Lymphoproliferative orbit: immunologicalapproach to ... · polyclonal lymphoproliferative masses, need radiotherapy. The finding of reduced numbers of circulating Tcellsinthe presence

British Journal of Ophthalmology, 1983, 67, 561-569

Lymphoproliferative disorders of the orbit:an immunological approach to diagnosisand pathogenesisA. GARNER, A. H. S. RAHI, AND J. E. WRIGHT

From the Department ofPathology, Institute ofOphthalmology, University ofLondon, andOrbital Clinic, Moorfields Eye Hospital, London

SUMMARY A prospective immunological study of patients with proptosis due to intraorbital lesionscomposed largely or exclusively of lymphoid tissue indicates that the use of antisera to specificheavy and light chain antibody components can be helpful in distinguishing between polyclonal andtruly neoplastic monoclonal disorders. On the basis of combined histological and immunologicalinformation it is possible to delineate 4 categories of patient: those with unequivocal chronicinflammation, those with a virtually pure lymphoproliferative lesion and a polyclonal profile, andthose with a monoclonal lymphomatous disorder which may be histologically comparable to theprevious category or, in a fourth group, be unmistakably malignant on cytological grounds. Thefirst of these groups will normally respond to corticosteroid treatment, but the others, including thepolyclonal lymphoproliferative masses, need radiotherapy. The finding of reduced numbers ofcirculating T cells in the presence of low plasma levels of IgA, and of autoantibodies in the serum ofa third of the patients, could mean that individuals developing lymphoproliferative lesions in theorbit, whether hyperplastic or neoplastic, are partially immunodeficient.

Is it a lymphoma or a pseudotumour? This is thequestion that has to be faced every time a patient'sproptosis is shown to be due to a localised prolifera-tion of lymphocytes. In that a continuous spectrum ofhistological appearances may be seen from un-mistakable chronic inflammation to undoubtedmalignancy, with many cases falling somewherebetween, precise diagnosis can be extremely difficult.Hence the attraction of an additional approach to thedifferential diagnosis of lymphoproliferative disorders.The traditional histopathological approach is

geared to morphological assessment, but, fortunately,lesions involving lymphocytes also lend themselves tofunctional evaluation. A hypothesis that is funda-mental to distinguishing between hyperplastic(reactive) and neoplastic lymphoid disorders is that areactive process will cause a number of lymphocytesto respond, whereas a neoplastic lesion stems from asingle abnormal cell. Thus both T and B lymphocytescan be expected to share in an inflammatory reaction,the latter cellsproducing avariety ofimmunoglobulins

Correspondence to Professor A. Garner, Department of Pathology,Institute of Ophthalnology, 17-25 Cayton Street, London EC1V 9AT.

which, while having the same antigenic specificity,will probably represent more than one class ofimmunoglobulin and almost certainly a mixture ofkappa and lambda light chains. Conversely, neo-plasia is considered to result from the mutation of asingle cell so that, should the relevant cell be a Blymphocyte, all the immunoglobulin produced will beof one type with identical light chains.To determine whether the lymphoproliferative

lesions with an overwhelmingly monoclonal profileare truly neoplastic, in that if not alreadydisseminated they do have a potential for metastasis,needs years of careful follow-up, and even then thediagnosis may remain unproved in patients who havereceived the treatment appropriate to such apresumption. Nevertheless, there are now severalstudies relating to the orbit and other sites whichindicate that it may be practicable to differentiatebetween neoplastic (monoclonal) and hyperplastic(polyclonal) lymphocytic lesions by immunologicalmethods.'s In the present paper we describe ourinterim findings in a long-term study of all patientspresenting with suspected lymphoid pseudotumoursto the Orbital Clinic of Moorfields Eye Hospital.

