Inhaled loxapine for agitation revisited: focus on effect sizes from 2 Phase III randomised controlled trials in persons with schizophrenia or bipolar disorder

Objective: To describe the efficacy of inhaled loxapine for the treatment of agitation associated with schizophrenia or bipolar disorder using different examples of effect size (ES). Data sources: Psychopharmacologic Drug Advisory Committee briefing documents as prepared by the product manufacturer and by the US Food and Drug Administration. Study selection: Phase III clinical trials. Data extraction: Effect size for primary and secondary efficacy outcomes. Data synthesis: Two similarly designed Phase III studies were completed with one conducted in patients with agitation associated with schizophrenia and one in patients with agitation associated with bipolar I disorder, manic or mixed episodes. In both studies, onset of anti-agitation effect was observed at 10 min (first time-point measured) for both the 5 mg and 10 mg doses, as evidenced by time to first statistically significant change from baseline on the Positive and Negative Syndrome Scale, Excited Component (PEC) as compared to placebo. Loxapine remained superior to placebo throughout the remainder of the study at all-time points measured. In the schizophrenia study, the ES difference from placebo on the PEC at 2 h was 0.45 for the 5 mg dose and 0.60 for the 10 mg dose (Cohen's d). ES differences for Clinical Global ImpressionsImprovement (CGI-I) were similar. Prior to rounding up, the number needed to treat (NNT) for PEC response and CGI-I response vs. placebo were 4.1 and 4.6, respectively, for loxapine 5 mg and 3.2 and 3.2, respectively, for the 10 mg dose. For the outcome of requiring only one dose of study medication and no rescue medication, the NNT vs. placebo for the 5 mg dose was not statistically significant but was statistically significant for the 10 mg dose vs. placebo with a value of 7. When plotting the PEC responders over time NNT becomes more robust as time after dosing elapses, with the 10 mg dose reaching a NNT of 4 by 20 min. For the bipolar disorder study, the ES difference from placebo on the PEC at 2 h was 0.73 for the 5 mg dose and 0.94 for the 10 mg dose. ES differences for the CGI-I were somewhat higher. Prior to rounding up, NNT for response vs. placebo for PEC and CGI-I criteria were 2.9 and 2.6, respectively, for the loxapine 5 mg dose and 2.2 and 2.1, respectively, for the 10 mg dose. For the outcome of requiring only one dose of study medication and no rescue medication, the NNT vs. placebo for the 5 mg dose was 7 and for the 10 mg dose was 3. The NNT difference between 5 mg vs. 10 mg was statistically significant in favour of the 10 mg dose and had a value of 5. When plotting the PEC responders over time NNT becomes more robust as time after dosing elapses, with the 10 mg dose reaching a NNT of 3 by 20 min. Additional information regarding pulmonary safety demonstrated low rates of pulmonary adverse events among those subjects in the efficacy trials, however, in separate Phase I safety studies conducted in persons with asthma and COPD, respiratory symptoms and / or changes in flow parameters were common. Conclusions: Inhaled loxapine is a non-invasive treatment option for the management of agitation associated with schizophrenia or bipolar disorder. Effect sizes for inhaled loxapine vs. placebo are robust and on par with those observed with intramuscular antipsychotics and benzodiazepines. Onset of action is rapid. The magnitudes of the effect sizes were generally larger for the 10 mg dose vs. the 5 mg dose, and the overall data supports the 10 mg dose as the dominant choice. The efficacy profile of inhaled loxapine will need to be viewed within the context of its pulmonary safety profile. The advisers to the Food and Drug Administration recommended that inhaled loxapine be restricted to a single dose in 24 h and be subject to a Risk Evaluation and Mitigation Strategy programme.