Proc Nat! Acad Sci USA 78:3848, 1981.

18. Boyd A W, Anderson KC, Frecdman AS et al: Studies of in vitro activation and differentia­tion of human B lymphocytes. I. Phenotypic and functional characterization of the B cell population responding to anti-Ig antibody. J Immunol 134:1516-23 , 1985.

280

19. Freedman AS, Boyd AW, Fisher DC, Schlossman SF, Nadler LM: Antigenic changes of the B cell panel of the Second International Workshop on Human Leukocyte Differentiation Antigens with in vitro activation. In: Reinherz EL, Haynes BF, Nadler LM, Bernstein ID (eds): Leukocyte Typing II, vol 2, New York, Springer Verlag, 1986, pp 443-454.

20. Freedman AS, Boyd AW, Anderson KC, Fisher DC, Schlossman SF, Nadler LM: B5, a new B cell restricted activation antigen. J Immunol 134:2228-35, 1985.

21. Thorley-Lawson, DA, Schooley RT, Bhan AK, Nadler LM: Epstein-Barr virus superin­duces a new human B cell differentiation antigen (B-LAST 1) expressed on transformed lymphoblasts. Cell 30:415-25, 1982.

22. Thorley-Lawson DA, Nadler LM, Bhan AK, Schooley RT: B-last-2 (EBVCS), an early cell surface marker of human B cell activation, is superinduced by Epstein-Barr virus. J Immunol 134:3007-12, 1985.

23. Yokochi T, Holly RD, Clark EA: B lymphoblast antigen (BB-1) expressed on Epstein-Barr virus-activated B cell blasts , B lymphoblastoid cell lines, and Burkitt's lymphomas. J Immunol 128:823-7, 1982.

24. Tsudo M, Uchiyama T, Uchino H: Expression of TAC antigen on activated normal human B cells. J Exp Med 160:612-7, 1984.

25. Boyd AW, Fisher DC, Fox D, Schlossman SF, Nadler LM: Structural and functional characterization of IL-2 receptors on activated B cells . J Immunol 134: 2387-92, 1985.

26. Kerhl JH, Muraguchi A, Fauci As: Differential expression of cell activation markers after stimulation of resting B lymphocytes. J Tmmunol 132:2857-61 , 19R4.

27. Freedman AS, Boyd AW, Berrebi A et al: Expression of B cell activation antigens on normal and malignant B cells. Leukemia 1:9, 1987.

2R. Ritz J, Pesando M, Notis-McConarty J, Lazarus H, Schlossman SF: A monoclonal anti­body to human acute lymphoblastic leukemia antigen. Nature 2R3:5R3, 1980.

29. Greaves MF, Hariri G, Newman RA, Sutherland DR, Ritter MA, Ritz J: Selective expression of the common acute lymphoblastic leukemia (gp100) antigen on immature lymphoid cells and their malignant counterparts. Blood 61 :628-39, 1983.

30. Ritz 1, Nadler LM, Bhan AK, Notis-McConarty J, Pesando JM, Schlossman SF: Expression of common acute lymphoblastic leukemia antigen (CALLA) by lymphomas of B cell and T cell lineage. Blood 58:648-52, 1981.

31. Kersey JH, LeBien TW, Abramson CS, Newman R, Sutherland R, Greaves M: p 24: A human hemopoietic progenitor and acute lymphoblastic leukemia-associated cell surface structure identified with a monoclonal antibody. 1 Exp Med 153:726-31, 1981.

32. Hercend T, Nadler LM, Pesando JM, Reinherz EL, Schlossman SF, Ritz 1: Expression of a 26,000 dalton glycoprotein on activated human T cells. Cell Immunol 64:192-9, 19R1.

33. Forster HK, Gudat FG, Girad MF, Albrecht R, Schmidt J, Ludwig C, Obrecht JP: Monoclonal antibody against a membrane antigen characterizing human leukemic B lym­phocytes . Cancer Res 42:1927-34, 1982.

34. Anderson KC, Park K, Bates M, Leonard RC, Hardy R, Schlossman SF, Nalder LM: Antigens on human plasma cells identified by monoclonal antibodies. 1 Tmmunol130:1132-R, 1983.

35. Terhorst C, van Agthoven A, LeClair K, Snow P, Reinherz EL, Schlossman SF: Bioche­mical studies in the human thymocyte antigens T6, T9, and TIO. Cell 23:771-80, 19R1.

36. Reinherz EL, Schlossman SF: The differentiation and functions of human T lymphocytes: A review. Cell 19:821-7, 1980.

37. Haynes BF: Summary of T cell studies performed during the Second International Work­shop and Conference on Human Leukocyte Differentiation Antigens. In: Reinherz EL, Haynes BF, Nadler LM, Bernstein TD (eds): Leukocyte Typing II, vol 1, New York, Springer Verlag, 1986, pp 3-30.

