rheumatoid synovitis and jointdisease

Ann. rheum. Dis. (1975), 34, 1

Rheumatoid synovitis and joint diseaseRelationship between arthroscopic and histological changes

D. B. YATES AND J. T. SCOTTFrom Charing Cross Hospital, London W6 8RF

Yates, D. B., and Scott, J. T. (1975). Annals of the Rheumatic Diseases, 34, 1. Rha-toid synovitis and joint disease. Relationship between arthroscopic and histologicalcanges.Arthroscopic and histological synovial features have been studied in forty-two patientswith classical or definite rheumatoid arthritis. A total index of disease activity as judgedarthroscopically correlates significantly with a total index of histological activity. Inthose patients who have dense, waxy looking villi, the intensity of the villus-proliferationis associated with lymphocyte infiltration of the synovium. No relationship betweensynovial lining cell proliferation and cartilage disease nor between sparsity oflymphocyteinfiltration and cartilage disease could be established.

While direct examination of the knee joint by arthro-scopy has been an established practice in Japan formany years (Watanabe, Takeda, and Ikeuchi, 1969),it is only recently that the technique has become morepopular in the Western world. Jayson and Henderson(1973) have shown the value of arthroscopy for thediagnosis of inflammatory joint disease, and inparticular the advantages of arthroscopy overarthrotomy for detailed examination ofthe synovium.The value of synovial biopsy material in assessing

the severity and prognosis of rheumatoid arthritishas been examined by many authors with mixedresults. Some of the problems have arisen from thefact that blind biopsy techniques were used, but evenwith adequate material major variations in histologyfrom area to area within any one joint have beenreported in up to 89% of patients (Cruickshank,1952). On the other hand, certain features ofrheuma-toid synovitis-synovial lining cell thickness andlymphocyte infiltration-are of importance in thedevelopment of subsequent damage to articularcartilage (Muirden and Mills, 1971). The presentinvestigation examined the relationship inrheumatoidarthritis betrn, arthroscopic appearances of thesynovium of the knee and histopathological changesin sections of synovial biopsies obtained under directvision; together with the relationships betweenhistology and visible damage in articular cartilage.

Patients, materials, and methods

Forty-two patients (33 seropositive and 9 seronegative)who fulfilled the A.R.A. criteria (Ropes, Bennett, Cobb,Jacox, and Jessar, 1958) for classical or definite rheumatoidAccepted for publication May 12, 1974.

arthritis have been studied. The duration of total diseaseactivity ranged from 4 weeks to 30 years (mean 7 years;seropositive 8-1 years; seronegative 3 years). Patients wereadmitted to hospital for one day only; clinical assessmentwas carried out, after which arthroscopy and simultaneoussynovial biopsy using a Stortz fibreoptic instrument wasperformed under local anaesthesia with full sterile pre-cautions in an operating theatre. On the day of theinvestigation blood was taken for rheumatoid factorserology (Rose, Ragan, Pearce, and Lipman, 1948) andESR.A full visual survey was made of the synovium of the

suprapatellar pouch, medial and lateral capsular walls,and over the infrapatellar fat pad. Normal synovium(Fig. la) appears as a pale yellow thin membrane throughwhich small vessels and the underlying white fibroustissue of the joint capsule may be seen. Abnormalitiesconsisting of synovial hyperaemia (Fig. lb), dense, waxy-looking 'active' villi (either squat, and appearing likecobblestones, or conical or pedunculated with clubbedends-Fig. lc), and white, avascular fibrous villi (Fig. ld)were each scored 0, no change; 1, slight; 2, moderate; or3, marked changes. A similar scoring system was appliedto fibrin floating free within the joint space (Fig. le) andto cartilage disease (either erosion, cartilage thinning,or pannus encroaching on to the cartilage (Fig. lf).After photographs had been taken, multiple synovialbiopsies were obtained. In order to minimize error due tosample variation, at least 6 biopsies (each measuring0-1 cm x 0-2 cm x 04 cm) were taken from an area highon the medical capsular wall; only these biopsy specimenswere used for subsequent statistical analysis.

