S. Chapel(1),Y. Chiu(2), J. Hsu(2), J. Xu(2), J. Cucchiaro(2), A. Loebel(2)
(1) Ann Arbor Pharmacometrics Group, Inc (A2PG), Ann Arbor, MI, USA (2) Sunovion Pharmaceuticals Inc., Fort Lee, NJ USA
Objectives: Characterization of dose-response relationships for psychotropic agents may be difficult to determine based on results of individual clinical trials, due to various confounds such as variability in attrition and placebo-response rates. As a consequence, post-marketing changes in recommended therapeutic dosing ranges for antipsychotic drugs are not uncommon. The goal of this exposure-response analysis was to more precisely clarify dose-response effects for lurasidone.
Methods: TeData were pooled from five 6-week, randomized, double-blind, placebo-controlled, once-daily, fixed-dose studies of lurasidone in the dosing range of 40-160 mg for the treatment of an acute exacerbation of schizophrenia. The PANSS and exposure data were fitted using the nonlinear mixed effects modeling methodology implemented in the NONMEM software (Version VI).xt regarding methods.
Results: In the final exposure-response model, LS mean change-from-baseline in PANSS exhibited a linear trend relative to dose of lurasidone. The 160 mg dose provided the greatest clinical benefit in terms of PANSS reduction relative to lower doses. In addition, the 120 mg dose produced improvement in PANSS that was intermediate between 80 mg and 160 mg. LS mean change in PANSS exhibited a linear trend relative to dose on treatment days 14, 28, 35, and 42. A time effect rate analysis indicated that 50% of the reduction in PANSS total score observed during acute treatment for each dose group occurred in the first 9 days after starting treatment. Between-study variability in clinical response was evident in the placebo group, but not in the lurasidone group, and was contributed to by demographic covariates (age, weight and race). A log-linear hazard model indicated that patients were more likely to drop out when baseline PANSS scores were higher, and during the initial hospitalization period. However, dropout rate was not correlated with dose of lurasidone.
Conclusions: In this pooled analysis, the effect of lurasidone was described using a linear dose-response model for drug effect, with increased treatment response observed at higher doses of lurasidone. Attrition was not correlated with lurasidone dose.
Reference: PAGE 21 () Abstr 2651 [www.page-meeting.org/?abstract=2651]
Poster: CNS