Accuracy in recognising happy facial expressions is associated with antidepressant response to a NOP receptor antagonist but not placebo treatment

Background: Clinical trials with putative antidepressants can be difficult to execute as it can take up to 8 weeks before differences emerge between drug and placebo, and long expensive trials often fail. Implementation of early response biomarkers could aid this process significantly with potential to identify new treatments. Aims: In a secondary analysis, we examined the association of early effects on emotional processing with later clinical outcome following treatment with the novel NOP antagonist LY2940094 versus placebo. We hypothesised that early induction of positive bias would be associated with reduced severity of depression after 8 weeks of treatment. Methods: This was a multicentre, randomised, double-blind, parallel-group, fixed-dose, placebo-controlled, 8 week study to assess sensitivity of the facial emotional recognition task (FERT) to early changes in emotional bias induced by LY2940094. Patients who met diagnostic criteria for major depression were randomised to receive LY2940094 (N = 70) or placebo (N = 66). At week 1 and 6, the FERT was completed by 33 patients in the LY2940094 group and 34 in the placebo group. Results: Patients identified happy faces with higher accuracy (Wald χ2(1,33) = 14.25, p