Endometrial lymphoid tissue: immunohistological study · Endometrial lymphoid tissue appeared to...

J Clin Pathol 1985;38:644-652

Endometrial lymphoid tissue: an immunohistologicalstudyHUGH MORRIS,* JOHN EDWARDS,t ANDREW TILTMAN,* MALCOLM EMMS*

From the Departments of *Pathology and tObstetrics and Gynaecology, University of Cape Town MedicalSchool and Groote Schuur Hospital, Cape Town, South Africa

SUMMARY Lymphoid tissue of the endometrium was analysed by histological, immunohistologi-cal, and electron microscopical methods in 10 healthy uteri. A panel of monoclonal antibodiesrecognising macrophages (OKMI), HLA-DR antigen, B lymphocytes, T lymphocytes and theirsubsets, and dendritic reticulum cells was used in a two stage indirect immunoperoxidase labellingtechnique. Endometrial lymphoid tissue showed a remarkably consistent pattern of labelling inall cases. Lymphoid tissue was present in three sites: namely, (i) intraepithelial lymphocytes(predominantly T lymphocytes with occasional macrophages) associated with periglandular andsub-epithelial HLA-DR+, OKMI+ macrophages; (ii) interstitial lymphocytes and macrophages;(iii) lymphoid aggregates in the stratum basalis. These were composed mainly of T lymphocyteswith a few B lymphocytes. Dendritic reticulum cells were found in those occasional lymphoidaggregates in which germinal centres were present.These features suggest that endometrial lymphoid tissue has many of the hallmarks of mucosal

associated lymphoid tissue as found elsewhere in the body-for example, the bronchus andintestine. Endometrial lymphoid tissue appears to be unique, however, in that most of thestratum functionalis in which it is situated shows cyclical shedding during the menstrual cycle.

The human endometrium is an immunologicallyreactive tissue which is capable of local antibodysynthesis in response to specific antigen exposure.'Immune responses of this type are normally impor-tant for maintaining the bacteriologically sterilemilieu of the endometrial lumen and in particularare concerned with the recognition and eradicationof ascending bacterial infections. Immune responseshave also been implicated in the pathogenesis ofinfertility in women suffering from endometriosis2and may be important in the process of early nida-tion.

Little is known of the microscopic organisation ofendometrial lymphoid tissue or of the functionalsubtypes of cells (B lymphocytes, T lymphocytes andtheir subsets, and macrophages) which participate inimmune responses. Nevertheless, it is well recog-nised that lymphoid aggregates and scattered inter-stitial lymphocytes are a constant feature of normalendometria.3We have studied the histological, immunohis-

Accepted for publication 17 February 1985

tological, and electron microscopical features ofendometrial lymphoid tissue in normal uteri in orderto shed light on the organisation and possible func-tion of this tissue under physiological conditions.This study has been greatly facilitated by recentdevelopments in immunohistological labellingmethods for detecting surface markers in lymphoidtissue as well as by the availability of several mono-clonal antibodies.4

Material and methods

MATERIALEndometrial tissue was obtained from 10 patientsundergoing hysterectomy for conditions confined tothe cervix alone. In each case a wedge of tissue wasexcised from the upper uterine segment and dividedfor histology, immunohistology, and electron mic-roscopy. Tissue for immunohistology was snap fro-zen in liquid nitrogen and stored at -20°C untilcryostat sections were prepared. Tissue for electronmicroscopy was fixed in 5% glutaraldehyde, whilethe remaining tissue was fixed in 10% bufferedformol-saline.

644

on January 12, 2021 by guest. Protected by copyright.