561

on June 8, 2020 by guest. Protected by copyright.

http://bjo.bmj.com

/B

r J Ophthalm

ol: first published as 10.1136/bjo.67.9.561 on 1 Septem

ber 1983. Dow

nloaded from

http://bjo.bmj.com/

A. Garner, A. H. S. Rahi, and J. E. Wright

Table 1 Clinical and radiologicalfindings

Case Age! History Physical signs Investigations ProvisionalSex diagnosis

1 44 1 Year painless progressive 6 mm Proptosisleft proptosis with diplopia

6 Weeks painless swellingright lower lid9 Months progressivepainless left ptosis2 Months painless swellingright upper eyelid4 Months painlessprogressive right proptosis4 Months painful proptosiswith diplopia3 Months painless swellingleft upper lid2 Months left orbitalswelling painful Vth nerve

2 Weeks painful swellingright upper lid7 Weeks painless swellingright upper lid18 Months painless leftproptosis

3 Years painless swellingright lower lid6 Months painless swellingright upper lid2-5 Years painlessprogressive left proptosis4 Months painlessprogressive swelling rightupper lid4 Months painless bilateralswollen lids progressiveproptosis4 Months painful swollenleft upper lid

6 Months painless swollenleft upper lid3 Months painless massleft upper lid

8 mm Proptosis, nontendermass floor of orbitNontender mass upper innerquadrantNontender mass upperouter quadrant6 mm Proptosis

4 mm Proptosis

3 mm Proptosis, tendermass superior orbit4 mm Proptosis,'orbital apex syndrome,'central scotoma2 mm Proptosis, mass

upper outer quadrantMass upper outer quadrant

9 mm Proptosis,subconjunctivalinfiltrationHard mass lower fornix

Mass upper outer quadrant

Large mass upper thirdorbit5 mm proptosis, mass

upper half orbit

Bilateral 9 mm proptosis,diffuse masses each orbit

4 mm Proptosis mass

upper orbit

2 mm Proptosis mass upperinner quadrant2 mm Proptosis tender massupper inner quadrant

T3, T4, TRH normal.CT: enlarged medial andlateral rectiCT: large mass floor of orbit

Radiographs normal

CT: large mass extendingposteriorlyCT: bilateral orbitalmasses, small on leftCT: intraconal mass incontact with globe mediallyCT: lacrimal gland mass

extending mediallyCT: enlarged lateral rectus,thickened optic nerve

CT: mass lacrimal fossa

Radiographs normal

CT: large diffuse mass

CT: small mass anteriororbital floorCT: small lacrimal glandmass

Radiographs normal

Radiographs normal

CT: bilateral diffuseorbital masses

CT: mass involving medialrectus muscle, extendingforwardCT: diffuse medial mass incontact with globeCT: localised mass near

levator

? Inflammatory? Dysthyroid

Lymphocytic mass

Lymphocytic mass

Lymphocytic mass

Lymphocytic mass

Inflammatory

Inflammatory

Inflammatory

Inflammatory

Lymphocytic mass

Lymphocytic mass

Inflammatory

Lymphocytic mass

Lymphocytic mass

Lymphocytic mass

Lymphocytic mass

Inflammatory

Lymphocytic mass

Inflammatory

Patients and methods

The 19 patients selected for this study had a

demonstrable orbital mass which was thought to beeither a lymphoproliferative lesion or an inflammatorypseudotumour after preliminary investigations hadexcluded other causes (Table 1). All patients hadradiographs of the skull, including selective views ofthe orbits. Computerised tomography (CT) scans inthe axial and coronal planes were performed in mostcases. Patients were thoroughly examined by a

physician to exclude a systemic cause for the orbitalmass. This assessment included examination of theperipheral blood.

A provisional diagnosis was made based on thehistory, physical signs and the results of investiga-tions. Patients with a painful, tender lesion were

diagnosed as having an inflammatory pseudotumour,whereas painless nontender masses were thought tobe lymphoproliferative lesions, the latter term beingused to describe lymphocytic masses which might or

might not be neoplastic.A biopsy was done in all cases. Anterior lesions

were approached through the orbital septumor the conjunctiva, more deeply placed lesionsthrough a lateral canthotomy or by disinserting themedial rectus muscle and entering the intraconalspace.

