Plasma serotonin level after 1 day of fluoxetine treatment:: a biological predictor for antidepressant response?

Rationale: Antidepressant treatments present a delayed onset of action. Objective: The present study investigated whether plasma or serum serotonin, 5-hydroxytryptamine (5-HT), could predict clinical improvement. Methods: Biological parameters were determined after a 4-week drug-free period (day 0) and 1, 14 and 28 days after the beginning of the treatment with fluoxetine 20 mg daily in depressed patients. Clinical evaluations were assessed on days 0, 14 and 28. Results: One day after a single dose, the mean values of plasma 5-HT (5.4 +/- 2.6 nmol/l) and serum 5-HT (484 +/- 215 nmol/l) were not statistically different from basal mean values (4.5 +/- 2.5 nmol/l and 523 +/- 263 nmol/l, respectively). The repeated treatment significantly reduced serum 5-HT to 34% (P = 0.002) and 17% (P = 0.0004) of pretreatment values after 14 and 28 days of treatment, respectively; plasma 5-HT was also reduced significantly to 28% and 15% of pretreatment values (P < 0.05 in both cases). At day 28, four of the eight patients responded by showing a reduction in MADRS score of at least 50 % of the baseline score. No correlation was found between pretreatment values of serum or plasma 5-HT and clinical evolution, even if a tendency (P < 0.07) to lower serum 5-HT pretreatment values was observed in responders. Plasma 5-HT after 1 day of treatment was significantly different between responders and non-responders: the plasma 5-HT concentration in responders was 3.4 +/- 1.7 nmol/l versus 7.4 +/- 1.6 nmol/l in non-responders (P = 0.02). Moreover, plasma 5-HT levels after 1 day of treatment were positively correlated to the final MADRS score (r = +0.89, n = 8, P = 0.003) and inversely correlated to its change from the initial score (r = -0.76, n = 8, P = 0.02). Conclusion. These preliminary data show that fluoxetine and norfluoxetine might influence 5-HT peripheral venous blood parameters and that plasma 5-HT after 1 day of treatment might be a biological predictor for antidepressant response.