Bioequivalence studies for three formulations of a recombinant human growth hormone: challenges and lessons learned.

TLDR: Two bioequivalence studies in healthy volunteers comparing new formulations of the recombinant human growth hormone (rhGH) Nutropin AQ (somatropin [rDNA origin] injection) with the currently marketed formulation were conducted to extend available dosing options, and positive results taught that demonstrating BE is an issue not only of pharmacology, but also of study methodology and execution.

Abstract: Two bioequivalence (BE) studies in healthy volunteers comparing new formulations of the recombinant human growth hormone (rhGH) Nutropin AQ (somatropin [rDNA origin] injection; Genentech, Inc., South San Francisco, CA) with the currently marketed formulation (5 mg/mL) were conducted to extend available dosing options. All formulations were administered by subcutaneous (SC) injection ranging in volume from 0.25 to 1.0 mL depending on the formulation concentration. Study A was a 2-period crossover design to assess the BE of 5 and 10 mg/mL. The estimate for relative bioavailability (AUC(0-24 h)) was within the prespecified BE interval (0.80-1.25). However, while the C(max) estimate (1.17) was contained within the range for BE, the 90% CI (0.986-1.38) extended beyond the prespecified BE interval. As a result, Study A failed to show BE between the 5 and 10mg/mL formulations. Review of the data showed unexpected increased variability in the observed C(max). Further review of individual data suggested that in 4 subjects, the GH concentration profile of 1 of the 2 injections closely resembled the absorption kinetics of an intramuscular injection rather than an SC injection. Because study conduct may have contributed to these results, we performed a second study, Study B. This study incorporated injection technique training, a defined injection site, and a larger sample size to accommodate variability. It also included a third formulation, creating a 3-period crossover design to assess the BE of 2.5, 5, and 10 mg/mL. Study B results demonstrated BE of the new 2.5- and 10-mg/mL formulations to the reference 5-mg/mL formulation, and BE to each other, with all 90% CIs within the BE range of 0.80 to 1.25. Thus the challenge of recognizing that design issues could affect outcomes gave us the tools to perform a second study, and the positive results taught us that demonstrating BE is an issue not only of pharmacology, but also of study methodology and execution.