214 Cancer Control July 2012, Vol. 19, No. 3

From the Department of Pathology and Microbiology at the Nebraska Medical Center, Omaha, Nebraska (AMP, WCC), and the Department of Internal Medicine, Clinical Hospital Dubrava, Zagreb, Croatia (ZM).

Submitted January 13, 2012; accepted April 30, 2012.

Address correspondence to Wing C. Chan, MD, Department of Pathology and Microbiology, 983135 Nebraska Medical Center, Omaha, NE 68198-3135. E-mail: jchan@unmc.edu

No signifi cant relationship exists between the authors and the companies/organizations whose products or services may be ref-erenced in this article.

clinical and laboratory investigations complemented by novel molecular methods such as gene expression profi ling (GEP) and other genome-wide investigations have helped to expand our understanding of the biol-ogy and diversity of different types of NHL. This is refl ected in an increase in the number of entities in the recent World Health Organization (WHO) classifi cation. The guiding principle of the WHO classifi cation is an attempt to defi ne “real” diseases that can be recognized by pathologists using all available information: mor-phology, immunophenotype, genetic characteristics, and clinical features.1 Multiple novel therapeutic op-tions have emerged in the treatment of NHL, including monoclonal antibodies and different classes of biological agents. Therefore, due to increased diagnostic sophis-

New prognostic markers that stratify

patients according to risk factors are

needed to provide the basis for

individually tailored treatment in

diffuse large B-cell lymphoma.

George Van Hook. Calm Boats. Oil on linen, 12˝ × 16 .̋

Biological Prognostic Markers in Diffuse Large B-Cell LymphomaAnamarija M. Perry, MD, Zdravko Mitrovic, MD, and Wing C. Chan, MD

Background: Multiple novel therapeutic options have emerged in the treatment of non-Hodgkin lymphoma, including monoclonal antibodies and different classes of biological agents. With this increased diagnostic sophistication, novel prognostic markers are needed to stratify patients according to risk factors, particularly those with a mechanistic underpinning, to provide the basis for individually tailored treatment.Methods: Numerous prognostic markers have been proposed in patients with diffuse large B-cell lymphoma (DLBCL), and this review discusses the more studied and the most widely used prognostic markers in DLBCL in the rituximab era.Results: Prognostic markers in DLBCL include a range of biomarkers assessed by morphology, immunohisto-chemistry, and relatively novel molecular methods including gene expression profi ling, high-resolution array comparative genomic hybridization, and next-generation sequencing. Most of these methods are not routinely used due to substantial cost, technical complexity, and the requirement for fresh or frozen tissue.Conclusions: Efforts are underway to translate previous microarray fi ndings to platforms that can be readily used in routine clinical practice with high reproducibility, precise measurements, and minimal loss of information. At the present time, there is no consensus on which biological prognostic markers should be routinely assessed in patients with DLBCL, and practices vary widely among different institutions. With more global approaches, the ability to assess biomarkers in the cellular or tumor context may be possible, resulting in a better understanding of their biological and prognostic signifi cance.

IntroductionNon-Hodgkin lymphoma (NHL) is composed of a hetero-geneous group of tumors arising from B or T/NK cells at various stages of differentiation.1 In the last decade,

July 2012, Vol. 19, No. 3 Cancer Control 215

tication and therapeutic armamentarium, novel prog-nostic markers are needed to stratify patients according to risk factors, particularly those with a mechanistic underpinning, to provide the basis for individually tai-lored treatment.

Prognostic markers in NHL encompass not only a wide range of traditional as well as novel biomarkers assessed by different methods, but also “pure” clinical markers represented by the International Prognostic Index (IPI) and its variants. The most extensively studied markers in NHL are pathological markers. Immunohisto-chemistry (IHC) is a commonly used method to separate B- from T-cell lymphomas and to further classify differ-ent entities that may have prognostic signifi cance. IHC staining is widely available in routine clinical practice, is easy to use, and is relatively inexpensive. However, the differences in tissue processing, antibody clones used, different staining protocols, and interobserver variability contribute to poor reproducibility of IHC among differ-ent institutions.2 At the other end of the diagnostic spec-trum are relatively novel molecular methods such as GEP, high-resolution array comparative genomic hybridization (aCGH), and next-generation genome sequencing. GEP has revealed novel, biologically distinct subtypes within NHL entities, providing the biological basis for targeted treatment. However, these methods are not widely avail-able due to substantial cost, technical complexity, and the requirement of fresh or frozen tissue.

Diffuse large B-cell lymphoma (DLBCL) constitutes 30% to 40% of all NHL cases in Western countries and represents a biologically heterogeneous group of tu-mors.1 One of the most important clinical predictors of survival in DLBCL patients is the IPI, which uses patient age, Ann Arbor tumor stage, serum lactate dehydro-genase, performance status, and number of involved extranodal sites to identify patients as low risk, low-intermediate risk, high-intermediate risk, and high risk.3 Subsequent studies have validated the prognostic value of the IPI in DLBCL patients.4,5 The addition of ritux-imab to the standard chemotherapy protocol of cyclo-phosphamide, doxorubicin, vincristine, and prednisone (R-CHOP) has signifi cantly improved the survival of patients with DLBCL.6-10 Some authors have questioned the validity of the IPI in the rituximab era and have pro-posed the revised IPI (R-IPI) for clinical stratifi cation of DLBCL patients. The R-IPI, compared to the “traditional” IPI, distinguishes three separate prognostic groups — very good, good, and poor — and allows for a simpler and more accurate prediction model.11

Numerous prognostic markers have been proposed in patients with DLBCL, and this review discusses the more studied and the most widely used markers. Prog-nostic markers are applicable primarily to DLBCL, not otherwise specifi ed. They have not been suffi ciently studied in special entities of this disease and therefore are not applicable to these entities. However, some spe-

cial entities of DLBCL, as defi ned by a variety of criteria, have unique clinical implications and are also discussed.

GEP and Distinct Subgroups of DLBCLGEP has defi ned at least three biologically and prognos-tically distinct subgroups of DLBCL: germinal center B-cell–like (GCB) DLBCL, activated B-cell–like (ABC) DLBCL, and primary mediastinal large B-cell lymphoma (PMLBCL). The GCB subtype is believed to be derived from germinal center B cells and maintains the GCB dif-ferentiation program, while the ABC subtype putatively arises from B cells that are arrested in their differentia-tion toward plasma cells. Furthermore, survival was signifi cantly better for patients in the GCB subgroup compared with those in the ABC subgroup.12-16

Since GEP requires fresh or frozen tissue and this application is not widely available, multiple ontogenetic biomarkers such as BCL6, GCET2 (HGAL), CD10, LMO2, FOXP1, and PKC-β were tested by IHC and correlated with survival.17 It is likely that at least part of the predic-tive power is related to their differential expression in the GCB and ABC subtypes of DLBCL, although singly, these markers are not suffi ciently specifi c to classify the two subtypes of DLBCL. Therefore, various immunohis-tochemical algorithms have been developed in paraffi n-embedded tissue to reproduce the GEP classifi cation.18 The most widely used is the Hans algorithm, which uses three markers (CD10, BCL6, and MUM1) to separate GCB DLBCL from non–GCB DLBCL.19 A more recent Choi algorithm, based on fi ve immunohistochemical markers (GCET1, CD10, BCL6, MUM1, and FOXP1), had concor-dance of 87% or higher with GEP results, which was superior to the Hans algorithm.18,20 Meyer et al18 recently re-examined a number of algorithms and also proposed a new “Tally” method. They found that most of the pub-lished algorithms perform well, with > 80% concordance with GEP-classifi ed cases (Fig 1). Some reports used differences in patient survival to assess the usefulness of the classifi cation algorithm; however, this is a fl awed approach as prognosis depends on many factors other than classifi cation, such as the size of the patient popula-tion, the characteristics of the different populations, and how well or how uniformly treatment is administered.

The prognostic usefulness of DLBCL subtyping has been questioned in the rituximab era. Some investiga-tors reported no difference in survival between GCB and ABC subtypes when rituximab was added to the chemo-therapy regimen,21-23 while others have shown persistent difference.10 All studies that showed no difference in survival between the GCB and ABC subtypes were based on IHC algorithms. A GEP study of 233 patients treated with R-CHOP showed that patients with GCB DLBCL still had signifi cantly higher overall and progression-free survival than patients with ABC DLBCL had.24

Bortezomib is a protease inhibitor that can inhibit the NF-κB pathway by blocking IκBα degradation.25 Since

216 Cancer Control July 2012, Vol. 19, No. 3

ABC DLBCL has constitutively activated NF-κB pathway,13 a study by Dunleavy et al25 investigated whether the addi-tion of bortezomib to doxorubicin-based chemotherapy would preferentially improve survival of patients with ABC DLBCL. They found that patients with the ABC subtype, compared to those with the GCB subtype of DLBCL, had signifi cantly higher response and median overall survival (OS) when bortezomib was combined with chemotherapy. However, this is a relatively small study of 49 patients with relapsed DLBCL, and the fi nd-ings need to be confi rmed in independent large cohorts of de novo DLBCL patients.

