British Journal of Rheumatology 1986;25:342-344 PEPTIC ULCER IN RHEUMATOID ARTHRITIS- INTRIIC OR RELATED TO DRUG THERAPY? BY D. E. MALONE, P. A. McCORMICK, L. DALY, B. JONES, A. LONG, B. BRESNIHAN, J. MOLONY AND D. P. O'DONOGHUE Departments of Gastroenterology and Rheumatology, St. Vincent's Hospital, and Department of Community Medicine and Epidemiology, University College Dublin, Eire SUMMARY The reported incidence of peptic ulcer disease (PUD) in patients with rheumatoid arthritis (RA) is higher than that of the general population. Unusual susceptibility to PUD in RA, independent of therapy, has been suggested. To compare RA patients with others who had similar drug exposure but no known predisposition to PUD, 120 patients hospitalized for treatment of severe arthritis (65 with RA, 55 with osteoarthritis) were assessed by questionnaire for PUD history, drug history and other relevant variables. The relationship of PUD to sex distribution, smoking, alcohol consumption and anti-inflammatory therapy followed expected patterns. We found high but similar PUD rates in RA and osteoarthritis (OA) patients (RA 15%, O A 18%). This suggests that a common factor (probably drugs) is responsible. We feel that the documented high incidence of PUD in RA is most probably related to drug therapy. Available methods cannot determine if PUD ever occurs as a primary manifestation of RA. KEY WORDS: Arthritis, Rheumatoid, Peptic ulcer. THE incidence of peptic ulcer disease (PUD) in patients with rheumatoid arthritis (RA) has been shown to be three to four times greater than in the general population [1]. It has been suggested that there is unusual susceptibility to PUD in RA, independent of the effects of therapy. To examine this hypothesis we performed a retrospective study comparing RA patients with osteoarthritis (OA) controls who could be expected to have similar drug exposure but no known predisposition to PUD. All patients were interviewed for history of PUD, drug history and other relevant variables. Should a non-iatrogenic association between PUD and RA exist, one would expect tofindan excess of PUD in RA patients. PATIENTS AND METHODS One hundred and twenty patients (65 with RA, 55 with OA) admitted to hospital for rheumatological assessment, rehabilitation or orthopaedic procedures were interviewed by questionnaire. Age, sex, disease duration, disease severity, treatment with nonsteroidal anti-inflammatory agents (AIDs), aspirin or corticosteroids, alcohol and cigarette consumption were recorded. PUD was defined as definite Submitted 27 June 1985; revised version accepted 1 March 1986. Address correspondence to Dr. D. E. Malonc, Department of Gastroenterology, St. Vincent's Hospital, Elm Park, Dublin 4, Eire 342 (previous gastric surgery or verifiable radiological or endoscopic evidence of PUD) or probable (compatible clinical features and history of radiological or endoscopic diagnosis). PUD occurring at any time in a patient's life was included. The activity of the arthritis was classified as mild, moderate or severe by assessing the degree of disability. Statistical analysis employed Student's t test and the chi squared test (or Fisher's exact test in 2x2 tables where expected numbers fell below 5). A logistic regression was also employed to examine the joint effect of a number of variables on the occurrence of PUD [2]. A 5% two-tailed level of statistical significance was taken. RESULTS Table I shows the characteristics of both groups. The RA patients were significantly younger but were not significantly different from the OA patients in regard to sex distribution. Disease duration and severity were similar. More RA patients had been exposed to AIDs (p < 0.05), aspirin (p < 0.05) and steroid therapy (p < 0.001). More detailed drug histories could not be obtained. The groups contain comparable proportions of smokers and teetotallers. No alcoholics or unusually heavy drinkers were identified. Table II shows that the percentages of RA and OA patients with PUD were similar. Table III compares the 20 patients who had PUD with the rest of the sample population.
MALONE ETAL.: PEPTIC ULCER IN RA 343 TABLE I CHARACTERISTICS OF RA AND OA GROUPS Mean age (SD) years Male sex Current smokers Non-drinkers Disease duration (mean) years Drug therapy Aspirin AID Steroids RA (n = 65) 56.4 (13) 21 25 39 6.6 18 63 32 = no significant difference. OA (n = 55) 65.9 (10) 27 24 31 5.7 6 48 0 Significance P <0.001 <0.001 Smoking was more prevalent in those with PUD (p < 0.05). No significant differences were found between these two groups for any other factor. When the joint effect on PUD of all factors was examined using logistic regression, smoking remained the only significant variable. DISCUSSION A three- to four-fold increase over the expected incidence of PUD has been found in RA [1], It has been suggested that this is because of an unusual susceptibility [3,4]. Whether this is a primary manifestation of the disease or is related to therapy has not been established. The purpose of our study was to test the hypothesis that there is an intrinsic predisposition to PUD in RA by comparing RA patients with a similarly-treated group. The surprising feature of our results is that the RA and OA groups have essentially the same PUD rate. Lax diagnostic criteria may produce spuriously highfigures [5] but our chosen criteria were rigid. When we examined the relationship of PUD to smoking, AIDs (including aspirin), steroid therapy, sex distribution and alcohol consumption