Urgent endoscopy for high-risk patients with acute upper gastrointestinal bleeding does not reduce mortality

Référence

Lau JYW, Yu Y, Tang RSY, et al. Timing of endoscopy for acute upper gastrointestinal bleeding. N Engl J Med 2020;382(14):1299-1308.

Sommaire

Current guidelines recommend endoscopy within 24 hours for patients who present with acute upper gastrointestinal bleeding. To study the potential benefit of earlier endoscopy, these investigators enrolled high-risk patients who presented to the emergency department or were admitted to a medical ward with hematemesis or melena or both. High risk was defined by a Glasgow-Blatchford score of at least 12 (scale = 0 – 23, with higher scores indicating a higher risk of further bleeding or death). Patients were randomized to receive either urgent endoscopy within 6 hours after consultation with gastroenterology (n = 258) or early endoscopy the next morning or within 24 hours (n = 258). Those patients who had persistent hemodynamic instability after initial resuscitation were excluded. Both study groups received an intravenous proton-pump inhibitor (PPI), followed by a high dose of continuous PPI infusion. If endoscopic treatment for bleeding ulcers was required, then an intravenous PPI infusion was continued for 72 hours; otherwise, the infusion was stopped. Analysis was by intention to treat. The 2 groups were similar at baseline: the mean Glasgow-Blatchford score was 14, approximately 60% of patients in both groups had bleeding peptic ulcers, and 7% to 10% had bleeding esophagogastric varices. The mean time from presentation to endoscopy was significantly shorter in the urgent endoscopy group than in the early endoscopy group (9.9 hours vs 24.7 hours). For the primary outcome of 30-day all-cause mortality, no significant difference was detected between the 2 groups (8.9% in the urgent group vs 6.6% in the early group; hazard ratio [HR] 1.35; 95% CI 0.72 - 2.54; P = 0.34). Additionally, there were no significant differences in further episodes of bleeding, duration of hospitalization, or percentage of patients who received transfusions. Patients with bleeding peptic ulcers, however, were more likely to require endoscopic hemostatic treatment in the urgent group (69% vs 51%; HR 1.35; 1.13 - 1.63). Of note, observed mortality in the early endoscopy group was lower than the 16% assumed in the sample-size calculation to show an 8 percentage-point mortality reduction in the urgent endoscopy group. As such, the trial was not powered for a smaller benefit in mortality, if such a benefit truly exists.