Prognostic Models Based on Gene Expression SignaturesA few prognostic models, based on the combination of expressions of several genes, were proposed in patients

with DLBCL treated with rituximab. Malumbers et al26 proposed a quantitative real-time polymerase chain reac-tion (Q-RT-PCR)-based model for prediction of outcome in DLBCL patients. The expression of six genes was measured in paraffi n-embedded tissues; LMO2, BCL6, and FN1 were associated with longer survival, while CCND2, SCYA3, and BCL2 were associated with shorter survival. Alizadeh et al27 recently proposed a two-gene model based on the expression of a tumor biomarker LMO2 and a tumor microenvironment marker TNFRSF9 in patients with DLBCL. The two-gene model was an independent predictor of survival in the multivariate analysis. Since it seems unlikely that the entire biol-ogy of tumor cells and host/tumor interaction can be captured by one transcript, efforts are underway to translate previous microarray fi ndings to a platform that can be readily used in routine clinical practice with high

Choi algorithm

Tally algorithm

“+” = 1, “–” = 0

ABC

GCB

GCB

MUM1(≥ 80%)

CD10(≥ 30%)

+

+

–

–

GCET1(≥ 80%)

+

–ABC

ABC

GCBBCL6(≥ 30%)

FOXP1(≥ 80%)+

+

–

–

GCB ABC CD10 (+ or –) Mum1 (+ or –) GCET1 (+ or –) FoxP1 (+ or –) Score (0, 1, 2) Score (0, 1, 2)

ScoreGCB > ABC

or

ABC > GCB

LMO2 ≥ 30% GCBLMO2 < 30% ABC

If GCB Score = ABC SCore:

A

B

Fig 1. — Choi (A) and Tally (B) immunohistochemical algorithms. In the Tally algorithm, antibody results are not examined in a particular order. Two antigens of germinal center B cells (GCB) and two antigens of activated B cells (ABC) are examined. The case is classifi ed according to expression of the higher number of GCB vs ABC-associated antigens. If an equal number of GCB-associated and ABC-associated antigens are positive, then LMO2 determines the phenotype. (A) Reproduced with permission of American Association for Cancer Research from Choi WW, Weisenburger DD, Greiner TC, et al. A new immunostain algorithm classifi es diffuse large B-cell lymphoma into molecular subtypes with high accuracy. Clin Cancer Res. 2009;15(17):5494-5502, permission conveyed through Copyright Clearance Center, Inc. (B) Reproduced with permission. © 2011 American Society of Clinical Oncology. All rights reserved. From Meyer PN, Fu K, Greiner TC, et al. Immunohistochemical methods for predicting cell of origin and survival in patients with diffuse large B-cell lymphoma treated with rituximab. J Clin Oncol. 2011;29(2):200-207.

July 2012, Vol. 19, No. 3 Cancer Control 217

reproducibility and better quantitation than IHC assays provide and with minimal loss of information.28,29

MicroRNA SignatureMicroRNAs (miRNAs) have been associated with out-come of DLBCL patients in a number of studies. Roehle et al30 described the global miRNA signature of B-cell lymphomas, including 58 cases of DLBCL. Eight miR-NAs were found to correlate with survival. Patients with downregulated miR-21, miR-23A, miR-27A, and miR-34A expression had an inferior OS, while patients with low levels of miR-19A, miR-195, and miR-LET7G had a shorter event-free survival (EFS). Patients with low expression of miR-127 had low OS and EFS. Alen-car et al31 studied expression of miRNAs in 176 DLBCL samples from rituximab-treated patients and found that increased expression of miR-18A was associated with shorter OS. Increased expression of miR-181A was seen in patients with longer EFS. In contrast, higher expres-sion of miR-222 was associated with shorter EFS.

MiRNA expression can also distinguish GCB and ABC subtypes of DLBCL. Several miRNAs, such as miR-155, miR-21, and miR-221, were found to be more highly expressed in the ABC subtype than in the GCB subtype.32,33 Furthermore, high miR-21 expression was associated with longer EFS in de novo DLBCL cases.32

Most of the reported series studies have been small and different platforms have been used with inclusion of variable number of miRNAs. The published fi ndings will need to be validated, refi ned, and extended in ad-ditional investigations that address these issues.

Tumor MicroenvironmentRecently, the tumor microenvironment has been shown to be an important prognostic factor in patients with DLBCL.24 Rosenwald et al13 identifi ed four gene-expres-sion signatures that predicted survival in CHOP-treated DLBCL patients: GBC, lymph node, major histocompat-ibility complex (MHC) class II, and proliferation.

Lymph node signature refl ected tumor microenvi-ronment, which was recently subdivided into two com-ponents: stromal-1 signature and stromal-2 signature.24 High stromal-1 signature identifi es tumors with vigor-ous extracellular-matrix deposition and infi ltration by monocytes/macrophages and predicts good prognosis, while the stromal-2 signature largely refl ects angiogen-esis and blood vessel density in the tumor stroma, and high expression portends poor prognosis. Meyer et al34 attempted to reproduce the stromal-1 signature using an antibody against secreted protein, acidic and rich in cysteine (SPARC) to evaluate its expression in the tumor microenvironment. Patients with high SPARC positivity in the tumor stroma had a signifi cantly lon-ger survival than those with low or no SPARC staining. Cardesa-Salzmann et al35 recently attempted to simu-late the stromal-2 signature by measuring microvessel

density (MVD) in DLBCL and found high MVD to be an unfavorable prognostic factor.

Genomic Aberrations in DLBCLA number of studies have investigated genomic aberra-tions in DLBCL and their infl uence on prognosis. Cer-tain genetic aberrations occur at different frequencies among DLBCL subtypes. The t(14;18) translocation and amplifi cation of 2p16 are associated with the GCB sub-type, while trisomy 3 or gain/amplifi cation of chromo-some arm 3q is associated with the ABC subtype. The ABC DLBCL subtype is further characterized by gain of 18q and loss of 6q.36

Scandurra et al37 analyzed samples from 124 ritux-imab-treated DLBCL patients using a high-density ge-nome-wide single nucleotide polymorphism-based array. They found 58 gains, 47 losses, 54 losses of heterozy-gosity, 5 recurrent amplifi cations, and 7 homozygous deletions. Twenty recurrent genetic lesions showed an impact on the clinical course, among which deletions affecting the short arm of chromosome 8 — del(8p23.1), del(8p), and del(8p23.1-21.2) — showed the strongest association with the poor outcome. Lenz et al36 analyzed 203 DLBCL samples using high-resolution array com-parative genomic hybridization (aCGH) and found two recurrently altered minimal common regions restricted to ABC DLBCL that predicted adverse survival: trisomy 3 and INK4A/ARF locus single/double deletion. Chigri-nova et al38 characterized DLBCL with chromosome 7q gain. The gain of 7q delineated a group of DLBCL with distinct biological and clinical characteristics. Most of the patients were females and had prolonged OS with no bone marrow involvement and signifi cantly lower in-volvement of extranodal sites. Salaverria et al39 recently found that t(6;14)(p25;q32) translocation that deregu-lates IRF4 is associated with GCB subtype of DLBCL, younger age at diagnosis, and a favorable outcome. It is important to include a suffi cient number of cases in each category and perform multivariate analysis to reach reliable conclusions.

Single Prognostic BiomarkersTP53 and TP21TP53 is a tumor suppressor gene that acts as a multi-functional transcription factor involved in cell cycle arrest, apoptosis, cell differentiation, replication, DNA repair, and maintenance of genomic stability. Muta-tions in TP53 have been described in 18% to 30% of patients with DLBCL.17 Young et al40,41 identifi ed TP53 mutations in 21% of DLBCL patients, and the OS was signifi cantly worse than that of patients with wild-type TP53. Mutations in TP53 DNA-binding domains were the strongest predictor of poor OS. Mutations in the Loop-Sheet-Helix and Loop-L3 were associated with signifi cantly decreased OS, but OS was not signifi cantly affected by mutations in Loop-L2 (Fig 2).

218 Cancer Control July 2012, Vol. 19, No. 3

Amino Acid Sequence of p53 Protein

Fre

qu

ency

of

Mu

tati

on

s

100

K132

R175 R213

R248

R273R280

R282

500

1

2

3

4

5

6

7

8

9

10

11

12

150 200 250 300 350

II III IV VConserved Regions

Number of Mutations 11

AA117-142

12

171-181

28

234-258

16

270-286

54 6 7 8 9Number ofMutations

TP53 Exons

1

AA125 AA186 AA224 AA261 AA306

33 17 30 21 0

p53 Protein

Structure

COOH H2 S10 S9

L3

S8 S7 S6

NH2S1

L1

S1 S2 S3

S1

L2

H1

H

S2

287 279 274 264 258 251 250 237 236 233 216 214 207 204199

195

194 181 177

167 170

164

163

IL-6

147141136132127124112

113 123

110

Loop-L1

Lo

op

-L2

Loop-L3LSH

A

B

C

D

Fig 2. — Schematic representation of the TP53 gene and its mutations in diffuse large B-cell lymphoma. (A) The distribution of TP53 mutations in ex-ons 4 to 9, (B) their relation to p53 protein structure, (C) the mutations in conserved regions, and (D) the distribution and frequency of TP53 mutations with peaks at known hot spot exons depicted. This research was originally published in Blood. Young KH, Leroy K, Møller MB, et al. Structural profi les of TP53 gene mutations predict clinical outcome in diffuse large B-cell lymphoma: an international collaborative study. Blood. 2008;112(8):3088-3098. © the American Society of Hematology.

