CYP3A4 and CYP3A5 Genotyping Recommendations

Abstract

The goals of the Association for Molecular Pathology Clinical Practice Committee's Pharmacogenomics (PGx) Working Group are to define the key attributes of pharmacogenetic alleles recommended for clinical testing and a minimum set of variants that should be included in clinical PGx genotyping assays. This document series provides recommendations for a minimum panel of variant alleles (tier 1) and an extended panel of variant alleles (tier 2) that will aid clinical laboratories when designing assays for PGx testing. The Association for Molecular Pathology PGx Working Group considered functional impact of the variant alleles, allele frequencies in multiethnic populations, the availability of reference materials, and other technical considerations for PGx testing when developing these recommendations. The goal of this Working Group is to promote standardization of PGx gene/allele testing across clinical laboratories. This document will focus on clinical CYP3A4 and CYP3A5 PGx testing that may be applied to all CYP3A4- and CYP3A5-related medications. These recommendations are not to be interpreted as prescriptive but to provide a reference guide.

Clinical providers can use pharmacogenomics (PGx) testing to facilitate medication selection and prescribe appropriate doses for their patients for certain drugs with sufficient evidence to support clinical implementation. There is a wide spectrum of variant alleles or variants that are interrogated by clinical PGx tests (https://www.ncbi.nlm.nih.gov/gtr, last accessed February 10, 2023).^1,2 Some tests use comprehensive genotyping panels or sequencing to detect many variants in a pharmacogene, whereas others are designed to detect only a limited number of variants and may not identify variants that are most relevant to clinical care. This can lead to an individual's genotype being reported as a default ∗1 normal function allele or no variants detected, although clinically relevant variants that were not interrogated may be unknowingly present. Ultimately, these differences in test design may cause discrepancies in interpretation, complicate advancement and clinical implementation of PGx testing, and ultimately impact patient care. It is also common for clinical laboratories to design assays without including variants that are preferentially present in population(s) with certain ancestry backgrounds, which potentially leads to underuse and/or incorrect interpretation of PGx testing in such populations. Until recently, there has been little effort to standardize the specific variants that should be included in clinical PGx tests.

The Association for Molecular Pathology (AMP) PGx Working Group has developed a series of documents that recommend a minimum set of variants and alleles to include in clinical PGx tests to facilitate standardization across laboratories and ensure that the most clinically relevant variants and alleles are included in clinical PGx tests. The previous documents covered CYP2C19,³ CYP2C9,⁴ genes important for warfarin PGx testing,⁵ CYP2D6,⁶ and TPMT and NUDT15.⁷ This document extends this series by focusing on two additional cytochrome P450 genes in the CYP3A subfamily of isoenzymes, CYP3A4 and CYP3A5.

This document is intended to provide guidance to clinical laboratorians and manufacturers who develop, validate, and/or offer clinical CYP3A4 and CYP3A5 genotyping assays. This document should be implemented together with other relevant clinical guidelines, including those published by the Clinical Pharmacogenetics Implementation Consortium (CPIC) and the Dutch Pharmacogenetics Working Group (DPWG), both of which focus primarily on the interpretation of PGx test results and therapeutic recommendations for specific drug-gene pairs (https://www.pharmgkb.org/guidelineAnnotations, last accessed February 10, 2023). The star (∗) allele definitions included in the AMP PGx Working Group recommendations are as defined by the Pharmacogene Variation Consortium.^8,9

The AMP PGx Working Group uses a two-tier strategy for selection criteria in recommending PGx variants for clinical testing.3, 4, 5, 6 Briefly, tier 1 recommended alleles are those that meet all the following criteria: i) have a well-characterized effect on the function of the protein and/or gene expression; ii) have an appreciable minor allele frequency (MAF) in a population/genetic ancestry; iii) have publicly available reference materials (RMs); and iv) are technically feasible for clinical laboratories to interrogate using standard molecular testing methods. Tier 2 recommended variant alleles include those that were considered due to having a well-characterized function and/or appreciable MAF but did not meet all the tier 1 criteria. The tier 2 alleles may be upgraded to tier 1 alleles in the future if additional information, RM(s), or advances in testing technology become available.

