Labs should be including a broad set of CYP2C9 alleles in their own genotype tests, according to new joint recommendations published May 7 by the Association for Molecular Pathology (AMP) and the College of American Pathologists (CAP) in the Journal of Molecular Diagnostics.
The recommendations set two standards for the design of laboratory tests for CYP2C9 alleles and their defining variants: Tier 1 includes the minimum alleles that experts believe should be included, while tier 2 alleles are optional -- at least for now, as the field is moving rapidly.
The AMP stressed the need to standardize testing for genetic deficiencies in the cytochrome P450 enzyme CYP2C9, which is routinely performed to identify patients who are unable to metabolize a broad range of commonly used medications such as the old-time anticoagulant warfarin, the antiplatelet drug clopidogrel (Plavix, Pfizer), and the antipsychotic aripiprazole (Abilify, Otsuka). The metabolism of antiepileptic drugs, antidepressants, and proton pump inhibitors is also affected.
"CYP2C9 testing has the potential to guide clinicians when considering the use of at least 12 different medications, some of which are among the top 200 most prescribed medications" in the U.S., wrote a team led by Indiana University molecular genetics professor Victoria Pratt, PhD.
"Currently, CYP2C9 tests can produce variable results due to factors such as the choice of tested alleles, targeted testing of populations with varying ethnic backgrounds, as well as the technical performance of the various platforms," Pratt, who is also president of the AMP, said in a statement. "These recommendations are intended to facilitate testing by laboratories and improve genotyping concordance across laboratories."
| CYP2C9 clinical testing recommendations | ||
| Allele | Allele functional status | Multiethnic allele frequency |
| Tier 1 variant alleles and their defining variants that meet three criteria. They have been well characterized and shown to significantly affect the function of the protein and/or gene, leading to an alteration in a drug response phenotype; have an appreciable minor allele frequency in a population/ethnicity group; and have publicly available reference materials. Should be included in all tests. | ||
| *2 | Decreased function | 0%-12% |
| *3 | Decreased function | 1%-11% |
| *5 | Decreased function | 0%-1% |
| *6 | No function | 0%-1% |
| *8 | Decreased function | 0%-5% |
| *11 | Decreased function | 0%-2% |
| Tier 2 variant alleles/variants meet at least one but not all three of the tier 1 criteria. | ||
| *12 | Decreased function | 0%-0.3% |
| *13 | Decreased function | 0%-0.2% |
| *15 | No function | 0%-0.01% |
Addressing variations in testing
The two-tiered guidance on CYP2C9 panel tests is part of a series of reports from the AMP Clinical Practice Committee Pharmacogenetics (PGx) Working Group addressing CYP2C9 alleles. The first AMP report advised on what to include in CYP2C19 panels and was published last year (J Mol Diagn, May 2018, Vol. 20:3, pp. 269-276). A third report with recommendations for CYP2C9 alleles and additional warfarin sensitivity-associated genes/alleles is slated for release early in 2020.
Clinical pharmacogenetic testing varies across different laboratories, and there are discrepant results, "mostly attributable to assay design," Pratt et al noted.
"According to the Genetic Testing Registry, and AMP Test Directory, the alleles included in CYP2C9 genotyping tests offered by clinical laboratories in the United States range from a few targeted alleles to interrogation of the entire coding region, and the techniques include targeted genotyping, bidirectional Sanger sequencing, next-generation sequencing (NGS), whole genome sequencing (WGS), or whole exome sequencing (WES), with or without deletion/duplication analysis," the authors wrote.
There are 10 sets of guidelines on CYP2C9 metabolized medications, but they "do not explicitly recommend specific variant alleles for clinical laboratories to include in CYP2C9 genotyping panels," and information about testing methods on U.S. Food and Drug Administration (FDA) labels is very limited, the authors noted. As for commercial platforms, Pratt and colleagues cited several available for CYP2C9 genotyping, including iPlex ADME (Agena Bioscience, research use only); PharmacoScan, Affymetrix (Thermo Fisher Scientfic, research use only); Infiniti (AutoGenomics, CE Mark); and eSensor (GenMark Diagnostics, FDA cleared).
Patients with African ancestry not well served
The CYP2C9*2 and *3 alleles are the most common ones interrogated by the commercial platforms. But while these two account for 98% to 100% of the variations associated with decreased function in European, Middle Eastern, and Asian populations, they catch only 25% in people with African ancestry, Pratt et al noted.
"Thus, genotyping for only the CYP2C9*2 and *3 variants will not detect the majority of CYP2C9 genomic variation leading to decreased enzymatic activity in populations with African ancestry," the authors wrote.
In the case of warfarin dosing, not getting the genetic testing right has implications for the occurrence of complications such as venous thromboembolism, major bleeding, supratherapeutic bleeding, and death. But FDA prescribing information doesn't include alleles that are important for warfarin response in patients of African descent.
"This is especially relevant for ethnically diverse and admixed populations," they wrote. "African alleles may be present in individuals that may not consider themselves of African descent; thus, determination of African ethnicity may not be required to test for these alleles."
"In addition, current CPIC [Clinical Pharmacogenetics Implementation Consortium] guidelines for warfarin dosing underscore the importance of accounting for African alleles and provide separate recommendations for patients of African versus non-African ancestry," the group wrote. "As such, we recommend that laboratories include all [tier] 1 alleles, including the major African alleles, in clinical CYP2C9 testing."
Of the commercially available platforms for CYP2C9 genotyping, only two had all tier 1 alleles recommended by the working group, including the alleles important for those with African ancestry -- *5, *6, *8, and *11 -- as well as the tier 2 alleles. However, some of the systems may be customized to include all alleles in both tiers, according to the working group. In January, the FDA approved a direct-to-consumer pharmacogenetic test from 23andMe that screens for CYP2C9, but while this covers most of the tier 1 alleles, it does not include *8 and *11, which are found in about 8% of people with African ancestry.
"The PGx [working group] is aware that the recommendations to include the alleles more prevalent among African and African-American populations may be difficult to implement with currently available genotyping platforms," they wrote. "However, the [working group] concluded that failure to include these alleles could lead to inaccurate CYP2C9 phenotype prediction among individuals with known or unknown African ancestry and may contribute to existing health care disparities in these populations."
"Implementation of this recommendation document is at the discretion of the laboratory," they added.
