Analysis of Nitrosamine Impurities in Drugs by GC-MS/MS

Abstract

Testing APIs and drug products for nitrosamine impurities requires reliable separation of target analytes and detection limits in the low ppb range. In method developed here, Pro EZGC modeling software was used to quickly identify a recommended column and conditions prior to beginning lab work. Subsequent testing in the laboratory demonstrated that the method provided good resolution and low detection limits in run times of less than 10 minutes.

Introduction

Recently, several “sartan” class drugs have been found to contain nitrosamines, which can be carcinogenic at low levels. This has led to the FDA to require testing of active pharmaceutical ingredients (APIs) and drug products when there is a risk of nitrosamine formation. The FDA has released example methods for both LC and GC for a variety of these nitrosamines, requiring detection limits in the low ppb range [1]. For GC, analysis of nitrosamine impurities can be carried out using either headspace sample introduction for a few of the more volatile nitrosamines, or direct liquid injection, which is effective over a wider volatility range. While direct injection may be preferred due to its simplicity over headspace, it also introduces more matrix into the system. Because of the low levels of detection required, combined with the potential for matrix interferences, most methods recommend using GC with triple quadruple mass spectrometric detection (GC-MS/MS).

In this work, we used the Pro EZGC chromatogram modeler to first identify an analytical column and conditions for the optimized separation of the nitrosamines in Table I. Method performance was then confirmed in the laboratory and resulted in good linearity (R²>0.996) and low ppb level LODs (below 3 ppb) for all analytes, except for N-nitroso-N-methyl-4-aminobutyric acid (NMBA) and N-nitrosodiethanolamine (NDELA).

Table I: Target Compounds for the Analysis of Nitrosamine Impurities in Drugs

Experimental

Standard Preparation

A mixed nitrosamines calibration standard (1000 ppm) prepared in dichloromethane was further diluted in methanol to make the working standard solutions. Seven-point calibration curves covered a range of 1.25–120 ppb for most analytes and 3.8–365 ppb for NMBA, which had much lower response than the other compounds. The response for N-Nitrosodiethanolamine (NDELA) was so minimal that this compound was not tested further.

Chromatographic Method

Pro EZGC chromatogram modeling software was used to select an analytical column and identify method parameters that would optimize the nitrosamine separations prior to performing lab work. The final chromatographic conditions generated by the Pro EZGC software are presented in Figure 1. Additional parameters used during confirmation testing in the laboratory are given in Figure 2 and Table II. Use of the chromatogram modeling software greatly reduced method development time in the lab.

Table II: Ion Transitions