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.38.6.644 on 1 June 1985. D

ownloaded from

http://jcp.bmj.com/

Endometrial lymphoid tissue: an immunohistological study

Primary monoclonal antibodies used in study

Antibody Specificity Source Dilution

OKT3 T lymphocyte common antigen Ortho Diagnostic Systems, USA 1/100OKT4 T helper lymphocytes (some cross reactivity Ortho Diagnostic Systems, USA 1150

with monocytes present)OKT8 T cytotoxic suppressor Ortho Diagnostic Systems, USA 1/50HLA-DR HLA-DR antigen Dakopatts a/s, Copenhagen 1/100B-cell B lymphocyte antigen Dakopatts a/s, Copenhagen 1150OKMI Monocytes Ortho Diagnostic Systems, USA 1/100DRCI Dendritic reticulum cell network in B cell Dakopatts a/s, Copenhagen 11100

follicles

In each case a detailed menstrual history wasavailable and the age of menarche, cycle length, andduration of menstruation were within normal limits.No hormone treatment had been given to thesepatients. The age range was 18-49 years with amean of 36 years.A total of five proliferative and five secretory

phase samples were studied. Each case was allocatedto one of three proliferative (early, mid, or late) orthree secretory (early, mid, or late) phase subgroupson the basis of standard histological criteria. Inaddition, the uteri of three infants were examined.These included a fresh stillborn infant, gestationalage 39 weeks; an infant of 3 weeks, who died ofdiarrhoea, vomiting, and pneumonia; and a childaged 2 years, who died of pneumonia.

were then washed in tap water, counterstained withhaematoxylin, and mounted for microscopicalexamination.

Electron microscopySpecimens fixed in 5% phosphate buffered glutaral-dehyde were washed overnight in 0-1 M phosphatebuffer (pH 7.2). Tissue was postfixed in osmium tet-roxide (Palades), dehydrated in acetone, andembedded in Spurr's resin. Sections were cut on anLKB Ultratome III and stained with uranyl acetateand lead citrate. Sections were viewed in a HitachiH-600 electron microscope. Further sections (1-5-2-0 um) were stained with basic fuchsin andmethylene blue.5

ResultsMETHODS

HistologyParaffin embedded sections were routinely stainedwith haematoxylin and eosin.

ImmunohistologyCryostat sections (6 ,um thick) were collected on toglass slides and fixed in acetone at room tempera-ture for 10 min.A two stage indirect immunoperoxidase staining

technique was used. The Table lists the primarymonoclonal antibodies used, their sources, and thedilutions used in the primary incubation step. Ineach case, after incubation with the primary mono-clonal antibody (30 min) the sections were washedin phosphate buffered saline (PBS) and incubatedfor 3 min with peroxidase conjugated rabbit anti-mouse immunoglobulin (Dakopatts a/s,Copenhagen, Denmark) to which normal humanserum (diluted 1/20 in PBS) had been added to a

final dilution of 1/50.Development of the peroxidase reaction was per-

formed by incubating sections with diaminoben-zidine (0-6 mg/ml) and hydrogen peroxide (0-01%)for 5-10 min at room temperature. The sections

HISTOLOGYEndometrial lymphoid tissue appeared to consist oflymphocytes present at three recognisable sites-namely, intraepithelial lymphocytes, interstitiallymphocytes, and basal lymphoid aggregates. Inaddition, macrophages were found in all areas of thestratum functionalis and stratum basalis.

Intraepithelial lymphocytes were found in all sec-tions and were present in both the surfaceepithelium and the glandular epithelium (Fig. la).These cells were recognised on the basis of a charac-teristic pericellular halo of cytoplasmic retractionand an oval or slightly convoluted nucleus with adense nuclear chromatin pattern. Lymphocytes inthe epithelium were readily distinguished from col-umnar lining cells, which had a more oval nucleuswith a finely stippled chromatin pattern. Lympho-cytes were also readily separated from pyknoticepithelial nuclei and apoptotic cells.

Interstitial lymphocytes were more difficult to dis-cern with certainty but were best seen in plasticembedded sections. These cells were often seenloosely arranged around spiral arterioles (Fig. 1).Plastic embedded sections clearly showed the fea-tures of capillaries and arterioles in relation to these

645


http://jcp.bmj.com

/J C


ownloaded from

http://jcp.bmj.com/

Morris, Edwards, Tiltman, Emms

41. I,t:

. j

* 4, 8 2 s~4b a f.:

