The incidence of obscure GI bleeding has been rising. This condition is now accepted to mainly originate from small bowel lesions and to be related to the wide use of low-dose aspirin and NSAIDs. We aimed to assess the efficacy of misoprostol, a prostaglandin analogue, in the healing of small bowel ulcers and erosions in patients using aspirin/NSAIDs and presenting with obscure GI bleeding (MASTERS Trial).Methods and design
This was a prospective, randomised, placebo-controlled trial, comparing misoprostol 200 ug with placebo taken 4-times daily for 8 weeks, while continuing on low-dose aspirin and/or NSAIDs. Patients were screened if they had one or more of the following: iron deficiency anaemia; a drop in Hb level of 2.0 gm/dL or more; and/or positive faecal occult blood test. Those without a potential source of bleeding on their upper GI endoscopy and colonoscopy but then found to have small bowel ulcers and erosions, using video-capsule endoscopy (OMOM Endoscopy System, China), were considered for randomisation. Small bowel lesions were classified as modified from Maiden et al, Clin Gastroenterol Hepatol 2007. Patients were clinically assessed 4 weeks after randomisation and with repeat capsule endoscopy after 8 weeks to check for mucosal healing. Capsule images were examined by two blinded assessors, and any discrepancies were resolved by consensus. The primary endpoint was full healing of small bowel ulcers and/or erosions. Secondary endpoints were changes in the numbers of ulcers and erosions, and haemoglobin (Hb) level.Results
Of 232 patients screened, 102 eligible patients were randomised: 50 took misoprostol and 52 took placebo. Complete healing of ulcers and erosions was observed in 27 (54%) misoprostol patients and 9 (17%) in the placebo group (p<0.001). Viewed separately, similar results were seen for healing of ulcers (68% vs 17%, p<0.001) and erosions (56% vs 21%, p<0.001). There were no significant changes in Hb levels over the study period. Compliance did not differ significantly between the two arms (76% misoprostol vs 90% placebo took >75% of study drugs, p=0.065). Mild, moderate, or severe adverse events of interest (abdominal pain, nausea/vomiting, diarrhoea, or other events) were observed in 48% of misoprostol patients vs 44% placebo (p=0.84), but tended to be more severe in the misoprostol group (p=0.017). There were no serious adverse events.Conclusions
Misoprostol is effective in the healing of small bowel ulcers and erosions in patients with obscure gastrointestinal bleeding while using low-dose aspirin and NSAIDs.