Annals of the Rheumatic Diseases, 1984, 43, 778-782

Synovitis with non-specific histological changes insynovium in chronic sarcoidosisDAVID G. PALMER AND H. RALPH SCHUMACHERFrom the Arthritis-Immunology Center, Veterans Administration Medical Center, the University ofPennsylvania School of Medicine, Philadelphia, PA, and the University of Otago Medical School, Dunedin,New Zealand

SUMMARY Polyarthritic episodes in seven patients during the course of chronic sarcoidosisinvolved the knees and less frequently other joints. Synovial effusions were non-inflammatory ormildly inflammatory. Needle synovial biopsies 10 days to seven weeks after the onset of jointsymptoms have revealed varying patterns, including mild lining cell proliferation, occasionalvascular congestion, diffuse infiltrates with lymphocytes and histiocytes, but no granulomas.These studies show distinctly less inflammation than in rheumatoid arthritis. Granulomas neednot be identifiable in synovium in all chronic sarcoidosis with arthritis.

Key words: synovial fluid, electron microscopy, granulomas, synovial biopsy.

Although an arthritis is a common clinical feature ofacute sarcoidosis (Lofgren's syndrome)' 2 it is lessfrequently recognised during the course of chronicdisease.3 When an arthropathy does appear underthe latter circumstances, a problem in differentialdiagnosis may arise if the nature of the underlyingdisorder has not been established, or if the possibil-ity of some coincidental arthritis has to be enter-tained. The arthritis which may accompany chronicsarcoid arthritis may be acute and evanescent,intermittent with relapses and remissions, or pro-gressive leading to joint destruction.f Differentia-tion from rheumatoid arthritis may not always beeasy, because systemic manifestations such asweight loss, tiredness, and sweating are common toboth, rheumatoid factor accompanies chronic sar-coidosis in up to 38% of patients,6 and bothdisorders may be accompanied by subcutaneousnodules.7 Further diagnostic confusion with othercauses of arthritis can occur as antinuclearantibodies,8 hyperuricaemia,3 4 and hypercalcae-mia9 1( have all been reported in some patients withsarcoidosis.

Accepted for publication 11 June 1984.Correspondence to H. Ralph Schumachcr, MD, Professor ofMedicine, University of PA School of Medicine, Director,Arthritis-Immunology Center, Veterans Administration MedicalCenter, University and Woodland Avenues, Philadelphia, PA19104, USA.

Since Sokoloff and Bunim described non-caseating granulomas in three of five surgical biopsyspecimens of synovium in cases of chronicsarcoidosis'1 it has been widely accepted'2 13 thatsuch changes would probably be found if involvedsynovium were to be examined histologically. Wehave reviewed synovial fluid analyses and the lightand electron microscopic appearances of synoviumobtained by needle biopsy from knee joints develop-ing signs of synovitis during the course of sarcoidosisand have found only non-specific changes in theseven patients studied.

Patients and methods

All patients (Table 1) had a well establisheddiagnosis of chronic sarcoidosis based on histologi-cal evidence of typical non-caseating granulomatouslymphadenitis associated with a compatible clinicalpicture. With the exception of concurrent systemiccorticosteroid therapy, specimens from patients withany additional problems which might have modifiedthe histological features were excluded from thesurvey. Synovial biopsy was performed with aParker-Pearson needle, specimens were fixed inbuffered formalin, and sections stained with haema-toxylin and eosin for light microscopy. For transmis-sion electron microscopy specimens from patients 1,

778

copyright. on F

ebruary 19, 2020 by guest. Protected by

http://ard.bmj.com

/A

nn Rheum

Dis: first published as 10.1136/ard.43.6.778 on 1 D

ecember 1984. D

ownloaded from

Synovitis with non-specific histological changes in synovium in chronic sarcoidosis 779