The human cytochrome P450 family 3 subfamily A (CYP3A) accounts for approximately 30% of the total CYP450 enzyme content in the human liver.¹⁰ CYP3A serves an important role in the metabolic transformation of a wide variety of compounds, including drugs, corticosteroids, xenobiotics, and carcinogens. CYP3A enzymes are important for metabolizing approximately 50% of marketed drugs,10, 11, 12 including fentanyl, midazolam, quetiapine, paclitaxel, statins, and immunosuppressants. CYP3A4 is the major isoform expressed in most individuals; however, CYP3A5 may contribute to total CYP3A activity because the two isoforms have overlapping substrate specificities (eg, cyclosporine and fentanyl).¹³ CYP3A5 is the primary extrahepatic CYP3A isoform and thus may contribute also to tissue-specific CYP3A metabolism.¹³

The CYP3A4 and CYP3A5 genes are located in a CYP3A gene locus on chromosome 7. The CYP3A4/5 genes are oriented on the negative strand of the chromosome, and therefore the coding DNA sequence is the reverse complement of the human reference genome sequence.

CYP3A4

Because of its abundant expression in both the liver and small intestine,¹⁴ CYP3A4 contributes significantly to the first-pass and systemic metabolism of substrate drugs. Hence, CYP3A4 is an important determinant for oral bioavailability and systemic clearance and thereby systemic drug exposure. There is substantial interindividual variability in CYP3A4 enzyme activity because of genetic variability, environmental factors, disease state, and comedications (ie, induction and inhibition) (https://medicine.iu.edu/internal-medicine/specialties/clinical-pharmacology/drug-interaction-flockhart-table, last accessed February 10, 2023).

Analyzing sequence variation in CYP genes to predict the phenotype or enzymatic activity and adjust dosing accordingly is applied clinically for several genes associated with altered drug response. Genetic variants in CYP3A4 were thought to have a limited contribution to the observed variability in activity because of the unimodal distribution of enzyme activity¹⁵ and wide range of hepatic protein expression.¹³ In 2011, the CYP3A4∗22 intron 6 variant (NM_017460.6:c.522-191C>T; rs35599367) was characterized using an allelic expression imbalance approach,¹⁶ which explained 12% of the variability in CYP3A4 enzyme activity among individuals. The CYP3A4∗22 allele predominantly occurs in individuals of European (MAF, 5%) and admixed American (MAF, 2.6%) descent and is less common among individuals of African (MAF, <0.1%) and/or Asian (MAF, <0.6%) descent.¹⁷ This variant was also shown to decrease the quantity of mRNA and protein expressed and was correlated with decreased enzymatic activity in vivo.¹⁷

The data for association of CYP3A4 genetic variants with drug response are most consistent for quetiapine, an atypical antipsychotic indicated for the treatment of schizophrenia and bipolar disorder. Quetiapine's pharmacologic activity is primarily provided by the parent compound, which is extensively metabolized via CYP3A4. One of its metabolites, N-desalkyl quetiapine (also referred to as norquetiapine), is also active and believed to provide antidepressant effects.¹⁸ Studies using various established clinical CYP3A4 substrates, such as midazolam and erythromycin, have shown that CYP3A4∗22 results in a 40% reduction in erythromycin clearance and a 21% lower midazolam metabolic ratio.¹⁹

CYP3A5

This CYP3A enzyme was found to be expressed in approximately 10% to 20% of individuals with European ancestry. The full-length coding DNA sequence of CYP3A5 was originally published in 1989.²⁰ CYP3A5∗3, defined by the intronic variant (NM_000777.5:c.219-237A>G; rs776746), is associated with poor metabolism (historically also known as the nonexpressor phenotype).^21,22 The gene sequence in genome reference consortium human build 37 (GRCh37) corresponds to the CYP3A5∗3 allele (C in the reference genome and G in the coding DNA sequence), whereas the genome reference consortium human build 38 (GRCh38) reference genome corresponds to the CYP3A5∗1 allele (T in the reference genome and A in the coding DNA sequence). Therefore, when using GRCh37 as reference, the CYP3A5∗3 allele is considered the reference sequence and is not reported as a variant.

The CYP3A5∗3 no function allele has a frequency of approximately 90% in individuals with European ancestry²³ and ranges widely in other populations, with the lowest frequencies of 24% to 32% being observed in individuals with African ancestry. Frequencies range between 67% and 75% in populations with Asian ancestry. Two other no function alleles, CYP3A5∗6 and CYP3A5∗7, occur predominately in individuals of African ancestry, with reported frequencies of 11% to 19% and 9% to 12%, respectively. These alleles have frequencies of <0.5% in populations of European and Asian ancestry. Approximately 85% of individuals with European ancestry, 50% of individuals with Asian ancestry, and 30% of individuals with African ancestry are CYP3A5 poor metabolizers.²³ The Pharmacogene Variation Consortium CYP3A5 GeneFocus review²⁴ provides an extensive overview of CYP3A5.