38. Reinherz EL, Geha R, Rappeport 1M, Wilson M, Penta AC, Hussey RE, Fitzgerald KA, Daley IF, Levine H, Rosen FS, Schlossman SF: Immune reconstitution in severe combined immunodeficiency following transplantation with T lymphocyte depleted HLA haplotype mismatched bone marrow. Proc Natl Acad Sci USA 79:6047-51, 1982.

281

39. Reinherz EL, Nadler LM, Sallan SE, Schlossman SF: Subset derivation of T cell acute lymphoblastic leukemia in man. J Clin Invest 64:392-7, 1979.

40. Bernard A, Boumsell L, Reinherz EL, Nadler LM, Ritz J, Coppin H, Richard Y, Valensi F, Dausset J, Flandrin G, Lemerle J, Schlossman SF: Malignant T cells from acute lymphoblastic leukemia and from malignant lymphomas exhibit different patterns of surface antigens. Blood 57:1105-10,1981.

41. Nadler LM, Reinherz EL, Weinstein HJ, D'Orsi CJ, Schlossman SF: Heterogeneity of T cell lymphoblastic malignancies. Blood 55:806-10, 1980.

42. Jaffe ES, Berard CW: Lymphoblastic lymphoma, a term rekindled with new precision. Ann Int Med 89:415-17,1978.

43. Reinherz EL, Nadler LM, Rosenthal DA, Moloney WC, Schlossman SF: T cell subset characterization of human T-CLL. Blood 53:1066-75,1979.

44. Boumsell L, Bernard A, Reinherz EL, Nadler LM, Ritz J, Coppin H, Richard Y, Dubertret L, Valensi F, Degos L, Lemerle J, Flandrin G, Dausset J, Schlossman SF: Surface antigens on malignant Sezary and T-CLL cells correspond to those of mature T cells. Blood 57:526-30, 1981.

45. Waldmann TA, Greene WC, Sarin PS, Saxinger C, Blayner DW, Blattner WA, Goldman CK, Bongiovanni K, Sharrow S, Depper JM, Leonard W, Uchiyama T, Gallo RC: Func­tional and phenotypic comparison of human T cell leukemia/lymphoma virus positive adult T cell leukemia with human T cell leukemia/lymphoma virus negative Sezary leukemia, and their distinction using anti-Taco J Clin Invest 73:1711-18, 1984.

46. Jansen J, Falkenburg JH, Stephan DE, LeBien TW: Removal of neoplastic cells from autologous bone marrow grafts with monoclonal antibodies. Sem Hem 21:164-81, 1984.

47. Bast RC Jr, Ritz J, Lipton J, Feeney M, Sallan SE, Nathan DG, Schlossman SF: Elimina­tion of leukemia cells from human bone marrow using monoclonal antibody and comple­ment. Cancer Res 43:1389-94, 1983.

48. Bast RC Jr, DeFabritiis P, Lipton J, Gelber R, Maver C, Nadler L, Sallan S, Ritz J: Elimination of malignant clonogenic cells from human bone marrow using multiple mono­clonal antibodies and complement. Cancer Res 45:499-503, 1985.

49. LeBien TW, Stephen DE, Bartholomew RM, Stong RC, Anderson JM: Utilization of a colony assay to assess the variables influencing eliminating of leukemic cells from human bone marrow with monoclonal antibodies and complement. Blood 65:945-50, 1985.

50. Baumgartner C, Imbach P, Luthy A, Odavi C: Autologous bone marrow transplantation for Stage IV abdominal non-Hodgkin 's lymphoma after in vitro purging with anti-Y 29/55 monoclonal antibody and complement. In: Monoclonal Antibodies and Cancer, Academic Press, 1983, pp 73-79.

51. Braine HG, Kaizer H, Yeager AM, Stuart RK, Burns WH, Saval R, Sensenbrenner U, Sautos GW: Treatment of refractory non-Hodgkin's lymphoma with intensive chemo­radiotherapy and autologous bone marrow transplantation. In: Cavalli F, Bonnadonna G, Rozencweig M (eds): Malignant Lymphomas and Hodgkin's disease: Experimental and therapeutic advances. Boston, Martinus Nijhoff Publishing 1985, pp 529-536.

52. Gorin NC, Douay L, Laporte JP, Lopez M, Zittoun R, Rio B, David R, Stachowiak J, Jansen J, Cazellas P, Poncelet P, Liance MC, Voisin GA, Salmon C, LeBlanc G, Deloux J, Najman A, Duhamed G: Autologous bone marrow transplantation with marrow deconta­minated by immunotoxin T101 in the treatment of leukemia and lymphoma: First clinical observations. Cancer Treat Rep 69:953, 1985.

