22ournalofNeurology, Neurosurgery, and Psychiatry 1995;59:322-325

SHORT REPORT

Immunohistological analysis of sarcoid myopathy

Dominique S Tews, Dieter E Pongratz

AbstractIn six cases of granulomatous myopathyimmunohistological analysis showed atypical pattern with macrophages and T4cells diffusely distributed throughout thecellular exudate. T8 lymphocytes wereinterspersed irregularly within the gran-ulomatous cellular infiltrate early ingranuloma maturation and in laterstages predominantly confined to a lym-phocytic mantle surrounding the granu-lomas. The cellular infiltrate displayednumerous activated HLA-DR and inter-leukin-2 receptor positive cells includingcell proliferation. Increased connectivetissue showed strong immunoreactivityfor fibronectin and hyaluronate. Musclefibres were negative for MHC class Imolecules. Atrophic muscle fibresexpressed desmin, a marker to distin-guish desmin positive myogenic giantcells from desmin negative Langhansgiant cells.

( Neurol Neurosurg Psychiatry 1995;59:322-325)

Keywords: myopathy; granulomas; sarcoidosis

Division ofNeuropathy,University ofMainz,GermanyD S TewsFriedrich-Baur-Institut, Ludwig-Maximilians-University, Munich,GermanyD E PongratzCorrespondence to:Dr Dominique S Tews,Division of Neuropathology,Mainz University MedicalCenter, Langenbeckstrat3e 1,D-55101 Mainz, Germany.Received 3 January 1995and in final revised form2 May 1995Accepted 9 May 1995

Generalised granulomatous inflammation ofskeletal muscle with giant cell reaction mayoccur in infectious diseases, in collagenoses,in inflammatory bowel diseases, as a remoteeffect of neoplastic processes, in associationwith thymoma or myasthenia gravis, idio-pathic especially in elderly woman, and as a

typical feature of sarcoidosis.'-4 Although theaetiology of sarcoidosis is still unknown sar-

coidosis is considered a disease of increasedimmunological activity.5

For organs other than muscle tissue,several publications describe the compositionand formation of granulomatous lesionsbased on immunohistochemical analysis.67Although there have been several reportsabout muscle involvement in sarcoidosis andCrohn's disease an immunohistologicalanalysis of the mononuclear cellular infiltrateand tissue in human skeletal muscle is stilllacking. l-

Materials and methodsMUSCLE TISSUE SPECIMENSWe studied limb muscle specimens from sixpatients with granulomatous lesions on rou-tine histological stains. Five patients weresubjected to biopsy for diagnosis of sarcoido-sis. One patient was suspected of a myopathyin association with an inflammatory boweldisease. The biopsy specimens were frozenimmediately in liquid nitrogen and stored at- 80°C. Serial transverse sections of 5 ,umthickness were cut at - 25°C on a cryostatmicrotome and left overnight at roomtemperature.

IMMUNOHISTOCHEMICAL PROCEDURESFor immunohistochemical analysis we usedthe alkaline phosphatase antialkaline phos-phatase (APAAP) technique. Monoclonalantibodies for CD 1 9, CD4, CD8,macrophage, HLA-DR, HLA-ABC, inter-leukin-2 receptor, Ki67, desmin, vimentin,fibronectin, collagen IV (all Dakopatts),CD11, Leu-7, Leu-l lb (all BectonDickinson), and hyaluronate (Dr Schleicher,Munich) were diluted with tris buffered saline(TBS) containing 10% fetal calf serum andincubated with the sections at room tempera-ture for 30 minutes. Primary antibodies fromrabbit required an additional antibody(monoclonal mouse antirabbit Ig,Dakopatts), which enabled these rabbit pri-mary antisera to react with the bridge anti-body (rabbit antimouse immunoglobulin,Dakopatts) in the second step. The bridgeantibody was connected with the APAAPenzyme complex (APAAP mouse mono-clonal, Dakopatts) in the third step. Finally,the substrate was applied by the fast redmethod. After immunostaining the sectionswere stained with haematoxylin and mountedin glycerol-gelatin medium. In controlsections, the primary antibody was eitheromitted or substituted with a non-specificrespective Ig subclass.

