Treatment of Early-Age Mania: Outcomes for Partial and Non responders to Initial Treatment

Objective: The Treatment of Early Age Mania (TEAM) study evaluated lithium, risperidone, and divalproex sodium (divalproex) in children with bipolar I disorder who were naive to antimanic medication, or were partial or nonresponders to 1 of 3 study medications. This report evaluates the benefit of either an add-on or a switch of antimanic medications for an 8-week trial period in partial responders and nonresponders, respectively. Method: TEAM is a randomized, controlled trial of individuals (N = 379) aged 6 to 15 years (mean +/- SD = 10.2 +/- 2.7 years) with DSM-IV bipolar I disorder (mixed or manic phase). Participants (n = 154) in this report were either nonresponders or partial responders to 1 of the 3 study medications. Nonresponders (n = 89) were randomly assigned to 1 of the other 2 antimanic medications and cross-tapered. Partial responders (n = 65) were randomly assigned to 1 of 2 other antimanic medications as an add-on to their initial medication. Adverse event (AE) rates are reported only for the add-on group. Results: Response rate for children switched to risperidone (47.6%) was higher than for those switched to either lithium (12.8%; p = .005; number needed to treat [NNT] = 3; 95% CI = 1.71-9.09) or divalproex (17.2%; p = .03; NNT = 3; 95% CI = 1.79-20.10); response rate for partial responders who added risperidone (53.3%) was higher than for those who added divalproex (0%; p = .0002; NNT = 2; 95% CI = 1.27-3.56) and trended higher for lithium (26.7%; p = .07; NNT = 4). Reported AEs in the add-on group were largely consistent with the known AE profile for the second medication. Weight gain (kg) was observed for all add-on medications: lithium add-on (n = 29 of 30) = 1.66 +/- 1.97; risperidone add-on (n = 15 of 15) = 2.8 +/- 1.34; divalproex add-on (n = 19 of 20) = 1.42 +/- 1.96. There was no evidence at the 5% significance level that the average weight gain was different by study medication for partial responders (p = .07, 1-way analysis of variance). Conclusion: Risperidone appears to be more useful than lithium or divalproex for children with bipolar I disorder and other comorbid conditions who are nonresponders or partial responders to an initial antimanic medication trial.