M

2 71M

3 72F

4 51M

5 70F

6 34M

7 14F

8 37F

9 80M

10 73F

11 65F

12 57F

13 75F

14 82F

15 66F

16 52M

17 54F

18 57M

19 63M

562


http://bjo.bmj.com

/B

r J Ophthalm


ber 1983. Dow

nloaded from

http://bjo.bmj.com/

Lymphoproliferative disorders ofthe orbit: an immunological approach to diagnosis and pathogenesis

Laboratory investigations

One part of each tissue sample was fixed in 10%formol-saline for routine light microscopy of 5 ,umsections stained by haematoxylin and eosin, methylgreen-pyronin Y, and the periodic acid-Schiffsequence. A second piece was fixed in bufferedglutaraldehyde for transmission electron microscopy(to be reported separately) and light microscopyusing 1 ,um plastic-embedded sections. The thirdpiece of tissue was used to produce cell monolayers,by pressing the unfixed tissue against a dry slide,and/or frozen sections; subsequently immuno-fluorescence by a conventional sandwich techniquefor the presence of IgG, IgM, and IgA antibodies andlight chains was performed. Parallel immunohisto-chemical studies were performed on paraffin-embedded sections using a horseradish peroxidasemarker. At the time of biopsy blood was collected forimmunological study and biochemical assessment ofthyroid function (T3, T4, and thyrotropin releasinghormone estimation). The immunological measure-ments included serum IgG, IgM, and IgA levels by

radial immunodiffusion, autoantibodies to nuclearand mitochondrial antigens, smooth muscle, gastricmucosa, and reticulin; T cell proportions as judged bya rosetting technique; and the lymphocyte trans-formation index in response to nonspecific mitogens(for example, phytohaemagglutinin) and recallantigens, for example, purified protein derivative(tuberculin) stimulation.

Treatment

Patients with lymphoproliferative lesions werereferred to a radiotherapist for evaluation andradiotherapy. A search for systemic involvement wasconfined to those under 65 years of age or if there wasphysical evidence of more generalised disease. Theamount of radiotherapy administered to the involvedarea was varied according to the response. Themajority were given 25 gray but a total dose as high as35 gray was needed in some cases (1 gray= 100 rads).

Systemic corticosteroids were given to most of thepatients with inflammatory pseudotumours. Someshowed evidence of spontaneous resolution or were

Table 2 Immunological and histopathologicalfindings

Case Circulating Tcells Serum immunoglobulins Auto- Tissue Histopathology(mgldl) andbodies Ig

% of total Trans.lympho- index IgG IgA IgMcytes

1 68 1240 232 110 - Poly Chronic inflammation2 67 1670 322 63 - Poly Indeterminate lymphoproliferative disease3 54 76PHA 822 118 706 SMA Poly Chronic inflammation4 44 5PHA 1130 110 178 - Mono Indeterminate lymphoproliferative disease5 82 9PHA 1430 250 63 - Poly Indeterminate lymphoproliferative disease6 53 3PPD 1230 413 122 SMA Poly Chronic inflammation with fibrosis7 73 21PHA 1700 319 112 - Poly Chronic inflammation

1OPPD8 57 7PHA 970 82 161 ANA Poly Chronic inflammation with fibrosis9 63 IIPHA 1560 411 84 - Poly Chronic inflammation10 60 803 279 65 ANA Mono Indeterminate lymphoproliferative disease11 65 38PHA 908 225 159 - Mono Indeterminate lymphoproliferative disease12 60 6PHA 781 154 64 - Poly Chronic inflammation with fibrosis13 71 22PHA 828 181 74 ANA Mono Indeterminate lymphoproliferative disease14 62 49PHA 985 499 109 SMA Mono Indeterminate lymphoproliferative disease15 62 35PHA 1450 185 233 AR Poly Indeterminate lymphoproliferative disease16 53 471 86 66 - Poly Indeterminate lymphoproliferative disease17 60 1390 482 264 - Poly Chronic inflammation (granulomatous)18 45 1030 184 376 - Poly Indeterminate lymphoproliferative disease19 66 8PHA 845 98 141 - Poly Chronic inflammation (granulomatous)

Chronic inflammation denotes infiltration with lymphocytes and other leucocyte types with or without fibrosis.Lymphoid hyperplasia indicates an overwhelming predominance of lymphocytes.Mono=antibody confined to a single type of light chain.Poly=antibody of more than one sort.PHA=phytohaemagglutinin.PPD=purified protein derivative (tuberculin).SMA=smooth muscle antibody.ANA=antinuclear antibody.AR=antireticulin antibody.SI conversion: immunoglobulins mg/dlxO 01 =g/1.

