Does antidepressant medication in patients with hepatitis C undergoing interferon alpha treatment reduce therapeutic efficacy?

Activation of the inflammatory immune system, as evidenced by the effects of interferon a (IFNa) treatment, induces depressive symptoms. Prescription of a selective serotonin reuptake inhibitor (SSRI) is now advocated in combination with antiviral therapy in patients exposed to IFNa when early symptoms of depression emerge. The recently published article written by Kraus et al (Gut 2008;57:531–6) gives further evidence for the efficacy of SSRIs to treat depressive symptoms during IFNa treatment. The mechanisms by which SSRIs reduce depressive symptoms remain unclear. Several hypotheses, such as impact on serotonin activity or neurotrophic actions, have been proposed. Another, that has good face validity in this context, is that the antiinflammatory properties of SSRIs mediate at least in part the decrease in depressive symptoms. Antidepressants have been shown in animals and humans to inhibit the production and/or release of proinflammatory cytokines and to stimulate the production of anti-inflammatory cytokines (such as interleukin 10). In addition, recent data demonstrate that antidepressant-like effects of desimipramine in the forced swimming test (an animal model frequently used to evaluate antidepressant efficacy) depend on the ability of the drug to inhibit central nervous system production of tumour necrosis factor a. Given the facts that (1) activation of the inflammatory immune system is responsible for increased antiviral activity and the development of depressive symptoms, (2) SSRIs reduce depressive symptoms during IFNa treatment and display anti-inflammatory effects, the hypothesis may be proposed that, whilst effectively reducing depressive symptoms, SSRIs may simultaneously reduce the efficacy of the treatment against hepatitis C virus (HCV), by reducing inflammatory activity. Based on this theory, the following two premises should apply. First, patients showing a good response to antiviral treatment have increased Th-1 cell activation and increased risk for depressive symptoms. Secondly, SSRIs, by reducing Th1 cell activation, reduce both depressive symptoms and the probability of viral clearance. With regard to the first premise, described above, few studies have been published reporting data on the association between viral clearance and depressive symptoms. Loftis and colleagues found that IFNa response rates (both end-of-treatment response (ETR) and sustained viral response (SVR)) were significantly higher in those patients who developed depression during treatment than in those who did not (ETR, 61.5% (8/13) and SVR, 38.5% (5/13) vs ETR, 26.9% (7/26) and SVR, 11.5% (3/26), respectively). Raison and colleagues, however, showed an association in the opposite direction: the lower the increase in depressive symptoms as measured with the Zung self-rating scale, the higher the probability of viral clearance. Two other studies found no associations. In order to resolve this apparent discrepancy, we analysed data from an IFNa-treated HCV patient sample, previously described in Wichers et al, in order to examine the association, using regression analysis, between alanine aminotransferase (ALAT) values and depressive symptoms. Increased ALAT values indicate the presence of an infected liver, and ALAT response to treatment is known to correlate strongly with a decrease in or negativity of HCV RNA. The results showed that increased response to treatment (as indicated by a decrease in ALAT) predicted an increase in depressive symptoms (B = 20.090, p,0.001). Concerning the second premise, no study designed to examine whether SSRIs impact negatively on the probability of viral clearance during IFNa treatment has been performed. One study that does give some information about this association shows that the SSRI group had less viral clearance than the placebo group, although numbers were very small. We conclude that although there are strong theoretical considerations that support the possibility that SSRI treatment may negatively influence viral clearance in IFNatreated patients, this assumption has not been investigated. There is an urgent need for studies designed and powered to examine this hypothesis.