IMMUNOHISTOCHEMICAL ANALYSISThe complete analysis was done on 16 con-secutive sections and in each section adefined area including an average of 500muscle fibres was followed and analysed by

322 on F

ebruary 21, 2020 by guest. Protected by copyright.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.59.3.322 on 1 Septem

ber 1995. Dow

nloaded from

on February 21, 2020 by guest. P

rotected by copyright.http://jnnp.bm

j.com/

J Neurol N

eurosurg Psychiatry: first published as 10.1136/jnnp.59.3.322 on 1 S

eptember 1995. D

ownloaded from

on F

ebruary 21, 2020 by guest. Protected by copyright.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.59.3.322 on 1 Septem

ber 1995. Dow

nloaded from

323Immunohistological analysis ofsarcoid myopathy

Distribution ofmononuclear cels in granuimatou myopathy

Cell type (%) Perimysial and pervascular Endomysial

Macrophages 51-5 (1-2) 39-2 (1 0)Leu-7 + cells 0 0 0-8 (0-2)Leu-llb+cells 00 00B cells 1 9 (0 9) 3-2 (0-8)T4 cells 31 6 (1-6) 44-6 (0 8)T8 cells 15-0 (2 2) 13-1 (1-2)T4/T8 ratio 2-1 (0-2) 3-6 (0-4)

HLA-DR + /T cells 0 9 (0 9) 25-3 (2 5)IL-2-R + /T cells 2-1 (1-3) 4-5 (0 4)CD11 + /T8 cells 49-2 (5-9) 62-5 (2-8)Ki67 + cells 0-7 (0-4) 3-6 (1-0)

All values are expressed as mean (SEM).

counting the positive cells. As some antibod-ies label more than one cell type, we did con-secutive sections to estimate the number ofthe "relevant" cell type by calculating the dif-ference between the total number of positivecells in one section and the number of posi-tive, but non-relevant cells in the consecutivesection: T4 lymphocytes = all CD4 + cells- (macrophages and monocytes); CD1 1 + Tcells = all CD11 + cells - (Leu-7 + and Leu-1 lb + cells); HLA-DR + T cells = all HLA-DR + cells - (macrophages and B cells).

STATISTICAL ANALYSISAll data were presented as means (SEM).

ResultsOn routine histological examination, numer-ous non-caseating granulomas consisting ofLanghans giant cells, epitheloid histiocytic

ZA

'-'4Y

~~~~-~~~~~~~-.T

cells, and lymphocytes were present inconnective tissue and within muscle bundles(figure A).The table shows the distribution of the

different cellular phenotypes. Within thegranulomas macrophages were more frequentthan lymphocytes, whereas lymphocytesdominated in the surrounding areas of thegranulomatous lesions with lymphorrhagesextending among adjacent muscle fibres. Allgranulomas contained numerous T4 positivecells distributed diffusely among themacrophages. Conversely, in most granulo-mata T8 positive cells were less numerous(figure B). In some granulomas T8 cellscould be demonstrated throughout the granu-lomatous infiltrates (figure C) whereas in asecond type of granulomata T8 cells werelocated as a rim around the granuloma butwere absent in the centre (figure D). Most ofthese T8 lymphocytes were CD 1 1 positivecells whereas T8 lymphocytes within thegranulomata were often CD 11 negative.Cytotoxic T cells invading non-necrotic mus-cle fibres were never seen. At all sites of mus-cle, T4 lymphocytes oumumbered T8 cells.

Leu-7 positive killer cells occurred in verysmall numbers only at endomysial sites.Likewise, activated and antigen presenting Tcells bearing the HLA-DR antigen or theinterleukin-2 (IL-2) receptor were found pre-dominantly at endomysial sites. Macrophagesalso exhibited HLA-DR antigen reactivity.Throughout the granulomas and the peri-granulomatous infiltrates, replicating Ki-67

t.S"~~~~~S4!1*nV'aB~~~~~~~

N$

~~~~~~s~ ~ ~ ~ A

LA.. W;! of4+

B s ss* _

. .Bea A X

4; t,4 .

