4462 Lack of effect of hormonal therapy in the prevention of rebleeding from gastrointestinal angiodysplasia: a multicenter, randomized controlled trial.

Angiodysplasia is a major cause of recurrent gastrointestinal bleeding. Several uncontrolled studies and one small randomized controlled trial with short follow-up have suggested that hormonal therapy may be useful in the prevention of rebleeding from gastrointestinal angiodysplasia. Aim: To investigate the efficacy of the long-term estrogen and progestagen therapy in the prevention of recurrent bleeding from gastrointestinal angiodysplasia. Patients and methods: Seventy-two patients bleeding from gastrointestinal angiodysplasia confirmed by endoscopy or angiography were included in three major teaching hospitals. Patients were randomized to receive, in strict double-blind conditions, continuous treatment with a combination of ethinilestradiol (0.05 mg) plus norethisterone (1mg) at a dose of 1 tab/day, or placebo (1 tab/day). Failure of treatment was defined as any gastrointestinal bleeding or persistence of chronic anemia during 6 months despite continuous iron therapy. Patients were followed up for a minimum period of 1 year(range 1-3 years). Results: No significant differences were found between boths groups regarding demographic, clinical data and severity of bleeding. Four patients were lost to follow-up. Failure of treatment occurred in 14 out of 33 patients (41%) in the treatment group and in 17 of 35 patients (48%) in the placebo group (p=NS). The actuarial probability of remaining free of rebleeding at 1 and 2 years of follow-up was 69% (CI:50-87) and 50% (CI:27-73) for the treatment group, and 55% (CI:36-74) and 36% (CI:14-58) for the placebo group (log-rank test, p=0.6949). In a multivariate analysis, previous bleeding from angiodysplasia, but not treatment received, was an independent predictor for failure (p=0.013). No significant differences between groups were found according to bleeding episodes (treatment: 0.7 ± 1.0 vs. placebo: 0.9 ± 1.5) and transfusional requirements (treatment: 0.9 ± 1.9 units vs. placebo: 0.7 ± 1.5 units). The incidence of metrorrhagia was higher in treatment group (5 of 34, 15%) than in placebo group (0 of 38, 0%)(p=0.02). Two patients in treatment group( pulmonary embolism, stroke) and 1 in placebo group (pulmonary embolism) had severe side-effects. There were no differences in mortality between hormonal and placebo group (0 vs. 1, respectively). Conclusion: Continuous estrogen+progestagen treatment is not useful in the prevention of rebleeding from gastrointestinal angiodysplasia.