For CYP3A5, a strong pharmacogenetic association exists for the metabolism of tacrolimus, which is a commonly prescribed immunosuppressant following solid organ transplant. The parent compound of tacrolimus is pharmacologically active and undergoes extensive metabolism by CYP3A enzymes. CYP3A5 normal and intermediate metabolizers (historically called expressors) have a higher rate of tacrolimus clearance, have lower dose-adjusted trough concentrations, and require higher tacrolimus doses to attain similar blood concentrations compared with poor metabolizers.²⁵ Given the narrow therapeutic index of tacrolimus, higher clearance rates have important implications for drug effectiveness as low systemic exposure correlates with increased risk for organ rejection,^26,27 although the correlation of CYP3A5 genotype with biopsy-confirmed acute rejection has not been proven. Studies have demonstrated earlier attainment of therapeutic tacrolimus concentrations with genotype-guided dosing versus empirical dosing.^28,29

Existing Clinical Guidelines and Recommendations

On the basis of pharmacokinetic data, the DPWG has developed recommendations for CYP3A4 genotype-based dosing for quetiapine (https://www.g-standaard.nl/risicoanalyse/B0005991.PDF, last accessed February 10, 2023). Practice guideline for CYP3A5 genotype-based dosing for tacrolimus is provided by CPIC.²⁵ In addition, the US Food and Drug Administration lists the CYP3A5/tacrolimus gene-drug pair in their pharmacogenetic association table for which the available data support therapeutic management recommendations (https://www.fda.gov/medical-devices/precision-medicine/table-pharmacogenetic-associations, last accessed February 10, 2023).

Testing Platforms

Selection of a molecular platform to use for testing PGx variants is based on many considerations that include but are not limited to technical feasibility of analysis of the genomic regions of interest, cost, laboratory workflow, test volume, and desired test turnaround time. Clinically relevant variants in the CYP3A4 and CYP3A5 genes are amenable to interrogation through molecular techniques commonly used in clinical laboratories, including targeted genotyping or sequencing (Sanger sequencing or next-generation sequencing) approaches. It is at the discretion of the specific laboratory to select a preferred testing platform. Unless long-read sequencing technology is used, almost all commonly used molecular testing platforms are unable to determine phasing information for detected variants. Therefore, as for other PGx genotyping assays, assigning diplotypes from genotyping or sequencing data is mostly empirical or inferred.

Materials and Methods

The AMP PGx Working Group is composed of subject matter experts from the College of American Pathologists, CDC, CPIC, DPWG, European Society for Pharmacogenomics and Personalized Therapy, Pharmacogenomics Knowledgebase (PharmGKB), Pharmacogene Variation Consortium, and the PGx clinical testing and research communities. CYP3A4 and CYP3A5 variant alleles were reviewed and classified into tiers based on four criteria that received equal weight during deliberations.

• i)Functional characterization of the allele (ie, whether it is known to affect expression of the gene or function of the encoded protein).
• ii)Presence at an appreciable MAF in a population/genetic ancestry.³⁰ In this CYP3A4 and CYP3A5 recommendation document, the Working Group used an MAF of ≥1% in at least one subpopulation as a criterion for tier 1 alleles, and ≥0.1% for tier 2 alleles based on currently available information from applicable resources (https://www.pharmgkb.org/page/cyp3a5RefMaterials, last accessed February 10, 2023).
• iii)Availability of RMs (Table 1).³¹
• iv)Technical feasibility for clinical laboratories to interrogate using standard molecular testing methods. This criterion was determined to not be relevant for these CYP3A4 and CYP3A5 recommendations, as none of the reviewed variant alleles was considered difficult to interrogate using standard methods.

Table 1.