we found that our results conformed to established patterns [6-9]. We found that 15% of RA patients and 18% of OA patients had at some time suffered from peptic ulcer. An exactly comparable rate for the general population in Ireland is not available. True rates can be difficult to define [5]. The PUD rate in the general population is usually estimated at about 5-10% [8]. It has been calculated that in England 5.8% of men and 1.9% of women may at some time suffer from peptic ulcer [10]. These figures suggest that our study shows an elevated PUD rate in RA due to a factor common to RA and OA patients. We found no evidence of an intrinsic predisposition to PUD in RA. Our study was not designed to examine the complex relationships between peptic ulcers and drug therapy. However, it has been shown that a similar incidence of mucosal abnormality occurs in OA and RA patients given AIDs [11] and RA patients are thought to be unusually prone to 'steroid ulcers' [4]. We believe it is reasonable to propose that AIDs are the 'common factor' between our two groups of patients and that PUD in RA is either secondary to, or becomes symptomatic during anti-inflammatory therapy. To show whether or not PUD occurs as a primary manifestation of RA would need a very large prospective study controlled for the effects of drugs. The crippling sequelae of inadequately treated RA would make this unethical. A properly-controlled retrospective study is impossible because most patients have been on many drugs and detailed drug histories cannot be obtained. A significant association between RA and a genetic marker of PUD (e.g. family history, pepsinogen levels) would suggest a non-iatrogenic association between the two diseases. However, since patients with any given disease are more likely to be aware of a positive family history than patients who do not have the disease in question [12], family history may be a difficult factor to evaluate. Even if it were proven that RA and 'genetic' PUD are associated, this would not rule out PUD as a primary manifestation of RA or make allowance for PUD becoming symptomatic on therapy. For this reason there seems little point in embarking on such a study. We suggest that the documented association between RA and an increased lifetime incidence of PUD [1,3,4] is related to drug therapy. It does not appear possible, using available methods, to show if PUD ever occurs as a primary manifestation of RA. TABLE II PEPTIC ULCER DISEASE IN RA AND OA GROUPS Peptic ulcer disease RA OA Definite Probable Absent Total X 2 = 1.007; df = 2;. 5 (7.7%} 5 (7.7%) 55 (84.6%) 3 (5.5%) 7 (12.7%) 45 (81.8%) 65 (100%) 55 (100%)
344 BRITISH JOURNAL OF RHEUMATOLOGY VOL. XXV NO. 4 TABLE III CHARACTERISTICS OF PATIENTS WITH PEPTIC ULCER DISEASE (DEFINITE AND PROBABLE) AND THOSE WITHOUT Mean age (SD) years Male sex Current smokers Non-drinkers Drug therapy Aspirin AIDs Steroids Rheumatoid arthritis PUD (n = 20) 60.6(12.5) 7 13 12 6 17 5 10 Non-PUD (n = 100) 60.8(13) 41 36 58 18 94 27 55 Significance P ACKNOWLEDGEMENTS We should like to record our thanks to the staff and patients of St. Joseph's Rehabilitation Unit, Harold's Cross and St. Mary's Orthopaedic Hospital, Cappagh, Dublin. REFERENCES 1. Bowen T, Mayne JG, Cain JC, Bartholomew L. Peptic ulcer in rheumatoid arthritis and relationship to steroid treatment. Proc Mayo Clinic 1960;35:537^t5. 2. Bourke GJ, Daly LE, McGilvray J. Interpretation and uses of medical statistics. 3rd ed. Oxford: Blackwell Scientific Publications, 1985;152^. 3. O'Brien WM. Ulcerogenic properties of antiinflammatory drugs. Pharmacology 1982; 25(suppl. l):9-lk 4. Spiro HM. Is the steroid ulcer a myth? N EnglJ Med 1983;309:45-7. 5. Donaldson RM. Factors complicating observed associations between peptic ulcers and other diseases. Gastroenterology 1975; 68:1608-14. 6. Soil AH, Isenberg I. Duodenal ulcer disease. In: Sleisenger MH, Fordtran JS, eds. Gastrointestinal disease (palhophysiology, diagnosis and management). 3rd ed. Philadelphia PA: W B Saunders, 1983;628-30. 7. Richardson CT. Gastric ulcer. In: Sleisenger MH, Fordtran JS, eds. Gastrointestinal disease (pathophysiology, diagnosis and management). 3rd ed. Philadelphia PA: W B Saunders. 1983;686-7. 8. Spiro HM, Milles SS. Clinical and physiologic implications of the steroid induced peptic ulcer. N Engl J Med 1960;263:286-94. 9. Atwater EC, Mongan ES, Wieche DR, Jacox RF. Peptic ulcer and rheumatoid arthritis. Arch Intern Med 1965;115:184-9. 10. Doll R, Jones FA, Buckatzsch M. Occupational factors in aetiology of gastric and duodenal ulcers with estimate of their incidence in the general population. London: His Majesty's Stationery Office, 1950 (Med Research Council, Spec. Rep. Series, No. 276). 11. Caruso I, Bianchi-Porro G. Gastroscopic evaluation of anti-inflammatory agents. Br Med J 1980;280:75-8. 12. Schull WJ, Cobb S. The intrafamilial transmission of rheumatoid arthritis. J Chron Dis 1969;22:217-22.
FILL THE KNOWLEDGE SUBSCRIBE TO THE JOURNAL OF RHEUMATOLOGY The Journal for Clinicians - Informative articles written by world renowned contributors. THE ONE THAT HAS IT ALL Reduced rates (or Trainees: iv oo CCOCDpamirti Idler from Regain-rain: v«>ou Bact fnuei are tvaoume il *2D4» per baac when In wock A nafor tupptemew iwhca panohcd I u mctwlrd with ucta *flh tbepncb Wca^e ewer ow tti^**.tiplnm t" Mail completed form lo Tbe JoanuU of Rbeamalok^ 250BloorS(.EnStdte401 Toroalo. OBlmrto Canxdj M4W3P2 NAME A DOR ESS CITY PROV ISCE COUNTRV tt>stalcode AMOUNT ENCLOSED» BILL ME