General Observations. All of the patients who were studied were seen within 5 days of the onset of the acute attack and the development of symptoms.

Active Treatment. Constitutional disturbances, which subsided after 2 to 5 days of treatments, were observed in nine members of each group. Six patients in the sulfasalazine and prednisolone group and four patients in each of the other groups required oral administration of ferrous sulfate to treat mild hypochromic anemia. During the first week after admission to the hospital, the conditions of eight patients in the sulfasalazine and prodnisolone group and six patients in each of the other groups deteriorated. Diarrhea increased to more than six times a day and was associated with moderate amounts of blood, but constitutional disturbances were not observed in any patient. These patients were treated by increased their prednisolone to 15 mg four times a day. However, two patients in each group continued to deteriorate and had constitutional disturbances; therefore, they underwent surgery.

Four patients in the group that was given sulfasalazine and prednisolone and six patients in each of the other groups reported drug-related side effects that consisted of headache, nausea, and general malaise. These effects, however, were only severe enough to warrant cessation of the treatment in three patients who were given allopurinol, and in two patients who were give DMSO experienced intolerance to the therapeutic regimen. Three patients in the allopurinol group (6%) reported mild allergic reactions, which were manifested by itching and or erythema and which did not require any treatment or necessitate termination of the allopurinol therapy.

Prophylactic treatment. During this period two patients who were given sulfasalazine alone had severe relapses that required surgery. Intolerance to the therapeutic regimen was reported by one of the patients who had been given DMSO. Adverse effects similar to those noted during active treatment were experienced by four members of each of the groups; however, they were only significant and led to withdrawal from the study in one of the patients who was given sulfasalazine alone. Four patients in the allopurinol group (9%) reported mild allergic reactions that were similar to those described above and that did not require any specific action or cessation of treatment.

(Table IV).

The remission rate after 2 weeks of active treatment was significantly (p < 0.01) higher in the allopurinol (40 patients, 87%) and DMSO (38 patients, 84%) groups than that in the sulfasalazine and prednisolone group (23 patients, 51%). In the remaining cases orally administered prenisolone was increased to 15 mg four times a day. All patients in the allopurinol and DMSO groups became free of symptoms with 3 weeks of treatment, a result that was superior (p < 0.01) to that of the sulfasalazine and prednisolone group in which only 66% of patients were free of symptoms at this stage. Remission in all members of the latter group was achieved after 4 weeks of treatment. In every patient the clinical recovery from an attack was associated with sigmoidoscopic appearances of remission and the histologic observation of an inflammatory bowel disease in an inactive phase. Thus the remission rate on the basis of symptoms coincided with that which was determined histologically.

During prophylactic treatment, 10 patients (25%) in the sulfasalazine alone group and two patients (5%) in each of the other groups had recurrent attacks of moderate proctosigmoiditis. All of these relapses were symptomatic and were confirmed histologically. There were no cases of histologically active disease in a patient who was free of symptoms (i.e., silent relapses). The results illustrate that addition of allopurinol of DMSO to sulfasalazine significantly (p < 0.5) enhances its protective effects. The colonoscopies and barium enema, which were performed during the maintenance therapy year, demonstrated no change in the extent of the disease.

A series of Cox proportional hazard models was fitted with all factors other than treatment with radical scavengers as covariates to obtain a group of patients and conditions that independently and significantly influence the rate of disease relapse and remission. Allopurinol or DMSO was then added to treatment as separate covariates. Increasing duration of symptoms, anemia, constitutional disturbances, and increasing number of daily bowel movements all had a significantly (p < 0,001) detrimental effect on the rate of disease relapse and remission at the 5% level. When these and all the other nonsignificant variables were allowed for, treatment with allopurinol or DMSO continued to exert a significant beneficial effect on the rate of remission after an acute attack (p < 0.01) and on reducing the maintance treatment (p < 0.05).

The influence of method of analysis on the remission and relapse rates of the disease was studied. Intention-to-treat analyses were carried out to determine what might have happened if all patients had been suitable for evaluation. This required postulating that some patients would have gone into remission and others into relapse at various time periods.

