HLA ANTIGENS IN PALINDROMIC RHEUMATISM AND PALINDROMIC ONSET RHEUMATOID ARTHRITIS

British Journal of Rheumatology 1986;25:345-348 HLA ANTIGENS IN PALINDROMIC RHEUMATISM AND PALINDROMIC ONSET RHEUMATOID ARTHRITIS BY L. R. FISHER 1, A. KIRK 2, J. AWAD 3, H. FESTENSTEIN 3, A. ALONSO 3, J. D. PERRY 2 AND M. SHIPLEY 1 'Department of Rheumatology, the Middlesex Hospital, London W1P 9PG, ^Department of Rheumatology, the London Hospital, London El IBB, and ^Department of Immunology, the London Hospital Medical School, London El, UK SUMMARY Fifty patients who presented with typical palindromic rheumatism of at least 6 months' duration were tissue-typed for HLA A, B, C antigens. DR typing was also performed but was not possible for technical reasons in three patients. Twenty-three patients who had progressed to definite or classical rheumatoid arthritis (RA) after a mean interval of 5 years were compared with 20 patients whose palindromic attacks had persisted over a similar period. Both groups showed a significantly higher frequency of DR4 antigen than a control population. The RA group also showed an increased frequency of DR1. There was no significant difference in the frequency of DR4 or of any other DR antigen between the two patient groups. The frequency of B27 antigen was significantly higher in the palindromic group compared with the controls. It is suggested that although DR4 may be associated with a tendency to inflammatory joint problems, environmental or other unrelated genetic factors may be more important in determining the progression of palindromic rheumatism to RA. KEY WORDS: HLA antigens, Palindromic rheumatism, Palindromic onset rheumatoid arthritis. IN 1941, Hench and Rosenberg [1] gave the name 'palindromic rheumatism' to an intermittent, unexplained and recurrent form of acute synovitis. The characteristic features of this condition include recurrent attacks of joint pain and swelling affecting one or more joints and lasting from a few hours to 7 days with variable symptom-free intervals. A similar clinical pattern can occur in systemic lupus erythematosus, crystal deposition disease, sarcoidosis, viral infections or seronegative arthropathy. In the absence of such denned disorders, palindromic rheumatism has a varied natural history; some remit, others continue to have intermittent attacks and some progress to typical rheumatoid arthritis (RA). There have been varying reports of the frequency of progression to RA. Hench and Rosenberg concluded that joint damage did not occur irrespective of the number of acute attacks [1]. Subsequent reports of patients followed up after means of 7 and 10 years from onset suggest that 40-50% of patients progress to typical RA, the majority within thefirst5 years [2-4]. Ansell and Bywaters have suggested that the syndrome will always progress to RA if patients are followed for long enough [5]. Studies attempting to identify factors which Submitted 22 January; revised version accepted 17 March 1986. Address correspondence to Dr. M. Shipley. 345 might predict progression to RA have suggested that presence of rheumatoid factor [3] or raised levels of Clq binding [6] may be of value. Serum complement and immune complex levels [2], analysis of synovial fluid and electron microscopy of synovial tissue during acute palindromic attacks [7] seem to have no such value. The aim of this collaborative study was to see whether genetic markers could differentiate patients with a palindromic onset of RA from those who remain persistently palindromic for 5 or more years. METHOD A group of 50 British caucasoid patients (29 female : 21 male) attending clinics at two hospitals were reviewed and typed for HLA A, B, C and DR antigens. All presented with typical attacks of palindromic rheumatism which had persisted for a minimum of 6 months. A full history and examination excluded individuals with psoriasis, ankylosing spondylitis, crystal deposition disease and systemic lupus erythematosus. Screening tests included a full blood count, ESR, serum uric acid, rheumatoid factor, antinuclear antibody and radiography of the hands and feet. Where appropriate, synovial fluid was examined for the presence of crystals. HLA A, B and C typing were performed using a modification of the standard NEH technique [8] testing for all recognized antigens. HLA DR