July 2012, Vol. 19, No. 3 Cancer Control 219

The cyclin-dependent kinase inhibitor TP21 nega-tively regulates cell cycle progression and inhibits cellu-lar proliferation. Although it is a downstream effector of the TP53, its expression is controlled by TP53-dependent and TP53-independent mechanisms.42 Multiple studies have quantifi ed expression of TP53 and TP21 by IHC. Strong nuclear staining for TP53 without TP21 staining has been associated with TP53 gene alterations and has been used as an imperfect surrogate for mutated TP53 in some studies.42 When assessed by IHC staining alone, TP53 has been shown to be an unreliable predictor of survival. Some investigators found an association be-tween high TP53 expression and adverse survival,43,44 while others failed to show this association.45,46 The addition of TP21 to the IHC panel somewhat improved the prognostic value of TP53 expression.42,47

MHC MoleculesLoss of MHC class I and class II (HLA-DP and HLA-DR) ex-pression has been reported to correlate with shortened survival in patients with DLBCL.48,49 The mechanism for

lost expression has been unclear. MHCII gene expres-sion is controlled by several transcription factors, includ-ing RFX, CREB, and NF-Y, which interact with a master transactivator protein class II transactivator (CIITA) to form an enhanceosome complex. Although overall in-frequent, decreases in CIITA expression appear to be the most prevalent mechanism of MHCII downregulation.50,51 In PMLBCL, CIITA translocation with a concomitant de-crease in MHCII expression is frequently observed.52

Pasqualucci et al53 recently performed massive parallel sequencing on DLBCL samples and found frequent inactivating mutations and deletions in the β2-microglobulin gene (B2M) (including homozygous deletions, biallelic mutations, or a combination of these). B2M encodes a polypeptide that associates with a 45-kD heavy chain to form the MHC class I molecule on the surface of all nucleated cells. Taken together, these inactivating mutations and deletions predict the loss of B2M, which is required for cell surface expression of HLA class I molecules and may impair the recognition of the tumor cells by cytotoxic T lymphocytes.

Fig 3. — Correlation of BCL2 protein expression with overall survival in (A) diffuse large B-cell lymphoma (DLBCL) as a single entity, (B) germinal center B-cell–like (GCB) subgroup, (C) activated B-cell–like (ABC) subgroup (30% cutoff), and (D) ABC subgroup (10% cutoff). BCL2 protein expression is predictive of survival in the ABC subgroup only (C and D). Reproduced with permission. © 2006 American Society of Clinical Oncology. All rights reserved. From Iqbal J, Neppalli VT, Wright G, et al. BCL2 expression is a prognostic marker for the activated B-cell–like type of diffuse large B-cell lymphoma. J Clin Oncol. 2006;24(6):961-968.

0 2 4 6 8 10

Ove

rall

Su

rviv

al

Time (years)

0.2

0.4

0.6

0.8

1.0

BCL2 protein negative (n = 71)

BCL2 protein positive (n = 67)

P = .095

0 2 4 6 8 10

Ove

rall

Su

rviv

al

Time (years)

0.2

0.4

0.6

0.8

1.0BCL2 protein negative (n = 28)

BCL2 protein positive (n = 27)

P = .69

0 2 4 6 8 10

Ove

rall

Su

rviv

al

Time (years)

0.2

0.4

0.6

0.8

1.0

BCL2 protein negative (n = 18)

BCL2 protein positive (n = 26)(30% cutoff)

P = .008

0 2 4 6 8 10

Ove

rall

Su

rviv

al

Time (years)

0.2

0.4

0.6

0.8

1.0BCL2 protein negative (n = 13)

BCL2 protein positive (n = 31)(10% cutoff)

P = .002

A

C

B

D

220 Cancer Control July 2012, Vol. 19, No. 3

BCL2BCL2, an antiapoptotic protein, was originally discov-ered due to its involvement in the t(14;18)(q32;q21) translocation, which juxtaposes the BCL2 gene (18q21) to the immunoglobulin (Ig) heavy-chain locus enhanc-ers and results in BCL2 overexpression. BCL2 protein is overexpressed in approximately 47% to 58% of DLBCL cases.17 BCL2 expression in the GCB subgroup of DLBCL is mainly through the presence of the translocation. However, BCL2 expression can also be upregulated by alternative mechanisms such as NF-κB activation and 18q21 gain/amplifi cation, as often observed in the ABC subgroup of DLBCL, which lacks the t(14;18).54 Numer-ous studies have investigated the correlation between BCL2 protein expression, BCL2 translocation, and out-come in patients with DLBCL with confl icting results. Iqbal et al55 detected t(14;18)(q32;q21) translocation in 17% of DLBCL cases based on fl uorescence in situ hybridization, and the great majority of cases were of the GCB subtype. However, there was no signifi cant difference in survival between the t(14;18)-positive and -negative patients in the GCB subgroup, in contrast to the signifi cantly poorer survival of ABC DLBCL with high BCL2 expression, when treated with CHOP (Fig 3).

Data regarding the BCL2 protein expression and its infl uence on survival in DLBCL patients are also contro-versial in the rituximab era. Some studies have found that the addition of rituximab to standard chemotherapy overcame the adverse prognostic infl uence of BCL2 ex-pression.56,57 Others have shown that BCL2 expression remained an adverse prognostic factor in the rituximab era, primarily in the non–GCB subgroup of patients.54,58 Iqbal et al59 recently evaluated a series of R-CHOP treated DLBCL patients who had GEP-defi ned DLBCL subsets and found that BCL-2 expression is a signifi cant predic-tor of survival in the GCB subgroup but not in the ABC subgroup (Fig 4). The addition of rituximab appears to have reduced the difference in survival between the BCL2-positive and -negative groups in the ABC subset. For GCB DLBCL, it may have improved the survival of

BCL2-negative patients to a signifi cantly greater extent than for the BCL2-positive subgroup. However, this latter fi nding needs to be confi rmed as the OS does not reach statistical signifi cance in the multivariate analysis. BCL2 mutations have been shown to occur most frequently in the GCB subtype of DLBCL.60 It is possible that some of the mutations that enhance the anti-apoptotic function of BCL2 may be selected and may add to the complexity of the analysis.

MYC and “Double-Hit” LymphomasMYC, located at chromosome band 8q24, encodes a tran-scription factor involved in the regulation of a variety of cellular processes that include proliferation, cell cycle control, metabolism, apoptosis, and cell migration.61 MYC is most commonly deregulated as a result of chro-mosomal translocation to an Ig gene locus in Burkitt lym-phoma (BL), but MYC translocations also occur in 7% to 10% of DLBCL cases.62,63 A number of studies reported an adverse prognostic impact of MYC on survival of patients with DLBCL who were treated with rituximab. Rimsza et al64 found that high level of MYC expression, assessed by quantitative nuclease protection assay (qNPA) in paraffi n-embedded tissue, was an independent indicator of poor survival. Several studies found that the presence of MYC rearrangements by fl uorescence in situ hybridization (FISH) studies was an independent predictor of survival in multivariate analysis.65-68 In addition, MYC transloca-tion was reported to be especially predictive of survival in the GCB subgroup of patients.65

B-cell lymphomas with concurrent IGH-BCL2 and MYC rearrangements, called “double-hit lymphomas,” are neoplasms with a spectrum of morphologic features overlapping with BL, DLBCL, and B-cell lymphomas, un-classifi able, with features intermediate between DLBCL and BL. These tumors, regardless of the histologic ap-pearance, are characterized by aggressive clinical behav-ior, often complex karyotypes, and poor outcome.1,69-72 Double-hit DLBCL usually shows a high proliferative index (average 80%) but lower than a typical BL, when

Fig 4. — Signifi cant correlation of BCL2 protein level with overall survival (OS) and event-free survival (EFS) in GCB DLBCL. Reproduced with per-mission of American Association for Cancer Research from Iqbal J, Meyer PN, Smith LM, et al. BCL2 predicts survival in germinal center B-cell–like diffuse large B-cell lymphoma treated with CHOP-like therapy and rituximab. Clin Cancer Res. 2011;17(24):7785-7795, permission conveyed through Copyright Clearance Center, Inc.