When all of the excluded patients were assumed to have recovered from their colitis attacks or when an attack persisted after 2 to 3 weeks of active treatment, addition of allopurinol or DMSO showed a significant (p < 0.01) advantage on the control of the disease and its remission rate afforded by sulfasalazine with prednisolone. When none of the excluded patients who were treated with allopurinol or DMSO had their disease controlled and the acute attack persisted after 3 weeks of treatment, whereas the excluded patients in the sulfasalazine and prednisolone group had the acute attack under control and went into remission within 2 or 3 weeks of treatment, allopurinol and DMSO continued to exert significant (p < 0.01 and p < 0.05, respectively) therapeutic advantages when added to the sulfasalazine and prednisolone regimen.

During prophylactic treatment, the significant (p < 0.05) advantage in terms of the protection against relapses that was afforded by addition of allopurinol or DMSO to sulfasalazine remained unchanged when all of the excluded patients were assumed to have relapsed or to have remained in remission. However, when all of the patients who were excluded from the allopurinol and DMSO groups were assumed to have relapsed, whereas those who were excluded from the sulfasolazine group remained in remission, neither allopurinol nor DMSO provided therapeutic advantages to prophylaxis with sulfasalazine alone.

One of the central issues in designing any clinical trial is the choice of the establishment at which the study would be carried out so that sufficient numbers of patients can be recruited to enable meaningful analysis of the results. The present investigation was conducted at the Medical City in Baghdad, which is the largest medical center in the country. It serves a population of well over one and a half million, hosts the various departments of Baghdad's University School of Medicine, and acts as the major referral hospital for the whole of Iraq. Consequently, a sizeable number of patients was seen during the study.

Corticosteroids and sulfasalazine are the only drugs that have been shown to shorten an acute attack of ulcerative colitis. ⁵ Corticosteroids achieve a rapid remission more frequently than sulfasalazine; however, the combination of these agents is widely used for treatment of colitis attacks. ⁵ This combination was employed in the present study to achieve the highest therapeutic efficacy against an acute attack and to determine whether addition of radical scavengers provides any further advantages.

Allopurinol is a potent scavenger of hydroxyl radicals^l3 and an inhibitor of the enzyme xanthine oxidase, which is responsible for the formation of superoxide radicals; DMSO is a scavenger of hydroxyl radicals. ⁶ Addition of these agents to the prednisolone/sulfasalazine regimen significantly enhanced control of the colonic inflammation produced by ulcerative colitis and thus afforded an earlier recovery from an acute attack (Table III).

Once patients have ulcerative colitis, it is virtually certain that they will experience further attacks. Sulfasalazine was the first drug to be shown to bring about a major reduction in the recurrence rate of colitis attacks. ¹⁴ The present study illustrates that the maintenance efficacy of sulfasalazine is significantly intensified by the addition of radical scavengers (Table IV).

It thus appears that scavengers of oxygen-derived free radicals are useful in the management of ulcerative colitis. Their addition to the corticosteroid/sulfasalazine regimen significantly reduces the amount of proctocolitis and causes rapid recovery from an acute attack. These scavengers also potentiate the prophylactic efficacy of sulfasalazine. It might be argued that allopurinol and DMSO produced the results that were noted in the present investigation in ways other than influencing oxygen-derived free radicals. The similarity in efficacy between these two agents and the fact that the only known action they share is scavenging oxygen-derived free radicals suggests that the results that were attributed to them in this study were achieved by scavenging oxygen-derived free radicals. It is therefore attractive to speculate that these radicals are implicated in the mechanism of development and recurrence of ulcerative colitis. On the basis of these considerations, it may be proposed that the etiologic factor or factors behind ulcerative colitis generate oxygen-derived free radicals, which then directly participate in the production of the histologic features that are typical of this disease.