:t :'.4Z 'v t} "S i *t**t,*4 e

.j 'd~~%

f s:. _ i

Fig. 1 (a) Endometrial surface epithelium showingintraepithelial lymphoid cells (arrows). Scattered dark cellsare also present in the stroma but withoutimmunohistological labelling it is not possible to separatelymphocytes from some stromal cells (stromalgranulocytes). Plastic section l pn; dibasic stain. x300. (b)Endometrial glands in the stratum basalis showing a clusterofstromal lymphoid cells with some intraepitheliallymphocytes in the adjacent epithelium (arrows). Plasticsection I ,um, dibasic stain. x400.

cells. The endothelial cells had relatively abundantgranular cytoplasm and in some areas migratinglymphoid cells were found apparently migratingacross the capillary wall.Lymphoid aggregates were present in all sections

and were found almost exclusively in the stratm-

Fig. 2 Endometrial surface epithelium.Immunoperoxidase labelling technique using monoclonalOKT8 (cytotoxic/suppressor cells) showing intraepitheliallymphocytes (arrows). Stromal lymphocytes of the samephenotype are also present. Cryostat section. X800.

basalis with only occasional aggregates in the deeperlayers of the stratum functionalis (Fig. lb). Basallymphoid aggregates also tended to be clusteredadjacent to the basal region of endometrial glandsand in some cases resulted in distortion of the glandmargin. In other areas lymphocytes were present inincreased numbers within the glandular epitheliumitself (epitheliotrophic spill over). Macrophageswere often difficult to distinguish from stromal cellsin haematoxylin and eosin stained sections but weremore easily recognised in plastic embedded sections.The columnar epithelial cells adjacent to

intraepithelial lymphocytes were carefully scrutin-ised in all sections in order to search for epithelialcells which showed similarity to the specialised mic-rovillous cells (or M cells) of the intestine.6 No con-vincing distinguishing features were seen on lightmicroscopical examination of paraffin or cryostatsections, but plastic embedded sections did showsome epithelial cells distorted around intraepitheliallymphocytes (see below, under electron micro-scopy).

IMMUNOHISTOLOGYThe above histological features were more clearlydelineated in sections stained by immunohistologicalmethods.

Intraepithelial lymphocytes were present in both..surface-~and glandular epithelium and consisted

4.

*

*.I

raw

646


http://jcp.bmj.com

/J C


ownloaded from

http://jcp.bmj.com/


Surface

Fig. 3 (a) Cryostat section ofsecretoryendometrium. Immunoperoxidase labellingtechnique using monoclonal HLA-DRantibody showing staining of the surfaceepithelium. The gland epithelium in thestratum functionalis did not stain. x200. (b)Here there is an abrupt change in staining inthe neck area of the gland (white arrows).Basal endometrial glands (not shown) alsoshow variable degrees ofHLA-DRexvpression. HLA-DR positive stromallymphoid cells are also shown (black arrows).x500.

IrA4_7A. ..

g W

Surface

af

w~~~~~~~::"voi_&

Fig. 4 Cryostat section ofbasal lymphoid aggregate.

Immunoperoxidase labelling technique using OKT8 (Tcytotoxic/suppressor lymphocytes) showing strong stainingoflymphoid cells. x450.

Fig. 5 Electron micrograph ofendometrial gland showingsingle intraepithelial lymphocyte (IEL) surrounded by a

retraction halo. x18 000.

647

.1 IN

...:

.1, I '*. 0.*-0 &..

I

I


http://jcp.bmj.com

/J C


ownloaded from

http://jcp.bmj.com/

648

almost entirely of T lymphocytes (OKT3+) withboth T cytotoxic/suppressor (OKT8+) and T helperlymphocytes (OKT4) present (ratio T cytotoxiclsuppressor to T helper about 4:1) (Fig. 2). Occa-sional intraepithelial cells with typical histologicalfeatures of lymphocytes failed to label with any ofthe T cell antibodies. The' nature of these cellsremains to be determined. In addition, HLA-DRpositive, OKMI positive macrophages showed astrong tendency to be dispersed around glands (par-ticularly basal glands) and in some cases appeared toextend into the epithelium between columnar cells.OKMI positive intraepithelial macrophages were arelatively small component of the total number ofintraepithelial cells found. No B lymphocytes werefound in the epithelium in any of the sections.A striking change seen in the surface epithelium

and in epithelium of glands in the stratum basaliswas the presence of HLA-DR antigen expression bycolumnar cells (Fig. 3). The intensity of stainingvaried from case to case, although the patternremained consistent. At the necks of the glands theepithelial cells showed a distinct change in the inten-sity of labelling with a complete absence of stainingin the main segments of the glands (Fig. 3). In thebasal region of each gland variable expression of theantigen was again found.