Reference Materials

CYP3A4 alleles	Coriell ID†	Diplotype‡	CYP3A5 alleles	Coriell ID†	Diplotype‡
CYP3A4∗2	HG00276	∗1/∗2	CYP3A5∗1	NA07439	∗1/∗1
CYP3A4∗3	NA12006	∗1/∗3	CYP3A5∗1	NA18564	∗1/∗1
CYP3A4∗4	HG00525	∗1/∗4	CYP3A5∗3	HG00436	∗3/∗3
CYP3A4∗4	HG01865	∗1/∗4	CYP3A5∗3	NA10856	∗1/∗3
CYP3A4∗5	HG01816	∗1/∗5	CYP3A5∗6	NA18518	∗1/∗6
CYP3A4∗5	NA18561	∗1/∗5	CYP3A5∗6	NA19819	∗3/∗6
CYP3A4∗6	NA18941	∗1/∗6	CYP3A5∗7	NA19143	∗6/∗7
CYP3A4∗7	HG00334	∗1/∗7	CYP3A5∗7	NA19920	∗7/∗7
CYP3A4∗7	NA20813	∗1/∗7	CYP3A5∗7	NA18484	∗1/∗7
CYP3A4∗8	HG00368	∗1/∗8	CYP3A5∗8	None	None
CYP3A4∗9	HG02146	∗1/∗9	CYP3A5∗9	None	None
CYP3A4∗10	HG00122	∗1/∗10
CYP3A4∗10	HG00734	∗10/∗22
CYP3A4∗11	HG00139	∗3/∗11
CYP3A4∗12	HG03159	∗1/∗12
CYP3A4∗12	NA19035	∗1/∗12
CYP3A4∗13	None	None
CYP3A4∗14	None	None
CYP3A4∗15	NA19109	∗1/∗15
CYP3A4∗15	NA19226	∗1/∗15
CYP3A4∗16	NA18966	∗1/∗16
CYP3A4∗16	NA18978	∗1/∗16
CYP3A4∗17	None	None
CYP3A4∗18	HG02134	∗1/∗18
CYP3A4∗18	HG00704	∗1/∗18
CYP3A4∗19	HG03885	∗1/∗19
CYP3A4∗19	NA21095	∗1/∗19
CYP3A4∗20	HG01275	∗1/∗20
CYP3A4∗21	NA18603	∗1/∗21
CYP3A4∗22	NA23313	∗1/∗22
CYP3A4∗22	NA24008	∗22/∗22
CYP3A4∗23	HG02054	∗1/∗23
CYP3A4∗24	NA19160	∗1/∗24
CYP3A4∗26	None	None
CYP3A4∗28	HG02029	∗1/∗28
CYP3A4∗29	None	None
CYP3A4∗30	None	None
CYP3A4∗31	None	None
CYP3A4∗32	None	None
CYP3A4∗33	None	None
CYP3A4∗34	None	None
CYP3A4∗35	NA12336	∗1/∗35
CYP3A4∗36	NA07439	∗1/∗1
CYP3A4∗38	NA18934	∗1/38

ID, identifier.

^†Available at https://www.coriell.org.

^‡According to Gaedigk et al.³¹

In addition, commercially available genotyping platforms (Supplemental Table S1) were reviewed for assessing the ability of laboratories to implement the Working Group recommendations; however, these data were not used as a determinant of tier assignment.

The AMP PGx Working Group used MAF and functional information from CPIC, PharmGKB, and the scientific literature. The PGx Working Group also used information curated by CPIC. Frequencies of unique sequence variants can be found in databases, such as the Genome Aggregation Database (https://gnomad.broadinstitute.org, last accessed February 10, 2023) and the International Genome Sample Resource (or 1000 Genomes; https://www.internationalgenome.org, last accessed February 10, 2023), although these resources do not report frequencies for haplotypes [or star (∗) alleles] that are composed of a combination of variants. For CYP3A5, CPIC relies on studies from the peer-reviewed published literature for allele frequencies (https://www.pharmgkb.org/page/cyp3a5RefMaterials, last accessed February 10, 2023). All CYP3A5 star (∗) alleles currently listed by Pharmacogene Variation Consortium are defined by a single sequence variant.²⁴ The CPIC allele clinical function assignment is based on literature reviews of allele function by guideline (https://www.pharmgkb.org/page/cyp3a5RefMaterials, last accessed February 10, 2023). However, CPIC assigned clinical function may not be the same as the biochemical function of the allele, which, therefore, may lead to discrepant variant classification. Because CPIC has not published any guidelines for CYP3A4, there are currently no CPIC or PharmGKB reference tables for allele frequency or allele function. As such, CYP3A4 allele frequencies were inferred on the basis of the unique variant frequency in Genome Aggregation Database version 2.1. To assess CYP3A4 allele function, the AMP PGx Working Group reviewed literature from PharmGKB regarding 13 different substrates comparing function across various CYP3A4 alleles (https://www.pharmgkb.org/gene/PA130/variantAnnotation, last accessed February 10, 2023) and the DPWG guideline for quetiapine (https://www.pharmgkb.org/guidelineAnnotation/PA166265421, last accessed February 10, 2023). CYP3A4 allele function was recently also reviewed by Zhai et al.³²

College of American Pathologists proficiency testing (PT) survey data were obtained from the Pharmacogenetics, PGX-B 2021 mailing¹ to determine which CYP3A4 and CYP3A5 alleles are currently being tested by clinical laboratories. Laboratories self-reported whether they clinically tested for CYP3A4 and/or CYP3A5 as well as which alleles their test is designed to detect. PT participants include both US-based and international laboratories. External assessment programs in Europe, such as the European Molecular Genetics Quality Network and the Reference Institute for Bioanalytics, were also reviewed.

