Peptic Ulcers in Rheumatoid Patients Receiving Corticosteroid Therapy

Abstract

This paper represents a portion of a clinical and experimental study (13) to determine the frequency with which peptic ulcers occur in patients with rheumatoid arthritis under treatment with various corticosteroid drugs and to investigate the effect of corticosteroids on gastric secretions. The present report is concerned only with the incidence of ulcers, their distribution, location, and roentgen appearance. Shortly after the introduction of corticosteroids for clinical use in 1950, reports began to appear on the increased occurrence of peptic ulcers in patients treated by these substances. Howell and Ragen (6) reported 18 ulcers in 68 patients receiving cortisone for five to thirty months. Other investigators placed the incidence anywhere from 5 per cent to as high as 37 per cent (7–10). Henderson (4) found that only 5.3 per cent of 1,440 patients treated with cortisone had ulcers. In another series, of 446 cases analyzed by the American Rheumatism Association (5), the figure was 6.6 per cent, which is approximately the incidence of peptic ulcers in the “normal” population. Sandweiss (3) experimented on Mann-Williamson dogs, administering corticotropin and cortisone in dosages comparable to those used in man and observed no peptic ulcer formation in the treated group. Whether there is a specific alteration in the composition of gastric secretions, a lower tissue resistance to the action of gastric juice, or other factors responsible for an increase in gastric ulcers has not been definitely determined. Clinical Study Over a period of twelve months, x-ray examination of the upper gastrointestinal tract was obtained in practically all of our patients with rheumatoid arthritis who had been maintained on corticosteroid therapy for periods of six months or more. These patients, numbering 114, were about equally divided between those attending the arthritis clinic of the Hospital for Special Surgery and private patients of one of the authors (W. H. K.). Thirty-five of the total, or 31 per cent, had peptic ulcers. Of a small control series of 33 patients with rheumatoid arthritis who were not receiving corticosteroid therapy, 3, or 9 per cent, had ulcers. Every effort was made to examine all the patients treated with corticosteroids, and gastrointestinal symptoms were not used as criteria for selection. Of the ulcers found, 4 represented reactivation of previously known ulcers and 31 were new. Ulcers developed in 6 of 22 men and in 29 of 95 women, a relatively equal incidence for the sexes, which is in sharp contrast to the far greater frequency of spontaneous peptic ulcers in males. Ulcers occurred as frequently in patients whose arthritis had existed less than ten years as in those with a longer history. A greater ratio of peptic ulcer was found in patients having moderately severe to severe rheumatoid arthritis (classified according to American Rheumatism Association criteria) than in those with milder disease.