.t .e ,+ ; .,{ sO.

.¢s

-_|Sr j_ i


Interstitial lymphocytes were predominantly Tlymphocytes with scattered B lymphocytes in a ratioof about 6:1. The T cells were seen in a periarterio-lar distribution as well as diffusely in the stroma ofthe stratum functionalis. T helper and T cytotoxic/suppressor cells were present in an apparently ran-dom scatter in the perivascular and interstitial areas.HLA-DR+, OKMI+ cells were commonly found inall areas of the stratum functionalis and stratumbasalis.

Basal lymphoid aggregates were composed mainlyof T lymphocytes and showed a T cytotoxic/suppressor to T helper ratio of about 5:1 (Fig. 4).Within each aggregate occasional B lymphocyteswere seen. In those basal lymphoid aggregates foundabutting on to basal endometrial glands a spill overof lymphocytes (also of T cell type) was found in theadjacent epithelium (epitheliotropism). In three ofthe 10 cases one to three lymphoid aggregates con-taining germinal centres were present. The mantlezone in these cases showed strong labelling with panB cell antibody, with a lesser degree of staining ofthe germinal centre itself. This pattern is typicallyseen in reactive lymphoidi aggregates in lymph nodesand tonsil.7 Immunolocalisation of dendriticreticulum cells showed that the larger basal lym-phoid aggregates contained dendritic cells. These

hvk5e~~~~~~~Iq_

Fig. 6 (a) Endometrial glandshowing a migrating large lymphoidcell with cytoplasm on both sides ofthe basement membrane (betweenwhite arrows). The presence ofphagocytic material and a bilobed

.5 nucleus suggests that this cell may beS~~~~a macrophage. The black arrow

shows the presence ofa cell junctionxiS 000. Inset: High power view of

-. the cell junction (black arrow).,... , x60 000.


http://jcp.bmj.com

/J C


ownloaded from

http://jcp.bmj.com/


Fig. 7 Endometrial glandshowing an intraepitheliallymphoid cell with evidence ofmigration across the basementmembrane between the areasmarked with dark arrows. x4500.

were present in the form of a prominent networksituated within and between the lymphocytes ofeach aggregate.

ELECTRON MICROSCOPYIntraepithelial lymphocytes were delineated in thesurface and gland epithelium in all cases examined.The nuclear chromatin pattern and the size of thenucleus as well as the presence of a retraction halosurrounding the cell readily separated intraepitheliallymphocytes from adjacent columnar cells (Fig. 5).A striking feature of intraepithelial lymphocytes andmacrophages was the presence of irregular cyto-plasmic processes which extended between colum-nar cells and in many cases indented adjacent col-umnar cells. In addition, at these sites of invagina-tion the plasma membrane of intraepithelial lym-phocytes and columnar cells were directly opposed.These contacts took the form of slender cytoplasmicprocesses, which in some areas showed blurring ofmembranes at the point of apposition (Fig. 6).Intraepithelial macrophages showed more abundantcytoplasm, secondary phagolysomes, mitochondria,and rough endoplasmic reticulum. These cells weremorphologically suggestive of macrophages and pre-sumably correspond to the light microscopicalappearance of OKMI positive cells in theepithelium. In the late secretory phase some mac-rophages appeared to contain apoptotic debris inphagosomes. These cells were also found in the adj-acent stroma.

Interstitial lymphocytes were difficult to delineatewithout the assistance of immunological markers.