<,$,,,*','0i, ..,,, ..5.*

,.,2*D

(A) Well defined granuloma surrounded by normal appearing muscle fibres. Haematoxylin-eosin, calibration bar = 140pm. (B) T8 cells spread within granulomatous tissue. Anti-T8, calibration bar = 53 pm. (C) Granuloma with T8lymphocytes distributed diffusely within and around the granuloma. Anti-T8, calibration bar = 28 ,um. (D) Sarcoidgranuloma encircled by T8 lymphocytes. Anti-T8, calibration bar = 28 pm.

on February 21, 2020 by guest. P

rotected by copyright.http://jnnp.bm

j.com/

J Neurol N

eurosurg Psychiatry: first published as 10.1136/jnnp.59.3.322 on 1 S

eptember 1995. D

ownloaded from

Tews, Pongratz

positive cells, predominantly lymphocytes,were scattered.The increased connective tissue showed

strong immunoreactivity for fibronectin andhyaluronate whereas muscle fibres displayedsarcolemmal labelling for collagen IV.Atrophic muscle fibres compressed by granu-lomas exhibited appreciable immunostainingfor desmin. Likewise, myogenic giant cellsreacted positively for desmin whereas multi-nucleated giant cells of the Langhans type didnot react with desmin or with any other anti-bodies used. Neither normal appearingmuscle fibres nor atrophic or degeneratedfibres or myogenic giant cells showedimmunoreaction for vimentin. Muscle fibresshowed no expression of HLA-ABC.

DiscussionThe immunophenotypic analysis of mononu-clear cells in granulomas showed the typicaldistribution of cell subtypes described ingranulomas for other organs.6 7 Granulomaformation is based on an inflammatory exu-date. Several factors-that is, antigen medi-ated signals involving membrane binding-are discussed as initiating accumulation ofinflammatory cells.8 Initially, numerous lym-phocytes, predominantly T4 cells, are inter-spersed diffusely in the granulomatousinfiltrate among macrophages and mono-cytes.5 These monocytes and macrophageshave a high potential to differentiate intoepitheloid and multinucleated giant cellsrepresenting maturation of the granuloma inwhich T8 cells are absent from central areas,but are confined to the periphery of the gran-ulomas. Like other authors we suggest thatthese two histological types of granulomas,also found in our muscle samples, may repre-sent different stages of granuloma matura-tion.5 9 Whereas T4 cells seem to beresponsible for induction and modulation ofgranuloma development T8 cells are intendedfor maturation and focusing of granulomas inlater stages. 101' The predominance of CD 1 1positive cells at the border of the granulomascould be an attempt to prevent further spreadof inflammatory infiltrates into surroundingareas.

Several unknown factors released byimmunocompetent cells control immunologi-cal events of granuloma formation and matu-ration. IL-2, released by T4 cells, binds toIL-2 receptor bearing antigen activated Tcells and selectively triggers their proliferationand differentiation into effector cells.5 Asshown in other studies, we also found in ourgranulomata numerous activated HLA-DRpositive as well as IL-2 receptor bearing Tcells.91213 Besides cellular redistribution andmigration from the bloodstream, a secondmajor mechanism for cellular accumulationentails cell proliferation at the sites of diseaseactivity indicated by numerous Ki-67 positivelymphocytes and macrophages.5611The role of several active mediators includ-

ing IL-2, collagen, and fibronectin in thedevelopment of fibrosis has not. been

clarified.5 1416 Fibronectin is regarded as acompetent factor in the early stage of fibro-blast migration and proliferation which stim-ulates fibroblast replication.'6 We found anenhanced expression of fibronectin as well asof hyaluronate within and around granulomastogether with an increase of connective tissue.Increases in both extracellular matrix compo-nents are also described for bronchoalveolarlavage fluid in sarcoidosis."7 Although thedestruction of muscle fibres seems to bemainly secondary by derangement and com-pressive destruction there is evidence thatmonocytes and macrophages are able torelease a type IV collagenase which mayimpair the muscle plasma membrane.'8

Atrophic and compressed muscle fibresdisplayed strong expression of desmin and itis under debate if this is a sign of fibre break-down or of regeneration. Desmin, however,seems to be a useful marker to distinguishmyogenic giant cells and multi-nucleatedgiant cells of the Langhans type. Myogenicgiant cells can be positively identified by theirexpression of desmin absent in Langhans typegiant cells.