53. Netzel B, Haas RJ, Rodt H, Kolb HJ, Thierfelder S: Immunologic conditioning of bone marrow for autotransplantation in childhood acute lymphoblastic leukemia. Lancet 1: 1332-4, 1980.

54. Ritz J, Sallan SE, Bast RC, Lipton JM, Clavell LA, Feeney M, Hercend T, Nathan DG, Schlossman SF: Autologous bone marrow transplantation in CALLA positive acute lym­phoblastic leukemia after in vitro treatment with J5 monoclonal antibody and complement. Lancet 2:60-3, 1982.

55. Johnson FL, Thomas ED, Clark BS, Chard RL, Hartmann JR, Storb R: A comparison of

282

marrow transplantation with chemotherapy for children with acute iymphoblastic leukemia in second or subsequent remission. N Engl J Med 305:846-51, 1981.

56. Ramsay N, LeBien TW, Nesbit M, McGlave P, WeisdorfD, Kenyon P, Hurd D, Goldman A, Kim T, Kersey J: Autologous bone marrow transplantation for patients with acute lymphoblastic leukemia in second or subsequent remission: Results of bone marrow treated with monoclonal antibodies BA-l, BA-2, and BA-3 plus complement. Blood 66:508-13, 1985.

57. Filipovich AH, Vallera DA, Youle RJ, Strong RC, Neville DM, Kersey IH: Marrow purging with immunotoxins in allogeneic and autologous bone marrow transplantation: Pilot clinical studies. In: Dicke KA, Spitzer G, Zander AR (eds): Autologous Bone Marrow Transplantation, Proceedings of the First International Symposium, 1985, pp 403-407.

58. Maraninchi D, Novakovitch G, Laurent G, Mascret B, Tubiana N, Derocq 1M, Casella P, Gastant lA, Perrimond H, Carcassone Y: Bone marrow transplantation after in vitro purging with TI01 Ricin A-chain immunotoxin a phase I clinical study. In: Dicke KA, Spitzer G, Zander AR (eds): Autologous Bone Marrow Transplantation, Proceedings of the First International Symposium, 1985, pp 425-427.

59. Appelbaum FR, Herzig GP, Ziegler lL, Graw RG, Levine AS, Deisseroth AB: Successful engraftment of cryopreserved autologous bone marrow in patients with malignant lympho­ma. Blood 52:85-95, 1978.

60. Gorin NC, David R, Stachowiak 1, Salmon C, Petit JC, Parlier Y, Najman A, Duhamel G: High dose chemotherapy and autologous bone marrow transplantation in acute leukemias, malignant lymphomas, and solid tumors. Eur 1 Cancer 17:557-68, 1981.

61. Appelbaum FR, Fefer A, Cheever MA, Buckner CD, Greenberg PD, Kaplan HG, Storb R, Thomas ED: Treatment of non-Hodgkin'S lymphoma with marrow transplantation in identical twins. Blood 58:509-13, 1981.

62. Appelbaum FR, Thomas ED: Review of the use of marrow transplantation in thc treat­ment of non-Hodgkin 's lymphoma. 1 Clin Oncol 1:440-7, 1983.

63. Phillips GL, Herzig RH, Lazarus HM, Fay JW, Wolff SN, Mill WB, Lin H, Thomas PRM, Glasgow GP, Shina DC, Herzig GP: Treatment of resistant malignant lymphoma with cyclophosphamide, total body irradiation, and transplantation of cryopreserved autologous marrow. N Engl J Med 310:1557-61,1984.

64. Baumgartner C, Brundel ReG, Forster HK, Bucher U, Delalen B, Hirt A, Imbach P, Luthy A, Stern AC, Wagner HP: Autologous bone marrow transplantation for pediatric non-Hodgkin'S lymphoma: In vitro purging of the graft with anti-Y 29/55 monoclonal antibody and complement. In: Dicke KA, Spitzer G, Zander AR (eds): Autologous Bone Marrow Transplantation, Proceedings of the First International Symposium, 1985, pp 377-381.

65. Philip T, Biron P, Maraninchi D, Goldstone AH, Herre P, Souillet G, Gastant JL, Plouviek E, Flesh Y, Philip I, Harousseau JL, LeMevel A , Rebatten P, Linch DC, Freycon F, Milan JJ, Souhami RL: Massive chemotherapy with autologous bone marrow transplan­tation in 50 cases of non-Hodgkin's lymphoma with poor prognosis. In: Dicke KA, Spitzer G, Zander AR (eds): Autologous Bone Marrow Transplantation, Proceedings of the First International Symposium, 1985, pp 89-107.

