Bioavailability of triple-bead mixed amphetamine salts compared with a dose-augmentation strategy of mixed amphetamine salts extended release plus mixed amphetamine salts immediate release.

OBJECTIVE: To compare the single-dose pharmacokinetics of triple-bead mixed amphetamine salts (MAS), an oral, once-daily, enhanced extended-release amphetamine formulation, with MAS extended release (MAS XR) (Adderall XR) + MAS immediate release (MAS IR) administered 8 h later. METHODS: This was a phase I, randomized, open-label, single-dose, single-center, two-period, crossover study in healthy adult volunteers designed to evaluate the bioavailability of triple-bead MAS over the course of a full day. Subjects were randomized to triple-bead MAS 37.5 mg or MAS XR 25 mg + MAS IR 12.5 mg administered 8 h later (MAS XR + MAS IR). The reference treatment was designed to mimic the clinical practice of providing extended coverage by supplementing a morning dose of MAS XR with a dose of MAS IR 8 h later in order to increase the duration of action. Plasma was assayed for d-amphetamine and l-amphetamine. Treatment-emergent adverse events (TEAEs), vital signs, electrocardiograms (ECGs), and laboratory data were also collected for safety evaluation. RESULTS: Exposure to d- and l-amphetamine was equivalent between triple-bead MAS and MAS XR + MAS IR based on maximum plasma concentration (Cmax) and area under the plasma concentration-time curve from time 0 to infinity (AUC(0-infinity)). For Cmax, least-squares mean ratios comparing triple-bead MAS with MAS XR + MAS IR were 101.0% and 90.9% for d-amphetamine and l-amphetamine, respectively, and for AUC(0-infinity) were 104.4% and 95.3% for d-amphetamine and l-amphetamine, respectively. Median time to maximum observed plasma concentration (Tmax) values for d-amphetamine and l-amphetamine were 8.0 h for triple-bead MAS and 10.0 h for MAS XR + MAS IR. There were no clinically meaningful differences between the study formulations for TEAEs or laboratory values. One subject experienced an ECG abnormality (asymptomatic premature ventricular contractions) leading to early termination from the study. CONCLUSIONS: In healthy adults, the exposure observed with triple-bead MAS 37.5 mg was bioequivalent to MAS XR 25 mg supplemented by MAS IR 12.5 mg administered 8 h later. These data demonstrate that a single morning dose of triple-bead MAS provides equivalent plasma concentrations to those observed with a dose-augmentation strategy of MAS XR in the morning followed by MAS IR in the afternoon, while minimizing peak-to-trough fluctuations. Triple-bead MAS was also generally well-tolerated in this study.

Dunnett's many-to-one test and least square means.

Dunnett's many-to-one test is used frequently today, especially in dose-finding studies. Using Dunnett's test, the Type I error level for the comparison between the raw mean of the control and the raw means of the study drug groups can be exactly calculated for the normal data. However, this computability depends on the independence of the raw means. Unfortunately, this independence does not exist for the model-based likelihood estimates (least square means) in the cases of ANCOVA and two-way ANOVA models without interaction for unbalanced data. This paper investigates this dependence between the least square means and derives some new procedures to calculate the joint distribution of the statistic for Dunnett's test.