Carbamazepine and valproate monotherapy: feasibility, relative safety and efficacy, and therapeutic drug monitoring in manic disorder

Abstract 

Rationale: Search for alternatives to lithium therapy for mood disorders commenced with anticonvulsants, carbamazepine (CBZ) and valproic acid (VPA), in the late 1970s. The comparative safety and efficacy data of CBZ and VPA monotherapy in patients with bipolar disorder remain to be established. Objectives: The main objectives of the study were to assess the relative anti-manic efficacy and safety of CBZ and VPA; to study the feasibility of using either, as a first line anti-manic agent; to investigate and generate clinically relevant parameters involving therapeutic drug monitoring of the two drugs. Methods: After a 2-day screening period, suitable patients (n=30) were randomly assigned to treatment with CBZ or VPA. Both the drugs were started with an average dose of approximately 20 mg/kg body weight per day. Further increment in dose was carried out at weekly intervals, guided by clinical improvement, serum levels and treatment emergent adverse events. The primary efficacy measure in the protocol was defined as a change from baseline to endpoint in total score on the Young Mania Rating Scale. A favourable clinical response was defined a priori as a decrease of more than 50% from baseline in Young Mania Rating Scale total score. Results: Both CBZ and VPA were found to be efficacious as single first-line anti-manic agents, however VPA proved to be better. Using the intent-to-treat analysis, the VPA group showed a significant fall in YMRS total scores after week 1 while the CBZ group showed a significant fall after week 2. In the primary efficacy analysis, valproate group experienced significantly greater mean improvement in Young Mania Rating Scale total score than the CBZ group. Of the VPA treated patients, 73% showed a favourable clinical response while 53% of the patients on CBZ responded favourably. In the CBZ group, significantly more patients received rescue medication during the week 2 and the requirement was quantitatively more as compared to the VPA group. The steady state serum concentration (Css) of CBZ ranged from 3 to 9 µg/ml; however, it did not appear to correlate with the dose or clinical response. The Css of VPA ranged from 50 to 100 µg/ml; a linear correlation was found between the dose and serum levels of VPA as well as between weekly rise in serum levels and clinical response. Weekly dose escalations of VPA also correlated positively with corresponding rise in serum levels. Significantly more patients in the CBZ group reported adverse events, including nausea, vomiting and dizziness, than VPA. Conclusions: The findings from this study suggest that both CBZ and VPA monotherapy is feasible for treatment of acute mania; however, VPA is more efficacious in terms of its early onset of action, lesser requirement for rescue medication and better tolerability. Further work needs to be undertaken to characterise the manic patients in terms of their differential psychopharmacologic response profile.