66. Nadler LM, Takvorian T, Botnick L, Bast RC, Finberg R, Hellman S, Canellos GP, Schlossman SF: Anti-Bl monoclonal antibody and complement treatment in autologous bone marrow transplantation for relapsed B-cell non-Hodgkin's lymphoma. Lancet 2:427-31, 1984.

67. Takvorian T, Canellos GP, Ritz 1, Freedman AS, Anderson KC, Manch P, Tarbell N, Coral F, Daley H, Yeap B, Schlossman SF, Nadler LM: Prolonged disease-free survival in patients with poor prognosis non-Hodgkin'S lymphoma following autologous bone marrow transplantation (submitted).

68. Anderson KC, Ritz 1, Takvorian T, Coral F, Daley H, Gorgone BC, Freeman AS, Canellos GP, Schlossman SF, Nadler LM: Hematologic engraftment and immune recon­stitution post-transplantation with anti-Bl purged autologous bone marrow. Blood 69:597-604, 1987.

283

69. Rohatiner AZS, Barnett Ml, Arnott S, Plowman N, Cotter F, Adams K, Batten EL, Outram S, Amess lAL, Horton MA, Lister T A: Ablative therapy supported by autologous bone marrow transplantation with in vitro treatment of marrow in patients with B cell malignancy. Blood 68:241a, 1986.

70. Hurd D, LeBien TW, Peterson B, Ramsay N, Levine E, Lasky L, Kim T, Bloomfield CD, Filipovich A, Vallera D, McGlave P, Kersey 1: Intensive therapy and autologous bone marrow support for the treatment of refractory non-Hodgkin's lymphoma. Proc Am Soc Clin Oncol 5:193, 1986.

71. Pilarski LM, Mant Ml, Ruether BA: Pre-B cells in peripheral blood of multiple myeloma patients. Blood 66:416-22, 1985.

72. Goldstein M, Hoxie 1, Zembryki D, Matthews D, Levinson AI: Phenotypic and functional analysis of B cell lines from patients with multiple myeloma. Blood 66:444-6, 1985.

73. Armitage 10, Gale RP: Bone marrow autotransplantation in man: Report of an interna­tional cooperative study. Lancet 2:960-962, 1986.

74. Armitage 10, lagannath S, Spitzer G, Bierman P, Kessinger A, Kumar P, Cabanillas F, Zander A, Vellekoop L, Dicke K: High dose therapy and autologous bone marrow transplantation as salvage treatment for patients with diffuse large cell lymphoma. Br 1 Hematol 22:871-877, 1986.

75. Fisher RI, DeVita VT lr, lohnson BL, Simon TL, Young RC: Prognostic factors for advanced diffuse histiocytic lymphoma following treatment with combination chemotherapy. Am 1 Med 63:177-83, 1977.

76. Fisher RI, Hubbard SM, DeVita VT lr, Berard CW, Wesley R, Cossman 1, Young RC: Factors predicting long-term survival in diffuse mixed, histiocytic, or undifferentiated lymphoma. Blood 58:45-51, 1981.

77. Armitage 10, Dick FR, Corder MP, Garneau SC, Platz CE, Slymen Dl: Predicting theraputic outcome in patients with diffuse histiocytic lymphoma treated with cyclopho­sphamide, adriamycin, vincristine, and prednisone (CHOP). Cancer 50:1695-1702, 1982.

78. Koziner B, Little C, Passe S, Thaler T, Sklaroff, Straus Dl, Lee Bl, Clarkson BD: Treatment of advanced diffuse histiocytic lymphoma: An analysis of prognostic variables. Cancer 49:1571-9, 1982.

79. Anderson T, DeVita VT lr, Simon RM, Berard CW, Canellos GP, Garvin Al, Young RC: Malignant lymphoma II. Prognostic factors and response to treatment of 473 patients at the National Cancer Institute. Cancer 50:2708-21, 1982.

80. Shipp MA, HArrington DP, Klatt MM, 10chelson MS, Pinkus GS, Marshall lL, Rosenthal DS, Skarin AT, Canellos GP: Identification of major prognostic subgroups of patients with large-cell lymphoma treated with m-BACOD or M-BACOD. Ann Int Med 104:757-65, 1986.

81. Clarkson B, Ellis S, Little C, Gee T, Arlin Z, Mertelsmann R, Andreeff M, Kempin S, Koziner B, Chaganti R, lhanwar S, McKenzie S, Cirrincione C, Gayner 1: Acute lympho­blastic leukemia in adults. Semin Hematol 12:160-179, 1985.

82. Ball ED, Mills LE, Coughlin CT, Beck JR, Cornwell GG: Autologous bone marrow transplantation in acute myelogenous leukemia: In vitro treatment with myeloid cell-specific monoclonal antibodies. Blood 68:1311-1315, 1986.