Results

Tier 1 CYP3A4 and CYP3A5 Variant Alleles

The CYP3A4 and CYP3A5 alleles recommended for inclusion in tier 1 include CYP3A4∗22 and CYP3A5∗3, CYP3A5∗6, and CYP3A5∗7 (Table 2). For Human Genome Variation Society nomenclature throughout, see nomenclature (https://www.ncbi.nlm.nih.gov/snp and http://www.ncbi.nlm.nih.gov/clinvar, last accessed February 10, 2023).

Table 2.

Tier 1 CYP3A4 and CYP3A5 Variant Alleles

Gene	CYP3A4	CYP3A5
Allele	CYP3A4∗22	CYP3A5∗3	CYP3A5∗6	CYP3A5∗7
Allele functional status	Decreased function	No function†	No function†	No function†
Defining variant‡	rs35599367	rs776746	rs10264272	rs41303343
RefSeqGene LRG nomenclature (GRCh37/GRCh38)	NG_008421.1:g.20493C>T	NG_007938.2:g.12083A>G (GRCh38)	NG_007938.2:g.19787G>A	NG_007938.2:g.32228dup
HGVS cDNA nomenclature	NM_017460.6:c.522-191C>T	NM_000777.5:c.219-237A>G (GRCh38)	NM_000777.5:c.624G>A	NM_000777.5:c.1035dup
HGVS protein nomenclature	Splicing defect	Splicing defect	Splicing defect; NP_000768.1: p.Lys208=	NP_000768.1: p.Thr346TyrfsTer3
Reference material available§	Yes	Yes	Yes	Yes
Multiethnic allele frequency, %	0–9	24–92	0–19	0–12

cDNA, coding DNA; GRCh37, genome reference consortium human build 37; GRCh38, genome reference consortium human build 38; HGVS, Human Genome Variation Society; LRG, Locus Reference Genomic.

^†Citations for CYP3A5 assignment of function can be found at https://www.pharmgkb.org/page/cyp3a5RefMaterials; and for HGVS nomenclature, at https://www.ncbi.nlm.nih.gov/snp and http://www.ncbi.nlm.nih.gov/clinvar (last accessed February 10, 2023).

^‡Defining star (∗) allele variants can be found at https://www.pharmvar.org (last accessed February 10, 2023).

^§According to Table 1 and Gaedigk et al.³¹

CYP3A4∗22

The decreased function CYP3A4∗22 allele is characterized by a splicing defect caused by a transition variant within intron 6 (NM_017460.6:c.522-191C>T; rs35599367). This intronic variant leads to the formation of a transcript with partial intron 6 retention and decreased levels of full-length functional mRNA and protein, leading to decreased enzyme activity.^16,33 This allele is found in European (non-Finnish and Finnish) and Ashkenazi Jewish populations at frequencies between 3.6% and 9.0%, respectively. It is also observed in African/African American and Latino/admixed American populations at frequencies of 0.9% and 2.5%, respectively. It is typically not found among East and South Asian populations (https://gnomad.broadinstitute.org/variant/7-99366316-G-A?dataset=gnomad_r2_1, last accessed February 10, 2023). The CYP3A4∗22 intronic variant can also be found in cis with NM_017460.6:c.1334T>C (rs4986910), the variant defining the CYP3A4∗3 allele whose function is unknown³²; this haplotype, CYP3A4∗37, was only recently discovered, and its function remains uncertain.³⁴

CYP3A5∗3

The no function CYP3A5∗3 allele is characterized by an intronic transition variant that generates a cryptic splice acceptor site within intron 3 (NM_000777.5:c.219-237A>G; rs776746). This substitution leads to the retention of parts of intron 3 that causes a shift in the reading frame, premature termination, and a truncated nonfunctional protein product.²² The CYP3A5∗3 allele is the major allele in all characterized human subpopulations, except sub-Saharan African and African American/Afro-Caribbean populations. It is observed in 24% to 92% across populations, with the highest frequency among individuals with European ancestry (https://www.pharmgkb.org/page/cyp3a5RefMaterials, last accessed February 10, 2023).

