Folate metabolism biomarkers from two randomised placebo-controlled clinical studies with paroxetine and venlafaxine.

OBJECTIVES: Based on the hypothesis of a role for folate and vitamin B12 in major depressive disorders (MDD), we aimed at validating the association between folate pathway biomarkers and depression or antidepressant response in clinical trial populations. METHODS: We investigated serum levels erythrocyte folate and serum levels of homocysteine, vitamin B12, and folate as disease and response biomarkers for MDD in two independent randomised, placebo-controlled clinical trials, where paroxetine or venlafaxine were used as active controls, for a total of 881 patients. RESULTS: Significant but weak correlations between depression severity and biomarker levels could be detected in the paroxetine study for serum folate and vitamin B12, with no correlations for any biomarker in the venlafaxine study. Besides a weak association for erythrocyte folate in the venlafaxine study, no significant associations were observed between treatment response and pre-treatment levels of any of the biomarkers tested. CONCLUSIONS: Notwithstanding the relatively large number of patients tested, we did not find consistent associations between folate biomarkers and MDD severity, or response to paroxetine and venlafaxine. Our results may be related to the particular study design or clinical population; however, our findings do not support the hypothesis of a dysfunction of one-carbon metabolism in MDD.

Biomarkers for response in major depression: comparing paroxetine and venlafaxine from two randomised placebo-controlled clinical studies.

The identification of biomarkers of response might speed drug development and set the premises to assist clinical practice in psychiatry. In this work, we evaluated a panel of peripheral biomarkers (including IL-6, IL-10, TNF-α, TNFRII, BDNF, CRP, MMP9 and PAI1) in depressed patients receiving paroxetine, venlafaxine, or placebo. Samples were obtained from two randomised placebo-controlled studies evaluating the efficacy and tolerability of a novel drug candidate, using either paroxetine or venlafaxine as active comparators. In both studies, the biomarker candidates were analysed in plasma collected at randomization and after 10 weeks of treatment with either placebo or active comparator (for a total of 106 and 108 subjects in the paroxetine and venlafaxine study, respectively). Data were obtained by multiplexing sandwich-ELISA system. Data were subjected to statistical analysis to assess their correlation with baseline severity and with response outcome. Increases in biomarker levels were correlated with reduction in depression severity for TNF-α, IL-6 IL-10 and CRP. Response to paroxetine treatment correlated with baseline IL-10, IL-6 and TNF-α levels, with the strongest signal being observed in males. In the venlafaxine study, a correlation was observed only between CRP level at randomisation and response, suggesting differences between the two active treatments and the two studies. Our investigations suggest that a combination of pro- and anti-inflammatory cytokines may predict response outcome in patients treated with paroxetine. The potential for IL-10, IL-6 and TNF-α as response biomarkers for a wider range of antidepressants warrants further investigations in clinical trials with other monoamine reuptake inhibitors.