Although the etiology of ulcerative colitis remains uncertain,^15, 16 available data suggest that phagocytic cells are the source of oxygen-derived free radicals in the bowel wall. Various types of cells that are derived from the immune system are prominent in the intestinal mucosa of patients with ulcerative colitis. ^{7, 15, 17} Such cells include T cells, activated T killer cells, macrophages, monocytes, polymorphonuclear leukocytes, basophils, and eosinophils. ¹⁸ Since the work of Metchnikoff,¹⁹ neutrophils and macrophages have been implicated in the pathogenesis of tissue injury at inflammatory sites. These phagocytes are capable of generating highly reactive metabolites of oxygen such as hydrogen peroxide and hypochlorite, as well as the superoxide and hydroxyl radicals. ^20-22 Although such products of oxygen are involved in bacterial killing, their formation outside the cell may also cause tissue damage. ²¹ Neutrophils and macrophages that migrate into the tissues originate from the circulating pool of neutrophils and monocytes, and recent evidence¹⁷ has shown that toxic oxygen metabolite production by circulating monocytes is significantly increased in patients with ulcerative colitis. It thus appears that activated phagocytic cells in the bowel wall produce toxic metabolites of oxygen coupled with the release of granular enzymes and the formation of highly reactive metabolites such as hypochlorite. In addition, patients with ulcerative colitis have other poorly defined immunologic abnormalities that may alter the gastrointestinal tract and make it a more prooxidative environment. ²³

Allopurinol, a xanthine oxidase inhibitor, is effective in ischemia-reperfusion models because ischemia leads to conversion of the native xanthine dehydrogenase to the superoxide generator xanthine oxidase. This transformation has not been shown to occur under inflammatory bowel conditions; it is therefore not clear that allopurinol exerts its actions via the inhibition of xanthine oxidase. DMSO is a well-known antiinflammatory agent that inhibits neutrophil function. ²⁴ Consequently, this agent could have employed some of its beneficial effects in the treatment of ulcerative colitis by inhibiting neutrophil-induced (and perhaps monocyte-induced as well) lysosomal enzyme release, which is known to perpetuate intestinal inflammation. ^{15, 16, 18, 23} The possibility also exists that DMSO may have prevented the inflammatory cells from generating oxygen-derived free radicals, thereby broadening the scope of its overall anti-free radical activity.

Neither of the radical scavengers that were employed in this study created additional inconveniences or side effects relative to those that were seen with sulfasalazine alone (Table I), which demonstrates their suitability for long-term use. Although DMSO can be smelled on the breath, it was never a serious inconvenience to any of the patients and cannot be considered a source of bias in favor of a particular group because the assessment parameters were objective ones.

The compliance of patients with their regimens was carefully monitored and any infringements led to exclusion. This compliance was similar among the study groups (Table I), and therefore the results cannot be ascribed to differences in patient compliance.

Some patients were excluded from this study because of drug intolerance and others because of side effects (Table I). It is well known that some 10% to 15% of patients cannot tolerate sulfasalazine in therapeutic doses and that many of the common side effects of this agent are dose-related. ²⁵ The use of mesalazine (5-aminosalicylic acid) as an alternative to sulfasalazine might have reduced the incidence of drug intolerance and the side effects that were encountered in this study.

It is not ethical to deny patients treatment of proven efficacy, and to allow moderate proctosigmoidal ulcerative colitis to go untreated may incur risks. For these reasons, it was not possible to have a placebo control group or to give patients radical scavengers without any additional treatment of proven value.

In conclusion, these findings suggest that oxygen-derived free radicals are directly implicated in the mechanism of ulcerative colitis and that removing them is of great help in the treatment of and protection against colitis.

Further studies are needed to determine:

1. the role of scavenging oxygen-derived free radicals in the management of severe or total colitis,
2. whether parenteral administration of radical scavengers is as efficacious as the oral route in the management of ulcerative colitis,
3. the role in this management of scavenging oxygen-derived free radicals, by retention enemas,
4. whether maintenance therapy of ulcerative colitis with radical scavengers influences the incidence of mucosal dysplasia and neoplasia that are produced by colitis, and
5. whether the number of daily doses of radical scavengers could be reduced without compromising efficacy.

[Author's note] I thank Dr. S. H. Alwash for giving me the opportunity to undertake this work at the Medical City. My thanks are also due to Mrs. Moira Caimey and to Mrs. Julia Gaskill for their skillful secretarial assistance.