CYP3A5∗6

The no function CYP3A5∗6 allele is characterized by a synonymous variant within exon 7 (NM_000777.5:c.624G>A, p.Lys208=; rs10264272). This substitution causes a splicing defect, which causes a shift in the reading frame and premature termination. The resulting truncated protein product is nonfunctional.²² The CYP3A5∗6 allele is most frequently observed in individuals with African American/Afro-Caribbean and sub-Saharan African ancestry at frequencies of 11% and 19%, respectively. It is also found among individuals of Latino and Near Eastern ancestry at an allele frequency of 4%, and it is rarely observed in individuals of East Asian and European ancestry at allele frequencies of <0.2% (https://www.pharmgkb.org/page/cyp3a5RefMaterials, last accessed February 10, 2023).

CYP3A5∗7

The no function CYP3A5∗7 allele is characterized by a single-nucleotide duplication within exon 11 (NM_000777.5:c.1035dup; p.Thr346TyrfsTer3; rs41303343). This variant leads to a shift in the reading frame and results in a truncated, nonfunctional protein product.²¹ The CYP3A5∗7 allele is most commonly observed among individuals with sub-Saharan African and African American/Afro-Caribbean ancestry at frequencies of 9% and 12%, respectively. It is also observed among individuals of Near Eastern and Latino ancestry at frequencies of 0.4% and 3%, respectively. It is not typically found in individuals of East Asian, Central/South Asian, and European ancestry (https://www.pharmgkb.org/page/cyp3a5RefMaterials, last accessed February 10, 2023).

Tier 2 CYP3A4 and CYP3A5 Variant Allele

A single CYP3A4 allele is recommended for inclusion in tier 2, CYP3A4∗20. No CYP3A5 alleles were identified as appropriate candidates for tier 2 (Table 3).

Table 3.

Tier 2 CYP3A4 Variant Allele

Gene	CYP3A4
Allele	CYP3A4∗20
Allele functional status†	Probably no function
Defining variant‡	rs67666821
RefSeqGene LRG nomenclature (GRCh37/GRCh38)	NG_008421.1:g.31002dup
HGVS cDNA nomenclature	NM_017460.6:c.1461dup
HGVS protein nomenclature	NP_059488.2: p.Pro488ThrfsTer7
Reference material available§	Yes
Multiethnic allele frequency, %	0–0.12

cDNA, coding DNA; GRCh37, genome reference consortium human build 37; GRCh38, genome reference consortium human build 38; HGVS, Human Genome Variation Society; LRG, Locus Reference Genomic.

^†Citations for assignment of function can be found at https://www.pharmvar.org; and for HGVS nomenclature, at https://www.ncbi.nlm.nih.gov/snp and http://www.ncbi.nlm.nih.gov/clinvar (last accessed February 10, 2023).

^‡Defining variant(s) can be found at https://www.pharmvar.org (last accessed February 10, 2023).

^§According to Table 1 and Gaedigk et al.³¹

CYP3A4∗20

The no function CYP3A4∗20 allele is characterized by a single-nucleotide duplication within exon 13 (NM_017460.6:c.1461dup; p.Pro488ThrfsTer7; rs67666821) that causes a frameshift and generation of a premature stop codon, resulting in truncation near the terminal end of the protein. Although the functional activity has not been as well characterized as the tier 1 CYP3A4∗22 allele, this truncated protein has been demonstrated to be nonfunctional in vitro because of its inability to incorporate heme.³⁴ In vitro studies have shown that the nonfunctional CYP3A4∗20 gene product was also associated with severe paclitaxel-induced neuropathy.^35,36 The CYP3A4∗20 allele is found among the Latino/admixed American population at a frequency of 0.1%, whereas the frequency is <0.03% in European (non-Finnish) and African/African American populations (https://gnomad.broadinstitute.org/variant/7-99355806-G-GT?dataset=gnomad_r2_1, last accessed February 10, 2023). One study reported a founder effect of the CYP3A4∗20 allele in the Spanish population with a frequency of 1.2%, with up to 3.8% in certain regions in Spain.³⁷

Discussion

In this document, the AMP PGx Working Group recommends inclusion of several CYP3A4 and CYP3A5 alleles in clinical PGx genotyping assays. The goals of this recommendation and other related Working Group recommendations are to promote standardization and to ensure that laboratories conducting PGx testing for CYP3A genes include the most clinically relevant alleles. The tier 1 recommended alleles were selected on the basis of having a well-characterized effect on functional activity, prevalence of >1% in at least one ancestral subpopulation, and the availability of RMs for assay validation. The CYP3A4∗20 allele was included as a tier 2 recommended allele for clinical laboratories wanting to offer a more comprehensive genotyping panel. Although the functional activity of CYP3A4∗20 is not as well documented as the tier 1 alleles, it is a frameshift variant that likely results in a nonfunctional protein and is present in >0.1% of Latino populations. CYP3A4 and CYP3A5 variants/alleles without a known or likely effect on enzymatic activity, and those that are present in <0.1% of any subpopulation, are not included in these recommendations.