346 BRITISH JOURNAL OF RHEUMATOLOGY VOL. XXV NO. 4 No. DR1 DR2 DR3 DR4 DR5 DRw6 DR7 DRw8 TABLE I HLA-DR TYHNG OF PATIENTS AND CONTROLS Palindromic rheumatism 20 2 (10%) 3 7 12 (60%)* u " 2.53 1 4 4 0 Rheumatoid arthritist 20 7 (35%)* rr = 2.84 7 4 14 (70%)" rr - 3.84 2 0* 2 0 Controls 161 26 (16.1%) 51 51 59 (36.6%) 23 31 44 12 p < 0.05; " p = 0.005. rr = relative risk. t Three of the total 23 RA patients were not DR-typed for technical reasons. typing was performed using the double immunofluorescence technique and testing for DR 1-10 [9]. The frequencies of the HLA antigens of the patient groups were compared with those of a random normal British caucasoid control panel. The significance of differences in the frequencies of the HLA antigens of the patient groups when compared with each other and with the control panel and the relative risk statistics were tested by Haldane's modification [10] of Woolfs method [11]. RESULTS Seven patients with palindromic rheumatism of less than 5 years' duration were excluded from the main analysis as it has previously been reported that most patients with palindromic rheumatism who eventually progress to RA do so within the first 5 years [2-4]. There were thus 20 patients (15 female : 5 male) with persistent palindromic attacks over more than 5 years: two were seropositive for rheumatoid factor at titres of 1:20 and 1:40 and in a further two there was a positive family history of RA. Twenty-three patients (11 female : 12 male) with a palindromic onset had progressed to 'classical' or 'definite' RA (ARA criteria); 20 were seropositive for rheumatoid factor and 18 were erosive. In this latter group the mean duration from onset to the development of RA was 5 years (range 1-10 years). Table I shows that there was a significant increase in the frequency of HLA DR4 both in patients with persistent palindromic rheumatism (12/20 = 60% +ve, p = 0.042: relative risk 2.53) and in those developing RA (14/20 = 70% + ve, p = 0.005: relative risk 3.84) when compared with the control population (36.6% DR4 +ve). There was no statistical difference in the frequency of DR4 between the two patient groups. Three RA patients were not tissue-typed for technical reasons. In the palindromic group the two seropositive patients and the two seronegative patients with a family history of RA were all DR4 positive. There was also a significantly higher frequency of DR1 in RA patients (7/20 = 35% DR1 +ve, p = 0.03, relative risk 2.84) compared with the controls (16.1% DR1 +ve), but not the palindromic group (2/20 = 10% DR1 +ve). Seven of 20 palindromic patients were positive for HLA B27 compared with three of 23 RA patients. The frequency of B27 in the palindromic group differed significantly from that in the control group (p = 0.008: relative risk 3.71). In the palindromic onset RA patients the frequency of HLA B27 was the same as in the controls (Table II). In the patients with palindromic rheumatism, the frequency of HLA DR4 was 30/47 (63.8%). This differed significantly from the controls (p = 0.0015). The p values have been quoted without correction for the number of observations. In the case of DR1 and DR4, correction is not necessary because they have already been reported as associations. In the case of B27, the corrected value would no longer be significant. The p value for the reduced frequency of DRw6 in the RA patients is also not significant after correction. DISCUSSION Palindromic rheumatism is an uncommon condition which may progress from intermittent self-limiting episodes of joint inflammation to a disease indistinguishable from RA. Investigations of palindromic rheumatism may provide insight into genetic or environmental factors involved in the development of chronic arthritis. This study attempted to determine whether there was an association between the major histocompatibility complex (MHC) and the pro- No. B27 TABLE II FREQUENCY OF B27 m PATIENTS AND CONTROLS Palindromic rheumatism 20 7 (35%, p = 0.008) rr = 3.71 Rheumatoid arthritis 23 3 (13%, NS) Controls 187 24 (13%)