Occasional cells, however, with the dense nuclearchromatin pattern characteristic of lymphocytes;were seen in the interstitium and surrounding smallblood vessels. Interstitial lymphocytes apparently inthe process of migration across the basement mem-brane into the surface or gland epithelium were seen(morphological evidence of migration was based onthe direction of basement membrane displacement,the amount of cytoplasm at the advancing front ofthe cell compared with the trailing end, and the posi-tion of nuclear pinching at the point of migrationacross the basement membrane) (Fig. 7).

ENDOMETRIAL LYMPHOID TISSUE INPROLIFERATIVE AND SECRETORY PHASESAll sections were categorised into groups forcomparison-that is, proliferative v secretory phase.In the early proliferative (early postmenstrual)phase basal aggregates as well as intraepitheliallymphocytes were present in all cases. In theremainder of the proliferative and early and mid-secretory phases no substantial alteration in thenumber of lymphoid cells was found. The late sec-retory phase, however, was characterised byincreased numbers of leucocytes in the stroma (rec-ognised on the basis of increased amounts ofendogenous peroxidase) and, although we have notquantitated the total number of lymphocytes in ourmaterial, there appeared to be increased numbers ofstromal lymphocytes in the late secretory phase.

INFANT UTERIHistologically, all three cases showed inactive

649

fs,


http://jcp.bmj.com

/J C


ownloaded from

http://jcp.bmj.com/

650

endometria with occasional halo cells in theepithelium of the surface epithelium and, to a lesserextent, the gland epithelium. Immunohistologyshowed the presence of occasional T lymphocytes inthe epithelium and the stroma, but no lymphoidaggregates were found in the basal regions.

Discussion

In this study we have documented the distribution ofT and B lymphocytes and macrophages in endomet-rial lymphoid tissue. The predominant intraepithel-ial lymphocytes are T lymphocytes with both Tcytotoxic/suppressor and T helper subtypes present.No intraepithelial B cells were seen. Occasionalintraepithelial macrophages (OKTM+, HLA-DR+) were also a feature of normal endometrialepithelium. These findings are similar to thoseobtained in studies of other mucosal sites exposed tothe environment-for example, bronchialepithelium,8 intestinal epithelium,6 skin,9oesophagus,'0 and cervix uteri," all of which showintraepithelial lymphocytes of mainly T cell type.A striking feature in all cases was the presence of

lymphoid aggregates in the stratum basalis, whichconsists predominantly of T lymphocytes with Tcytotoxic/suppressor cells exceeding T helper cells.This was an unexpected finding as we anticipatedthat lymphocyte aggregates which were devoid ofgerminal centres would consist largely of B lympho-cytes (as is usually seen in unstimulated primary fol-licles in lymph nodes). This difference may beexplained on the basis of the limited antigenic stimu-lation which appears to exist in the endometrial cav-

ity since this area is normally a bacteriologicallysterile environment. It is well recognised, however,that it is posslble for particulate matter to be trans-ported from the vaginal and the endocervical luminato as high as the fallopian tube lumen.'2 Hence thereis a potential channel for bacterial antigens to gainaccess to both the endometrial cavity and the tuballumen. In those lymphoid aggregates which showedgerminal centres the mantle zone showed charac-teristic B lymphocyte staining, suggesting thatunstimulated lymphoid aggregates may transform tosecondary follicles with germinal centres on expos-ure to appropriate antigens. An analogous situationis found in the ontogeny of Peyer's patches in ani-mals. In the rat, neonatal Peyers patches consistmainly of T lymphocytes but on exposure to feeds,and therefore various antigens, B lymphocyte pro-liferation occurs and'germinal centres are formed.The above findings (intraepithelial lymphocyte of

a particular lymphocyte type, stromal macrophagesand lymphocytes in the interstitial areas, and basallymphoid aggregates) suggest that endometrial lym-


phoid tissue has many of the hallmarks of mucosalassociated lymphoid tissue (MALT) as seen at theother mucosal sites. In general, the MALT systemforms part of the peripheral lymphoid tissue and ispresent at epithelial sites which are fully or poten-tially exposed to environmental antigens-forexample, bronchus, intestine, salivary ducts, cervixuteri.The characteristic features of the MALT system

have been partially elucidated and include many ofthose presented in this report. It is thought that theintraepithelial lymphocytes are responsible for anti-gen detection in columnar epithelia (facilitated insome areas by specialised columnar M cells),6 whileother accessory cells are important in the uptake ofantigen in squamous epithelia-for example,Langerhans' cells in the oesophagus'° and cervix."Antigen primed lymphocytes are thought to migrateto regional lymphoid tissue (lymph nodes) and thento enter the systemic circulation (via the thoracicduct) en route to the central lymphoid tissue. Trans-formed lymphocytes re-entering the circulationfrom the central lymphoid tissue show a strong ten-dency to home back to the area of original antigenencounter.4