OS

0.0

0.1

2 3 54 6 7 8 9 10 11

Pro

po

rtio

n

Years

0.2

0.3

0.4

0.5

0.6

0.7

0 1

P = .009

0.8

0.9

1.0

BCL2 protein expression

< 50%≥ 50%

n = 94n = 74

Pro

po

rtio

n

0.0

0.1

2 3 54 6 7 8 9 10 11Years

0.2

0.3

0.4

0.5

0.6

0.7

0 1

P = .001

0.8

0.9

1.0

EFS

A B

July 2012, Vol. 19, No. 3 Cancer Control 221

assessed by IHC stain for Ki-67.70,72 The presence of MYC translocation and expression level of MYC per se may not be the best prognosticators. Other modifi ers of MYC activity, cooperative oncogenic pathways, and MYC mutations have to be determined to provide a more complete picture of MYC as a biomarker.

Ki-67Ki-67 is a nuclear antigen expressed by cycling cells. The percentage of Ki-67 expressing cells refl ects the proportion of the tumor cells that are actively cycling.17 The prognostic signifi cance of Ki-67 expression in DLBCL is controversial. Several studies conducted in rituximab-treated patients showed that elevated Ki-67 expression was associated with inferior OS and EFS.73,74 However, the cutpoints used to defi ne “high” vs “low” Ki-67 have differed among authors, thus making the comparison of individual studies diffi cult. In the study by Lenz et al,24 Ki-67 expression or the proliferative index was not an independent predictor of survival in rituximab-treated patients.

CD43The CD43 molecule is a multifunctional type I trans-membrane glycoprotein expressed in a variety of he-matopoietic cells.75 The role of CD43 in B cells is not completely clear, but coexpression of CD43 and CD20 on peripheral B cells is suggestive of malignancy.76 CD43 is expressed in 16% to 28% of DLBCL.77,78 Mitrovic et al79 found that patients with CD43-positive DLBCL had signifi cantly lower complete response, OS, and EFS compared with CD43-negative DLBCL patients. Inter-estingly, the effect of CD43 was signifi cant in patients treated with R-CHOP, while the signifi cance was not observed in the CHOP-treated cohort.

Special Entities of DLBCLPMLBCL is a distinct subtype of DLBCL of putative thy-mic B-cell origin. Recent studies support the late ger-minal center or postgerminal center stage of differentia-tion. Most patients are in the third decade of life, with a slight female predominance. The majority of patients present at early stage of disease with mediastinal involve-ment, and bone marrow involvement is rare. Morpho-logically, tumor cells are characteristically associated with compartmentalizing alveolar fi brosis. PMLBCL expresses pan-B cell antigens and is positive for CD30 in the majority of cases. Tumor cells are also frequently positive for IRF4/MUM1 and CD23 and are variably posi-tive for BCL2 and BCL6.1,80 Several IHC markers were proposed to aid in differentiating PMLBCL from other types of DLBCL. These markers include, but are not limited to, NFκB family member c-Rel, NFκB target gene TRAF1, MAL antigen, dendritic cell marker TNFAIP2, and a member of the TP53 family TP73L.1,81-84 PMLBCL shows clonally rearranged Ig heavy- and light-chain

genes, but most cases are surface Ig-negative. PMLBCL shows a unique profi le of chromosomal abnormalities in-cluding frequent gains of chromosomes 2p, 9p, 12q, Xq, 7q, and 9q, and losses involving 1p. Gains in 9p include JAK2, PDL1, PDL2, and SMARCA2 genes,1,15,16,85 while 2p gains include the REL proto-oncogene. Interestingly, PMLBCL shows a unique GEP, and it shares many ex-pressed transcripts with classical Hodgkin lymphoma cell lines and has been recently shown to have a high frequency of translocations involving CIITA similar to Hodgkin lymphoma.15,16,52 Patients with PMLBCL have a similar survival to those with GCB DLBCL, with current cure rates up to 80%.1

T-cell/histiocyte-rich large B-cell lymphoma (THRLBCL) is a variant of DLBCL associated with a prominent component of reactive T cells and also fre-quently histiocytes. The median patient age is the sixth to seventh decade, with a slight male predominance. It has been suggested that at least a proportion of cases are pathogenetically related to, or derived from, nodular lymphocyte predominant Hodgkin lymphoma. Com-pared to conventional DLBCL, THRLBCL more common-ly presents with advanced-stage disease and bone mar-row involvement. Morphologically, the large neoplastic cells usually account for less than 10% of the cellular population and are dispersed singly in a background of small lymphocytes. Tumor cells express pan-B cell markers and are usually negative for CD30 and CD15. BCL2, BCL6, and EMA are variably expressed. The small cells in the background are CD3-positive T cells of pre-dominantly CD8-positive cytotoxic type. THRLBCL has clonally rearranged Ig genes. BCL2 rearrangement is present in approximately one-fourth of cases. THRLBCL is often an aggressive lymphoma, with a 3-year OS rate of 46%. Frequently advanced clinical stage at diagnosis contributes to the aggressiveness of this lymphoma. However, when matched for the IPI, THRLBCL and con-ventional DLBCL have similar outcomes.1,80,86

Intravascular large B-cell lymphoma (IVLBCL) is a rare type of large B-cell lymphoma characterized by selective growth of lymphoma cells within the lumina of small blood vessels, particularly capillaries, but not larger arteries and veins. This tumor is derived from peripheral B cells, with the majority of cases showing non–GCB phenotype. It occurs most commonly in the sixth to seventh decade of life. Tumor cells lack the expression of CD29 (β1 integrin) and CD54 (ICAM-1), which are molecules important for transvascular lym-phocyte migration. This might explain the propensity of tumor cells to be localized inside the vessel lumens. IVLBCL is a clinical mimicker of many diseases, and two clinical variants are recognized: Western and Asian. The Western form is most commonly characterized by nonspecifi c, nonlocalizing neurologic symptoms or skin lesions. However, any organ can be involved.1,80 The Asian variant, mostly reported by Japanese authors,

222 Cancer Control July 2012, Vol. 19, No. 3

Other rare subtypes/variants of DLBCL with adverse prognostic implications include plasmablastic lympho-ma, ALK-positive large B-cell lymphoma, primary cuta-neous DLBCL-leg type, DLBCL associated with chronic infl ammation, and primary effusion lymphoma.1

DLBCL in Immune-Privileged Sites: CNS and TestisPrimary DLBCL of the CNS is relatively rare, represent-ing < 1% of all NHL and approximately 2% to 3% of all brain tumors. It occurs in both immunocompetent and immunosuppressed individuals. Most immunocompe-tent patients are older, with a median age of 60 years and a slight preponderance in males. Approximately 60% of all CNS DLBCL cases are located supratentori-ally, and multiple lesions are often present. Patients most commonly present with focal neurological defi -cits. Most primary CNS DLBCL cases are of non–GCB subtype and are usually negative for Epstein-Barr virus when occurring in immunocompetent patients. The most common genetic abnormality is BCL6 translocation (30% to 40%). Commonly, there are deletions at 6q and gains at 12q, 22q, and 18q21, with amplifi cation of BCL2 and MALT1.1,91,92 The prognosis has been improved by novel chemotherapeutic protocols that include metho-trexate and high-dose cytarabine. The International Extranodal Lymphoma Study Group (IELSG) reported a complete remission rate of 46% in patients treated with methotrexate and cytarabine compared with 18% in patients treated with methotrexate alone. The 3-year OS rates were 46% and 32% in the patients treated with and without the addition of cytarabine, respectively.93 Most relapses occur in the CNS but can also involve breast and testis.1

Primary DLBCL of the testis usually presents in adults with median age in the sixth decade. The most common clinical presentation is painless testicular en-largement with rapid progression. Local involvement of the adjacent structures, as well as involvement of the regional lymph nodes, can occur in the course of disease. Most testicular DLBCL, like CNS types, are of non–GCB subtype and have high proliferative activity.94 Genetic alterations in testicular DLBCL often comprise complex abnormalities, including translocations, triso-mies, amplifi cations, and deletions. The more common alterations are abnormalities of 3q27 and 6q deletions. Primary testicular DLBCL is an aggressive disease, with frequent relapses and in general a poorer outcome than that seen in “classic” DLBCL.95-97 Gundrum et al95 re-ported a median OS of 4.6 years, whereas the disease-specifi c survival rates at 3, 5, and 15 years were 71.5%, 62.4%, and 43%, respectively.

