ABSTRACT
This article presents a case study in dealing with robustness investigations and attempts by our analytical laboratory to address these issues without sacrificing valuable time in revamping the method validation prior to submission. A liquid chromatographic method is developed for the analysis of a novel triazinetrione anticoccidial product. The method effectively separates the active pharmaceutical ingredient (API), impurities, and preservatives in the API and product formulation. For much of the validation, the method holds up to the rigorous guidelines of the International Conference of Harmonization, the International Cooperation on Harmonization of Technical Requirements for Registration of Veterinary Medicinal Products, and the Good Manufacturing Practices. However, in analyzing a base-degraded sample one of the impurity peaks yields inconsistent retention times (RTs) during a series of injections. When switching the system to another analytical column from the same supplier, this impurity peak elutes at a different retention window and the remaining peaks in the chromatographic profile remain essentially the same. This RT variation of a single peak in the chromatographic profile is observed with additional columns from the same supplier and from different manufacturing lots. This suitability problem is not encountered with the columns used in the method development stage. The method no longer meets the robustness criteria established for pharmaceutical methods. An investigation is commenced and it is discovered that with the addition of tetrabutylammonium hydroxide to the mobile phases, the impurity peak gives a consistent RT in relation to the active peak. The peak shows comparable RTs relative to that of the API peak with columns of different silica lots and bond lots. All peaks, including the aforementioned impurity peak, are well-resolved under the revised high-performance liquid chromatographic conditions. This temporary solution enables continued submission work for FDA, but the robustness of this method is still a concern. After further investigation, it is determined that inhomogeneity of the active sites on the column's stationary phase is the likely culprit. Fortunately, a new column is found to be more suitable for this method and a column qualification study is initiated.