83. Griffin lD, Linch D, Sabbath KD, Larcom P, Schlossman SF: A monoclonal antibody reactive with normal and leukemic human myeloid progenitor cells. Leuk Res 4:521-534, 1984.

Index

Acquired immune deficiency syndrome (AIDS), 109, 115,223

Acute lymphoblastic leukemia (ALL) bone marrow transplantation (BMT) in ,

265 , 269, 270-273 , 278 central nervous system (CNS)

involvement in, 149-150, 151, 152, 153 , 154

classification of, 5-6 immunological phenotypes of, 153 interferon therapy in , 239 lymphoid differentiation scheme for,

16-18 monoclonal antibody therapy in , 253 residual disease monitoring in, 175-176,

177,178,179,180,182,187-189,199 terminal deoxynucleotidyl transferase

immunofluorescence (TdT-IF) assay in, 157, 158-160

Acute lymphocytic leukemia , 109 Acute myeloblastic leukemia (AML), and

bone marrow transplantation (BMT), 265,278

Acute myelogenous leukemia (AML) interferon therapy in, 241 monoclonal antibody therapy in, 253

Acute myeloid leukemia (AML) central nervous system (CNS)

involvement in, 149-150, 153 residual disease monitoring in, 198 surface membrane markers for, 157 terminal deoxynucleotidyl transferase

immunofluorescence (TdT-IF) assay in , 157

Acute nonlymphoblastic leukemia, and interferon therapy, 239

Adult T-celileukemia/lymphoma autologous bone marrow transplantation

(BMT) in, 268 , 278 classification of, 12-13 monoclonal antibody therapy in , 253

Aneuploid cells, in residual disease monitoring, 179

Angioimmunoblastic lymphadenopathy (AILD) , 83, 125-126

Antibody-dependent cellular cytotoxicity, and interferon therapy, 242-243

Antiferritin therapy, in advanced Hodgkin's disease, 224-228

Atypical blastoid peripheral T-cell lymphoma (PTL), 120

Atypical follicular lymphoid hyperplasia, 73 Autologous bone marrow transplantation

(BMT),265-278 B-cell antigens in, 266-268 clinical studies of, 270-273 future directions for, 276-278 non-Hodgkin's lymphomas and, 273-276 preclinical studies of, 268-270 residual disease monitoring and, 177 T-cell antigens in, 268 90-yttrium antiferritin therapy in

advanced Hodgkin's disease and, 227-228

B-cell acute lymphoblastic leukemia (B-ALL)

cerebrospinal fluid (CSF) cell analysis and, 153, 154

residual disease monitoring in, 188-189 B-cell immunoblastic lymphoma (IBL),

89-94 B-celllymphoblastic lymphoma (LBL),

103-105 B-celllymphomas

autologous bone marrow transplantation (BMT) and, 269, 270

monoclonal antibody therapy and, 255-258

B-cell non-Hodgkin's lymphoma autologous bone marrow transplantation

(BMT) in, 273, 275 , 276 monoclonal antibody therapy and, 259

B-cell small lymphocytic lymphoma (B­SLL),53-56

j3-glucuronidase, in cerebrospinal fluid

286

(CSF) cell analysis, 152 f)-2-microglobulin (f)2m), in cerebrospinal

fluid (CSF) cell analysis, 152 Blastic transformation of follicular small

cleaved cell lymphoma (FSCL), 73 B-lymphocytes

bone marrow transplantation (BMT) and,266-268

cell markers for, 1-3 interferon therapy and, 243 lymphoid differentiation scheme with,

16-18 non-Hodgkin's lymphoma phenotyping

with,34-37 residual disease monitoring and, 180-181

Bone marrow transplantation residual disease monitoring and, 175 see also Autologous bone marrow

transplantation (BMT) Breast cancer

radioimmunoscintigraphy (RIS) with monoclonal antibodies in, 216-217

residual disease monitoring of, 174, 177, 179

Burkitt's-like lymphoma (BLL), 107-116 follicular variants of, 73 histology of, 107 immunologic characteristics of, 109-112

Burkitt's lymphoma (BL), 107-116 bone marrow transplantation (BMT) in,

265,267 classification of, 10-11 follicular variants of, 73 histology of, 107 immunologic characteristics of, 109-112 interferon therapy in, 241, 244 pathogenesis of, 112-116

