Perioperative β-Blockers: Fewer MIs but Increased Mortality

In several small trials published in the late 1990s, perioperative β-blockade reduced the incidence of cardiac ischemic complications associated with noncardiac surgery. However, two medium-sized trials published in 2006 failed to duplicate the previous findings (JW Nov 29 2006). This uncertainty set the stage for the large POISE trial, conducted in 23 countries.

All 8351 enrolled patients (age ≥45) were scheduled for noncardiac surgery and met at least one of the following criteria: history of coronary disease, stroke, or peripheral vascular disease; recent hospitalization for heart failure; undergoing major vascular surgery; or any three of seven other risk criteria outlined in the protocol. Patients were randomized to receive extended-release metoprolol (100 mg) or placebo 2 to 4 hours before surgery. Postoperatively, patients received metoprolol (maximal dose, 200 mg daily) or placebo for 30 days, titrated according to blood pressure and pulse parameters.

At 30 days, the incidence of the primary composite endpoint (cardiovascular death, nonfatal myocardial infarction, or nonfatal cardiac arrest) was lower with metoprolol than with placebo (5.8% vs. 6.9%; P=0.04). This difference was driven by a significantly lower incidence of nonfatal MI in the metoprolol group (3.6% vs. 5.1%; P=0.001). However, overall mortality was higher in the metoprolol group than in the placebo group (3.1% vs. 2.3%, P=0.03), as was stroke incidence (1.0% vs. 0.5%; P=0.005). Clinically important hypotension and bradycardia occurred significantly more often with metoprolol than with placebo, and post hoc multivariable analysis suggested that hypotension partly explained the link between metoprolol and mortality and metoprolol and stroke. Additional analyses did not identify any subgroup that benefited from metoprolol.