Is it fair to the children? Asenapine trials for acute manic or mixed episodes in children with bipolar disorder were misrepresented

Objective: In 2015, the FDA approved the antipsychotic asenapine for acute manic/mixed episodes in pediatric patients (10-17 years) with bipolar I disorder, based on one trial lasting three weeks, followed by a 50-week openlabel extension (OLE) phase. The authors claimed that “asenapine was generally well tolerated”. Method: One unblinded observer extracted data on harms and quality of life from the paper and compared these to data from the trial register and the FDA clinical review. A second observer validated the data. Results: Of 404 participants, 303 were randomized to one of three asenapine doses and 101 to placebo, 321 entered OLE and 140 completed it. Non-serious adverse events (AEs) were reported with a 5% threshold. During the randomised phase, extrapyramidal symptoms were reported for 13 (4.3%) asenapine-treated children versus two placebo-recipients (2%), weight gain above 7% occurred in 26 (8.6%) patients on asenapine versus one on placebo, and there were 414 versus 43 non-serious AEs. During the OLE (described to assess safety), 613 serious AEs in 23 participants were registered in the trial register; self-harm was reported 192 times. Extrapyramidal symptoms events were reported 20 times and 109 experienced >7% weight gain. Non-serious AEs totaled 3437 events in 200 participants (62%) while the paper reported selected AEs. The pediatric Quality of Life Enjoyment and Satisfaction Questionnaire is a self-reported 70-point scale. Results were only reported in the trial register. During the randomized phase, the change from baseline was 3.4 (standard deviation 9.8) and 1.5 (8.2) for patients on asenapine and placebo, respectively. For OLE, the change from baseline was 1.4 (8.6). Thus, we saw minimal clinical difference and high variation in the data. Conclusion: Comparing published results to those in a trial register for the one trial that led to approval of asenapine in children showed that the authors’ conclusions were misleading. The extent of harms was understated, as is commonly seen. The participating children experienced a vast amount of AEs, and they did not feel better. A debate on the justification of OLEs and reporting and interpretation of harms data in general is clearly needed.