Many cases of testicular and CNS DLBCL show de-creased or no expression of HLA class I and II proteins, thus allowing the tumor cells to escape immune at-tack. These tumors were found to have small deletions

is characterized by fever, hepatosplenomegaly, hemo-phagocytic syndrome with cytopenias, marrow involve-ment, and disseminated intravascular coagulation.1,80,87 IVLBCL expresses CD45 and pan-B cell markers. CD5, CD10, or BCL6 is expressed in some cases, with about 20% frequency. Cytogenetic abnormalities involving 8p21, 19q13, 14q32, and chromosome 18 have been re-ported in the Asian variant. This tumor was invariably fatal in the past, but more recent reports suggest that aggressive chemotherapy can lead to complete remis-sion and long-term survival in some patients. The Asian variant has an aggressive clinical course, with a median survival of 7 months.1,80

Epstein-Barr virus (EVB)–positive DLBCL of the elderly is an EBV-associated clonal B-cell proliferation occurring in patients older than 50 years without any known immunodefi ciency or prior lymphoma. It is pos-tulated that this lymphoma results from immunologic deterioration associated with aging. This entity has been reported most commonly in Asians, with a frequency of 8% to 10% of all DLBCL cases among patients without a documented predisposing immunodefi ciency. Data in the Western population are scarce, but the overall incidence is about 3% in this patient population.88 The median age of reported cases at diagnosis is 71 years, with a slight male predominance. About 70% of patients present with extranodal disease, with or without nodal involvement, while 30% of patients have only nodal disease. Morphologically, two subtypes are recognized: polymorphic and large-cell lymphoma. Tumor cells usually express pan-B cell markers, although they occa-sionally may lack CD20 expression. CD30 expression is variable, and CD10 and BCL6 are usually negative, while IRF4/MUM1 is commonly positive. The tumor cells contain EBV, and EPV-encoded RNA (EBER) positivity is demonstrated in the majority of tumor cells. Ig genes are usually clonally rearranged. The clinical course is aggressive, with a median survival of 2 years and a 5-year survival rate of approximately 25%.1,80

De novo CD5-positive DLBCL is a subtype with CD5 expression. Most of the reports concerning this sub-type are from Japan, where approximately 10% of all de novo DLBCL cases express CD5. The median age of patients is the seventh decade, with a slight female predominance. Patients most commonly present in higher clinical stages, and the majority have extranodal involvement. The tumor cells are usually positive for BCL2 and BCL6 and negative for CD10. CD23 and cyclin D1 are negative. The majority of cases are classifi ed im-munophenotypically as non–GCB. BCL6 is rearranged in 40% of cases. Described genetic aberrations include gains of 10p14-15, 19q13, 11q21-24, and 16p and losses of 1q43-44 and 8p23. Compared with conventional DLBCL, de novo CD5-positive DLBCL is associated with a more aggressive clinical course, an overall worse prognosis, and central nervous system (CNS) recurrence.1,89,90

July 2012, Vol. 19, No. 3 Cancer Control 223

of 6p21.3 affecting the HLA region that contributes to the loss of HLA class I and II proteins expression.1,98 Booman et al98 showed that loss of expression of HLA-DR at the mRNA level in testicular DLBCL is associated with a signifi cantly lower expression of many immune-regulated genes such as markers for T cells, NK cells, macrophages, and antigen-presenting cells. The coordi-nate downregulation of these genes with HLA-DR levels indicates a severe disruption of the immune response in testicular DLBCL.

A summary of the different biological subtypes of DLBCL and their impact on prognosis is provided in the Table.

DLBCL in HIV Infection/AIDSThe association between HIV infection and the develop-ment of lymphoma has been observed since the early phases of the AIDS epidemic. In 1986, the Centers for Disease Control and Prevention recognized NHL as an AIDS-defi ning illness. In the era prior to the introduc-tion of highly active antiretroviral therapy (HAART), NHL represented the second most frequent cancer as-sociated with AIDS, after Kaposi sarcoma.99 DLBCL is the most common type of AIDS-related lymphoma. Following the introduction of HAART, the incidence of HIV-related lymphomas has decreased, most promi-nently in primary CNS lymphoma. BL incidence also decreased, with a relative increase in DLBCL.80,100,101 A number of studies have shown improved survival in AIDS-related NHL, including DLBCL, after the introduc-tion of HAART therapy.99-102 Navarro et al103 found that HIV-infected DLBCL patients treated with HAART and chemotherapy had similar response rates to chemo-therapy, OS, and EFS as HIV-negative DLBCL patients receiving CHOP therapy.

Bone Marrow Involvement in DLBCLApproximately 10% to 25% of DLBCL patients exhibit bone marrow involvement by lymphoma at the time of diagnosis. Many have histologically concordant in-volvement with large B cells; however, 40% to 72% of patients have discordant marrow infi ltrates consisting of mainly small B cells. In these cases, it is presumed

that the DLBCL developed from an occult small B-cell lymphoma or that two unrelated lymphomas are pres-ent. Concordant bone marrow involvement has been associated with the poorer outcome, while the data regarding discordant involvement and its infl uence on prognosis have been controversial. Sehn et al104 ana-lyzed a series of 795 rituximab-treated DLBCL patients and found that 67 (8.4%) had concordant and 58 (7.3%) had discordant bone marrow involvement. The pa-tients with concordant bone marrow involvement had lower OS, while EFS was inferior in both concordant and discordant involvement. In a multivariate analy-sis, concordant involvement remained an independent predictor of EFS.

Gray Zone LymphomasThe 2008 WHO classifi cation introduced two new enti-ties in which features of DLBCL overlap with BL or with classical Hodgkin lymphoma (CHL).1,105

B-cell lymphomas, unclassifi able, with features intermediate between DLBCL and BL, are aggressive lymphomas that have overlapping genetic, morphologi-cal, and IHC features of DLBCL and BL. These relatively infrequent tumors usually present with widespread, extranodal disease. Some cases resemble BL morpho-logically but have one or more immunophenotypic or molecular genetic deviations that would exclude it from the BL category. On the contrary, some cases have im-munophenotypic and/or genetic features of BL but are morphologically too atypical for BL. These cases tend to have more complex genetic abnormalities than BL has, and they are far more likely to have non–Ig-MYC translocations. Some cases have concomitant BCL2 translocation (“double-hit” cases). B-cell lymphomas, unclassifi able, with features intermediate between DLBCL and BL, generally have an aggressive clinical course and poor response to standard chemotherapy regimens, with “double-hit” lymphomas having an es-pecially poor prognosis.1,105

B-cell lymphoma, unclassifi able, with features inter-mediate between DLBCL and CHL, demonstrates over-lapping clinical, morphological, and/or immunophe-notypic features between CHL and DLBCL, especially

PMLBCL. These tumors usually occur in young men and present as a medi-astinal mass, with or without involve-ment of supraclavicular lymph nodes. This diagnosis should be restricted to cases showing signifi cant overlapping features with marked diagnostic discor-dance between the morphology and the immunophenotype. Recent evidence from methylation analysis and genetic studies note that this group of cases does have intermediate features be-tween typical CHL and PMLBCL.106,107

Table. — Different Biological Subtypes of Diffuse Large B-Cell Lymphoma and Their Impact on Prognosis

Good Prognosis Intermediate Prognosis Poor Prognosis

DLBCL, GCB subtypePMLBCL

DLBCL, non-GCB subtypeTHRLBCLCD5-positive DLBCL

IVLBCLEBV-positive DLBCL of the elderlyPrimary CNS DLBCLPrimary testicular DLBCL

CNS = central nervous system, DLBCL = diffuse large B-cell lymphoma, GCB = germinal center B-cell–like, EBV = Epstein-Barr virus, IVLBCL = intravascular large B-cell lymphoma, PMLBCL = primary mediastinal large B-cell lymphoma, THRLBCL = T-cell/histiocyte-rich large B-cell lymphoma.

224 Cancer Control July 2012, Vol. 19, No. 3

These cases also may relapse with more typical CHL or PMLBCL compared to the original biopsy. These lymphomas have an aggressive clinical course and a poorer outcome than either PMLBCL or CHL has.1,105 There is currently no consensus on optimal treatment of this entity, although some authors propose that CD20-positive gray zone lymphomas should be treated with immunochemotherapy with rituximab followed by ra-diation treatment.108

ConclusionsNumerous biological prognostic markers have been pro-posed in patients with diffuse large B-cell lymphoma (DLBCL), and the signifi cance of many of these that were studied before the rituximab era need to be reassessed. Prognostic markers are assayed by a variety of methods, most commonly by morphology and immunohistochem-istry. Although widely available and relatively cheap, the immunohistochemistry method suffers from poor reproducibility and diffi culty in quantifi cation due to differences in tissue processing, staining protocols, and interobserver variability. Molecular methods such as gene expression profi ling (GEP), high-resolution array comparative genomic hybridization, and next-generation sequencing hold great promise in elucidating the patho-genesis and prognosis of DLBCL, but these methods are not widely available due to substantial cost, technical complexity, and requirement for fresh and frozen tis-sue. However, more focused assays can be designed for further studies, and the ability to apply these assays to formalin-fi xed, paraffi n-embedded tissue would al-low the inclusion of large patient cohorts to improve statistical power.