Accurate and comprehensive CYP3A4 and CYP3A5 allele frequencies are still emerging, particularly for rare variants; however, the AMP PGx Working Group tier 1 recommended variants account for most of the reported alleles with decreased or no function across all populations.

Laboratories performing sequencing may identify additional sequence variants. Exome sequencing may miss the tier 1 CYP3A4∗22 and CYP3A5∗3 intronic variants, depending on the target capture probes. Clinical laboratories should follow best practices for assay validation and adhere to the applicable regulatory requirements for their location.³⁸

CYP3A4 and CYP3A5 Alleles Not Included in Tier 1 or 2 Recommendations

The CYP3A4∗1B (CYP3A4∗1.001) allele corresponds to the G nucleotide in the current reference sequences NM_017460.6 and NG_008421.1 (GRCh38) but is the minor allele for rs2740574. It is associated with normal function and is therefore not included in the tier 1 or 2 recommendations for routine clinical testing. Also, the c.-392G>A variant (NG_008421.1:g.4713 G) defining CYP3A4∗1.002 (formerly CYP3A4∗1A; rs2740574) also occurs on numerous other haplotypes and is therefore not included in the tier 1 or 2 recommendations and will not be further discussed. The CYP3A4∗6 (NM_017460.6:c.830dup; p.Asp277GlyfsTer9; rs4646438), CYP3A4∗26 (NM_017460.6:c.802C>T; p.Arg268Ter; rs138105638), and CYP3A4∗30 (NM_017460.6:c.388C>T; p.Arg130Ter; rs778013004) alleles are all protein truncating variants (nonsense or frameshift) and therefore likely encode a nonfunctional protein; however, these were considered too rare to be included in the tier 1 or 2 recommendations for routine clinical testing.

Given that the function of CYP3A4 alleles has not been previously curated by CPIC, several variant alleles were not recommended for inclusion in either tier 1 or tier 2 because of unknown clinical effects, although they may meet the allele frequency cutoffs for consideration. These include the following alleles. CYP3A4∗2 (NM_017460.6:c.664T>C; p.Ser222Pro; rs55785340) occurs with a frequency of approximately 0.1%. CYP3A4∗3 (NM_017460.6:c.1334T>C; p.Met445Thr; rs4986910) has a frequency range of 0% to 1.8%; its allele-defining variant has been found in cis with two other variants, the CYP3A4∗22-defining variant NM_017460.6:c.552-191C>T (designated CYP3A4∗37) and the CYP3A4∗11-defining variant NM_017460.6:c.1088C>T (p.Thr363Met; rs67784355; designated CYP3A4∗38). CYP3A4∗4 (NM_017460.6:c.352A>G; p.Ile118Val; rs55951658) has an allele frequency of 0% to 0.5%. CYP3A4∗5 (NM_017460.6:c.653C>G; p.Pro218Arg; rs55901263) has a frequency range of 0% to 0.1%. CYP3A4∗10 (NM_017460.6:c.520G>C; p.Asp174His; rs4986908) occurs largely in individuals of European ancestry with a frequency of approximately 0.3%. CYP3A4∗15 (NM_017460.6:c.485G>A; p.Arg162Gln; rs4986907) occurs in approximately 2.6% of individuals of African ancestry. CYP3A4∗18 (NM_017460.6:c.878T>C; p.Leu293Pro; rs28371759) has a frequency of approximately 1.9% of East Asian ancestry. CYP3A4∗23 (NM_017460.6:c.484C>T; p.Arg162Trp; rs57409622) occurs primarily in individuals of African ancestry, with a frequency of approximately 0.3% (https://gnomad.broadinstitute.org, last accessed February 10, 2023). Publicly available RMs are available for all of these alleles (Table 1).³¹ When the function of these alleles is further defined, one or more of these alleles may be recommended for inclusion in either tier 1 or tier 2.

