Analysis of Alkyl, Benzyl and Organohalides as Potential Genotoxic Impurities in Active Pharmaceutical Ingredients (APIs)
While removing all traces of genotoxic impurities from APIs would be optimal, it is often not possible. Therefore the maximum limitation is the next best option. Alkyl halides, benzyl halides and other reactive organohalides have structural characteristics which can result in mutagenicity. Analytical techniques for these compounds have gained a great deal of momentum recently, with new information surfacing on the different methods of analysis. Separation, identification and quantification of these compounds have proven to be difficult because of matrix interference from the API, degradation of the impurity, and the possible generation of false-positive impurity compounds in some analytical methods. Benzyl chloride, for example, if shown to be Ames negative it could be treated as a normal impurity and a 500 ppm detection limit would be suitable. However if benzyl chloride was shown to be Ames positive would require a technique that had low detection limits, around 10 ppm. The lower detection limits is a common difficulty faced by many researchers today.
Gas chromatography (GC) coupled with mass spectroscopy (MS), flame ionization detection (FID) or electron capture detection (ECD) have enhanced detection of these lower quantities. When direct injection lacks concentration or results in matrix interference, headspace or solid phase extraction (SPE) are commonly employed. High performance liquid chromatography (HPLC) plays a key role when the analyte is temperature sensitive or insufficiently volatile. Thin layer chromatography (TLC) and capillary electrophoresis (CE) are a few techniques that have shown to have low value resulting from low sensitivity. Supercritical fluid chromatography (SFC) could be a good method but there is little review of this technique.
Strategies on Controlling Alkylating Agents
Chemists and chemical engineers can try to eliminate or minimize the use of reactive reagents, starting materials and intermediates. The FDA chose a computational toxicological approach for evaluating the genotoxicity and carcinogenicity of an impurity based on its structure-activity relationship. A recommendation that structurally related groups should be categorized under the same exposure limit as if it were a single compound. This would lower the LOQ needed of each of the individual impurities in that group. Ames bacterial mutagenicity test is a valuable tool in evaluating the effects of these impurities.
Gas Chromatography (GC)
GC can be coupled with multiple detectors such as FID, ECD or MS. Direct injection, headspace and SPE are a few sample preparation methods that also lend to the versatility of GC. Direct injection method was proved beneficial because of less chance of sample lost.
- Headspace GC/ECD was used to evaluate 23 different alkyl/aryl halides because of its usefulness in determination of levels of volatile compounds. The liquid phase had a ratio of 70/30 (v/v) of water and dimethylsulphoxide. The headspace was optimized at a ratio of 20/1 (v/v) gas/liquid phase to create a saturated environment. Accuracy of this generic technique showed recoveries of 80-120% at 1000 ng/ml, with detection limits ranging from 2.5-290 ng/ml. Iodide giving the best results, paralleling other research data.
- Analysis of trace levels of carbonic acid chloromethyl tetrahydropyran-4-yl ester (CCMTHP) in a beta-lactam API was unsuccessful when combined with SPE and direct injection because of similar physiochemical properties between the impurity and the API. Also derivatization was not used because of fear of low or multiple yields.
- When analyzing multiple detection techniques such as MS, FID and ECD, one researcher found that SIM at m/z of 49 amu provided the best sensitivity at 10 ppm using external standards. Linearity was demonstrated over 10-1000 ppm.
- GC/FID was combined with automated headspace and solid phase micro-extraction to concentrate the analytes chosen. This method was linear from 0.1-0.5 µg/ml for chloroethyl methyl ether (CEME) and 0.01-10.0 µg/ml for benzyl chloride (BC). LOQs were calculated at 10 and 0.9 ng/ml for CEME and BC respectively. LODs were 4.0 and 0.3 ng/ml for CEME and BC respectively. Recoveries ranged from 84.8-90.8% for CEME and from 74.1-77.2% for BC both over the concentration range of 0.1-10.0 µg/ml. Low recoveries were found to be a result of matrix interference.
- Liquid/liquid extraction was successfully combined with GC and direct injection for the determination chlorohydrins levels in an API. LOD and LOQ were determined to be 0.09 and 0.31 ppm respectively. On column formation of the epoxide from chlorohydrin was lessened by increasing the initial column pressure.
HPLC is a useful method for the determination of alkylating impurities, predominantly on reverse phase with MS detection. Detection modes can vary with increased selectivity using SIM, single quadropole and selective reaction monitoring (SRM). Ionization modes range from electron spray ionization (ESI) and atmosphere pressure chemical ionization (APCI). ESI has shown to have better response for ionic compounds.
- HPLC/ESI/MS was shown to have better response compared to APCI when measuring N,N-dimethyl aminoethyl chloride (DMC) in diltiazem hydrochloride with SIM at m/z 108. Interfering API was removed in reverse phase HPLC and the polar analyte was determined using ion exchange chromatography. Linearity was shown from 0.2-10 ppm.
- HPLC/ESI/MS in reverse phrase with a linear gradient at 40 over 18 minutes was used to determine residual levels of an alkyl bromide in a secondary amine API. Only one out of the three analytes showed acceptable recoveries (88%) which was attributed to high reactivity of the other two impurities. To reduce reactivity a chilled auto sampler at 15, mixing the sample by vortexing. Recoveries were increased to 90% for all analytes with an LOD of 10 ppm.
- Negative ion APCI with SIM was used for determination of three residual alkyl halides. Alkyl iodides again showed the best response.
- HPLC/MS/SIM with a m/z of 689.6 in reverse phase and gradient mobile phase at 60 was employed. The method showed linearity over 0.2-18 ppm with an LOD and LOQ of 0.3 and 0.1 ppm respectively. Recoveries ranged from 93.4-104.2% over 0.2-25 ppm.
- HPLC/UV was used by several different research groups, many using single wave detection. This technique showed linearity, high recoveries and useful LOD and LOQ values. Some of the analytes chosen were DACT, CAPA, a chloromethyl alkylating impurity and an impurity in nefadzone hydrochloride, an antidepressant drug
Thin Layer Chromatography (TLC)
There are few publications describing the use of TLC for determining residual alkylating agents at low levels. Befits range from minimum clean up procedures required and having a range of specific derivatization techniques available. TLC was used to determine residual sulfur and nitrogen mustards in multiple substrates. Achievable results were reported in the low microgram range.
Capillary Electrophoresis (CE) and other electrical separation techniques
Several different electrically driven separation techniques were used to test the selectivity of alkyl chloride and alkyl bromide. The LOQ for the analytes was determined to be 0.05%. Linearity was also achieved a correlation coefficient of 0.985. Low sensitivity proved to be a problem.
While initially GC analysis of alkylating agents would rely on their volatility combined with FID, more researchers are using MS methods for improved selectivity. Reverse phase HPLC is commonly used for compounds with low volatility and high temperature sensitivity, preventing them from being analyzed with GC. A shift from single wave UV detection towards MS has been seen, accompanied be the need for lower detection limits. Extraction and/or pre-concentration techniques such as headspace have been used to reduce substrate interference. Reduced energy mixing such as vortexing and reduced temperature storage has shown to improve stability. TLC, CZE, MEKC, MEEKC and CE have shown to be incapable of producing the sensitivity and selectivity required in the determination of the low levels of analyte.
Elder, D. P., Lipczynski, A. M., & Teasdale, A. (2008). Control and analysis of alkyl and benzyl halides and other related reactive organohalides as potential genotoxic impurities in active pharmaceutical ingredients (APIs). Journal of Pharmaceutical and Biomedical Analysis, 48(3), 497-507.
Summarized by R. Jarvis