Quantitative structure–retention relationships of cannabimimetic aminoalkilindole derivatives and their metabolites

Development of a method of analysis for profiling of the impurities in phenoxymethylpenicillin potassium based on the analytical quality by design concept combined with the degradation mechanism of penicillins

Accurate analysis of all of the impurities present in a substance is critical for controlling the impurity profiles of drugs. Penicillins can easily yield a formidable array of degradation-related impurities (DRIs) with significantly different polarities and charge properties, which renders identifying each one a complicated matter. In this work, phenoxymethylpenicillin potassium (Pen V) was selected to find a way to quickly establish a robust analysis method for the impurity profiling of penicillin. Based on the analytical quality by design (AQbD) concept and the degradation mechanism of the drug, structures of all of the DRIs were first proposed. Then Pen V and its detected DRIs were separated and identified by liquid chromatography-tandem mass spectrometry method (LC-MS). Characteristic fragment ions and mass fragmentation process of Pen V and its detected DRIs were summarized. In addition, a quantitative structure-retention relationship (QSRR) model was constructed to predict the retention times of undetected impurities and to evaluate whether the chromatographic system can separate them. Finally, a stability-indicating high-performance liquid chromatography (HPLC) method was developed that can separate all of the DRIs of Pen V.

Identification of impurities in macrolides by liquid chromatography-mass spectrometric detection and prediction of retention times of impurities by constructing quantitative structure-retention relationship (QSRR)

Macrolides are multicomponent drugs whose impurity control is always a challenge demanding analysis method with good sensitivity and selectivity. Three separate, sensitive, accurate liquid chromatography tandem mass spectrometry methods (LC-MS) were developed for the measurement of three 16-membered ring macrolides (josamycin, josamycin propionate and midecamycin acetate) and related substances in commercial samples. The characteristics of impurities in macrolides were summarized as useful guidance for the impurity analysis of this class of drugs. For each drug, a large number of unknown components have been detected with the high-sensitive MS detector and possible structures of the majority of them were postulated based on the summarized fragmentation rules of 16-membered ring macrolides. A QSRR model was constructed by multilinear regression to predict the retention times of identified impurities which were not detected by the LC-MS methods, without obtaining their reference standards. Satisfactory performance was obtained during leave-one-out cross-validation with a predictive ability (Q²) of 0.95. The generalisation ability of the model was further confirmed by an average error of 2.3% in external prediction. The best QSRR model, based on eight molecular descriptors, exhibited a promising predictive performance and robustness.