Next-generation sequencing has detected numerous mutations in patients with DLBCL, but whether any of those mutations are important for prognosis, alone or in combination, is not yet answered. However, such global studies allow us to examine markers in the context of other modifying factors and hence overcome the prob-lems of single-marker studies. For example, BCL2 may be an important biomarker, but its clinical signifi cance is infl uenced by other factors such as other biological activities of the pathway that leads to its expression, the coexisting factors such as MYC translocation, and possibly even mutations that may affect its biological activities. Thus, studying BCL2 expression alone as a biomarker may not generate reproducible results from different populations. A more global approach may al-low all of these factors to be included in the analysis, thereby producing a more meaningful biomarker profi le.

At the present time, there is no consensus on which biological prognostic markers should be routinely as-sessed in patients with DLBCL, and practices vary widely among different institutions. With more global approaches, such as those noted above, the ability to assess biomarkers in the cellular or tumor context may

be possible, resulting in a better understanding of their biological and prognostic signifi cance.

References 1. Swerdlow SH, Campo E, Harris NL, et al, eds. WHO Classifi ca-tion of Tumours of Haematopoietic and Lymphoid Tissues. 4th ed. Lyon, France: IARC; 2008. 2. de Jong D, Rosenwald A, Chhanabhai M, et al. Immunohisto-chemical prognostic markers in diffuse large B-cell lymphoma: validation of tissue microarray as a prerequisite for broad clinical applications. A study from the Lunenburg Lymphoma Biomarker Consortium. Clin Oncol. 2007;25(7):805-812. 3. A predictive model for aggressive non-Hodgkin’s lymphoma. The International Non-Hodgkin’s Lymphoma Prognostic Factors Project. N Engl J Med. 1993;329(14):987-994. 4. Nicolaides C, Fountzilas G, Zoumbos N, et al. Diffuse large cell lymphomas: identifi cation of prognostic factors and validation of the In-ternational Non-Hodgkin’s Lymphoma Prognostic Index. A Hellenic Coop-erative Oncology Group Study. Oncology. 1998;55(5):405-415. 5. Wilder RB, Rodriguez MA, Medeiros LJ, et al. International prog-nostic index-based outcomes for diffuse large B-cell lymphomas. Cancer. 2002;94(12):3083-3088. 6. Coiffi er B, Lepage E, Briere J, et al. CHOP chemotherapy plus rituximab compared with CHOP alone in elderly patients with diffuse large-B-cell lymphoma. N Engl J Med. 2002;346(4):235-242. 7. Sehn LH, Donaldson J, Chhanabhai M, et al. Introduction of combined CHOP plus rituximab therapy dramatically improved outcome of diffuse large B-cell lymphoma in British Columbia. J Clin Oncol. 2005;23(22):5027-5033. 8. Feugier P, Van Hoof A, Sebban C, et al. Long-term results of the R-CHOP study in the treatment of elderly patients with diffuse large B-cell lymphoma: a study by the Groupe d’Etude des Lymphomes de l’Adulte. J Clin Oncol. 2005;23(18):4117-4126. 9. Habermann TM, Weller EA, Morrison VA, et al. Rituximab-CHOP versus CHOP alone or with maintenance rituximab in older patients with diffuse large B-cell lymphoma. J Clin Oncol. 2006;24(19):3121-3127. 10. Fu K, Weisenburger DD, Choi WW, et al. Addition of rituximab to standard chemotherapy improves the survival of both the germinal center B-cell-like and non-germinal center B-cell-like subtypes of diffuse large B-cell lymphoma. J Clin Oncol. 2008;26(28):4587-4594. 11. Sehn LH, Berry B, Chhanabhai M, et al. The revised International Prognostic Index (R-IPI) is a better predictor of outcome than the standard IPI for patients with diffuse large B-cell lymphoma treated with R-CHOP. Blood. 2007;109:1857-1861. 12. Alizadeh AA, Eisen MB, Davis RE, et al. Distinct types of diffuse large B-cell lymphoma identifi ed by gene expression profi ling. Nature. 2000;403(6769):503-511. 13. Rosenwald A, Wright G, Chan WC, et al. The use of molecular profi ling to predict survival after chemotherapy for diffuse large-B-cell lym-phoma. N Engl J Med. 2002;346(25):1937-1947. 14. Wright G, Tan B, Rosenwald A, et al. A gene expression-based method to diagnose clinically distinct subgroups of diffuse large B cell lym-phoma. Proc Natl Acad Sci U S A. 2003;100(17):9991-9996. 15. Rosenwald A, Wright G, Leroy K, et al. Molecular diagnosis of primary mediastinal B cell lymphoma identifi es a clinically favorable sub-group of diffuse large B cell lymphoma related to Hodgkin lymphoma. J Exp Med. 2003;198(6):851-862. 16. Savage KJ, Monti S, Kutok JL, et al. The molecular signature of mediastinal large B-cell lymphoma differs from that of other diffuse large B-cell lymphomas and shares features with classical Hodgkin lymphoma. Blood. 2003;102(12):3871-3879. 17. Lossos IS, Morgensztern D. Prognostic biomarkers in diffuse large B-cell lymphoma. J Clin Oncol. 2006;24(6):995-1007. 18. Meyer PN, Fu K, Greiner TC, et al. Immunohistochemical methods for predicting cell of origin and survival in patients with diffuse large B-cell lymphoma treated with rituximab. J Clin Oncol. 2011;29(2):200-207. 19. Hans CP, Weisenburger DD, Greiner TC, et al. Confi rmation of the molecular classifi cation of diffuse large B-cell lymphoma by immunohisto-chemistry using a tissue microarray. Blood. 2004;103(1):275-282. 20. Choi WW, Weisenburger DD, Greiner TC, et al. A new immunostain algorithm classifi es diffuse large B-cell lymphoma into molecular subtypes with high accuracy. Clin Cancer Res. 2009;15(17):5494-5502. 21. Nyman H, Adde M, Karjalainen-Lindsberg ML, et al. Prognostic impact of immunohistochemically defi ned germinal center phenotype in diffuse large B-cell lymphoma patients treated with immunochemotherapy. Blood. 2007;109(11):4930-4935. 22. Ilić I, Mitrović Z, Aurer I, et al. Lack of prognostic signifi cance of the germinal-center phenotype in diffuse large B-cell lymphoma patients treated with CHOP-like chemotherapy with and without rituximab. Int J Hematol. 2009;90(1):74-80. 23. Seki R, Ohshima K, Fujisaki T, et al. Prognostic impact of immuno-

