Whole-genome sequencing-based profiling of CYP2D6 structural variants in Japanese patients with solid tumors

Abstract

Cytochrome P450 2D6 (CYP2D6), a key enzyme involved in the metabolism of a wide range of pharmaceuticals, exhibits extensive genetic variants, including copy number variations (CNVs) and structural alterations. Hybrid alleles formed between CYP2D6 and its highly homologous pseudogene CYP2D7 frequently result in reduced or absent enzymatic activity and show marked interethnic variability. However, accurate characterization of CNVs and hybrid structures in the Japanese population remains technically challenging. In this study, we performed a comprehensive analysis of CYP2D6 genetic and structural variants in 1,287 Japanese patients with solid tumors using whole-genome sequencing (WGS). CNVs and diplotypes were inferred using Sequenza and Aldy, visually inspected with Integrative Genomics Viewer (IGV), and validated using quantitative PCR and Sanger sequencing. A total of 68 distinct diplotypes were identified. The most frequent diplotype was *1/36+*10 (0.2176), followed by *1/*1 (0.1919), *1/*2 (0.0940), *1/*10 (0.0653), *2/*36+*10 (0.0645), and *36/*36+*10 (0.0513). Complete gene deletion (*5/*5) was detected in three cases, and heterozygous deletion (*5/–) in 86 cases. Given that functional annotations for CYP2D6 alleles are well established, WGS-based comprehensive detection represents a practical and scalable strategy for pharmacogenomic profiling. This approach is expected to support optimized drug selection and advance personalized medicine.

INTRODUCTION

Cytochrome P450 2D6 (CYP2D6) is expressed in multiple tissues, including the liver, brain, and intestine, and it is responsible for the metabolism of approximately 20–25% of commonly prescribed drugs, including antidepressants, antipsychotics, antiarrhythmics, opioids, and β-blockers (Taylor et al., 2020; Molden and Jukić, 2021).

The gene exhibits extensive genetic variation and copy number variations (CNVs), resulting in substantial inter-individual variability in enzymatic activity. The functional consequences of many CYP2D6 alleles have been systematically catalogued by the Pharmacogene Variation Consortium (PharmVar.; https://www.pharmvar.org/gene/CYP2D6; Nofziger C et al., 2020).

Accurate identification of CYP2D6 genetic variants is essential for interpreting variability in drug responses, predicting adverse drug reactions, and guiding appropriate therapeutic selection. However, conventional approaches, including targeted genotyping, long-range PCR, and quantitative real-time PCR (qPCR), have limited ability to resolve complex structural variants, particularly in the highly homologous region shared with the adjacent pseudogene CYP2D7 (Charnaud et al., 2022). Hybrid alleles formed between CYP2D6 and CYP2D7 often lead to reduced or absent enzyme function, underscoring the importance of their precise characterization.

Although next-generation sequencing has advanced pharmacogenomic research, whole-exome sequencing (WES) remains insufficient for resolving CNVs and structural alterations in highly homologous loci such as CYP2D6 (Ohnami et al., 2017; Ohnami et al., 2022). To address these limitations, we conducted a comprehensive analysis of CYP2D6 and its structural variants in Japanese patients with solid tumors using whole-genome sequencing (WGS).

MATERIALS AND METHODS

This study was conducted using 1,287 blood samples collected during surgery at the Shizuoka Cancer Center Hospital between January 2014 and December 2022 (Yamaguchi et al., 2014; Nagashima et al., 2020). The tumor types of patients with solid tumors are summarized in Supplementary Table 3. Genomic DNA was extracted from the buffy coat isolated from whole blood using a QIAamp DNA Blood Midi Kit (Qiagen, Hilden, Germany) and quantified using a Qubit fluorometer (ThermoFisher, Waltham, Massachusetts, USA).

WGS was performed on an Illumina NovaSeq 6000 platform (Illumina, San Diego, CA, USA) following protocols described in previous reports (Nagashima et al., 2024). Germline variants, including single-nucleotide variants (SNVs) and insertions or deletions (indels), were identified using the DRAGEN Structural Variant Caller (Illumina).