1. Svartz N. The pathogenesis and treatment of ulcerative colitis, Acta Med Scand 1951;141:172-9.
2. Lennard-Jones JE. Ulcerative colitis. In: Hadfield J, Hobsley M, eds. Current surgical practice. London: Arnold, 1976;1:185-91.
3. Kirsner JB. Inflammatory bowel disease: considerations of etiology and pathogenesis. Am J Gastroenterol 1978;69:253-71.
4. Truelove SC, Witts LJ. Cortisone in ulcerative colitis: preliminary report on a therapeutic trial. Br Med J 1954;2:375-7.
5. Truelove SC. Medical management of ulcerative colitis and the indications for colectomy. World J Surg. 1988;12:142-7.
6. Itoh M, Guth P. Role of oxygen-derived free radicals in hemorrhagic shock-induced gastric lesions in the rat. Gastroenterology 1985:88 :162-67.
7. Bartnik W, Shorter RG. Inflammatory bowel disease: immunological developments. In: Berk JE, ed. Developments in digestive diseases. Philadelphia: Lea and Febiger, 1980:109-27.
8. Sam AS. The significance of removing oxygen-derived free radicals in the treatment of acute and chronic duodenal ulceration in the rat. J Pharm Pharmacol 1990:42:64-7
9. Salim AS. Role of oxygen-derived free radicals in mechanism of acute and chronic duodenal ulceration in the rat. Dig Dis Sci 1990:35:73-9.
10. Salim AS. Role of oxygen-derived free radicals in the mechnism of chronic gasric ulceration in the rat. Implications for cytoprotection. Digestion 1989:43:113-19.
11. Edwards FC, Truelove SC. The course and prognosis of ulcerative colitis. Part IV. Carcinoma of the colon. Gut 1964-5:15-18.
12. Lee ET. Statistical methods for survival analysis. Belmont, California, Wadsworth, 1980.
13. Moorhouse PC. Grootveld M, Halliwell B, Quinlan JG, Guteridge JMC. Allopurinol and oxypurinol are hydroxyl radical scavengers. FEBS Lett 1987 213:23-8.
14. Miziewicz JJ, Lennard-Jones JE, Connell AM, Baron JH, Jones F. Controlled trial of sulphasalazine in maintence therapy for ulcerative colitis. Lancet 1969: 1:648-51.
15. Strober W, James SP. The immunological basis of inflammatory bowel disease. J Clin Immunol 1986:6:415-32.
16. Kirsner JB, Shorter RG. Recent developments in nonspecific inflammatory bowel disease. N. Engl. J Med 1982:306:237-46.
17. William JG, Hughes LE, Hallett MB. Toxic oxygen metabolite production by circulating phagocytic cells in inflammatory bowel disease. Gut 1990:31:187-93.
18. Kagnoff MF. Inflammatory bowel disease--The search for an etiology. In: Rachmilewitz D, ed. Inflammatory bowel disease. Proceedings of the International Symposium of Inflammatory Bowel Diseases. Jerusalem, Sept. 7-9, 1981. The Hague: Martinus Nijhoff Publishers, 1982:59-67.
19. Metchnikoff E. Immunity in infective diseases. Cambridge, England: Cambridge University press, 1905.
20. Klebanoff SJ. Oxygen metabolism and the toxic properties of phagocytes. Ann lntern Med 1980;93:480-9.
21. Weiss SJ. Oxygen, ischaemia, and inflammation. Acta Physiol Scand. 1986;(suppl. 548):9-37.
22. Grisham MB, Granger DN. Neutrophil-modiated mucosal injury: role of reactive oxygen membolites. Dig Dis Sci 1988;33(suppl. 3):6S-15S.
23. Jewell DP, Patel C. Immunology of inflammatory bowel disease. Scand J Gastroenterol 1985;114(suppl):119-26.
24. Beilke MA, Collins-Lech C, Solnle PG. Effects of dimethyl sulfoxide on the oxidative function of human neutrophils. J LAB CLIN MED 1987; 110:91-6.
25. Sandberg-Gertzen H, Järnerot G, Kraaz W. Azodisal sodium in the treatment of ulcerative colitis. A study of the tolerance and relapse-prevention properties. Gastroenterology 1986;90:1024-8.

From the University Department of Surgery, The Medical City, Baghdad, Iraq.

Submitted for publication July 8. 1991; revision submitted Oct. 16. 1991; accepted Oct. 17. 1991.

Reprint requests: Aws S. Salim. PhD (Surg), FRCS Ed. FRCS, Glasg, FICS, FCICD, The Department of Surgery, Ward 6, Stobhill General Hospital. Glasgow G21 3UW. U.K. 5/1/36404

DMSO Organization thanks the publisher for allowing us to put this article on our World Wide Web Site. The publisher retains all copyright. To copy any portion of this article, please submit your request to the publisher.