Statistical considerations on real time and extended controlled temperature conditions (ECTC) stability data analysis of vaccines

• Published in: Vaccine 2023-10, Vol.41 (42), p.6206-6214
• Main Authors: Kim, Deok Ryun, You, Young Ae, Ahn, Hyeon Seon, Park, Eun Lyeong, Lim, Jacqueline KyungAh, Song, Katerina Rok, Chon, Yun, Lynch, Julia
• Format: Article
• Language: English
• Subjects: Antigens; Cold; Cold storage; Confidence intervals; Data analysis; Data collection; Drug Stability; Drug Storage - methods; Extended controlled temperature conditions; Manufacturers; Material requirements planning; Mathematical models; Minimum release potency; Real time; Real-time real conditions; Refrigeration; Regression analysis; Regression models; Samples; Source programs; Stability; Stability analysis; Statistical analysis; Statistical models; Temperature; Vaccine; Vaccines
• ISSN: 0264-410X; 1873-2518; 1873-2518
• DOI: 10.1016/j.vaccine.2023.08.012

Although maintaining vaccines in a strict cold chain has cost and logistical implications in low- and middle-income countries, only a few vaccines have obtained approval for extended controlled temperature conditions (ECTC) application, which permits the administration of vaccines after storage outside of the cold chain for a defined period. We developed a methodology to evaluate stability data and calculate minimum release potency (MRP) in support of ECTC application. The methodology is focused on statistical considerations consisting of stability data collection, statistical analysis plan, statistical modelling, and statistical report. It uses mock stability data from a hypothetical product and may serve as a helpful guide for other products. The statistical data analysis is performed using the R program which is an open-source program and validated using the SAS software. We developed a stability data testing scheme that included 24 lots with six-time points for up to 24 months under real-time and real condition (RT) in the cold chain samples stored at 2–8 °C and 12 lots with six timepoints for 14 days under ECTC samples stored at 40 °C. The log-transformed stability data met the linear regression assumptions and were poolable from representative lots with no significant lot variation. The linear regression analysis model with a common slope and intercept confirmed the stable antigen content over time under RT and ECTC by the mean regression line and 95% confidence interval. Based on the fitted models and the estimated coefficients, the antigen content value of 966 was derived as the MRP under RT for 24 months followed by 14 days under ECTC. The presented framework of statistical considerations, with practical methods and R program codes to perform statistical analysis, may serve as a guide for developing the CTC data for a vaccine’s stability evaluation prospectively.