July 2012, Vol. 19, No. 3 Cancer Control 225

histochemical biomarkers in diffuse large B-cell lymphoma in the rituximab era. Cancer Sci. 2009;100(10):1842-1847. 24. Lenz G, Wright G, Dave SS, et al. Stromal gene signatures in large-B-cell lymphomas. N Engl J Med. 2008;359(22):2313-2323. 25. Dunleavy K, Pittaluga S, Czuczman MS, et al. Differential effi cacy of bortezomib plus chemotherapy within molecular subtypes of diffuse large B-cell lymphoma. Blood. 2009;113(24):6069-6076. 26. Malumbres R, Chen J, Tibshirani R, et al. Paraffi n-based 6-gene model predicts outcome in diffuse large B-cell lymphoma patients treated with R-CHOP. Blood. 2008;111(12):5509-5514. 27. Alizadeh AA, Gentles AJ, Alencar AJ, et al. Prediction of survival in diffuse large B-cell lymphoma based on the expression of 2 genes re-fl ecting tumor and microenvironment. Blood. 2011;118(5):1350-1358. 28. Payton JE, Grieselhuber NR, Chang LW, et al. High throughput digital quantifi cation of mRNA abundance in primary human acute myeloid leukemia samples. J Clin Invest. 2009;119(6):1714-1726. 29. Williams PM, Li R, Johnson NA, et al. A novel method of amplifi ca-tion of FFPET-derived RNA enables accurate disease classifi cation with microarrays. J Mol Diagn. 2010;12(5):680-686. 30. Roehle A, Hoefi g KP, Repsilber D, et al. MicroRNA signatures char-acterize diffuse large B-cell lymphomas and follicular lymphomas. Br J Haematol. 2008;142(5):732-744. 31. Alencar AJ, Malumbres R, Kozloski GA, et al. MicroRNAs are in-dependent predictors of outcome in diffuse large B-cell lymphoma patients treated with R-CHOP. Clin Cancer Res. 2011;17(12):4125-4135. 32. Lawrie CH, Soneji S, Marafi oti T, et al. MicroRNA expression distin-guishes between germinal center B cell-like and activated B cell-like sub-types of diffuse large B cell lymphoma. Int J Cancer. 2007;121(5):1156-1161. 33. Jung I, Aguiar RC. MicroRNA-155 expression and outcome in dif-fuse large B-cell lymphoma. Br J Haematol. 2009;144(1):138-140. 34. Meyer PN, Fu K, Greiner T, et al. The stromal cell marker SPARC predicts for survival in patients with diffuse large B-cell lymphoma treated with rituximab. Am J Clin Pathol. 2011;135(1):54-61. 35. Cardesa-Salzmann TM, Colomo L, Gutierrez G, et al. High mi-crovessel density determines a poor outcome in patients with diffuse large B-cell lymphoma treated with rituximab plus chemotherapy. Haematolog-ica. 2011;96(7):996-1001. 36. Lenz G, Wright GW, Emre NC, et al. Molecular subtypes of diffuse large B-cell lymphoma arise by distinct genetic pathways. Proc Natl Acad Sci U S A. 2008;105(36):13520-13525. 37. Scandurra M, Mian M, Greiner TC, et al. Genomic lesions asso-ciated with a different clinical outcome in diffuse large B-Cell lymphoma treated with R-CHOP-21. Br J Haematol. 2010;151(3):221-231. 38. Chigrinova E, Mian M, Shen Y, et al. Integrated profi ling of diffuse large B-cell lymphoma with 7q gain. Br J Haematol. 2011;153(4):499-503. 39. Salaverria I, Philipp C, Oschlies I, et al. Translocations activating IRF4 identify a subtype of germinal center-derived B-cell lymphoma affect-ing predominantly children and young adults. Blood. 2011;118(1):139-147. 40. Young KH, Weisenburger DD, Dave BJ, et al. Mutations in the DNA-binding codons of TP53, which are associated with decreased ex-pression of TRAILreceptor-2, predict for poor survival in diffuse large B-cell lymphoma. Blood. 2007;110(13):4396-4405. 41. Young KH, Leroy K, Møller MB, et al. Structural profi les of TP53 gene mutations predict clinical outcome in diffuse large B-cell lymphoma: an international collaborative study. Blood. 2008;112(8):3088-3098. 42. Winter JN, Li S, Aurora V, et al. Expression of p21 protein predicts clinical outcome in DLBCL patients older than 60 years treated with R-CHOP but not CHOP: a prospective ECOG and Southwest Oncology Group cor-relative study on E4494. Clin Cancer Res. 2010;16(8):2435-2442. 43. Zhang A, Ohshima K, Sato K, et al. Prognostic clinicopathologic factors, including immunologic expression in diffuse large B-cell lympho-mas. Pathol Int. 1999;49(12):1043-1052. 44. Ichikawa A, Kinoshita T, Watanabe T, et al. Mutations of the p53 gene as a prognostic factor in aggressive B-cell lymphoma. N Engl J Med. 1997;337(8):529-534. 45. Sohn SK, Jung JT, Kim DH, et al. Prognostic signifi cance of bcl-2, bax, and p53 expression in diffuse large B-cell lymphoma. Am J Hematol. 2003;73(2):101-107. 46. Maartense E, Kramer MH, le Cessie S, et al. Lack of prognos-tic signifi cance of BCL2 and p53 protein overexpression in elderly pa-tients with diffuse large B-cell non-Hodgkin’s lymphoma: results from a population-based non-Hodgkin’s lymphoma registry. Leuk Lymphoma. 2004;45(1):101-107. 47. Visco C, Canal F, Parolini C, et al. The impact of P53 and P21(waf1) expression on the survival of patients with the germinal center phenotype of diffuse large B-cell lymphoma. Haematologica. 2006;91(5):687-690. 48. Miller TP, Lippman SM, Spier CM, et al. HLA-DR (Ia) immune phe-notype predicts outcome for patients with diffuse large cell lymphoma. J Clin Invest. 1988;82(1):370-372. 49. Rimsza LM, Roberts RA, Miller TP, et al. Loss of MHC class II gene and protein expression in diffuse large B-cell lymphoma is related to decreased tumor immunosurveillance and poor patient survival regard-less of other prognostic factors: a follow-up study from the Leukemia and

Lymphoma Molecular Profi ling Project. Blood. 2004;103(11):4251-4258. 50. Cycon KA, Rimsza LM, Murphy SP. Alterations in CIITA consti-tute a common mechanism accounting for downregulation of MHC class II expression in diffuse large B-cell lymphoma (DLBCL). Exp Hematol. 2009;37(2):184-194. 51. Rimsza LM, Chan WC, Gascoyne RD, et al. CIITA or RFX cod-ing region loss of function mutations occur rarely in diffuse large B-cell lymphoma cases and cell lines with low levels of major histocompatibility complex class II expression. Haematologica. 2009;94(4):596-598. 52. Steidl C, Shah SP, Woolcock BW, et al. MHC class II transactiva-tor CIITA is a recurrent gene fusion partner in lymphoid cancers. Nature. 2011;471(7338):377-381. 53. Pasqualucci L, Trifonov V, Fabbri G, et al. Analysis of the coding genome of diffuse large B-cell lymphoma. Nat Genet. 2011;43(9):830-837. 54. Iqbal J, Neppalli VT, Wright G, et al. BCL2 expression is a prog-nostic marker for the activated B-cell-like type of diffuse large B-cell lym-phoma. J Clin Oncol. 2006;24(6):961-968. 55. Iqbal J, Sanger WG, Horsman DE, et al. BCL2 translocation de-fi nes a unique tumor subset within the germinal center B-cell-like diffuse large B-cell lymphoma. Am J Pathol. 2004;165(1):159-166. 56. Mounier N, Briere J, Gisselbrecht C, et al. Rituximab plus CHOP (R-CHOP) overcomes bcl-2-associated resistance to chemotherapy in elderly patients with diffuse large B-cell lymphoma (DLBCL). Blood. 2003;101(11):4279-4284. 57. Wilson KS, Sehn LH, Berry B, et al. CHOP-R therapy overcomes the adverse prognostic infl uence of BCL-2 expression in diffuse large B-cell lymphoma. Leuk Lymphoma. 2007;48(6):1102-1109. 58. Nyman H, Jerkeman M, Karjalainen-Lindsberg ML, et al. Bcl-2 but not FOXP1, is an adverse risk factor in immunochemotherapy-treated non-germinal center diffuse large B-cell lymphomas. Eur J Haematol. 2009;82(5):364-372. 59. Iqbal J, Meyer PN, Smith LM, et al. BCL2 predicts survival in germi-nal center B-cell-like diffuse large B-cell lymphoma treated with CHOP-like therapy and rituximab. Clin Cancer Res. 2011;17(24):7785-7795. 60. Schuetz JM, Johnson NA, Morin RD, et al. BCL2 expression in dif-fuse large B-cell lymphoma. Leukemia. 2011 Dec 22. Epub ahead of print. 61. Schrader A, Bentink S, Spang R, et al. High myc activity is an inde-pendent negative prognostic factor for diffuse large B cell lymphomas. Int J Cancer. 2011 Sep 12. Epub ahead of print. 62. Dalla-Favera R, Bregni M, Erikson J, et al. Human c-myc onc gene is located on the region of chromosome 8 that is translocated in Burkitt lymphoma cells. Proc Natl Acad Sci U S A. 1982;79(24):7824-7827. 63. Kramer MH, Hermans J, Wijburg E, et al. Clinical relevance of BCL2, BCL6, and MYC rearrangements in diffuse large B-cell lymphoma. Blood. 1998;92(9):3152-3162. 64. Rimsza LM, Leblanc ML, Unger JM, et al. Gene expression pre-dicts overall survival in paraffi n-embedded tissues of diffuse large B-cell lymphoma treated with R-CHOP. Blood. 2008;112(8):3425-3433. 65. Yoon SO, Jeon YK, Paik JH, et al. MYC translocation and an in-creased copy number predict poor prognosis in adult diffuse large B-cell lymphoma (DLBCL), especially in germinal centre-like B cell (GCB) type. Histopathology. 2008;53(2):205-217. 66. Savage KJ, Johnson NA, Ben-Neriah S, et al. MYC gene rear-rangements are associated with a poor prognosis in diffuse large B-cell lymphoma patients treated with R-CHOP chemotherapy. Blood. 2009;114(17):3533-3537. 67. Barrans S, Crouch S, Smith A, et al. Rearrangement of MYC is as-sociated with poor prognosis in patients with diffuse large B-cell lymphoma treated in the era of rituximab. J Clin Oncol. 2010;28(20):3360-3365. 68. Zhang HW, Chen ZW, Li SH, et al. Clinical signifi cance and progno-sis of MYC translocation in diffuse large B-cell lymphoma. Hematol Oncol. 2011;29(4):185-189. 69. Niitsu N, Okamoto M, Miura I et al. Clinical features and prognosis of de novo diffuse large B-cell lymphoma with t(14;18) and 8q24/c-MYC translocations. Leukemia. 2009;23(4):777-783. 70. Snuderl M, Kolman OK, Chen YB, et al. B-cell lymphomas with concurrent IGH-BCL2 and MYC rearrangements are aggressive neo-plasms with clinical and pathologic features distinct from Burkitt lymphoma and diffuse large B-cell lymphoma. Am J Surg Pathol. 2010;34(3):327-340. 71. Aukema SM, Siebert R, Schuuring E, et al. Double-hit B-cell lym-phomas. Blood. 2011;117(8):2319-2331. 72. Li S, Lin P, Fayad LE, et al. B-cell lymphomas with MYC/8q24 re-arrangements and IGH@BCL2/t(14;18)(q32;q21): an aggressive disease with heterogeneous histology, germinal center B-cell immunophenotype and poor outcome. Mod Pathol. 2012;25(1):145-156. 73. Yoon DH, Choi DR, Ahn HJ, et al. Ki-67 expression as a prognostic factor in diffuse large B-cell lymphoma patients treated with rituximab plus CHOP. Eur J Haematol. 2010;85:149-157. 74. Broyde A, Boycov O, Strenov Y, et al. Role and prognostic sig-nifi cance of the Ki-67 index in non-Hodgkin’s lymphoma. Am J Hematol. 2009;84(6):338-343. 75. Remold-O’Donnell E, Zimmerman C, Kenney D, et al. Expression