Carcinoembryonic antigen, in cerebrospinal fluid (CSF) cell analysis, 152

Cell markers, 1-5 Cerebrospinal fluid (CSF) malignant cells

central nervous system (CNS) and detection of, 149-150

immunological marker analysis of, 153-158

methods for detection of, 150-153 terminal deoxynucleotidyl transferase

immunofluorescence (TdT-IF) assay on, 158-162

Chemotherapy advanced Hodgkin's disease and, 226-227 monoclonal antibody therapy and, 253 residual disease monitoring for drug

resistance in, 174-175 Childhood acute lymphoblastic leukemia

(ALL), and residual disease monitoring, 177, 178

Chronic lymphocytic leukemia (CLL)

autologous bone marrow transplantation (BMT) in, 266

central nervous system (CNS) involvement in, 153

classification of, 13-14 differential diagnosis of, 9, 59 interferon therapy in, 237, 241, 245 lymphoid differentiation scheme for,

16-18 monoclonal antibody therapy in, 253 radioimmunoscintigraphy (RIS) of, 210,

214,218 residual disease monitoring in, 178

Chronic myelogenous leukemia (CML) bone marrow transplantation (BMT) in,

265 interferon therapy and, 238, 241

Chronic myeloid leukemia (CML) central nervous system (CNS)

involvement in, 153 residual disease monitoring in, 181

Clusters of differentiation (CD) for monoclonal antibodies, 3, 5 (table)

for B lymphocytes, 2 (table) residual disease monitoring and, 188-189 for T lymphocytes, 4 (table)

Common acute lymphoblastic leukemia (ALL)

cerebrospinal fluid (CSF) cell analysis and, 153

residual disease monitoring in, 188-194, 199,200-201

Composite lymphomas, 132-136 Cutaneous T-cell lymphoma (CTCL)

classification of, 12 interferon therapy in, 236-237, 241 lymphoid differentiation scheme for, 16 monoclonal antibody therapy in, 253 radioimmunoscintigraphy (RIS) of, 210,

211,214,217

Diffuse large cell lymphoma (DLCL), 80-83,86-100

autologous bone marrow transplantation (BMT) in, 267, 274

classification of, 9-10 differential diagnosis of, 100 extranodal sites of, 94-98 histologic transformation of, 56-57,

71-73 lymphoid differentiation scheme for, 16

Diffuse mixed cell lymphoma (DMxL), 9, 80,84-86

Diffuse poorly differentiated lymphocytic lymphoma (DPDL), and autologous bone marrow transplantation (BMT), 266-267,274

Diffuse small cleaved cell lymphoma (DSCL),80

classification of, 9, 120 differential diagnosis of, 59, 100 immunologic characteristics of, 83-85 lymphoid differentiation scheme for, 16

Drug resistance , and residual disease monitoring, 174-175

Epithelial neoplasm, poorly differentiated, 116

Epstein-Barr (EB) virus, with Burkitt's lymphoma (BL), 112-115

Essential thrombocytopemia, and interferon therapy, 238-239

Extramedullary plasmacytoma, 129-130 Extranodal pseudolymphoma, 73

Ferritin Hodgkin's disease and levels of, 223-224 therapy in advanced Hodgkin's disease

with radiolabeled antibody to, 224-228 Follicular large cell lymphoma (FLCL) , 63 ,

71 blastic transformation of, 73 classification of, 8-9, 61 lymphoid differentiation scheme for, 16

Follicular mixed cell lymphoma (FML), 61 , 63 , 64-71

Follicular small cleaved cell lymphoma (FSCL),59-61

classification of, 8-9, 59 differential diagnosis of, 76 immunologic characteristics of, 61-64 large cell lymphoma (LCL) with, 134 lymphoid differentiation scheme for, 16 monoclonal antibody therapy in, 256

Graft versus host disease (GVHD), 26 Granulocytic sarcoma, 100, 112, 116, 126

Hairy cell leukemia (HCL), 122, 124 autologous bone marrow transplantation

(BMT) in, 267 classification of, 15 interferon therapy in , 235, 245-246

Heavy chain disease classification of, 15 lymphoid differentiation scheme with, 16

Hodgkin's cells (He), 18-21 Hodgkin's disease

classification of, 18-21 differential diagnosis of, 100 ferritin as a tumor antigen in, 223-224 interferon therapy in , 236 iodine-131 antiferritin therapy in

advanced, 225-226 lymphocyte-depleted form of, 100, 126-

127 monoclonal antibody therapy in, 259

90-yttrium antiferritin therapy in advanced, 226-228

287

Human T-cell leukemia/lymphoma virus-I (HTLV-I) , 12

Human T-cell leukemia/lymphoma virus-I (HTL V-I)-associated T-cell leukemia/lymphoma, 119, 124,268

Human T-cell leukemia/lymphoma virus-II (HTLV-II), 124

Human T-cell leukemia/lymphoma virus-III (HTLV-III),115

Immunoblastic lymphoma (IBL), 80, 89-94, 118

lllindium anti ferritin therapy, in advanced Hodgkin's disease, 226, 228

lllindium radiolabeled antibody studies, with lymphoma, 210, 211, 213-214 , 217,219

