Development and validation of a GCMS method for the quantitation of potential genotoxic impurities in pharmaceutical substances

Abstract

The company LOBA Feinchemie GmbH is specialized in the production of naphazoline hydrochloride and naphazoline nitrate. Those salts are used as active pharmaceutical ingredients in nose sprays and eye drops. The used raw material 1-naphthylacetonitrile is commonly synthesized from the potential genotoxic impurity 1-chloromethylnaphthalene. In order to guarantee, that the amount of the potential genotoxic impurity is at or below the acceptable limit of 48 ppm in the product, a gas chromatography-mass spectrometry (GCMS) method was developed at TU Wien, which allows the absolute quantitation of 1-chloromethylnaphthalene in the raw material 1-naphthylacetonitrile. A five-point standard addition procedure was used for the quantitation. The analytes of interest cover a broad range of polarity, which is why a non-polar stationary phase had to be combined with a polar solvent. Since 1-chloromethylnaphthalene reacts in the presence of methanol and ethanol, isopropyl alcohol was chosen as the solvent. In order to compare different individual sample measurements and to compensate for all possible volume errors and variations in the function of the instrument, naphthalene-d8 was implemented as internal standard. The used raw material 1-naphthylacetonitrile contains several non-volatile impurities, which create active sites in the inlet liner and the beginning of the column. Those active sites must be saturated prior to each standard addition procedure to establish comparable measurement conditions. The developed method was transferred to and validated at the receiving laboratory. A quantitation limit of 0.97 ppm for 1-chloromethylnaphthalene was determined and the linearity of the standard addition curve was shown in the range of 2 - 10 ppm, with a coefficient of determination (R2) 0.9939 (Acceptance criteria: R2 0.98). The obtained relative standard deviation (RSD) of the repeatability experiments was 6.13 % (Acceptance criteria: RSD 20 %) and recovery rates between 92.5 % - 100.5 % were determined in the course of the accuracy experiments (Acceptance criteria: recovery rate between 80 % - 120 %). Specificity and robustness of the method were also successfully shown in the course of the validation.