adalimumab vs golimumab

NCT00385736 · Phase 3 · completed

Design is strong (randomized, quadruple-blind, placebo-controlled, n=576, hard endoscopic/clinical Mayo remission endpoint), but the primary endpoint result was discordant across the two pre-specified dose-vs-placebo comparisons: one significant, one not. A well-designed trial with a mixed result is capped at moderate rather than strong.

NCT01863771 · Phase 3 · completed

Design meets the 'strong' bar: randomized, double-blind (participant and investigator masked), placebo-controlled, phase 3, with a hard clinical endpoint (Mayo score-based clinical response); the primary comparison was clearly met in the favorable direction (18/32 vs 6/31, non-overlapping 95% CIs).

NCT00195702 · Phase 3 · completed

This was a large, randomized, double-blind, placebo-controlled trial. More patients on adalimumab responded on the ACR20 joint-response measure than on placebo (about 61-63% vs about 30%, p<0.001 for both doses), and adalimumab also significantly slowed radiographic joint damage and improved physical function compared with placebo (p<0.001 for both).

NCT00264537 · Phase 3 · completed

This trial was well designed -- randomized, fully blinded, and placebo-controlled with over 600 patients -- but its two main outcomes told different stories: the joint-damage (structural) outcome clearly favored golimumab plus methotrexate, while the joint-response (ACR50) outcome for the combined golimumab groups narrowly missed statistical significance against placebo, with only one of the individual dose comparisons reaching significance on its own. A well-designed trial with a split result on its co-primary outcomes does not qualify as strong evidence of benefit.