CNVs and diplotypes were inferred using the Copy Number Variant Pipeline (Illumina) with Sequenza ver. 3.0.0 (Favero et al., 2015) and the Aldy software package (https://algo-cancer.github.io/projects/aldy.html), respectively. During this process, variant allele frequencies (VAFs), CNVs, and the presence or absence of hybrid alleles were visualized using the Integrative Genomics Viewer (IGV) (Robinson et al., 2011). To further validate structural alterations, including CNVs and hybrid variants, qPCR and Sanger sequencing were performed.

The study protocol (Project HOPE) was approved by the Institutional Review Board of the Shizuoka Cancer Center (authorization number 25-33). Pathogenic germline alterations were predicted using blood-derived DNA. Written informed consent, including consent for the disclosure of potential secondary findings arising from constitutional genomic analysis, was obtained from all participants with approval from the ethics review board. All procedures involving human participants adhered to the ethical principles outlined in the Declaration of Helsinki. The genotype data referenced in this study are available from public repositories, including UCSC (https://genome.ucsc.edu/), jMorp (https://jmorp.megabank.tohoku.ac.jp/202109/), ClinPGx(https://www.clinpgx.org/) and PharmVar. The germline data generated in this study are available from the NBDC Human Database (temporary accession ID: hum0127.v4) or from the corresponding author upon reasonable request.

RESULTS

The frequencies of CYP2D6 variants in 1,287 Japanese patients with cancer are summarized in Table 1. The diplotype frequencies determined by WGS and Aldy analyses were classified into 68 distinct diplotypes. Among them, the wild-type genotype (*1/*1) had a frequency 0.1919, highlighting the substantial diversity of CYP2D6 variants, including several previously unreported genotypes. The *1/*36+*10 diplotype was the most frequent CYP2D6 genotype, with a frequency of 0.2176 (280/1287). The frequencies of other major genotypes were 0.1919 (247/1287) for *1/*1 (wild-type), 0.0940 (121/1287) for *1/*2, 0.0653 (84/1287) for *1/*10, 0.0645 (83/1287) for *2/*36+*10, and 0.0513 (66/1287) for *36/*36+*10 (Table 1). In contrast, whole-gene deletion (*5/*5) was observed in three cases, and single-allele deletion (*5/*1, /*2, /*10, /*21, or /*41-) were identified in 86 cases. Classification based on the CYP2D6 activity score defined by ClinPGx/CPIC showed that 19.0% (244/1287) of patients had reduced metabolizer status (activity score ≤0.5), and 19.7% (254/1287) were had reduced metabolizer status when 26 cases with unverified diplotypes were included.

Table 1. CYP2D6 variant frequencies in Japanese patients with solid tumors.

As shown in Fig. 1, CYP2D6 variants identified by WGS using the DRAGEN Structural Variant Caller and Aldy analysis, followed by visual inspection with IGV, were validated using multiplex PCR and Sanger sequencing. In particular, the *10, *14, *4, *131, *21, *51, and *18 alleles, each associated with markedly reduced metabolic activity, were confirmed using three uniquely designed PCR primer sets (Supplementary Table 2), followed by Sanger sequencing. Furthermore, for hybrid alleles involving CYP2D7, which are associated with reduced metabolic activity, breakpoint sequences were confirmed using Sanger sequencing (Fig. 1C). In contrast, the *13, *63 and *79 alleles, which represent CYP2D6 and CYP2D7 hybrid structures, as well as the *49, *52, *69, and *132 haplotypes consisting of multiple variants, could not be fully validated at the sequences level (Supplementary Table 1). Therefore, these haplotypes were excluded from the analysis presented in Table 1. CNV also plays an important role in determining CYP2D6 enzymatic activity (Jarvis et al., 2019). Fig. 2 shows the results of real-time PCR validation of CNVs identified by IGV using two commercially available probe sets (Fig 2A). The qPCR-based CNV results were consistent with those obtained using IGV (Fig. 2B).

Fig. 1

Validation of identified CYP2D6 alleles and CYP2D6/CYP2D7 hybrid alleles. (A) Schematic representation of the genomic structure and breakpoint regions of the *36 hybrid alleles between CYP2D6 and CYP2D7. The closed triangle denotes the estimated breakpoint position. Numbers (PCR1-4) indicate the PCR amplicon lengths generated using primers, as listed in Supplementary Table 2. (B) Raw gel images obtained using the TapeStation system. The lower panel shows an enlarged view of PCR bands corresponding to the PCR1-3 depicted in Fig. 1A. (C) Sequence verification of the breakpoint region corresponding to the *36 hybrid alleles shown in Fig. 1A.