-cto ~(1' 1 - -

U~~~~

0~~~~~~~~~~~~~~~~

_ : r-g 1 N

U E

-2 IQ Q Q Q 7 Q~~~~~~~~~~~~ot,JQ -Nb ^

_~ ~~E ~'~- .3 >8 °O =

_ t a: ' _-2 U

2 Ut: tc-s;:r-rm_- -._

cU

0 X ns

copyright. on F

ebruary 19, 2020 by guest. Protected by

http://ard.bmj.com

/A

nn Rheum

Dis: first published as 10.1136/ard.43.6.778 on 1 D

ecember 1984. D

ownloaded from

780 Palmer, Schumacher

2, 3, 4, and 7 were fixed in 1/2 strength Karnovsky'sparaformaldehyde-glutaraldehyde, and processed aspreviously described.'4 The gross appearances ofknee synovial fluid aspirates, leucocyte counts, andcell differentials on Wright's stained smears wererecorded. All fluids were examined with compen-sated polarised light and no crystals were found.

Results

CLINICAL PRESENTATION (Table 1)Four of the seven patients had initially presentedwith constitutional symptoms including anorexia,weight loss, or fever. All had pulmonary involve-ment, but this varied from asymptomatic hilarlymphadenopathy (without erythema nodosum) tosevere respiratory impairment with productivecough and shortness of breath. Two patients (cases 3and 7) had biopsy proved hepatic granulomas andpatient 4 had hepatomegaly. These three hadsplenomegaly. Case 3 had conjunctival granulomasand case 7 had radiological evidence of boneinvolvement. Compatible skin lesions were presentin five patients with biopsy proof of granulomas inpatients 1 and 3.

All patients had at least two joints involved.Lower limb joint involvement, particularly of the

knees, predominated and was usually symmetrical.Upper limb joint involvement was less common andtended to be asymmetrical. Morning stiffness wasonly an occasional minor complaint. Arthritis hadpersisted as long as one year before examination,though most patients were seen during the first 2-7weeks of the arthritis.The interval between the appearance of symp-

toms attributed to the onset of sarcoidosis and initialjoint manifestations varied from three weeks to 13years. The long-term course of the arthritis in mostof our patients is not known, as many of them havehad limited follow-up visits. Patient 1 has hadrecurrent non-destructive arthritis over 3 years. Noevidence of articular damage was detected in radio-graphs of the involved joints in any patient.

All synovial fluid samples had non-inflammatoryor mildly inflammatory characteristics, being re-latively clear, colourless or straw-coloured, and ofgood viscosity. The total white cell count rangedfrom 0.25 to 2-65 x 109/1 with 56-100% mononuclearcells. There were 9-82% lymphocytes and manycells that appeared to be synovial lining cells orother large mononuclear cells.

LIGHT MICROSCOPYThe histological appearances were heterogeneous.

Fig. 1 Patient 1. Hyperplasia andhypertrophy of the synovial lininglayer (at top and bottom) isapparent. There is a diffuseinfiltrate of tissue histiocytes andlymphocytes. Vessels (V) arecongested. (H and E, x 30).

copyright. on F

ebruary 19, 2020 by guest. Protected by

http://ard.bmj.com

/A

nn Rheum

Dis: first published as 10.1136/ard.43.6.778 on 1 D

ecember 1984. D

ownloaded from

SYnovitis with nioni-specific histological changes in synovium in chronic sarcoidosis 781

Fibrin deposition on the synovial lining ceils wasusually present but was inconspicuous and discon-tinuous. Synovial hypertrophy and hyperplasia werelikewise variable, being absent in some specimensbut forming a mantle some 5-7 cells deep in others(Fig. 1). The size of specimen available did not allowfor an adequate evaluation of villous hypertrophy,but some increased complexity of these structureswas seen.There was a rather diffuse increase in interstitial

histiocytes and lymphocytes (Figs. 1. 2) andalthough perivascular localisation was encounteredcompact foci of cells were not seen. Diffuse infiltra-tion was best seen in patients 1, 2, and 7. Closelypacked epithelioid cells or multinucleated cells werenot found (Figs. 1. 2). Plasma cells were relativelysparse, and polymorphonuclear leucocytes wereusually seen onlv in association with debris.Where histiocytic infiltration was marked, vascu-

larity appeared to be decreased. In other areas smallblood vessels were numerous (Figs. 1, 2). Perivascu-lar and interstitial fibrosis of marked degree waspresent in part of the specimen from patient 3.Necrosis was the dominant finding in specimensfrom patients 3 and 5 with loss of intimal cells,necrosis of interstitial cells, and hyaline change inthe interstitium itself.