Biopsy specimens were fixed in neutral phosphatebuffered 10% formalin and sections stained with haema-toxylin and eosin, Periodic Acid Schiff, and Lendrum'sMartius Scarlet Blue trichrome. These sections wereexamined, without knowledge of the identity of thepatient, in three ways.

copyright. on N

ovember 27, 2021 by guest. P

rotected byhttp://ard.bm

j.com/

Ann R

heum D

is: first published as 10.1136/ard.34.1.1 on 1 February 1975. D

ownloaded from

http://ard.bmj.com/

2 Annals ofthe Rheumatic Diseases

(aI

r:(IC.)

FIG. 1 (a) Normal synovium. (b) Synovial hyperaemia. (c) 'Active' villus formation; pedunculated large and waxylooking villi arising from synovium. (d) Fibrous villus formation; white avascular villi arising from synovium. (e) Ricebody; large mass offibrinfloating above hypertrophic synovium. (f) Cartilage disease; pitted, irregular surface ofmedialfemoral condyle with loss ofarticular cartilage in the centre

copyright. on N



j.com/

Ann R

heum D


ownloaded from

http://ard.bmj.com/

Rheumatoid synovitis and joint desease 3

A Histological features consistent with rheumatoidarthritis according to the A.R.A. criteria (surface fibrindeposition, synovial cell hyperplasia, subsynovial mono-nuclear cell infiltration, perivascular cellular infiltration,and perivascular oedema) were each graded 0-4 dependingon the severity of each feature and the sum used as anindex of total histological activity.B Sections were assessed for the dominant phase of thefour phases of synovitis categorized by Britton, Ruddy,Corson, Sosman, Schur, and Austen (1970). These phasesare as follows.

(i) Exudative Synovial congestion of vessels andoedema with some exudation of neutrophilic leucocytesand fibrin deposition.(ii) Infiltrative and proliferative Infiltration by lym-phocytes, plasma cells, and histiocytes, proliferation offibroblasts, synovial cell hyperplasia, and conspicuouscapillaries.(iii) Necrotic Areas of necrosis of lining cells andalso in subsynovial connective tissue together with somefibrin deposition.(iv) Fibrotic Predominant fibrosis without mono-nuclear cell infiltration or necrosis.

C Finally, in order to correlate cartilage damage withtwo specific histological features, the average thicknessof the synovial lining layer was calculated in terms ofnumbers of cells; and the intensity of lymphocyticinfiltration was graded 0-8 from a count of the averagenumber of lymphocytes seen in three random high powerfields of subsynovium.

Arthroscopy caused minimal discomfort only and therewere no complications.

Results

Correlations between arthroscopic and histologicalfeatures are summarized in Table I; correlations

between clinical features and arthroscopic andhistological features are summarized in Table II. Itcan be seen (Fig. 2) that by comparing the totalhistological activity with the total visible synovitis(the score derived from the sum of the four arthros-copic synovial features but excluding cartilagedisease) a significant (P < 0 01) relationship exists.However, if individual histological features are eachassessed against the total visible synovitis, there areno significant correlations. Where 'active' villusformation is seen on arthroscopy, however, there issignificant relationship (P < 0 01) between theseverity of the villus formation and the intensityof lymphocyte infiltration (Fig. 3). No relation-ship, direct or inverse, is found between villusformation and the thickness of the synovial lininglayer.When the severity of cartilage disease is assessed

against histological features, again there is no signifi-cant correlation. In particular, there is no relation-ship between cartilage disease and lymphocyteinfiltration (Fig. 4), nor between cartilage diseaseand thickness of the synovial lining layer (Fig. 5).From Table II it is seen that there is no relation-

ship between any of the clinical or serological featuresmeasured and arthroscopic appearance. The domi-nant phase of synovitis histologically does howevervary with the duration of activity within the joint(Fig. 6). In the majority of patients with an exudativesynovitis the disease had been active for 6 months orless (X2 = 7-33; P < 0 01), whereas a fibrotic histologywas seen in synovium where the disease had beenactive for more than 18 months (X2= 6-33; P <0-025).

