2015 QPRC

An Innovative Simulation Method to Predict the Effectiveness of Viral Inactivation for a Vaccine Product Account for Variability in Multiple Process Steps

Abstract
In a viral vaccine product development, gamma irradiation will be used to inactivate the live virus in the final product. The goal of the inactivation is to achieve <1 virus/vial in the final product. Many factors can affect the inactivation process: initial viral load, infectivity loss during lyophilization (lyo) at different shelves in lyo cabinet, variability of rate of inactivation of gamma radiation, and the radiation dose each vial actually receives due to their position in the gamma tote. A simulation method was used to simulate data from the initial bulk viral load, then the distribution was fed into next process step, the resulting distribution was fed into its next process step until the end process. Three lots of bulk product were simulated to represent low, middle, and high initial viral load. The loss of infectivity is simulated based on 2% loss shelf in lyo cabinet. The gamma dose received was a bimodal mixture normal distribution with average inactivation rate of 0.42 log unit/kGy. The remaining virus in each vial was calculated after all the process steps. This approach provided an empirical method to simulate the uncertainty at each step without assuming the worst case, thus provided more real estimate of pass rate for the final product.