ELECTRON MICROSCOPYUltrastructural studies allowed identification ofsmall numbers of synovial lining cells with nounusual features. The majority of infiltrating cellsseen seemed to be histiocytes, and in four of the fivepatients studied by EM these contained consider-able finely granular material in their phagocyticvacuoles. Vessels occasionally had multilaminatedbasement membranes, and there were occasionalpale endothelial cells or gaps between them. Therewas some interstitial necrotic debris. Mast cells wereseen to be degranulated in two specimens. Specifi-cally not found were any virus-like or bacterialparticles, tubuloreticular structures, or electrondense deposits in vessel walls.

Discussion

Involvement of joints in the course of acute sar-coidosis and a chronic persistent arthritis withhistological evidence of a non-caseating granuloma-tous reaction accompanying involvement of otherorgans in chronic disease are closely identifiedwith these two seemingly distinct forms of sarcoidreaction. The duration of the synovitis before biopsyseems not to be entirely critical in determining thepresence of granulomas in that these have been

Fig. 2 Patient 7. Synovial bloodvessels (V) are numerous. Thehistiocytic and lymphocytic cellularinfiltrate shows little evidence offocal accumulation (H andE, x 60).

copyright. on F

ebruary 19, 2020 by guest. Protected by

http://ard.bmj.com

/A

nn Rheum

Dis: first published as 10.1136/ard.43.6.778 on 1 D

ecember 1984. D

ownloaded from

782 Palmer, Schumacher

recognised by the fifth week.9 In contrast, thenature of the mild and often transient arthralgiasand arthritides which may also punctuate the courseof chronic sarcoidosis have not been welldefined.3 4 15 Sporadic reports which include two ofthe five patients reported on by Sokoloff andBunim'1 have indicated that the histological changesmay be of a non-specific nature.5 9 13A6As chronic sarcoidosis may run a course of many

years in patients who may be immunologicallycompromised (both as a result of their disease 7 andthrough long continued corticosteroid therapy),other types of arthritis such as infective arthritismight well be occasionally encountered. It was ourexperience in reviewing the case notes of all patientswith sarcoidosis and arthritis that documented unre-lated disorders which might have contributed to thesynovial pathology were in fact common. Such caseswere excluded from this study. The earlier diagnosisand treatment of the arthritis as rheumatoid arthri-tis, the diagnosis and treatment of the illness at itsonset as tuberculosis, a history of non-specificurethritis, and the culture of Aspergillus fumigatusfrom a granulomatous node in patients with other-wise typical sarcoid disease were among associatedproblems encountered. Although it was not practi-cal to exclude patients treated with corticosteroidtherapy from this series, the possibility of corticos-teroid induced osteonecrosis was considered in thedifferential diagnosis.18Our principal conclusions are that a non-caseating

granulomatous reaction is not necessarily a readilydetectable histological feature of the episodicpolyarthritis which may evolve in the course ofchronic systemic sarcoidosis, and that synovial fluidfrom the involved joints is usually of a non- or mildlyinflammatory type even though the onset of thearthropathy may be acute and dramatic. The pre-ponderance of mononuclear cells in the synovialfluid is similar to the findings in most previousreports." 13 19 The diffuse synovial infiltrate consist-ing principally of histiocytic cells and lymphocytesmight be representative of the change precedingepithelioid transformation and granulomaformation21 or it might be due to other mechanismsin these patients. Although some inflammatory cellinfiltration was common, it was less marked thanthat seen in typical chronic rheumatoid disease, sothat adequate biopsies might help separate this fromsarcoidosis even in the absence of granulomas.