At the present time, we are not able to viewour results of immunohistological analysis asspecific for sarcoidosis. Perhaps, morespecific markers will be of diagnostic help.Nevertheless, there are some features specificfor granulomatous myopathy and of diagnos-tic use for differential diagnosis in inflamma-tory myopathies.'9 Besides the typicalformation of T8 lymphocytes in granulo-matous infiltration, no other inflammatorymyopathy shows such a high percentage ofT4 cells and T4/T8 ratio. Conversely, bothpolymyositis and inclusion body myositisshow clear dominance of T8 lymphocytesinvading non-necrotic muscle fibres.

Dermatomyositis displays, likewise, a highamount ofT4 cells but also a distinct increasein B cells both predominantly located at per-imysial sites. Other inflammatory myopathiesshow HLA-ABC expression of muscle fibres,which is lacking in granulomatous myopathy.This study, supported by the Friedrich-Baur-Stiftung, waspart of a thesis submitted to the Ludwig-Maximilians-University, Munich, Germany, 1992, and was read at the 9thmeeting of the Scientific Council of the German Society forthe Treatment of Muscle Diseases, Hamburg, Germany,September 1989. We are grateful to Dr Schleicher for provid-ing the antibody against hyaluronate and thank Eva Wiens forroutine histological preparations, Walther Meffert for photo-graphy, and Astrid Woeber for linguistic assistance.

1 Menard DB, Haddad H, Blain JG, et al. Granulomatousmyositis and myopathy associated with Crohn's colitis.NEnglJMed 1976;295:818-9.

2 Cheng KH, Brinkman CJJ, Rothova A. An unusual caseof neurosarcoidosis confirmed by a muscle biopsy speci-men. Am J Ophthalmol 1990;110:574-5.

3 Hewlett RH, Brownell B. Granulomatous myopathy: itsrelationship to sarcoidosis and polymyositis. J NeurolNeurosurg Psychiatry 1975;38: 1090-9.

4 Lynch PG, Bansal DV. Granulomatous polymyositis.J Neurol Sci 1973;18:1-9.

5 Semenzato G. The immunology of sarcoidosis. SeminRespirMed 1986;8: 17-29.

6 Chilosi M, Menestrina F, Campelli P, et al. ramnuno-histochemical analysis of sarcoid granulomas. Am JPathol 1988;131:191-8.

7 Munro CS, Campell DA, Du Bois RM, et al. Suppressionassociated lymphocyte markers in lesions of sarcoidosis.Thorax 1988;43:471-2.

8 Strausz J, Mueller-Quernheim J. In-vitro models for theanalysis of inflammatory processes in sarcoidosis.Pneumologie 1990;44:694-8.

324 on F

ebruary 21, 2020 by guest. Protected by copyright.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.59.3.322 on 1 Septem

ber 1995. Dow

nloaded from

Immunohiswlogical analysis ofsarcoid myopathy

9 Haslam PL, Parker DJ, Townsent PJ. Increases inHILA-DQ, DP, DR, and transferrin receptors on alveo-lar macrophages in sarcoidosis and allergic alveolitiscompared with fibrosing alveolitis. Chest 1990;97:651-61.

10 Chensue SW, Boros DL, David CS. Regulation of granu-lomatous inflammation in murine schistosomiasis: Invitro characterization of T lymphocyte subsets involvedin the production and suppression of migration inhibit-ing factor. JExp Med 1980;151:1398-413.

11 Semenzato G, Pezzutto A, Pizzolo G, et al. Immuno-histochemical study in sarcoidosis: evaluation at differ-ent sites of disease activity. Clm Immunol Immunopathol1984;30:29-40.

12 Spurzem JR, Saltini C, Kirby M, et al. Expression ofHIAclass II genes in alveolar macrophages of patients withsarcoidosis. Am Rev Respir Dis 1989;140:89-94.