Fig. 2

Validation of the identified CYP2D6 copy number variations (CNVs). (A) Genomic positions of the two commercially available TaqMan probes (a and b). (B) CNVs results for the CYP2D6 variants.

As shown in Table 1, reduced metabolizer status (activity score ≤0.5) attributable to CYP2D6 variants accounted for 19.0% of all cases, with minimal variation across tumor types (Supplementary Table 3). Notably, 28.8% (17/59) of patients with breast cancer, in which CYP2D6 activity is known to influence tamoxifen metabolism (Kiyotani et al., 2013; He et al., 2020), were classified as reduced metabolizers. However, a limitation of the present study is that the proportion of CYP2D6 reduced metabolizers among patients with breast cancer remains uncertain, as our findings were derived from a relatively small hospital-based cohort. Therefore, the significance of the findings observed in this study needs to be validated in a statistically more powerful, well-designed study.

DISCUSSION

In this study, we used WGS to comprehensively characterize the spectrum and frequency of CYP2D6 variants in Japanese patients with solid tumors. The frequency of the CYP2D6 wild-type (*1/*1) genotype was 19.2%, which is consistent with previously reported rates (16.2-19.6%) in healthy Japanese individuals (Hosono et al., 2009; Kisoi et al., 2020; Fukunaga et al., 2021). CYP2D6 is a CYP family gene that exhibits pronounced interethnic variability (Petrović et al., 2020; Kurose et al., 2012).

CYP2D6 variants associated with decreased enzyme activity accounted for approximately 19% of cases in our cohort. Notably, CYP2D6/CYP2D7 hybrid alleles with an activity score of 0.5 or less, characterized by duplicated sequences or regions of high homology, comprised 16.5% of all samples. These hybrid alleles are known to confer reduced or absent enzyme activity (Gaedigk et al., 2006) and therefore have substantial clinical relevance as predictors of drug efficacy and adverse drug event risk.

In routine clinical practice, only null-function CYP2D6 variants are typically assessed because of the technical limitations of conventional genotyping methods. The ability of WGS to detect a broad range of CYP2D6 variants, including complex hybrid alleles, is a major methodological advantage (Yang et al., 2017).

However, among the CYP2D7/CYP2D6 hybrid alleles identified in this study, *13, *63, and *79 could not be validated by subsequent PCR assays, indicating the need for further investigation into the accuracy and interpretation of these structural variants.

Fukunaga et al. (2021) characterized CYP2D6 haplotypes in 990 Japanese individuals using targeted sequencing followed by long-read sequencing (LRS), and further published their predicting phenotypes on the haplotypes based on the in vitro metabolic activities. These functional data have been incorporated into PharmVar. (https://www.pharmvar.org/gene/CYP2D6).

Compared with short-read sequencing (SRS), LRS enables more precise resolution of structural variants and facilitates the identification of highly homologous gene polymorphisms, such as those in CYP2D6 and CYP2A6, as well as the discovery of novel structural variants and diplotypes. Such capabilities are expected to advance future pharmacogenomic research.

In this study, the frequencies of CYP2D6 diplotypes were generally consistent with those previously reported using LRS, except for the *10/*36 diplotype. Although this diplotype was confirmed by IGV and Sanger sequencing following SRS-based analysis, the reason for the discrepancy in its reported frequency between the two approaches remains unclear. In contrast, in our previous CYP2A6 study using LRS (Ohnami et al., 2025), we identified several structural variants, including novel CYP2A7/CYP2A6 hybrids. These variants were also detectable by SRS-based BLAST search, although confirmation by methods such as Sanger sequencing was required for both approaches.

In Japan, the implementation of short-read WGS (SRS) in routine clinical practice is currently progressing under the Japan Genomic Medicine Initiative (GeMJ). Although LRS provides more comprehensive structural information, its high cost and requirement for high-molecular-weight DNA currently limit its use to research setting. The CYP2D6 diplotype analysis performed using SRS in this study yielded results that were validated by orthogonal methods, such as Sanger sequencing and CN assay. These findings are expected to serve as a practical reference for routine clinical applications in Japan and contribute to the advancement of personalized medicine.