Electron microscopy provided only very non-specific findings and did not yield any positiveevidence for mechanisms that might be involved inthis type of arthritis. Whether any of the interstitialor phagocytosed granular material might containantigenic material important in pathogenesis would

require immunomorphological studies that were notdone on these specimens.While better surgical biopsy specimens might

possibly have yielded tissue containing focal granu-lomas in some patients, it is doubtful if thisprocedure can ordinarily be justified in patients witha diagnosis of sarcoidosis alreadly well substanti-ated, so that most clinical evaluations of arthritis insarcoidosis may well depend on interpretation ofneedle biopsies as used here. Study of additionalpatients in which long-term follow-up can beobtained may clarify whether or how often this mildsynovitis progresses to a destructive granulomatousarthritis.

References

I Lofgren S. Primary pulmonary sarcoidosis. 1. Early signs andsymptoms. Acta Med Scand 1977; 145: 424-31.

2 Kitridou R P, Schumacher H R. Arthritis of acute sarcoidosis.Arthritis Rheum 1970; 13: 328 (abstr).

3 Gumpel J M, Johns C J, Shulman L E. The joint disease ofsarcoidosis. Ann Rheum Dis 1967; 26: 194-205.

4 Spilberg 1, Siltzbach L E, McEwen C. The arthritis ofsarcoidosis. Arthritis Rheum 1969; 12: 126-37.

5 LeGoff P, Jaffres R, Schwarzberg C. Brousse A. LeRoy J P.Arthrite chronique destructrice du genou d'origine sarcoido-sique associee a des geodes des os longs. Rev Rhuin MalOsteoartic 1982; 49: 647-52.

6 Oreskes I, Siltzbach L E. Changes in rheumatoid factor activityduring of sarcoidosis. Am J Med 1968; 44: 60-7.

7 Boyd R E, Andrews B S. Sarcoidosis presenting as cutaneousulceration, subcutaneous nodules and chronic arthritis. JRheumatol 1981; 8: 31 1-36.

8 Veien N K, Hardt F, Bendixen G0 et al. Humoral and cellularimmunity in sarcoidosis. Acta Med Scand 1978; 203: 321-6.

9 Bianchi F A, Keech M K. Sarcoidosis with arthritis. AnnRheum Dis 1964; 23: 463-79.

10 David N J, Verner J V, Engel F L. The diagnostic spectrum ofhypercalcemia. Am J Med 1962; 33: 88-110.

11 Sokoloff L, Bunim J J. Clinical and pathological studies of jointinvolvement in sarcoidosis. N Engl J Med 1959; 260: 842-7.

12 Arnold W J. The rheumatic manifestations of sarcoidosis. In:Kelley W N, ed. Textbook of rheumatology. Philadelphia:Saunders, 1981: 1531-7.

13 Schumacher H R. Sarcoidosis. In: McCarty D J. ed. Arthritisand allied conditions. 9th ed. Philadelphia: Lea and Febiger,1979: 918-24.

14 Schumacher H R, Newton C, Halliwell R E W. Synovialpathologic changes in spontancous caninc rheumatoid-likearthritis. Arthritis Rheum 1980; 23: 412-23.

15 Kaplan H. Sarcoid arthritis with a response to colchicine. NEngl J Med 1960; 263: 778-81.

16 Ferguson R H. Paris J. Sarcoidosis. Study of twenty-nine caseswith a review of splenic, hepatic, mucous-membrane, retinal,and joint manifestations. Arch Intern Med 1958; 101: 1065-84.

17 Daniele R P, Dauber J H, Rossman M D. Immunologicabnormalities in sarcoidosis. Ann Intern Med 1980; 92: 406-16.

18 Heimann W G, Freiberger R H. Avascular necrosis of thefemoral and humeral heads after high-pose corticosteroidtherapy. N Engl J Med 1960; 263: 672-5.

19 Varkey B. Synovial fluid in sarcoidosis. Ann Intern Med 1974;81: 557.

20 Crystal R G, Roberts W C, Hunninghake G W, Gadek J E,Fulmer J D, Line B R. Pulmonary sarcoidosis: a diseasecharacterized and perpetuated by activated lung T-lymphocytes. Ann Intern Med 1981; 94: 73-94.

copyright. on F

ebruary 19, 2020 by guest. Protected by

http://ard.bmj.com

/A

nn Rheum

Dis: first published as 10.1136/ard.43.6.778 on 1 D

ecember 1984. D

ownloaded from