FISHER ETAL.: HLA ANTIGENS IN PALINDROMIC RHEUMATISM 347 gression of palindromic rheumatism to chronic arthritis. A close relationship has been demonstrated between RA and the MHC in hospital-based studies [12,13]. It has been suggested that the presence of HLA DR4 and Dw4 both predispose to the development of RA and are associated with more severe erosive disease (14-17] although this is disputed [18]. Furthermore, some have demonstrated a close association between HLA DR4 and the presence of rheumatoid factor [17,19], but others have not [16]. In our study there was an increased frequency of HLA DR4 in all patients with palindromic rheumatism of more than 6 months' duration irrespective of whether they progressed to chronic arthritis. This suggests that DR4 is either a marker for intermittent arthritis and that other factors are involved in disease progression, or that most patients will eventually develop progressive disease if followed up for longer than 5 years. Seropositivity appeared to be a better guide to the development of persistent joint disease, and was independent of HLA DR4. An increased frequency of DR1 was first reported in Asian Indian RA patients [20] and has also been found in a British study [21]. The increased frequency of DR1 in our patients who progressed from palindromic rheumatism to RA could indicate a difference between the two patient groups. However, an increased frequency of DR1 is not found invariably in RA [18] and the difference in the DR1 frequency between the two patient groups examined in this study was not statistically significant (p = 0.06). There has been one other study investigating the relationship between the MHC and palindromic rheumatism in which no significant increase in frequency of HLA DR4 was seen [16]. Patient selection differed from ours by the inclusion of subjects with intermittent acute arthritis on a background of mild but constant symptoms. This study did not demonstrate any increase of HLA-B27 in the palindromic group. Dahlquist et al. [22] suggested that HLA-B27 may be associated with more severe and nodular RA. Our results suggested that the presence of this antigen with DR4 confers some degree of protection from the development of RA, or at least may delay its development. There was no clinical or radiological evidence that any of the patients suffered from a B27-related spondarthropathy although sacro-iliac radiographs were not available in all cases. In conclusion, our study demonstrated that the progression of acute palindromic attacks of joint inflammation to chronic arthritis is generally associated with seropositivity but is independent of HLA DR4. While genetic factors may have a role in the onset and course of palindromic rheumatism, as is implied by the relationships demonstrated with DR4 and B27, other, perhaps environmental factors are also likely to be important. REFERENCES 1. Hench PS, Rosenberg EF. Palindromic rheumatism: a 'new' oft recurring disease of joints apparently producing no articular residues. Proc Mayo Clinic 1941;16:8O8-15. 2. Wajed MA, Brown DL, Currey HLF. Palindromic rheumatism clinical and serum complement study. Ann Rheum Dis 1977;36: 56-61. 3. Mattingly S, Jones DW, Robinson WM, Williams RA, Dunn EC. Palindromic rheumatism. J R Coll Physicians Lond 1981 ;15: 119-23. 4. Williams MH, Sheldon PJHS, Torrigiani G, Eisen V, Mattingly S. Palindromic rheumatism clinical and immunological studies. Ann Rheum Dis 1971;30:375-9. 5. Ansell BM, Bywaters EGL. Palindromic rheumatism [Letter]. Ann Rheum Dis 1959;18: 331. 6. Thompson B, Mohammed I, Holborow EJ, Currey HLF. Palindromic rheumatism: II. Failure to detect circulating immune complexes during acute episodes. Ann Rheum Dis 1979;38:329-31. 7. Schumacher R. Palindromic onset of rheumatoid arthritis clinical, synovial fluid and biopsy study. Arthritis Rheum 1982;25: 361-9. 8. Festenstein H, Adams E, Burke J, Oliver RTD, Sacks JA, Wolf E. Histocompatibility testing. EdJ Dansoct 1972;175-8. 9. Van Rood JJ, Van Leeuwer A, Phoem JS. Simultaneous detection of two cell populations by two colour fluorescence and application to the recognition of B-cell determinants. Nature 1976;262:795-7. 10. Haldane JBS. The estimation and significance of the logarithm of a ratio of frequencies. Ann Hum Genet 1955;20:309-ll. 11. Woolf B. On estimating relation between blood group and disease. Ann Hum Genet 1955;19:251-3. 12. Stastny P. Association of the B cell alloantigen DRw4 with rheumatoid arthritis. N Engl J Med 1978;298:869-71.

