Prediction of saxagliptin stability using a new approach based on Partial Least Squares and Design of Experiments.

The objective of this study was to assess the possibility of applying Partial Least Squares (PLS) statistics with the use of experimental design approach towards stability evaluation of the Saxagliptin drug product. The influences of temperature, time, dose, packaging, batch, and oxygen protection were analyzed for identification of critical factors responsible for degradation of saxagliptin and prediction of impurity levels at various storage conditions. Predicted levels of the impurity DP-2 were lower for at least 0.2 % when the drug product was protected from oxygen after its manufacture. Additionally, the PLS model revealed that the lower strength is at least twice less stable concerning impurity DP-1. Based on this analysis shelf life for Zone II was proposed at 24 months with high reliability. Comparison of the PLS model estimates with the measured stability data at shelf life revealed good predictive ability of the developed model. Moreover, PLS predictions of DP-1 and Total impurities were more accurate than those obtained with a standard linear least squares regression, while DP-2 predictions were at least as accurate. We can thus propose a more extensive use of this approach for stability evaluation of pharmaceuticals.