Water Activity as a Primary Indicator of Moisture in Lyophilized Products: Linking Non-Destructive Headspace Analysis...

Recently, the pharmaceutical industry has recognized the benefits of assessing moisture by measuring the water activity of the product instead of its total water content. USP <922> has introduced water activity as a metric for determining the free, or "active", water that is available in a pharmaceutical product that can participate in undesired reactions such as hydrolysis or create an environment that supports microbiological growth. Traditional water content determination methods for lyophilized products, such as Karl Fischer titration, are time-consuming and measure the total water content which includes both the "active" and "inactive" water populations. In addition, the destructive nature of Karl-Fischer titration means that the moisture analysis of statistically relevant sample sets of lyophilized drug products is typically avoided, resulting in limited insight into the drying process and creating challenges for accurate determination of moisture specifications for product stability. Moisture assessment using water activity measurements, as mentioned in USP <922>, can be accomplished using newer non-destructive methods such as frequency modulated spectroscopy (FMS) headspace analysis. A moisture sorption isotherm can be used to bridge the measurements between the traditional Karl-Fischer total water content and the headspace water activity by establishing the relationship between the two for a particular drug product.In this study, a lyophilized product already in production was used as a model to evaluate this approach. Using a moisture sorption isotherm generated using a Dynamic Vapor Sorption (DVS) instrument, good correlation of the total water content of the lyophilized product was observed between the experimental Karl Fischer (KF) results and the values inferred from the FMS-Water Activity measurements. Due to the non-destructive nature of the headspace water activity analysis, rapid acquisition of a statistically relevant number of samples from three production batches were compared to demonstrate the robustness of this method for assessing moisture.