In the endometrium an immune response of thistype is also consistent with the data on localimmunoglobulin production by endometrium.' Themechanisms and possible routes of lymphocytetraffic from the areas of antigen encounter-forexample, intraepithelial lymphocytes or mac-rophages in surface or gland columnar epithelium-to regional lymphoid tissue (regional pelvic lymphnodes) are not, however, apparent from our study.Nevertheless, two possible mechanisms deservefurther consideration. Firstly, it is possible that somecolumnar cells act as accessory cells and enhanceantigen uptake by intraepithelial lymphocytes andmacrophages. This suggestion is based on ourfinding on electron microscopy of close appositionand, in some cases, blurring of the plasma mem-brane between intraepithelial lymphocytes and col-umnar cells. Secondly, it is possible that these anti-gen primed lymphocytes and macrophages migrateto the lymphoid aggregates in the stratum basalis. Itis well recognised that dendritic reticulum cells areantigen presenting cells and are important in theprocess of antigen specific B lymphocyte transfor-mation.7 In this way amplification of the immuneresponse could occur within the basal lymphoidaggregates. This would in turn be followed by themigration of lymphocytes to regional and then cen-tral lymphoid tissue.

Since the stratum basalis remains after sheddingof the stratum functionalis during menstruation it islikely that the basal lymphoid aggregates are a rela-


http://jcp.bmj.com

/J C


ownloaded from

http://jcp.bmj.com/

651Endometrial lymphoid tissue: an immunohistological study

Exogenous antigenEfferentcomponent

Recirculatedlymphocytes

Interstitial lymphoidtissue:

* T-lymphocytes

* B-lymphocytes

* Macrophages

Intraepithelial lymphocytes:Tc/s > m >> Th

Regional and centrallymphoid tissue

Fig. 8 Diagram summarising the organisation ofendometrial lymphoid tissue. Lymphoid cells are found at three mainsites-namely, (i) intraepithelial lymphocytes, in which T cytotoxic/suppressor cells exceed T helper cells and macrophages;(ii) interstitial lymphoid tissue; (iii) basal lymphoid aggregates, in which most cells are ofT cytotoxiclsuppressor phenotypeapart from those aggregates which show germinal centre formation (in which B cells predominate and dendritic reticulumcells are found). Lymphocyte traffic remains speculative but is likely to entail the migration ofintraepithelial lymphocytesand basal lymphocytes to regional lymph nodes (afferent component). Amplification ofthe immune response may occur inthe central lymphoid tissue and recirculated lymphocytes probably home back to the area ofthe endometrium in whichantigen was originally encountered (efferent component). Tcls = T cytotoxiclsuppressor lymphocyte phenotype. M =macrophages. Th = T helper lymphocyte phenotype. MYO = myometrium.

tively constant component of endometrial lymphoidtissue and indeed may act as replenishment centresduring the early proliferative (early postmenstrual)phase. These speculative views are summarised inFig. 8.

It is also of interest to compare the pattern ofHLA-DR staining by columnar epithelial cells withthe distribution of lymphoid tissue. HLA-DR stain-ing was present (albeit of variable intensity) in thesurface epithelial cells and in the epithelium of theglands situated in the stratum basalis. Although orig-

inally thought to be confined to lymphoid cells (mac-rophages, B lymphocytes, and activated T cells),expression of HLA-DR antigen by normal epithelialcells is now documented at several sites includingductal epithelium of the breast'3 and glandepithelium of the endometrium (Natali etal'3 andpresent observations). HLA-DR antigen is alsoexpressed in epithelia in which there is an activelocal immune response (such as graft versus hostdisease in the epidermis'3). The findings in our studytherefore suggest that areas of increased HLA-DR


http://jcp.bmj.com

/J C


ownloaded from

http://jcp.bmj.com/

652

expression in epithelial cells are also areas ofenhanced immune response. The basal glands are inan area of increased lymphoid tissue and perhapsplay an important, but as yet unrecognised, part inimmune responses in the endometrium.