Beyond CYP2D6, the capacity of WGS to comprehensively investigate the CYP gene family, including clinically important genes such as CYP2A6 and CYP3A4, is of particular significance. We recently demonstrated that long-read WGS enables the precise identification of the switching point within CYP2A7/CYP2A6 hybrid alleles, further highlighting the utility of WGS in resolving complex genomic structures (Ohnami et al., 2025).

In addition, we previously reported a WES-based assay for simultaneous profiling of germline variants in major drug-metabolizing enzyme genes in Japanese patients with cancer (Ohnami et al., 2017; Ohnami et al., 2022). At that time, accurate analyses of structural variants and CNVs of CYP family genes, including CYP2D6, remained a challenge, an issue also noted in the germline variant data of the Japanese Multi Omics Reference Panel, jMorp.

In the present study, many CYP2D6 variants were successfully detected using WGS, demonstrating clear improvements in analytical performance. Unfortunately, although the results of our study may provide valuable pharmacogenomic insights, we did not evaluate direct or indirect associations with drug exposure, treatment response, adverse events, or concomitant medication history. Therefore, determining whether the findings of this study can be applied to pharmacogenomic practice remains a key challenge for the future.

Funding

This study was supported by the Japan Agency for Medical research and Development (AMED), grant number: JP22k0106689.

Conflict of interest

The authors declare that there is no conflict of interest.

Data availability

The data in this study are included in the article/supplementary materials. Contact the corresponding author(s) directly to request the underlying data.

Author contributions

Conceptualization: Sumiko Ohnami, Takeshi Nagashima, Keiichi Hatakeyama, Kouji Maruyama, Shumpei Ohnami

Investigation: Sumiko Ohnami, Takeshi Nagashima, Keiichi Hatakeyama, Shumpei Ohnami

Supervision: Shumpei Ohnami, Ken Yamaguchi

Visualization: Sumiko Ohnami, Takeshi Nagashima, Koji Maruyama, Maki Mizuguchi, Sou Nakatani, Fukumi Kamada, Yuji Shimoda, Keiichi Ohshima, Kenichi Urakami, Shumpei Ohnami

Writing – original draft: Sumiko Ohnami, Shumpei Ohnami

Writing – review and editing: Sumiko Ohnami, Takeshi Nagashima, Keiichi Hatakeyama, Shumpei Ohnami, Ken Yamaguchi

Ethical approval and consent to participate

Not applicable.

Patient consent for publication

Not applicable.

