Determining the Sources of Variance in the Preparation of Analytical Standards for Chromatographic Analysis of a Lyophilized Peptide Drug Substance by Nested ANOVA Statistical Analysis.

Synthetic peptides used as therapeutic medicines is continuing to grow as an area of focus within the pharmaceutical industry due to specificity and potency. As such, quality control areas need to continue to advance their capabilities to ensure that appropriate analyses are being performed, and that the data generated are both accurate and precise. One area which poses a significant challenge compared with traditional small molecule drug products is having a highly robust, low variability method of quantifying the assay of the active substance. As many peptide therapeutics are formulated as liquid drug products for injection and preparation procedures to make these samples amenable to traditional chromatographic analysis are inherently low variability (i.e., a simple dilution), potential sources of variance derived from the preparation of the analytical standards used to quantify the assay of the product must be investigated. Here, a fully nested ANOVA experimental design was utilized to examine this process. Such a design allowed for multiple variables to be interrogated as well as the potential interplay of such differences. It was determined that sonication of the standards contributed the most variance, while the balance used and scale on which the standard preparation procedure was performed also contributed significantly. Finally, different procedures for introducing the material into a coulometric Karl Fischer (KF) titration device to quantify the water content of the drug substance were compared and showed that indirect quantification by anhydrous methanol extraction is a significantly more variable method than using an Oven KF autosampler.