226 Cancer Control July 2012, Vol. 19, No. 3

on blood cells of sialophorin, the surface glycoprotein that is defective in Wiskott-Aldrich syndrome. Blood. 1987;70(1):104-109. 76. Knowles DM. Immunophenotypic markers useful in the diagno-sis and classifi cation of hematopoietic neoplasms. In: Knowles DM, ed. Neoplastic Hematology. 2nd ed. Philadelphia, PA: Lippincott Williams & Wilkins; 2001:93-226. 77. Lai R, Weiss LM, Chang KL, et al. Frequency of CD43 expres-sion in non-Hodgkin lymphoma. A survey of 742 cases and further characterization of rare CD43+ follicular lymphomas. Am J Clin Pathol. 1999;111(4):488-494. 78. Gelb AB, Rouse RV, Dorfman RF, et al. Detection of immunophe-notypic abnormalities in paraffi n-embedded B-lineage non-Hodgkin’s lym-phomas. Am J Clin Pathol. 1994;102(6):825-834. 79. Mitrovic Z, Ilic I, Nola M, et al. CD43 expression is an adverse prog-nostic factor in diffuse large B-Cell lymphoma. Clin Lymphoma Myeloma. 2009;9(2):133-137. 80. Jaffe Es, Harris NL, Vardiman JW, et al, eds. Hematopathology. Philadelphia, Pa: Elsevier Saunders; 2011. 81. Copie-Bergman C, Plonquet A, Alonso MA, et al. MAL expres-sion in lymphoid cells: further evidence for MAL as a distinct molecu-lar marker of primary mediastinal large B-cell lymphomas. Mod Pathol. 2002;15(11):1172-1180. 82. Zamò A, Malpeli G, Scarpa A, et al. Expression of TP73L is a help-ful diagnostic marker of primary mediastinal large B-cell lymphomas. Mod Pathol. 2005;18(11):1448-1453. 83. Rodig SJ, Savage KJ, LaCasce AS, et al. Expression of TRAF1 and nuclear c-Rel distinguishes primary mediastinal large cell lymphoma from other types of diffuse large B-cell lymphoma. Am J Surg Pathol. 2007;31(1):106-112. 84. Kondratiev S, Duraisamy S, Unitt CL, et al. Aberrant expression of the dendritic cell marker TNFAIP2 by the malignant cells of Hodgkin lymphoma and primary mediastinal large B-cell lymphoma distinguishes these tumor types from morphologically and phenotypically similar lym-phomas. Am J Surg Pathol. 2011;35(10):1531-1539. 85. Rui L, Emre NC, Kruhlak MJ, et al. Cooperative epigenetic modula-tion by cancer amplicon genes. Cancer Cell. 2010;18(6):590-605. 86. Achten R, Verhoef G, Vanuytsel L, et al. T-cell/histiocyte-rich large B-cell lymphoma: a distinct clinicopathologic entity. J Clin Oncol. 2002;20(5):1269-1277. 87. Murase T, Yamaguchi M, Suzuki R, et al. Intravascular large B-cell lymphoma (IVLBCL): a clinicopathologic study of 96 cases with spe-cial reference to the immunophenotypic heterogeneity of CD5. Blood. 2007;109(2):478-485. 88. Caponetti GC, Bhagavathi S, Torabi A, et al. EBV-driven diffuse large B-cell lymphoma in the elderly: a diagnostic entity? Mod Pathol. 2010;23(suppl 1). Abstract 1295. 89. Yamaguchi M, Nakamura N, Suzuki R, et al. De novo CD5+ diffuse large B-cell lymphoma: results of a detailed clinicopathological review in 120 patients. Haematologica. 2008;93(8):1195-1202. 90. Westin J, McLaughlin P. De novo CD5+ diffuse large B-cell lympho-ma: a distinct subset with adverse features, poor failure-free survival and outcome with conventional therapy. Leuk Lymphoma. 2010;51(1):161-163. 91. Hattab EM, Martin SE, Al-Khatib SM, et al. Most primary central nervous system diffuse large B-cell lymphomas occurring in immunocom-petent individuals belong to the nongerminal center subtype: a retrospec-tive analysis of 31 cases. Mod Pathol. 2010;23(2):235-243. 92. Gualco G, Weiss LM, Barber GN, et al. Diffuse large B-cell lympho-ma involving the central nervous system. Int J Surg Pathol. 2011;19(1):44-50. 93. Ferreri AJ, Reni M, Foppoli M, et al. High-dose cytarabine plus high-dose methotrexate versus high-dose methotrexate alone in pa-tients with primary CNS lymphoma: a randomised phase 2 trial. Lancet. 2009;374(9700):1512-1520. 94. Booman M, Douwes J, Glas AM, et al. Primary testicular diffuse large B-cell lymphomas have activated B-cell-like subtype characteristics. J Pathol. 2006;210(2):163-171. 95. Gundrum JD, Mathiason MA, Moore DB. Primary testicular diffuse large B-cell lymphoma: a population-based study on the incidence, natu-ral history, and survival comparison with primary nodal counterpart before and after the introduction of rituximab. J Clin Oncol. 2009;27(31):5227-5232. 96. Mazloom A, Fowler N, Medeiros LJ, et al. Outcome of patients with diffuse large B-cell lymphoma of the testis by era of treatment: the MD Anderson Cancer Center experience. Leuk Lymphoma. 2010;51(7):1217-1224. 97. Horne MJ, Adeniran AJ. Primary diffuse large B-cell lymphoma of the testis. Arch Pathol Lab Med. 2011;135(10):1363-1367. 98. Booman M, Douwes J, Glas AM, et al. Mechanisms and effects of loss of human leukocyte antigen class II expression in immune-privileged site-associated B-cell lymphoma. Clin Cancer Res. 2006;12(9):2698-2705. 99. Tirelli U, Spina M, Gaidano G, et al. Epidemiological, biological and clinical features of HIV-related lymphomas in the era of highly active

antiretroviral therapy. AIDS. 2000;14(12):1675-1688. 100. Kirk O, Pedersen C, Cozzi-Lepri A, et al. Non-Hodgkin lymphoma in HIV-infected patients in the era of highly active antiretroviral therapy. Blood. 2001;98(12):3406-3412. 101. Hoffmann C, Wolf E, Fätkenheuer G, et al. Response to highly ac-tive antiretroviral therapy strongly predicts outcome in patients with AIDS-related lymphoma. AIDS. 2003;17(10):1521-1529. 102. Mounier N, Spina M, Gabarre J, et al. AIDS-related non-Hodgkin lymphoma: fi nal analysis of 485 patients treated with risk-adapted inten-sive chemotherapy. Blood. 2006;107(10):3832-3840. 103. Navarro JT, Lloveras N, Ribera JM, et al. The prognosis of HIV-in-fected patients with diffuse large B-cell lymphoma treated with chemother-apy and highly active antiretroviral therapy is similar to that of HIV-negative patients receiving chemotherapy. Haematologica. 2005;90(5):704-706. 104. Sehn LH, Scott DW, Chhanabhai M, et al. Impact of concordant and discordant bone marrow involvement on outcome in diffuse large B-cell lymphoma treated with R-CHOP. J Clin Oncol. 2011;29(11):1452-1457. 105. Hasserjian RP, Ott G, Elenitoba-Johnson KS, et al. Commentary on the WHO classifi cation of tumors of lymphoid tissues (2008): “Gray zone” lymphomas overlapping with Burkitt lymphoma or classical Hodgkin lymphoma. J Hematop. 2009;2(2):77-81. 106. Eberle FC, Rodriguez-Canales J, Wei L, et al. Methylation profi ling of mediastinal gray zone lymphoma reveals a distinctive signature with ele-ments shared by classical Hodgkin’s lymphoma and primary mediastinal large B-cell lymphoma. Haematologica. 2011;96(4):558-566. 107. Eberle FC, Salaverria I, Steidl C, et al. Gray zone lymphoma: chro-mosomal aberrations with immunophenotypic and clinical correlations. Mod Pathol. 2011;24(12):1586-1597. 108. Grant C, Dunleavy K, Eberle FC, et al. Primary mediastinal large B-cell lymphoma, classic Hodgkin lymphoma presenting in the mediastinum, and mediastinal gray zone lymphoma: what is the oncologist to do? Curr Hematol Malig Rep. 2011;6(3):157-163.