Presentation of coefficient of variation for bioequivalence sample-size calculation

• Published in: International journal of clinical pharmacology and therapeutics Vol. 55; no. 7; pp. 633 - 638
• Main Authors: Lee, Yi Lin, Mak, Wen Yao, Looi, Irene, Wong, Jia Woei, Yuen, Kah Hay
• Format: Journal Article
• Language: English
• Published: Germany Dustri - Verlag Dr. Karl Feistle GmbH & Co. KG 01.07.2017
• Subjects: Area Under Curve; Clinical Trials as Topic - methods; Clinical Trials as Topic - statistics & numerical data; Data Interpretation, Statistical; Delayed-Action Preparations; Drug Compounding; Humans; Male; Metabolic Clearance Rate; Pharmacokinetics; Sample Size; Studies; Therapeutic Equivalency
• ISSN: 0946-1965
• DOI: 10.5414/CP202731

The current study aimed to further contribute information on intrasubject coefficient of variation (CV) from 43 bioequivalence studies conducted by our center. Consistent with Yuen et al. (2001), current work also attempted to evaluate the effect of different parameters (AUC , AUC , and C ) used in the estimation of the study power. Furthermore, we have estimated the number of subjects required for each study by looking at the values of intrasubject CV of AUC and have also taken into consideration the minimum sample-size requirement set by the US FDA. A total of 37 immediate-release and 6 extended-release formulations from 28 different active pharmaceutical ingredients (APIs) were evaluated. Out of the total number of studies conducted, 10 studies did not achieve satisfactory statistical power on two or more parameters; 4 studies consistently scored poorly across all three parameters. In general, intrasubject CV values calculated from C were more variable compared to either AUC and AUC . 20 out of 43 studies did not achieve more than 80% power when the value was calculated from C value, compared to only 11 (AUC ) and 8 (AUC ) studies. This finding is consistent with Steinijans et al. (1995) [2] and Yuen et al. (2001) [3]. In conclusion, the CV values obtained from AUC and AUC were similar, while those derived from C were consistently more variable. Hence, CV derived from AUC instead of C should be used in sample-size calculation to achieve a sufficient, yet practical, test power.
.