CONCLUSIONWe conclude from this study that endometrial lym-phoid tissue has a uniform immunohistologicalorganisation and that it appears to be a unique formof mucosal associated lymphoid tissue which is cap-able of rapid replenishment after shedding of thestratum functionalis during menstruation. Althoughmany aspects of local immune responses in theendometrium remain to be elucidated, the organisa-tion of endometrial lymphoid tissue suggests thatthis tissue has an important physiological role inmaintaining the mileu interior of the uterine cavity.We hope that the present morphological study mayserve as a basis for future investigations of the pat-tern of migration of lymphocytes under normal anddisease conditions.

We thank the Medical Research Council for finan-cial support.

References

Ogra PL, Ogra SS. Local antibody response to polio vaccine inthe human female genital tract. J Immunol 1973; 110: 1307.

2 Weed JC, Arquembourg PC. Endometriosis: can it produce anautoimmune response resulting in infertility? Clin Obstet

Morris, Edwards, Tiltman, EmmsGynecol 1980;23:885-93.

Hendricksen M, Kempson RL. The normal endometrium. In:Bennington JL, Saunders WB, eds. Surgical pathology of theuterine corpus. London: WB Saunders, 1980:36-98.

4 Mason DY, Naiem M, Abdulaziz, et al. Immunohistologicalapplications of monoclonal antibodies. In: McMichael AJ,Fabe J, eds. Monoclonal antibodies in clinical medicine. Lon-don: Academic Press, 1983:585-635.

Agnese PA, Jensen K. Dibasic staining of large epoxy tissuesections and applications to surgical pathology. Am J ClinPathol 1984;80:25-9.

6 Parrott DMV. The gut as a lymphoid organ. Clin Gastroenterol1976;5:211-28.

7Stein H, Gerdes J, Mason DY. The normal and malignant germi-nal centre. Clin Haematol 1982; 11:531-59.

Bienenstock J, Johnson M, Perey DYE. Bronchial lymphoidtissue. (I) Morphological characteristics. Lab Invest1973;28:686-92.

9 Patterson JAK, Edelson RL. Interaction of T-cells with theepidermis. Br J Dermatol 1982; 107:107-22.

'° Geboes K, De Wolf-Peeters C, Rutgeerts P, Janssen J, van Trap-pen G, Desmet V. Lymphocytes and Langerhans' cells in thehuman oesophageal epithelium. Virchows Arch (Pathol Anat)1983;401:45-55.

Morris HHB, Gatter KC, Stein H, Mason DY. Langerhans' cellsin human cervical epithelium: an immunohistological study. BrJ Obstet Gynaecol 1983; 90:400-11.

i2 Egli GE, Newton M. The transport of carbon particles in thehuman female reproductive tract. Fertil Steril 1961; 12:151-5.

"3Natal PG, Martino C, Quaranta V, et al. Expression of Ia-likeantigens in normal human non-lymphoid tissues. Transplanta-tion 1981;31:75-78.

4 Lampert IA, Suitters AJ, Chisolm PM. Expression of Ia-likeantigen on epidermal keratinocytes in graft-versus-host dis-ease. Nature 1981;293:149-50.

Requests for reprints to: Dr HB Morris, Department ofPathology, University of Cape Town Medical School,Observatory, Cape Town, Republic of South Africa 7925.


http://jcp.bmj.com

/J C


ownloaded from

http://jcp.bmj.com/

Endometrial lymphoid tissue: immunohistological study · Endometrial lymphoid tissue appeared to...

Documents

Transcript of Endometrial lymphoid tissue: immunohistological study · Endometrial lymphoid tissue appeared to...