Clinical features of patients with treatment-emergent suicidal behavior following initiation of paroxetine therapy

Abstract Background Understanding suicidal behavior is an important component of assessing suicidality in psychiatric patients. GlaxoSmithKline (GSK) conducted a meta-analysis of randomized, placebo-controlled trials to compare suicidality in adult patients treated with paroxetine vs. placebo. The goal was to identify emergent clinical characteristics of patients with definitive suicidal behavior (DSB: preparatory act, suicide attempt, completed suicide). Methods The dataset comprised 14,911 patients from 57 placebo-controlled paroxetine trials. Possible cases of suicidality were identified and were blindly reviewed by an expert panel, which categorized cases as suicidal or non-suicidal. DSB incidences were compared between paroxetine and placebo. Clinical narratives and case report forms for major depressive disorder (MDD) and anxiety disorder patients with DSB were reviewed. For MDD, rating scale items relating to suicidality, insomnia, agitation, and anxiety were examined. Results Overall (all indications) there were no differences between paroxetine and placebo for DSB (50/8958 [0.56%] vs. 40/5953 [0.67%], respectively; OR = 1.2 [CI 0.8, 1.9]; p = 0.483). However, in patients with major depressive disorder (MDD), the incidence of DSB was greater for paroxetine (11/3455 [0.32%] vs. 1/1978 [0.05%], OR = 6.7 [CI 1.1, 149.4]; p = 0.058). Review of the 11 paroxetine MDD cases revealed common clinical features: symptomatic improvement; younger age (18–30 years); psychosocial stressors; overdose as method; and absent/mild suicidal ideation at the visit prior to the event. There was no evidence for a consistent adverse event profile or onset of akathisia/agitation or a manic/mixed state. Anxiety disorder patients with DSB had a heterogeneous clinical picture. Limitations Limitations to the study include the relatively small number of cases and the retrospective nature of the study. Conclusions DSB incidence was similar between paroxetine and placebo overall, but a higher frequency of DSB was found for paroxetine in MDD patients, driven by young adults aged ≤ 30 years. Most MDD patients with DSB improved prior to the attempt and experienced a psychosocial stressor. Patients should receive careful monitoring for suicidality during paroxetine therapy.