San Fran,

Your point is well taken. Unless the two drugs were compared head to head in the same trial, as you emphasized, one cannot claim superiority of one vs. the other. That being said, the patient populations in both LEADER and EMPA-REG OUTCOME were patients with type 2 diabetes and high-risk for cardiovascular disease and are more similar than different. But yes, different trial lengths, different patient populations, and other uncontrolled differences prevent a direct comparison of the two trials. One can appreciate this by observing that the 3-point MACE event rates in LEADER (14.9% placebo; 13.0% treatment) were greater than that in EMPA-REG OUTCOME (12.1% placebo; 10.5% treatment).

One can still learn something by comparing the trials by looking at the statistics and the 3-point MACE sub-components (cardiovascular death, non-fatal MI, non-fatal stroke). In EMPA-REG OUTCOME, Empagliflozin (SGLT2 inhibitor) BARELY squeaked by with a p-value of 0.04 (cut-off is 0.05) for showing 14% reduction in 3-point MACE. LEADER trial with Liraglutide (GLP-1 receptor agonist) fared better with a p-value of 0.01 for its 13% reduction in 3-point MACE.

Which sub-component(s) were driving the MACE reduction and helping (or hurting) the statistics? For both EMPA-REG OUTCOME and LEADER, the primary driver was a 38% (p<0.001) and 22% (p=0.007) reduction, respectively, in cardiovascular death. So for CV death, both are good. For non-fatal MI, both trials had about a 12% reduction, but p-value was 0.11 for LEADER and 0.22 for EMPA-REG OUTCOME. So for non-fatal MI, both are trending in the same direction but failed to meet statistical significance. For non-fatal stroke, LEADER showed an 11% decrease (p=0.30) but EMPA-REG OUTCOME showed a 24% INCREASE (p=0.16). While both p-values failed to meet the p=0.05 cut-off, there is a troubling trend for increased stroke with Empagliflozin but an encouraging trend for reduced stroke with Liraglutide.

Lessons learned are that 1) significant reduction in CV death is the main driver of the reduced 3-point MACE metric in both trials; 2) non-significant reduction in non-fatal MI also positively contributes to the overall 3-point MACE metric in both trials; and 3) non-significant, but opposing effects of Liraglutide and Empagliflozin on non-fatal stroke partially explain the stronger p-value for 3-point MACE metric in LEADER than EMPA-REG OUTCOME.

These two drugs act via distinct mechanisms, and it wouldn't be too surprising in the future to see a cardiovascular outcomes trial directly comparing Liraglutide alone vs. Empagliflozin alone, as well as in combination. But hopefully by that time BETonMACE will be in the spotlight :-)

BearDownAZ