Systematic review with network meta-analysis: dual therapy for high-risk bleeding peptic ulcers

Keda Shi, Zeren Shen, Guiqi Zhu, Fansheng Meng, Mengli Gu, Feng Ji

Adding a second endoscopic therapy to epinephrine injection might improve hemostatic efficacy in patients with high-risk bleeding ulcers but the optimum modality remains unknown. We aimed to estimate the comparative efficacy of different dual endoscopic therapies for the management of bleeding peptic ulcers through random-effects Bayesian network meta-analysis. Different databases were searched for controlled trials comparing dual therapy versus epinephrine monotherapy or epinephrine combined with another second modality until September, 30 2016. We estimated the ORs for rebleeding, surgery and mortality among different treatments. Adverse events were also evaluated. Seventeen eligible articles were included in the network meta-analysis. The addition of mechanical therapy (OR 0.19, 95% CrI 0.07-0.52 and OR 0.10, 95% CrI 0.01-0.50, respectively) after epinephrine injection significantly reduced the probability of rebleeding and surgery. Similarly, patients who received epinephrine plus thermal therapy showed a significantly decreased rebleeding rate (OR 0.30, 95% CrI 0.10-0.91), as well as a non-significant reduction in surgery (OR 0.47, 95% CrI 0.16-1.20). Although differing, epinephrine plus mechanical therapy did not provide a significant reduction in rebleeding (OR 0.62, 95% CrI 0.19-2.22) and surgery (OR 0.21, 95% CrI 0.03-1.73) compared to epinephrine plus thermal therapy. Sclerosant failed to confer further benefits and was ranked highest among the 5 treatments in relation to adverse events. Mechanical therapy was the most appropriate modality to add to epinephrine injection. Epinephrine plus thermal coagulation was effective for controlling high risk bleeding ulcers. There was no further benefit with sclerosants with regard to rebleeding or surgery, and sclerosants were also associated with more adverse events than any other modality.