Table I Correlations between arthroscopic and histologicalfeatures

Total arthroscopic score v. Total histological score r = 0-46 P < 0.01Total arthroscopic score v. Lymphocvtic infiltration r = 0-31 NSTotal arthroscopic score v. Lining ceil thickness r = 0-24 NSActive villus formation v. Lymphocytic infiltration r = 0-56 P < 0 01Active villus formation v. Lining cell thickness r = 0-12 NSVisible hyperaemia v. Lymphocytic infiltration r = 0-20 NSVisible hyperaemia v. Lining cell thickness r = 0-12 NSCartilage disease v. Lymphocytic infiltration r = 0-07 NSCartilage disease v. Lining cell thickness r = 0-05 NS

Table II Correlations between clinicalfeatures and arthroscopic and histologicalfeatures

Level ofDAT v. Total arthroscopic score r = 0-01 NSLevel ofDAT v. Total histological score r = 003 NSDuration of disease activity v. Total arthroscopic score r = 0-12 NSDuration of disease activity v. Total histological score r = 0-04 NSSynovial fluid protein level v. Total arthroscopic score r = 0-02 NSSynovial fluid protein level v. Total histological score r = 0-12 NSDisease activity <6 m v. Exudative synovitis Y.'= 7-33 p < 0 01Disease activity >18 m v. Fibrotic synovitis x2 = 6-33 P < 0 025

2

copyright. on N



j.com/

Ann R

heum D


ownloaded from

http://ard.bmj.com/

Annals ofthe Rheumatic Diseases

FIG. 2 Correlation between total arthro-scopic score and total histological score.Mean linear regression ±2 standarderrors ofthe estimate; r = 0-46; P < 0 01

2 3 4 5 6 7 8 9 10

Total Synovial Activity (Arthroscopic

R 0. 56P< 0.01

FIG. 3 Correlation between 'active'villus formation and subsynovial lympho-cyte infiltration. Mean linear regression±2 standard errors of the estimate;r=056;P <001

1 2 3 4 5 6 7 8 9Lymphocytes

1++-* R" 0.07

FIG. 4 Correlation between severity of cartilage diseaseand intensity of subsynovial lymphocyte infiltration; r=0 07: not significant

..I

1 2 3 4 5 6 7 8

Lymphocytes

0

0 00 e e a

9

+

copyright. on N



j.com/

Ann R

heum D


ownloaded from

http://ard.bmj.com/

Rheumatoid synovitis andjoint disease 5

R - 0. 047

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Lining Cell Thickness

FIG. 5 Correlation between severity of cartilage diseaseand the average number ofcells in the synovial lining layer;r = 0O05: not significant

Fibrotic

Infiltrative &

Proliferative

Exudative

::-;. -::;-:.;. -: ;;s:;x; :;;............, ..,,,

0-3 3-6 6-9 9-12 12-15 15-18

Months

FIG. 6 Distribution of dominant phase of synovitisagainst duration ofactive disease in thejoint studied

Discussion

Most pathological studies of patients with activeor chronic rheumatoid synovitis have failed to showa significant association between synovial histo-pathological changes and clinical disease activity

(Schumacher and Kitridou, 1972). This may in padbe due to problems in interpretation of synovialbiopsy material which may vary from area to area inone joint. Similar variation has been observed inarthroscopic appearances (Robles-Gil, 1973; Watan-abe and others, 1969). It is not uncommon, forexample, to see relatively normal synovium in thesuprapatellar pouch, with gross abnormalities in thesynovium covering the medial and lateral capsularwalls. In this study, therefore, multiple synovialbiopsies were taken from a similar area of the medialcapsular wall in all patients.

Since considerably more synovium can be exam-ined with an arthroscope than with a micrsocope,useful details of severity and activity of synovialdisease may be obtained by arthroscopy. The presentstudy has shown a significant correlation between totalvisible synovial changes and those seen histologically,with a further association between 'active' villusformation and the number of lymphocytes presentin the subsynovium, a finding in accordance with thearthroscopic study of Takeda (1960) of 384 cases ofvarious knee joint disorders. The 'active' villi corres-pond with those described by Palmer (1967) as bulkyand dense and characteristic of RA-being eithersquat and sessile or elongated and pedunculated-but nothing like the delicate form of normal villi.There appears to be an overall correlation between

'active' villus formation, subsynovial lymphocyteinfiltration, reduced synovial fluid complementlevel (Britton and others, 1970) and a severe clinicalcourse with increased joint damage (Hedberg, 1963;Pekin and Zvaifler, 1964).