563


http://bjo.bmj.com

/B

r J Ophthalm


ber 1983. Dow

nloaded from

http://bjo.bmj.com/


minimally inconvenienced and received no treatment.Others were treated with radiotherapy alone. A dailydose of60-80 mg of prednisolone was given orally for5-7 days; the dosage was then reduced progressivelyduring the ensuing weeks. Ifsteroids failed to producecomplete resolution, or a maintenance dose wasneeded to control the patient's signs and symptoms,or the use ofsystemic steroidswas judged inadvisable,the lesion was treated with radiotherapy in a mannersimilar to that employed for lymphoproliferativelesions.

Results

This interim report deals with some of theimmunological aspects of the study and a morecomplete account covering clinical, therapeutic,immunological, and histopathological parameterswill be published later.The findings in individual cases are given in Tables

2 and 3.

Histology. Two broad categories of cellular infiltra-tion were identified. Nine patients presented lesionsin which the lymphocytic population was diffuse anduneven and mainly in the form of small maturelymphocytes. Germinal centres were seen in 2 cases(Fig. 1). Appreciable numbers of other types ofleucocyte (chiefly macrophages but also occasionaleosinophils and neutrophils) as well as plasma cellswere also seen, and the lesions were classified aschronic inflammation. In 3 patients large amounts offibrous tissue were observed in conjunction with arelatively milder degree of chronic inflammation, andin 2 granuloma formation was seen. The lesion in theremaining 10 patients consisted of marked lympho-cytic proliferation (Figs. 2 and 3) with small numbersof transformed cells with recognisable but smallnucleoli or cleaved nuclei (centrocytes). In a singlecase (case 19) germinal centre formation wasidentified, the distribution of immature cells beingrandom in the other 9 patients. Plasma cells werereadily identified in 2 cases but were present to some

Table 3 Course ofdisease and response to treatment

Case no. Final diagnosis Treatnent Response Follow-up

1 Chronic inflammation None Gradual resolution of 18 Months still proptosis 5 mmdiplopia

2 Polyclonal lymphoproliferative Prednisolone 40 mg daily Partial response 18 Months no recurrencelesion 25 gray Good resolution

3 Polyclonal lymphoproliferative 25 gray Resolution 18 Months no recurrencelesion

4 Monoclonal lymphoproliferative 30 gray Resolution 15 Months no recurrencelesion (lymphoma)

5 Polyclonal lymphoproliferative 30 gray (right) Resolution 1 Year mass left orbit requiredlesion 26 gray (left) radiotherapy; 2 Years no

recurrence6 Chronic inflammation Prednisolone 80 mg Resolution above 15 mg 1 Year no recurrence

reducing 25 gray7 Chronic inflammation Prednisolone 60 mg Partial response Subsequentrightlacrimalglandmass

25 gray Partial response8 Chronic inflammation Prednisolone 60 mg Resolution 18 Months no recurrence

4 days, then reducing to10 mg for 6 months

9 Chronic inflammation None Resolved spontaneously 18 Months no recurrence10 Monoclonal lymphoproliferative 30 gray Resolution 18 Months no recurrence

lesion (lymphoma)11 Monoclonal lymphoproliferative 30 gray Resolution 18 Months no recurrence

lesion (lymphoma)12 Chronic inflammation None Mass remains 15 Months no change13 Monoclonal lymphoproliferative 20 gray Resolution 12 Months no recurrence

lesion (lymphoma)14 Monoclonal lymphoproliferative 25 gray Resolution 14 Months no recurrence

lesion (lymphoma)15 Polyclonal lymphoproliferative 33 gray Resolution 13 Months no recurrence

lesion16 Polyclonal lymphoproliferative 20 gray Resolution 10 Months no recurrence

lesion17 Chronic inflammation 25 gray Resolution 10 Months no recurrence18 Polyclonal lymphoproliferative 25 gray Resolution 8 months no recurrence

lesion.19 Chronic inflammation None Resolved spontaneously 2 Years no recurrence

Modal dose of x-rays in SI units: 1 gray= 100 rads.