REFERENCES

•  Charnaud,  S.,  Munro,  J.E.,  Semenec,  L.,  Mazhari,  R.,  Brewster,  J.,  Bourke,  C.,  Ruybal-Pesántez,  S.,  James,  R.,  Lautu-Gumal,  D.,  Karunajeewa,  H.,  Mueller,  I. and  Bahlo,  M. (2022): PacBio long-read amplicon sequencing enables scalable high-resolution population allele typing of the complex CYP2D6 locus. Commun. Biol., 5, 168.
•  Favero,  F.,  Joshi,  T.,  Marquard,  A.M.,  Birkbak,  N.J.,  Krzystanek,  M.,  Li,  Q.,  Szallasi,  Z. and  Eklund,  A.C. (2015): Sequenza: allele-specific copy number and mutation profiles from tumor sequencing data. Ann. Oncol., 26, 64-70.
•  Fukunaga,  K.,  Hishinuma,  E.,  Hiratsuka,  M.,  Kato,  K.,  Okusaka,  T.,  Saito,  T.,  Ikeda,  M.,  Yoshida,  T.,  Zembutsu,  H.,  Iwata,  N. and  Mushiroda,  T. (2021): Determination of novel CYP2D6 haplotype using the targeted sequencing followed by the long-read sequencing and the functional characterization in the Japanese population. J. Hum. Genet., 66, 139-149.
•  Gaedigk,  A.,  Bradford,  L.D.,  Alander,  S.W. and  Leeder,  J.S. (2006): CYP2D6*36 gene arrangements within the cyp2d6 locus: association of CYP2D6*36 with poor metabolizer status. Drug Metab. Dispos., 34, 563-569.
•  He,  W.,  Grassmann,  F.,  Eriksson,  M.,  Eliasson,  E.,  Margolin,  S.,  Thorén,  L.,  Hall,  P. and  Czene,  K. (2020): CYP2D6 genotype predicts tamoxifen discontinuation and prognosis in patients with breast cancer. J. Clin. Oncol., 38, 548-557.
•  Hosono,  N.,  Kato,  M.,  Kiyotani,  K.,  Mushiroda,  T.,  Takata,  S.,  Sato,  H.,  Amitani,  H.,  Tsuchiya,  Y.,  Yamazaki,  K.,  Tsunoda,  T.,  Zembutsu,  H.,  Nakamura,  Y. and  Kubo,  M. (2009): CYP2D6 genotyping for functional-gene dosage analysis by allele copy number detection. Clin. Chem., 55, 1546-1554.
•  Jarvis,  J.P.,  Peter,  A.P. and  Shaman,  J.A. (2019): Consequences of CYP2D6 copy-number variation for pharmacogenomics in psychiatry. Front. Psychiatry, 10, 432.
•  Kisoi,  M.,  Imai,  M.,  Yamamura,  M.,  Sakaguchi,  Y.,  Murata,  S.,  Ichikawa,  A. and  Kinoshita,  K. (2020): Unique genotyping protocol of CYP2D6 allele frequency using real time quantitative PCR from Japanese healthy women. Biol. Pharm. Bull., 43, 904-907.
•  Kiyotani,  K.,  Mushiroda,  T.,  Zembutsu,  H. and  Nakamura,  Y. (2013): Important and critical scientific aspects in pharmacogenomics analysis: lessons from controversial results of tamoxifen and CYP2D6 studies. J. Hum. Genet., 58, 327-333.
•  Kurose,  K.,  Sugiyama,  E. and  Saito,  Y. (2012): Population differences in major functional polymorphisms of pharmacokinetics/pharmacodynamics-related genes in Eastern Asians and Europeans: implications in the clinical trials for novel drug development. Drug Metab. Pharmacokinet., 27, 9-54.
•  Molden,  E. and  Jukić,  M.M. (2021): CYP2D6 reduced function variants and genotype/phenotype translations of CYP2D6 intermediate metabolizers: implications for personalized drug dosing in psychiatry. Front. Pharmacol., 12, 650750.
•  Nagashima,  T.,  Yamaguchi,  K.,  Urakami,  K.,  Shimoda,  Y.,  Ohnami,  S.,  Ohshima,  K.,  Tanabe,  T.,  Naruoka,  A.,  Kamada,  F.,  Serizawa,  M.,  Hatakeyama,  K.,  Matsumura,  K.,  Ohnami,  S.,  Maruyama,  K.,  Mochizuki,  T.,  Kusuhara,  M.,  Shiomi,  A.,  Ohde,  Y.,  Terashima,  M.,  Uesaka,  K.,  Onitsuka,  T.,  Nishimura,  S.,  Hirashima,  Y.,  Hayashi,  N.,  Kiyohara,  Y.,  Tsubosa,  Y.,  Katagiri,  H.,  Niwakawa,  M.,  Takahashi,  K.,  Kashiwagi,  H.,  Nakagawa,  M.,  Ishida,  Y.,  Sugino,  T.,  Takahashi,  M. and  Akiyama,  Y. (2020): Japanese version of The Cancer Genome Atlas, JCGA, established using fresh frozen tumors obtained from 5143 cancer patients. Cancer Sci., 111, 687-699.
