Individual and mean plasma concentrations, as well as the plots of the plasma levels for all subjects versus time, were graphically displayed for three treatments. Ln-transformed AUC0–t , AUC0–inf and C max were analysed using general linear model (GLM) procedure selleck products in SAS® following the method A recommended by the EMA (CHMP Pharmacokinetics Working Party [PKWP] EMA/618604/2008 Rev. 3). The statistical model included sequence, period, treatment and subject within sequence as fixed factors. The sequence effect was tested using the subject-within-sequence effect as the error term. The treatment

and period effects were tested against the residual mean square error. Within-subject coefficient of variation (CVWR) was calculated for the reference Adavosertib product using analysis of variance (ANOVA), on reference data only, with sequence, subject within sequence, and period as fixed effects. The point estimate and the 90 % geometric confidence interval INCB024360 concentration for the test-to-reference geometric mean ratio (T/R) were calculated for AUC0–t , AUC0–inf and C max using the least-squares means statement. K el and T ½ el were also analysed using the GLM Procedure. Wilcoxon’s test was performed on the mean T max for both treatments. All statistical tests

were performed at the alpha level of 0.05. According to the regulatory requirements [4] translated into the study protocol, the hypothesis of bioequivalence between a generic medicinal product and a reference medicinal product is accepted if the 90 % geometric confidence intervals of the ratio of least-squares means of the test to reference product of ln-transformed AUC0–t is within the acceptance range of Non-specific serine/threonine protein kinase 80.00–125.00 %. For C max, the protocol established a scaled average bioequivalence approach. This approach is based on the CVWR: if the CVWR is inferior or equal to 30 % (≤30 %), the 90 % geometric confidence intervals of the ratio T/R of least-squares means of the ln-transformed C max should be within the acceptable range of 80.00–125.00 % to conclude bioequivalence. On the other hand, if the CVWR for the

reference product was superior to 30 % (>30 %) for C max, the bioequivalence acceptance limits for this pharmacokinetic parameter had to be scaled to the within-subject variability of the reference product (to a maximum of 69.84–143.19 %). For scaled average bioequivalence, the applicant should justify that the calculated CVWR is a reliable estimate and that it is not the result of outliers. Therefore, a box plot analysis using the studentized intra-subject residuals from the ANOVA model including only data for the reference treatment was done using the univariate procedure in SAS®. A box plot was constructed from studentized intra-subject residuals corresponding to the first administration of reference product in each subject. Values that were further away from the box by more than three interquartile ranges were considered outlying observations and these values are indicated by an asterisk in the box plot.