Interferon-a antibody-dependent cellular cytotoxicity

and,243 antiproliferative effect of, 241 cell surface antigen modulation with, 244 characteristics of, 231-233 clinical experience with, 235, 236 , 237-

238,239,241,245-246 mode of action of, 239-241 monocyte function and , 243 natural killer activity and, 242 oncogene expression and, 245 T-cells and, 244

Interferon-~ cell surface antigen modulation with, 244 characteristics of, 231-233 mode of action of, 239-241 monocyte function and, 243

Interferon-y B-Iymphocytes and, 243 cell surface antigen modulation with, 244 characteristics of, 231-233 monocyte function and, 243

Interferon therapy, 231-246 characteristics of human interferon

species used in, 231-233 clinical experience with , 232 (table),

233-239 immunomodulatory activity of, 241-245 mode of action in, 239-241 specific diseases with, 245-246

Intermediate lymphocytic lymphoma (IDL), 127-129

differential diagnosis of, 58-59, 100, 129 123iodine radio labeled antibody studies, with

lymphoma, 218-219 125iodine radio labeled antibody studies, with

lymphoma , 210, 213-214, 218 131 iodine radiolabeled antibody studies,

with lymphoma, 216

288

Large cell immunoblastic lymphoma, 9-10 Large cell lymphoma (LCL), 118

differential diagnosis of, 65 follicular small cleaved cell lymphoma

(FSCL) with, 134 histologic transformation of, 71-73

Lennert's lymphoma (LEL), 80, 84-86, 100,124

Lethal midline granuloma (LMG), 125-126 Leukemic reticuloendotheliosis, 15 Lymphoblastic lymphoma (LBL), 100-107

classification of, 10 differential diagnosis of, 76-78, 105-107 histologic features of, 101 immunologic features of, 102-105 lymphoid differentiation scheme for, 16 residual disease monitoring and, 182

Lymphocyte-depleted form of Hodgkin's disease (LDHD), 100, 126-127

Lymphocytic thymoma, 105-107 Lymphomatoid granulomatosis (LG), 125-

126 Lymphomatoid papulosis (LP), 125-126 Lymphosarcoma cell leukemia, 9

Malignant histiocytosis (MH), 98-100 Malignant lymphoma, 10 Mantle zone lymphoma (MZL), 127-129

differential diagnosis of, 58-59, 129 immunologic features of, 127

Metastatic carcinoma, 100 radioimmunoscintigraphy (RIS) with

monoclonal antibodies of, 216-217 residual disease monitoring in occult, 177

Monoclonal antibodies (MoAB) autologous bone marrow transplantation

(BMT) and , 272, 276, 277, 278 B-cell associated antigens and, 1-3 cluster designations for , 3, 5 (table) human antimurine antibody response

(HAMA) and, 217-218 non-Hodgkin's lymphomas and, 33

(figure), 34, 39 radioimmunoscintigraphy (RIS) of

lymphoma with, 209-219 residual disease monitoring with, 182-199 T-cell associated antigens and, 3, 4 (table)

Monoclonal antibody therapy, 253-261 anti-idiotype antibodies in clinical trials

in , 255-258 labeled monoclonal antibodies in, 259,

260 (table) toxicity in, 259 unlabeled antibodies in clinical trials in,

254 (table), 255, 258-259 Monocytes, and interferon therapy, 243 Monocytic leukemia, and residual disease

monitoring, 178 Multiple myeloma

classification of, 15 interferon therapy in, 237-238, 241 residual disease monitoring in, 178

Mycosis fungoides (MF), 12, 132

Natural killer cell activity, and interferon therapy,241-243,245

Neuroblastoma, and residual disease monitoring, 179

Nodular, lymphocyte predominant Hodgkin's disease (NLPHD) , 78

Nodular lymphomas autologous bone marrow transplantation

(BMT) in , 274 (table) classification of, 8-9 residual disease monitoring in, 177

Nodular poorly differentiated lymphocytic lymphoma (NPDL), and autologous bone marrow transplantation (BMT), 266-267,274,277

Nodular sclerosis Hodgkin's disease, 132 Non-Hodgkin's lymphoma (NHL)

autologous bone marrow transplantation (BMT) in, 265, 269, 270, 273-276, 277

central nervous system (CNS) involvement in, 149-150, 151, 152, 154

classification of, 8, 31-136 composite lymphomas forms of, 132-136 immunologic principles of, 32-38 interferon therapy in, 236 monoclonal antibodies (MoAB) used

with, 33 (figure), 34, 39, 259 phenotyping of, 34-38 terminal deoxynucleotidyl transferase

immunofluorescence (TdT-IF) assay on, 158-159

working formulation (WF) of, 31-32 Nonlymphoid leukemia, and residual

disease monitoring, 198 Non-T-acute lymphoblastic leukemia (non­

T-ALL) autologous bone marrow transplantation

(BMT) in, 267 classification of, 6, 7 (table)