•  Nagashima,  T.,  Yamaguchi,  K.,  Urakami,  K.,  Shimoda,  Y.,  Ohnami,  S.,  Ohshima,  K.,  Tanabe,  T.,  Naruoka,  A.,  Kamada,  F.,  Serizawa,  M.,  Hatakeyama,  K.,  Ohnami,  S.,  Maruyama,  K.,  Mochizuki,  T.,  Mizuguchi,  M.,  Shiomi,  A.,  Ohde,  Y.,  Bando,  E.,  Sugiura,  T.,  Mukaigawa,  T.,  Nishimura,  S.,  Hirashima,  Y.,  Mitsuya,  K.,  Yoshikawa,  S.,  Kiyohara,  Y.,  Tsubosa,  Y.,  Katagiri,  H.,  Niwakawa,  M.,  Takahashi,  K.,  Kashiwagi,  H.,  Yasunaga,  Y.,  Ishida,  Y.,  Sugino,  T.,  Kenmotsu,  H.,  Terashima,  M.,  Takahashi,  M.,  Uesaka,  K. and  Akiyama,  Y. (2024): Evaluation of whole genome sequencing utility in identifying driver alterations in cancer genome. Sci. Rep., 14, 23898.
•  Nofziger,  C.,  Turner,  A.J.,  Sangkuhl,  K.,  Whirl-Carrillo,  M.,  Agúndez,  J.A.,  Black,  J.L.,  Dunnenberger,  H.M.,  Ruano,  G.,  Kennedy,  M.A.,  Phillips,  M.S.,  Hachad,  H.,  Klein,  T.E. and  Gaedigk,  A. (2020): PharmVar GeneFocus: CYP2D6. Clin. Pharmacol. Ther., 107, 154-170.
•  Ohnami,  S.,  Hatakeyama,  K.,  Nagashima,  T.,  Maruyama,  K.,  Isaka,  M.,  Kamada,  F.,  Nakatani,  S.,  Shimoda,  Y.,  Mizuguchi,  M.,  Naruoka,  A.,  Miyake,  K.,  Ohshima,  K.,  Ohnami,  S.,  Akiyama,  Y.,  Urakami,  K. and  Yamaguchi,  K. (2025): Novel CYP2A7/CYP2A6 germline hybrids associated with mutation burden in lung cancer revealed by whole-genome long-read sequencing. Sci. Rep., 15, 44989.
•  Ohnami,  S.,  Nagashima,  T.,  Urakami,  K.,  Shimoda,  Y.,  Kamada,  F.,  Saito,  J.,  Naruoka,  A.,  Serizawa,  M.,  Masuda,  Y.,  Ohnami,  S.,  Kusuhara,  M. and  Yamaguchi,  K. (2017): Whole exome sequencing detects variants of genes that mediate response to anticancer drugs. J. Toxicol. Sci., 42, 137-144.
•  Ohnami,  S.,  Naruoka,  A.,  Isaka,  M.,  Mizuguchi,  M.,  Nakatani,  S.,  Kamada,  F.,  Shimoda,  Y.,  Sakai,  A.,  Ohshima,  K.,  Hatakeyama,  K.,  Maruyama,  K.,  Ohde,  Y.,  Kenmotsu,  H.,  Takahashi,  T.,  Akiyama,  Y.,  Nagashima,  T.,  Urakami,  K.,  Ohnami,  S. and  Yamaguchi,  K. (2022): Comparison of genetic susceptibility to lung adenocarcinoma and squamous cell carcinoma in Japanese patients using a novel panel for cancer-related drug-metabolizing enzyme genes. Sci. Rep., 12, 17928.
•  Petrović,  J.,  Pešić,  V. and  Lauschke,  V.M. (2020): Frequencies of clinically important CYP2C19 and CYP2D6 alleles are graded across Europe. Eur. J. Hum. Genet., 28, 88-94.
•  Robinson,  J.T.,  Thorvaldsdóttir,  H.,  Winckler,  W.,  Guttman,  M.,  Lander,  E.S.,  Getz,  G. and  Mesirov,  J.P. (2011): Integrative genomics viewer. Nat. Biotechnol., 29, 24-26.
•  Taylor,  C.,  Crosby,  I.,  Yip,  V.,  Maguire,  P.,  Pirmohamed,  M. and  Turner,  R.M. (2020): A review of the important role of CYP2D6 in pharmacogenomics. Genes (Basel), 11, 1295.
•  Yamaguchi,  K.,  Urakami,  K.,  Ohshima,  K.,  Mochizuki,  T.,  Akiyama,  Y.,  Uesaka,  K.,  Nakajima,  T.,  Takahashi,  M.,  Tamai,  S. and  Kusuhara,  M. (2014): Implementation of individualized medicine for cancer patients by multiomics-based analyses—the Project HOPE—. Biomed. Res., 35, 407-412.
•  Yang,  Y.,  Botton,  M.R.,  Scott,  E.R. and  Scott,  S.A. (2017): Sequencing the CYP2D6 gene: from variant allele discovery to clinical pharmacogenetic testing. Pharmacogenomics, 18, 673-685.