Alors que les autorités publiques françaises pointent la nécessité d’une prescription rationnelle des médicaments, et en particulier des psychotropes, peu de données ont été synthétisées à l’usage du clinicien sur la prescription des antidépresseurs de seconde génération (SGA).

Proposer pour chaque SGA une description de l’efficacité et de la tolérance/acceptabilité selon les données des méta-analyses les plus récentes.

Nous avons recherché pour chaque SGA la méta-analyse de meilleure qualité et la plus récente dans les bases Medline (1966–présent), Web of Science (1975–present) et Cochrane (toutes les revues).

Le traitement de première intention de l’épisode dépressif majeur, en incluant le rapport efficacité–tolérance et le prix (générique), serait actuellement la sertraline (50mg/j). Elle serait cependant plus mal tolérée que d’autres ISRS sur le plan digestif (par l’action sérotoninergique), ce qui appellerait de la prudence dans l’augmentation des posologies. La fluoxétine semble intéressante dans le traitement des symptômes négatifs de la schizophrénie (20mg/j) et de la boulimie (60mg/j). La fluvoxamine semble intéressante en cas d’effets secondaires sexuels indésirables lors d’un traitement par un autre ISRS, ainsi que dans le traitement des dépressions anxieuses (son affinité pour les récepteurs sigma-1 pourrait lui conférer une action spécifique), des troubles anxieux et des dépressions psychotiques. La mirtazapine serait un traitement intéressant lorsque le clinicien souhaite un amendement plus rapide des symptômes.

Win 2009 French people consumed nearly 36 billion euros worth of drugs, 18% more than in 2004, resulting in growing burden for French Social Security. Expenses refunded by public health insurance amount to 26.8 billion euros or 74.5% of the total cost of drug purchase. While French authorities point to the need for rational prescribing, especially concerning psychotropic drugs, few data on the prescription of second-generation antidepressants (SGA) are synthesized for clinicians’ use.

To offer a description of effectiveness and tolerability/acceptability for each SGA according to data from the most recent meta-analyzes.

For each SGA we searched for the latest meta-analysis in Medline (1966–present), Web of Science (1975–present) and Cochrane (all journals).

Considering the benefit–risk ratio and the cost (generic), the first-line treatment for a major depressive episode may be currently sertraline (50mg/day). It may however have more digestive side effects than other SSRIs (due to the serotonin action), which calls for caution while increasing doses. The four most effective antidepressants may be, in this order, mirtazapine, escitalopram, venlafaxine and sertraline. Escitalopram and sertraline may show a better acceptability profile than mirtaapine, duloxetine, fluvoxamine, paroxetine, and venlafaxine. Fluoxetine seems relevant in treatment of negative symptoms of schizophrenia (20mg/day) and in bulimia (60mg/day). Fluvoxamine seems relevant in the case of sexual side effects with a previous SSRI, in treatment of anxiety disorders (it’s affinity for sigma receptors may confer a specific action) and in psychotic depression. Mirtazapine may be a treatment of interest when a fast remission of depressive symptoms is warranted but its tolerance profile makes it difficult to use. We found no meta-analysis comparing the efficacy and tolerability of duloxetine with other antidepressants, probably due to the recent availability of the molecule. We found no arguments in favor of first-prescription of tianeptine except perhaps a better safety profile but data were scarce. The presence of generalized anxiety disorder is an indication to start treatment at half doses over a few days, whatever the molecule, to avoid a resurgence of anxious symptomatology. Prescription of antidepressants has demonstrated its efficacy in major depressive episode, in anxiety disorders (for most molecules), in dysthymia but not in seasonal affective disorder (SAD). Efficiency in dysthymia may explain the prescription of antidepressants that exceed the scope of the marketing authorization.

We summarize in this article all current data from recent meta-analyzes regarding the effectiveness, tolerance and acceptability of SGA to help practitioners’ choices of prescription.