Null acute lymphoblastic leukemia (ALL) cerebrospinal fluid (CSF) cell analysis

and, 153 residual disease monitoring in, 188-189

Occult metastatic disease , and residual disease monitoring, 177

Oncogenes , and interferon therapy, 244-245

Osteogenic sarcoma, 174

Peripheral T-cell lymphoma (PTCL, PTL), 94, 112, 116-127

classification of, 11

differential diagnosis of, 124-126 histologic variants of, 120-122 immunologic features of, 116-120 lymphoid differentiation scheme for, 16

Philadelphia (Ph l ) chromosome in chronic myelogenous leukemia; and interferon therapy, 238

Plasmacytoma, 100, 129-130 Pre-B-cell acute lymphoblastic leukemia

(pre-B-ALL), 153 residual disease monitoring in, 188-189,

200 Prolymphocytic leukemia (PL)

classification of, 14 lymphoid differentiation scheme for, 16

Pseudofollicular growth centers (PFGCs), 53-56,57,76

Pseudofollicular lymphoblastic lymphoma, 105

Pseudofollicular T-cell lymphoma, 76 Pseudolymphoma, extranodal, 73

Radioimmunoscintigraphy (RIS) of lymphoma, 209-219

factors in tumor targeting in, 209-210 future developments in, 218-219 human anti murine antibody response

(HAMA) and, 217-218 intravenous delivery in, 210-214 lymphatic delivery in , 214-217

Reed-Sternberg cells (RSC) classification of Hodgkin's disease and,

18-21,78 lymphocyte-depleted form of Hodgkin's

disease (LDHD) with , 100, 126-127 Residual disease monitoring, 173-201

autologous bone marrow transplantation and, l77

chemotherapeutic drug resistance with, 174-175

clinical usefulness of, 199-200 criteria for success in, 177 current status of, 198-200 future directions for, 200-201 immunologic markers used in, 182-198 induction chemotherapy response and,

175-176 methods used in, 177-182 occult metastatic disease and, 177 rationale for , 173-177 relapse detection with, 173-175

Reticulum cells, in Hodgkin 's disease (HD) , 19

Sclerosing variant, diffuse large cell lymphoma (DLCL), 94

Sezary cell leukemia, 12 Sezary cells, and radioimmunoscintigraphy

289

(RIS),212-213 Sezary syndrome, and T-cell antigens and

bone marrow transplantation (BMT), 268

Signet ring cell lymphoma, 71 Signet ring cells, in peripheral T-cell

lymphoma (PTL), 124 Single photon emission computed

tomography (SPECT) , 219 Small lymphocytic lymphoma (SLL), 53-59

B-cell origin, 53-56 classification of, 9, 122 differential diagnosis of, 57-59, 76 histologic transformation of, 56- 57, 71,

73 immunologic diagnosis of, 43-45 lymphoid differentiation scheme for, 16 T-cell variants of, 56

Small noncleaved cell lymphoma, 10-11

T-cell acute lymphoblastic leukemia (T-ALL), 7-8,102-103 , 153

autologous bone marrow transplantation (BMT) in , 268, 276

residual disease monitoring in, 187, 188-189,198-199

T-cell chronic lymphocytic leukemia (CLL), 268

T-cell immunoblastic lymphoma (IBL), 91-94

T-cell lymphoblastic lymphoma (LBL) autologous bone marrow transplantation

(BMT) in, 267, 268 immunologic features of, 103-105

T-cell non-Hodgkin's lymphoma , and monoclonal antibody therapy , 259

T-cell receptor genes, 3-5 T-cell small lymphocytic lymphoma, 56 99mtechnetium radio labeled antibody

studies, with lymphoma, 217, 219 Terminal deoxynucleotidyl transferase

(TdT) cerebrospinal fluid (CSF) cell analysis

with immunofluorescence (IF) assay of, 149-150, 151 , 152, 153, 154

residual disease monitoring with, 179, 186,189-194,198-199;200-201

Ty lymphoproliferative disease classification of, 11-12 lymphoid differentiation scheme for, 16

T-lymphocytes cell markers for, 3 interferon therapy and, 243-244 lymphoid differentiation scheme with,

16-18 non-Hodgkin's lymphoma phenotyping

with,37-38 residual disease monitoring and, 180-181

True histiocytic malignancies, 98-100

290

Undifferentiated lymphoma, and residual disease monitoring, 182

Waldenstriim's macroglobulinemia classification of, 15

Working formulation (WF) of non­Hodgkin's lymphomas, 31

90-yttrium antiferritin therapy, in advanced Hodgkin's disease, 226-228