The functional impact of the CYP3A4∗1 G allele, which is characterized by a variant in intron 10 (NM_017460.6:c.1026+12G>A; rs2242480), remains uncertain regarding its impact on CYP3A4 and CYP3A5 expression levels. Furthermore, NM_017460.6:c.1026+12G>A is also part of multiple other CYP3A4 haplotypes (https://www.pharmvar.org/gene/CYP3A4, last accessed February 10, 2023), further complicating the functional characterization of this allele.

Proficiency Testing/External Quality Assessment

Several PT or external quality assessment programs are commercially available for CYP3A5 genotyping; however, programs that include CYP3A4 are currently not widely available. College of American Pathologists PT data were queried to understand the testing practices of clinical laboratories, including which alleles are currently included in clinical testing. A total of 247 laboratories participated in the College of American Pathologists Pharmacogenetics PGX-B 2021 Survey.¹ Of these, 132 laboratories responded to questions related to CYP3A4 and CYP3A5 testing. Of these, 107 laboratories (81.1%) indicated testing both CYP3A4 and CYP3A5, whereas 12 laboratories (9.1%) indicated only testing CYP3A5, and 3 laboratories (2.3%) indicated only testing CYP3A4. Ten (7.6%) laboratories responded that they do not perform clinical testing for either gene. Laboratories were also asked whether their test includes specific alleles. Of the laboratories that indicated that they tested for each gene, the number and percentage that reported testing for CYP3A4 and CYP3A5 alleles are presented in Table 4. In Europe, there are two major proficiency testing vendors, the Reference Institute for Bioanalytics and the European Molecular Genetics Quality Network. The alleles that are tested in these European PT surveys are also listed in Table 4; however, the number and percentage of laboratories that test each allele and subscribe to either survey are not available.

Table 4.

Summary of CYP3A4 and CYP3A5 Proficiency Testing Programs

Gene	Allele	Tier	CAP participants, n (%)	European Molecular Genetics Quality Network	Reference Institute for Bioanalytics
CYP3A4 (n = 110)	∗1B	None	56 (50.9)
	∗17	None	65 (59.1)
	∗22	1	105 (95.5)		X
CYP3A5 (n = 119)	∗3	1	118 (99.2)	X	X
	∗6	1	107 (89.9)	X
	∗7	1	106 (89.1)	X
	∗8	None	44 (37.0)
	∗9	None	47 (39.5)

The number/percentage of laboratories that report testing for specific CYP3A4 and/or CYP3A5 alleles is provided by the CAP Biochemical and Molecular Genetics Committee: PGX-B, 2021 Proficiency Testing Survey, CAP, 2021.¹ The alleles included in the European Molecular Genetics Quality Network and Reference Institute for Bioanalytics surveys are indicated with an X.

CAP, College of American Pathologists.

Limitations

This document focuses only on recommendations of alleles to include in clinical laboratory genotyping assays for CYP3A4 and CYP3A5. This document does not include, for example, correlation of genotypes with phenotype, clinical interpretation of genotypes, or recommendations for changes of medications or dosing based on genotype, as these were deemed to be out of scope for this document and/or available from other sources, such as CPIC, PharmGKB, and DPWG. Although technical challenges related to interrogating CYP3A4 and CYP3A5 were discussed in this document, the Working Group does not recommend or endorse any specific molecular testing platform(s) for clinical CYP3A4 and CYP3A5 genotyping.

Conclusions

This document provides recommendations on which CYP3A4 and CYP3A5 alleles to include in clinical genotyping assays. These recommendations are intended to facilitate the design and implementation of pharmacogenomic testing by clinical laboratories. In addition, these recommendations are intended to promote test standardization and improve genotype concordance between clinical laboratories.

Disclaimers

The Association for Molecular Pathology (AMP) Clinical Practice Guidelines and Reports are developed to be of assistance to laboratory and other health care professionals by providing guidance and recommendations for particular areas of practice. The Guidelines or Reports should not be considered inclusive of all proper approaches or methods, or exclusive of others. The Guidelines or Reports cannot guarantee any specific outcome, nor do they establish a standard of care. The Guidelines or Reports are not intended to dictate the treatment of a particular patient. Treatment decisions must be made on the basis of the independent judgment of health care providers and each patient's individual circumstances. The AMP makes no warranty, express or implied, regarding the Guidelines or Reports and specifically excludes any warranties of merchantability and fitness for a particular use or purpose. The AMP shall not be liable for direct, indirect, special, incidental, or consequential damages related to the use of the information contained herein.

The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the CDC/the Agency for Toxic Substances and Disease Registry. Use of trade names and commercial sources is for identification only and does not imply endorsement by the CDC, the Public Health Service, or the US Department of Health and Human Services.