564


http://bjo.bmj.com

/B

r J Ophthalm


ber 1983. Dow

nloaded from

http://bjo.bmj.com/


Fig. 1 Chronic orbitalinflammation. A granulomatousfocus in top left-hand corner isassociated with markedlymphocytic infiltration in whichgerminal centres can be recognised.(Haematoxylin and eosin, x80).

extent in all 10. This latter group were designated asindeterminate lymphoproliferative disease. Therewere no patients who presented unequivocal histo-logical evidence of (malignant) neoplasia.Immunology. Immunofluorescence of the mono-

layer cell impressions and immunoperoxidase

Fig. 2 Polyclonal lymphoproliferative lesion. The cells aremostly small lymphocytes with no cytological irregularityand there are very few nonlymphoid cells. (Haematoxylinand eosin, x 430).

staining of histological sections showed anunequivocal monoclonal antibody staining pattern in2 cases (Figs. 4 and 5) and a predominantly mono-clonal pattern in another (mainly X light chainimmunoglobulins). In each of the remaining 14 casesthere was an admixture of K and X light chain-containing plasma cells and B lymphocytes (Fig. 6),which approximated to the K:X ratio of 2:1 pertainingin the circulating blood of the general population.The IgG levels were not significantly abnormal

save in one patient (case 16) in whom there was ageneral reduction in serum immunoglobulin levels.Five of the 19 patients had serum IgA levels almost 2standard deviations below the usual mean of 200mg/dl (2 g/l) and 4 had serum IgM values that wereconsiderably elevated above the normal mean of 130mg/dl (1P3 g/l); in 2 the increase was far beyond 2standard deviations. Autoantibodies were detectedin the serum of 7 patients: in 3 the antibodies weredirected against smooth muscle, in 3 they wereantinuclear, and in 1 antireticulin.The proportion ofT cells in the peripheral blood is

normally over 65%, but in 6 of the 19 cases it wasbelow 60%. In no case where the capacity forlymphoblastic transformation was tested was thereany evidence of functional deficiency in the T cells.

Discussion

In the absence of long-term follow-up it is impossibleto substantiate the accuracy of diagnosis whetherclinical, histological, or immunological criteria areused, particularly where the course of the disease hasbeen modified by treatment. Irradiation is an effectivetreatment for both lymphoma and presumed reactivehyperplasia.3 It is probably significant, nevertheless,

565


http://bjo.bmj.com

/B

r J Ophthalm


ber 1983. Dow

nloaded from

http://bjo.bmj.com/


Fig. 3 Monoclonal lymphoproliferative lesions(lymphoma). The uniform mass ofpredominantly smalllymphocytes is histologically indistinguishablefrom thepolyclonal lesion depicted in Fig. 2. In this case, however,there was a monoclonal immunohistochemical profile whichtogether with the bland morphology suggests a lymphoma oflow-grade malignancy. (Haematoxylin and eosin, x430).

that of 9 cases diagnosed on histological grounds aslikely to have inflammatory disorders none presenteda monoclonal immunoglobulin profile in the tissue,whereas of 10 lesions diagnosed as lymphoidhyperplasia 5 were characterised by monoclonal or

* .AFig. 4 Monoclonallymphoproliferative lesion

(lymphorna). Directimmunofluorescence on a cellimpression shows a strong reactionfor X light chains on the surfacemembranes of the proliferatinglymphocytes. There was a negativereaction for K light chains. FurtherimmunofluorescenceshowedapureIgM synthesis. (The use ofmonolayer cell impressions gives amuch more distinctstainingresponse than can be obtainedfromfrozen sections.) (x800).

near monoclonal antibody production. The close cor-relation of polyclonality with clinical and histologicalevidence of chronic inflammation, while expected, isreassuring, but it is in the context of the lymphoidhyperplasia group that the chief interest lies. In adetailed morphological study of 22 adnexal lymphoidlesions, most of them involving the orbit, Jakobiec etal.3 found that 5 gave a polyclonal antibody profileand showed no evidence of the systemic involvementthat might typify a lymphomatous process in a follow-up of up to 3 years. The follow-up period in our seriesis even shorter and does not warrant comparison, butit will be interesting to determine whether the 5patients coming into this polyclonal lymphopro-liferative category continue to be free of disease. For,while monoclonality is the characteristic ofmalignancy, there is growing evidence that non-metastasising lymphocytic proliferations inalternative sites to the orbit can be polyclonal.67Furthermore such hyperplastic polyclonal lesions mayproceed on occasion to yield a monoclonal lymphomaas a result of subsequent mutation.6 The finding thatthe 3 patients with substantial fibrosis showed a poly-clonal antibody picture is in keeping with the viewthat scar tissue formation is a postinflammatory event.