13 Keicho N, Kitamura K, Takaku F, et al. Serum concen-tration of soluble interleukin-2 receptor as a sensitiveparameter of disease activity in sarcoidosis. Chest 1990;98:1123-9.

14 Postlethwaite AE, Snyderman R, Kang AH. The chemo-tactic attraction of human fibroblasts to a lymphocyte-derived factor. JExp Med 1976;144:1188-203.

15 Chiang TM, Postiethwaite AE, Beachey EH, et al.Binding of chemotactic collagen-derived peptides tofibroblasts. Clin Invest 1978;62:916-22.

16 Lewandowska K, Haing UC, Rosenberg LC. Fibronectin-mediated adhesion of fibroblasts: inhibition bydermatan sulfate proteoglycan and evidence for a cryp-tic glycosaminoglycan-binding domain. Cell Biol 1987;105:1443-54.

17 Blaschke E, Eklund A, Hernbrand R. Extracellular matrixcomponents in bronchoalveolar lavage fluid in sarcoido-sis and their relationship to signs of alveolitis. Am RevRespir Dis 1990;141:1020-5.

18 Garbisa S, Balmn M, Daga-Gordini D, et al. Type IVcollagenolytic activities by human macrophages:Purification and characterization of a specific 68kDprotease. JBiol Chem 1986;261:2369-75.

19 Engel AG, Arahata K. Mononuclear cells in myopathies.Hum Pathol 1986;17:704-21.

325 on F

ebruary 21, 2020 by guest. Protected by copyright.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.59.3.322 on 1 Septem

ber 1995. Dow

nloaded from

Journal of Neurology, Neurosurgery, and Psychiatry 1996;60:125

NOTICE

Announcement from the British Neuro-psychiatry Association: 1996 summer

meeting

The 1996 Summer meeting will be heldon 14-16 July at Robinson College,Cambridge. It will include topics on neuro-

development, language, and the presenta-tion of short scientific papers and single case

videos by members. The Association's AGMwill be held on 16 July.

For further details of these meetings pleasecontact: Sue Garratt, AdministrativeAssistant, BNPA, 17 Clocktower Mews,London NI 7BB. Telephone/Fax: 0171 2265949.

For details of membership of the BNPA, whichis open to medical practitioners in psychiatry,neurology, and related clinical neurosciences,please contact: Dr J7onathan Bird, SecretaryBNPA, Burden Neurological Hospital, StokeLane, Stapleton, Bristol, BS16 1QT.Telephone: 01179 701212 ext 2925/2929 or

Sue Garratt at the address given above.

CORRECTION

Clinicopathological case conference.Neurol Neurosurg Psychiatry 1995;59:322-7.

Dr CC Tijssen should be included as an

author for this case conference, whichshould read:Harrison MJG, Teepen JLJM, Tijssen CC.A case of recurrent headache and neurologi-cal deficit. Neurol Neurosurg Psychiatry1995;59:322-7.

BOOKREVIEWS

All titles reviewed here are available fromthe BMJ Bookshop, PO Box 295, LondonWC1H 9TE. Prices include postage in theUnited Kingdom and for members of theBritish Forces Overseas, but overseas

customers should add £2 per item forpostage and packing. Payment can be madeby cheque in sterling drawn on a UnitedKingdom bank, or by credit card

(Mastercard, Visa or American Express)stating card number, expiry date, and yourfull name.

Neurovascular Surgery. Edited by LPHILIP CARTER and ROBERT F SPETZLER.Associate Editor MARK G HAMILTON. (Pp1446; £325 00). Published by BlackwellScience, Oxford 1994. ISBN Hb 0-070-11020-4.