348 BRITISH JOURNAL OF RHEUMATOLOGY VOL. XXV NO. 4 13. Panayi GS, Wooley P, Batchelor JR. Genetic basis of rheumatoid disease. Br MedJ 1978; 2:214. 14. Young A, Jaraquemada J, Awad J, et al. Association of HLA DR4/DW4 with radiologic changes in a prospective study of patients with RA. Arthritis Rheum 1984;27: 20-5. 15. Walker DJ, Griffiths M, Dewar P, et al. Association of MHC antigens with susceptibility to and severity of rheumatoid arthritis in multicase families. Ann Rheum Dis 1985 ;44:519-25. 16. Gran JT, Husby G, Thorsby E. HLA antigens in palindromic rheumatism, non-erosive rheumatoid arthritis and classical rheumatoid arthritis. J Rheumatol 1984;11:136-40. 17. Jaraquemada D, Oilier W, Awad J, et al. HLA and RA; a combined analysis of 440 British patients. (In press.) 18. Walton K, Dyer PA, Grennan DM, Haeney M, Harris R. Clinical features, autoantibodies and HLA-DR antigens in rheumatoid arthritis. J Rheumatol 1985;12: 223-6. 19. Bardin T, Legrand L, Naveau B, et al. HLA antigens and seronegative rheumatoid arthritis. Ann Rheum Dis 1985;44:50-3. 20. NicholFE, WoodrowJC. HLA-DR antigens in Indian patients with rheumatoid arthritis. Lancet l981;i:220-l. 21. Oilier W, Silman A, Gosnell N, etal. HLA and rheumatoid arthritis: an analysis of multicase families. Disease Markers 1986;4:(In press). 22. Dahlquist SR, Nordmark LG, Bjelle A. HLA B27 and involvement of sacroiliac joints in RA. / Rheumatol 1984;ll:27-32.

THE GENTLE ART OFARTHRITIC CONTROL A unique mode of action. Acceptability in arthritis. Mobility and relief of pain. Prescribing information can be found overleaf

azapropazone dihydrate Prescribing iiu otiiu tion UKS Osteoarthrids, rheumatoid arthritis, ankytosing spondylitis and painful musculo-skeleta] conditions. Acute and chronic gout Douge 1200mg daoy in divided doses, in cute gout 24O0mg m divided doses during the first 24 hours reducing to 1200mg until symptoms have disappeared. To control nyperurkaemla. 1200mg daily in divided doses. Elderly patients, 600mg daffy, m divided doses, for ajl indications except acute gout Contra-indications Rheumox should not be given to patients receiving an anti-coagulant phenytoin or hypogh/caemk agent nor to patients with a history of peptic ukeratfon or wood dyscrasias. Precautions Avoid use in pregnancy wjten possible. Use with caution in renal insufficiency. Side EBectsSEnrash (sometimesphotoallergic), gastrointestinal effects including haemorrhage, flukj retention and angio-neurofac oedema. Very rarely, arveobhs and haemoryrk anaemia have been reported Product Licence Nranben Rheumcw 600 0100/0059; Wwumox 0100/0037. Bask NHS Price Rheumox capsules 100 13 94; Rheumox 600 tablets 100 2738. Rheumox is a registered trade mark. Further information and full data sheet available from: /1-H-ROBINS A H Robins Co. limited, Langhurst, Horcham, West Sussex RH155QP. ARTINPRJNT A H Robins has produced a full colour print of the dancer depicted in the advertisement. If you are interested m obtaining a copy of this print please send your name and address to 'Rheumox pnnf, A H Robins Co. Limited, Unghurst, Horsham. West Sunex RH13 5QP.