Jakobiec and his colleagues subdivided theirmonoclonal lesions on the basis ofcytology and foundthat those characterised by well-differentiated smalllymphocytes were free of systemic involvement afterbrief follow-up, whereas of those showing cytologicalirregularity slightly more than halfpresented evidenceof disseminated disease. The morphology of thelymphoid infiltrates in the 5 patients in our own serieswith a monoclonal antibody reaction corresponds tothe well-differentiated category and suggests afavourable outcome. Even so it is entirely reasonableto suppose that such 'benign' lymphoproliferativetumours may suffer further mutational events and

566


http://bjo.bmj.com

/B

r J Ophthalm


ber 1983. Dow

nloaded from

http://bjo.bmj.com/


Fig. 5A

Fig. 5 Monoclonallymphoproliferative lesion(lymphoma). A: Conventionalhistology shows a proliferationof lymphocytes withscattered immature fqrns.B: Immunoperoxidase stainingforK light chains is negative, but (C) theimmunoperoxidase reaction for Xlight chains has stained the surfaceof the majority of cells. (x 430).

Fig. SB

Fig. 5C

develop into metastasising lymphomas.8 An increasein immature forms observed in repeat orbital biopsiesin patients with lymphocytic proliferations has beenreported.9 From the standpoint of management itseems appropriate to regard all monoclonal lesions as

potential if not actual malignant lymphomas and totreat them accordingly. Indeed, unless the clinicaland histological findings indicate an unequivocal in-

flammatory process, there is every reason to subjectnot only the monoclonal lesions but also polyclonallymphoproliferative disorders to radiotherapy fromthe outset, the results in the 6 cases so treated in thepresent study being uniformly satisfactory (Table 3).IgM levels were elevated to above 200 mg/dl (2 g/l)

in 4 patients, 2 ofwhom also had reduced proportionsof circulating T cells. The mean IgM concentration in

567


http://bjo.bmj.com

/B

r J Ophthalm


ber 1983. Dow

nloaded from

http://bjo.bmj.com/

A. Garner, A. H. S. Rahi, andJ. E. Wright

Fig. 6A

Fig. 6 Polyclonallymphoproliferative lesion. A: Thehistological appearance is ofauniform proliferation ofmainlysmall lymphocytes. The cytoplasmofrandomly distributedplasmacytoid cells stainedfor (B) Klight chains and (C) X light chains insimilar proportions. (x430).

Fig. 6B

Fig. 6C

patients with low T cell counts was 268 mg/dl (2 7 g/l),whereas in patients with T cell counts of60% or morethe mean IgM level was 119 mg/dl (1 2 g/l). This couldmean that the reduction in T cell numbers relatedprincipally to the suppressor subset, a deficiency ofsuch cells having a permissive effect on IgMproduction. Now that specific antisera are availableto enable the T cell subsets to be identified, steps to

investigate this possibility are being taken and will bereported later.Apart from one patient with low values for all the

immunoglobulin fractions, IgG levels were essentiallywithin normal limits.

However, in contrast to the IgM levels the IgAvalues were reduced in patients with low proportionsof circulating T cells to a mean which was less than

568


http://bjo.bmj.com

/B

r J Ophthalm


ber 1983. Dow

nloaded from

http://bjo.bmj.com/


one-third of that for patients with normal T cellnumbers. It is possible that the association isattributable to a deficiency of helper T cell function,and it may be relevant to note that Knowles andJakobiec'° demonstrated just such a reduction in 2 of6 patients with ocular lymphomas, although thiswould be to postulate a relative reduction ofsuppressor activity in respect of one class of antibody(IgM) and of the opposite in respect of IgA. Analternative is that both the T cell and the IgAdeficiencies are separate manifestations of a morewidely based inadequacy of immune function.