Variations in disease activity are not only found indifferent situations but also during the course of thedisease. At any one time synovial membrane shows amixed picture of acute and chronic changes togetherwith signs of tissue repair and fibrosis. It has beenshown that only in the very early stages ofrheumatoiddisease (Kulka, Bocking, Ropes, and Bauer, 1955) isthere any one consistent histologicdl pattern, namelycongestion, oedema, and fibrin deposition withoutmuch cellular infiltration. It is of interest thatexudative synovitis is now seen to occur relativelyearly in an involved knee joint while fibrosis is a laterfeature.Much interest has recently been aroused by the

report of Muirden and Mills (1971) who showed in aseries of 42 surgical synovectomies that extensivejoint damage as estimated by articular cartilageerosion and thinning or by pitted erosions alongbone-cartilage junctions was associated with ahistological picture ofmarked lining cell proliferationand a sparseness of subsynovial lymphocytes. Themethod of study carried out by these workers usingsurgical synovectomy specimens was different fromthe present arthroscopic study, and most of thepatients in the latter had little or no cartilage disease

++

t:

copyright. on N



j.com/

Ann R

heum D


ownloaded from

http://ard.bmj.com/

6 Annals of the Rheumatic Diseases

(Figs. 4 and 5). A direct comparison of results may the previous findings lack confirmation. Althoughnot therefore be justified, but in so far as no relation- slightly less articular cartilage is visualized arthro-ship, either direct or inverse, has been found between scopically than at arthrotomy, the correlationcartilage damage observed at arthroscopy and either coefficients (both less than 01) are so low that anylining cell proliferation or lymphocyte infiltration, correlation seems highly unlikely.

References

BRITTON, M. C., RUDDY, S., CORSON, J. M., SOSMAN, J. L., SCHUR, P. H., AND AUSTEN, K. F. (1970) 'Thecomplement system in rheumatoid synovitis', in 'Immune Reactions and Experimental Models in RheumaticDiseases. Proceedings of the 4th Canadian Conference on Research in the Rheumatic Diseases', Toronto,ed. D. Gordon, p. 40

CRUICKSHANK, B. (1952) Ann. rheum. Dis., 11, 137 (Interpretation of multiple biopsies of synovial tissue inrheumatic diseases)

HEDBERG, H. (1963) Acta rheum. scand., 9, 165 (Studies on the depressed haemolvtic complement activity ofsynovial fluid in adult rheumatoid arthritis)

KULKA, J. P., BOCKING, D., ROPES, M. W., AND BAUER, W. (1955) Arch. Path., 59, 129 (Early joint lesions ofrheumatoid arthritis)

JAYSON, M. I. V., AND HENDERSON, D. R. F. (1973) Rheumatol. and Rehab., 12, 195 (Arthroscopy in the diagnosisin inflammatory joint disease)

MUIRDEN, K. D., AND MILLS, K. W. (1971) Brit. med. J., 4, 219 (Do lymphocytes protect the rheumatoid joint ?)PALMER, D. G. (1967) Arthr. and Rheum., 10, 451 (Synovial villi: an examination of these structures within the

anterior compartment of the knee and metacarpo-phalangeal joints)PEKIN, T. J., AND ZVAIFLER, N. J. (1964) J. clin. Invest., 43, 1372 (Hemolytic complement in synovial fluid)ROBLES-GIL, J. (1973) Arthroscopy monographROPES, M. W., BENNETT, G. A., COBB, S., JACOX, R., AND JESSAR, R. A., (1958) Bull. rheum. Dis., 9, 175 (1958

Revision of diagnostic criteria for rheumatoid arthritis)ROSE, H. M., RAGAN, C., PEARCE, E., AND LIPMAN, M. 0. (1948) Proc. Soc. exp. Biol. (N. Y.), 68, 1 (Differential

agglutination of normal and sensitized sheep erythrocytes by sera of patients with rheumatoid arthritis)SCHUMACHER, H. R., AND KITRIDOU, R. C. (1972) Arthr. and Rheum., 15, 465. (Synovitis of recent onset)TAKEDA, S. (1960) J. Jap. Orth. Ass., 34, 373 (Relationship between arthroscopic and histological observations for

the synovial membrane)WATANABE, M., TAKEDA, S., AND IKEUCHI, H. (1969) In 'Atlas of Arthroscopy', 2nd ed. Springer, Berlin

copyright. on N



j.com/

Ann R

heum D


ownloaded from

http://ard.bmj.com/

rheumatoid synovitis and jointdisease

Documents