BRITISH MEDICAL JOURNAL VOLUME 290 23 FEBRUARY 1985 Double blind comparative study of omeprazole 10 mg and 30 mg daily for healing duodenal ulcers P J PRICHARD, D RUBINSTEIN, D B JONES, F J DUDLEY, R A SMALLWOOD, W J LOUIS, N D YEOMANS Abstract Healing of duodenal ulcers was assessed in 66 patients who received omeprazole either 10 mg or 30 mg daily for four weeks in a double blind study. Healing was rapid in both groups. At two weeks the ulcers in 15 of the 30 patients taking 10 mg daily had healed compared with 28 of the 36 (78%) taking 30 mg daily (p <0 03). At four weeks the respective proportions had risen to 83% (25/30) and 94% (33/35) (p > 0-05). In non-smokers the proportion of ulcers healed did not differ significantly with the two doses, although there was a trend for less healing at two weeks with 10 mg daily; in smokers significantly fewer ulcers (p <0 05) were healed with 10 mg than 30 mg daily at two weeks (7/16 (44%) v 17/21 (81%)) and at four weeks (12/16 (75%) v all 21 (100%)). Adverse reactions were few and transient and were considered unlikely to be due to omeprazole. Introduction Omeprazole is a substituted benzimidazole that acts selectively at the secretory surface of the parietal cell.1 2 Although this particular compound is being reassessed because of a recent suggestion of toxicity in animals at high dosage, this new class of gastric acid inhibitors could be a potent alternative to H2 receptor blocking drugs in the treatment of peptic ulcers. With all forms of antisecretory treatment problems may arise when doses that totally suppress acid secretion are used for a long time. Thus studies that define the clinical response to smaller doses are important. The optimal regimen for treatment with omeprazole is still uncertain. Three recent comparative dose studies in small groups of patients showed rapid healing of duodenal ulcers treated with 20, 30, 40, or 60 mg daily for four weeks.3 We report a double blind study comparing a lower dose of omeprazole than has been studied previously (10 mg daily) with an intermediate dose (30 mg daily). Patients and methods The study was conducted at two teaching hospitals in Melbourne. Patients included in the study had active duodenal ulceration of at least 5 mm diameter that had been verified at endoscopy withinthe previous three days. We excluded patients aged below 18 or above 80; Departments of Gastroenterology and Clinical Pharmacology, Austin Hospital, Heidelberg, Victoria 3084, Australia P J PRICHARD, MB, BS, fellow in clinical pharmacology D B JONES, MRCP, senior research officer R A SMALLWOOD, MD, FRACP, director of gastroenterology W J LOUIS, MD, FRACP, professor of clinical pharmacology N D YEOMANS, MD, FRACP, first assistant in medicine Gastroenterology Service, Alfred Hospital, Melbourne, Victoria 3081, Australia D RUBINSTEIN, MB, FRACP, gastroenterology registrar F J DUDLEY, MD, FRACP, director of gastroenterology Correspondence to: Dr N D Yeomans, Department of Medicine, Austin Hospital. women of childbearing potential; patients treated in the previous three days with either anticholinergic agents or histamine H2 receptor antagonists; and patients with concurrent gastric ulceration, pyloric stenosis, severe reflux oesophagitis, previous gastric surgery, chronic alcoholism, or drug abuse (including analgesic abuse). Each patient gave signed, informed consent before inclusion, and approval for the study was obtained from the human research committee of each hospital. Sixty seven patients entered the study and were randomised to receive either 10 mg or 30 mg omeprazole (formulated as enteric coated granules in hard gelatin capsules) each morning for four weeks. With this number of patients we calculated that a 3500 difference in the proportions of ulcers that healed with the two dosages would be detected (x=0-05, power=0-8). As the 10 mg and 30 mg capsules differed in appearance a double dummy technique was used to ensure blinding (each patient received one active and one placebo each day). Separate randomisation lists generated by computer were prepared for each centre and put in groups of six to minimise the influence of any variables related to time. Treatment was allocated so that there were similar proportions of smokers in each group, and patients were advised to maintain their previous intake of tobacco throughout the trial. Antacid tablets (Mylanta, buffering capacity 12 5 mmol for each tablet) were allowed ad libitum to relieve ulcer pain. Endoscopy was done by a trained observer who was unaware of the treatment given. Healing of ulcers was assessed by endoscopy at two weeks and again at four weeks if the ulcers had not healed. Healing was defined as complete re-epithelialisation of the site of the ulcer. During the trial patients were reviewed weekly to assess symptoms, compliance, and possible adverse events. Laboratory investigations were carried out before and on days eight and 29 of treatment. These included assessment of serum electrolyte, urea, and creatinine concentrations; liver function (bilirubin concentration and alkaline phosphatase, serum aspartate aminotransferase, and serum alanine aminotransferase activities); glucose and haemoglobin concentrations; white cell and platelet counts; and analysis of urine. Thyroid function tests (measure- ment of serum triiodothyronine, thyroxine, and thyroid stimulating hormone concentrations) were performed before the trial started and on day 29. Measurement of serum alanine aminotransferase activity was not included in the initial group of tests but was included after the trial began because of a report of raised activity in some patients.4 Results in the two treatment groups were compared with Fisher's test of exact probability (two tailed).6 When p exceeded 0-05, 95°Z confidence limits for differences in the proportions of healed ulcers were also calculated.7 Results Of the 67 patients initially entered into the trial, one, who was treated with 10 mg omeprazole, was lost to follow up before the review on day eight and could not be contacted. One further patient (taking 30 mg omeprazole) refused endoscopy on day 29 and was therefore not included in the analysis of healing at four weeks. Table I shows the initial characteristics of the patients after rando- misation. The differences in these characteristics between the two groups were not significant (p > 005). Table II gives the proportions of patients whose ulcers healed. At two weeks significantly more ulcers had healed in patients treated with 30 mg daily than in those treated with 10 mg daily (78% v 5001; p < 0-03). The difference in healing between groups at four weeks was not significant (94%o v 83°0). Even large ulcers (> 15 mm) healed rapidly. Duodenitis was present at initial endoscopy in 25 (83%) of the group taking 10 mg omeprazole and in 24 (67°V ) of the group taking 30 mg (p = 0-2). At endoscopy after two weeks duodenitis was still evident in 22 (730o ) of those given 10 mg compared with only 14 (39%/) given 30 mg (p < 0-05). 601