Autoantibodies were demonstrated in 7 patients,but the only tentative association was with IgA levels,2 of the 5 such patients having circulating autoanti-bodies. A correlation between selective IgAdeficiency and autoimmune disease has been reportedpreviously,'1 and it appears that low levels of IgAcarry an increased risk of developing inflammatorydisease, autoimmunity, and neoplasia. Selective IgAdeficiency is a fairly common abnormality in thepopulation at large, and in a study of more than 4000patients with a variety of ocular disorders (mainlyinflammatory) Addison and Rahi'2 found that serumIgA levels below 60 mg/dl (0-6 g/l) occurred in 15%.In the same study 17% of patients with malignantmelanoma of the uvea had low IgA values. It isperhaps worth examining further the possibility thatindividuals with a reduced output of IgA and/or lownumbers of circulating T cells have an enhanced riskof developing lymphoproliferative disease which maybe inflammatory or neoplastic in nature.

Conclusions at this stage in our study must betentative, but on the basis of the foregoing evidence,both our own and that of other workers, we suggestthat 4 broad categories of orbital lymphoid prolifera-tion can be identified:

(1) Unequivocal chronic inflammation. This mayor may not be granulomatous but it will show a varietyof types of leucocyte and present a polyclonalimmunological profile.

(2) Polyclonal lymphoproliferative disease. Analmost pure mass of mainly mature lymphocytes isassociated with a mixture of antibody types.

(3) Lymphoma (monoclonal lymphoproliferativedisease). Histologically similar to the previouscategory but identified as neoplastic by virtue of amonoclonal antibody profile.

(4) Unequivocal malignant lymphoma. Cytologicalfeatures of malignancy are allied to monoclonalantibody formation.

We are indebted to Dr G. Lloyd for his radiological analyses, toDr J. M. Henk for undertaking the radiotherapy, to Mr J. Prasad forperforming the immunological tests, to Mr R. A. Alexander for thehistological material, and to Mrs P. Goodwin for typing themanuscript.

References

1 Knowles DM, Jakobiec FA, Halper JP. Immunologic char-acterization of ocular adnexal lymphoid neoplasms. Am JOphthalmol 1979; 87: 603-19.

2 Saraga P, Hurlimann J, Ozzello L. Lymphomas and pseudo-lymphomas of the alimentary tract: an immunohistochemicalstudy with clinico-pathologic correlations. Hum Pathol 1981; 12:717-23.

3 Jakobiec FA, Iwamoto T, Knowles DM. Ocular adnexallymphoid tumors: correlative ultrastructural and immunologicmarker studies. Arch Ophthalmol 1982; 100: 84-98.

4 Hautzer NW, Nikolai V. The application of immunoenzymatictechniques to lymphoreticular lesions of the orbit. IRCS MedicalScience: The Eye 1982; 10: 229.

5 Orcutt JC, Garner A, Henk JM, Wright JE. Treatment of idio-pathic inflammatory orbital pseudotumours by radiotherapy. BrJOphthalmol 1983; 67: 570-74.

6 Banerjee D, Ahmad D. Malignant lymphoma complicatinglymphocytic interstitial pneumonia: a monoclonal B-cellneoplasm arising in a polyclonal lymphoproliferative disorder.Hum Pathol 1982; 13: 780-2.

7 Abo W, Takada K, Kamada M, et al. Evolution of infectiousmononucleosis into Epstein-Barr virus carrying monoclonalmalignant lymphoma. Lancet 1982; 1: 1272-6.

8 Freeman C, Berg JW, Cutler SJ. Occurrence and prognosis ofextranodal lymphomas. Cancer 1972; 29: 252-6.

9 Chavis RM, Gamer A, Wright JE. Inflammatory orbital pseudo-tumor: a clinicopathologic study. Arch Ophthalmol 1978; 96:1817-22.

10 Knowles DM, Jakobiec FA. Ocular adnexal neoplasms: clinical,histopathogic, electron microscopic and immunologic charac-teristics. Hum Pathol 1982; 13:148-62.

11 Doniach D, Bottazo GF. In: Franklin EC, ed. Clinicalimmunology update. Amsterdam: Elsevier, North Holland, 1981:95.

12 Addison DJ, Rahi AHS. Immunoglobulin A (IgA) deficiencyand eye disease. Trans Ophthalmol Soc UK 1981; 101: 9-11.

569


http://bjo.bmj.com

/B

r J Ophthalm


ber 1983. Dow

nloaded from

http://bjo.bmj.com/

Documents

Lymphoproliferative orbit: immunologicalapproach to ... · polyclonal lymphoproliferative masses, need radiotherapy. The finding of reduced numbers of circulating Tcellsinthe presence