Comparative study on the bioequivalence of two formulations of pravastatin. Data from a crossover, randomised, open-label bioequivalence study in healthy volunteers.

An open-label, randomised, crossover single-dose study, using two periods, two sequences, with a minimum washout period of 7 days, was conducted in order to assess the comparative bioavailability of a pravastatin (CAS 81131-70-6) 40 mg formulation and that of a reference formulation. Blood samples were collected up to +14 h post dosing, the plasma was separated and pravastatin concentrations were determined by high-performance liquid chromatographic method with tandem mass spectrometry detection (HPLC-MS/MS), with a lower limit of quantification of 0.40 ng/mL. Bioequivalence analyses were conducted using two models. The main analysis was done according to a general linear model (model I) using formulation, period and sequence as fixed model effects, and subject(sequence) and residual as random effects; gender-related differences were investigated using ANOVA (model II), including formulation, period, sequence, gender, sequence-by-gender interaction and gender-by-formulation interaction as fixed model effects and subject within sequence-by-gender interaction and residual as random effect. Mean values of the individual Cmax were 126.73 +/- 61.99 ng/mL and 123.52 +/- 52.78 ng/mL for the test and reference, spectively. Mean +/- SD total area under the curve up to the last measurable concentration (AUClast) was 224.26 +/- 104.74 ng x h/mL for the test formulation and 216.55 +/- 80.30 ng x h/mL for the reference formulation. Mean +/- SD total area under the curve (AUCinf) was 226.72 +/- 104.69 ng x h/mL for the test formulation and 218.81 +/- 79.95 ng x h/mL for the reference. Terminal elimination half-life was 2.15 +/- 0.85 h for test and 1.97 +/- 0.54 h for the reference. Ninety percent confidence intervals were comprised within the bioequivalence acceptance criteria (80-125 %) for all of the parameters analysed, both using model I and model II. Model II returned a statistically significant gender effect (p < 0.05) for Cmax, AUClast and AUCinf but the effect became non-significant (p > 0.05) when the dose was adjusted per body weight for all three parameters. The comparison between male and female mean body weight showed a significant difference: p = 0.03, 95 % confidence intervals 68.27-76.93 kg (men), 56.84-60.61 kg (women). These results suggest that the effect of gender in the bioequivalence analysis in model I could be due to a difference in body weight between males and females. Both formulations were shown to be bioequivalent in terms of rate and extent of absorption, irrespectively of the model used.