This is a highly impressive and wellresearched text book covering every aspectof modem neurovascular surgery. The edi-tors have assembled 125 contributors whoare all well respected in their fields. As aresult, the text book is extensive with globalreference. The book is divided into six mainparts consisting of general principles, occlu-sive disease, haemorrhagic conditions, vas-cular compression, spinal vascular disease,and vascular injuries. All sections are wellcovered and include essential medical neu-roanaesthetic and interventional aspects aswell as the direct surgical descriptions.Despite multiauthorship, the editors appearto have kept overlap down to an absoluteminimum.What I found particularly impressive was

the detail in which the key trials and clinicalpapers have been documented. As a result, Ihave found the book useful in extracting keymaterial and figures with ease. The extensivecoverage makes this book suitable for inter-ested parties other than surgeons. I justwonder whether it may have been moreappropriately entitled "The Treatment ofNeurovascular Conditions" so as not todeter the non-surgical community. I wouldpersonally recommend this book to neuro-intensivists and neuro-radiologists in addi-tion to neurosurgeons of all grades.

In summary, this really is an excellentbook which has been put together by highlyrespected workers in the discipline of neuro-vascular conditions. I congratulate them onwhat must have been an enormous effort.

PETER KIRKPATRICK

Etiology of Parkinson's Disease. Editedby JONAS H ELL.ENBERG, WILLIAM C KOLLERand J WILLiAM LANGSTON. (Pp 600;$195.00). Published by Marcel Dekker,New York 1995. ISBN 0-8247-8823-0.

This book, inspired by the fortuitous discov-ery of MPTP induced parkinsonism in asmall group of Californian drug addicts, is ajoy to read. It represents an attempt toreview and explore all the theories that havebeen put forward to account for the cause ofParkinson's disease (PD); a review thatextends to a bibliography of 2413 references!In addition to presenting the possible theo-ries, it openly criticises and highlights theshortcomings of studies set up to unravel theaetiology of this common yet incurable dis-ease.

The book opens with an excellent discus-sion on the causes of parkinsonism and howthese relate to idiopathic PD (IPD). Thediscussion then goes on to highlight thedifficulties in identifying IPD from otherforms of parkinsonism, especially otherneurodegenerative conditions such as multi-system atrophy. The difficulty in establish-ing and verifying a diagnosis of IPDantemortem seems to have been helped bythe advent of functional imaging with thePET scanner. Although this technique isnot widely available, problems are alreadyappearing on the horizon as pointed out byGolbe in his chapter on the genetics of PD.He makes the point that some twin studieshave shown abnormal fluorodopa PET scanssimilar to those in PD, in twins who do nothave PD but only a postural tremor; a pointfurther discussed in later chapters on theneuroepidemiology and comorbidity in PD.

This initial discussion on what constitutesPD is fundamental to understanding whatmay cause it. However, in addition thepathology of the condition has to beexplained and Fomo gives an excellentaccount on the neuropathology of PD. Thischapter makes another important point-namely, that although the dopaminergicnigral neurons bear the brunt of the pathol-ogy-they are not the only neurons to beinvolved in the disease process. This pointmust therefore be taken into account whenany theory purporting to explain the aetio-pathology of this condition is put forward.The diagnosis and pathology of IPD

having been established, the question thenarises as to what causes the neurons to belost. Irwin and Langston begin by presentingpossible mechanisms of cell death, althoughno discussion on the ontogeny of the nigros-triatal pathway is given. This is a shame as itmay be relevant to the mechanism of celldeath in the disease state. Nevertheless, thepossible cellular mechanisms that cause thedopaminergic neurons to be lost is thentaken up in later chapters that specificallyaddress the possibilities of endogenous andexogenous toxins as aetiological agents,including a discussion of MPTP itself. Thisdiscussion on toxins raises the further ques-tion as to whether levodopa itself is toxic tothe nigral dopaminergic neuron and socatalyses the disease process. A significantamount of in vitro work is omitted from thediscussion but the overwhelming work fromthose studies agrees with the conclusion putforward in this book-namely, that althoughthis is a theoretical risk there is no convinc-ing evidence that it is a dominant factor dri-ving the pathogenesis of this condition.The only criticisms I have of this book are

minor ones, in that some of the chaptersread rather too much like lists (for example,chapter 7) whilst others are unnecessarilytechnical (for example, chapter 11 on assess-ment of predictors). Overall, though, thisbook is good fun. It presents a large amountof information in an interesting and criticalway and I would therefore strongly recom-mend it to all who wonder at the cause ofthis major neurodegenerative disorder.

ROGER BARKER

125