Skip to main content
Download PDF

Discovering novel germline genetic variants linked to severe fluoropyrimidine-related toxicity in- and outside DPYD

Genome Medicine volume 16, Article number: 101 (2024) Cite this article

Abstract

Background

The Alpe-DPD study (NCT02324452) demonstrated that prospective genotyping and dose-individualization using four alleles in DPYD (DPYD*2A/rs3918290, c.1236G > A/rs75017182, c.2846A > T/rs67376798 and c.1679 T > G/rs56038477) can mitigate the risk of severe fluoropyrimidine toxicity. However, this could not prevent all toxicities. The goal of this study was to identify additional genetic variants, both inside and outside DPYD, that may contribute to fluoropyrimidine toxicity.

Methods

Biospecimens and data from the Alpe-DPD study were used. Exon sequencing was performed to identify risk variants inside DPYD. In silico and in vitro analyses were used to classify DPYD variants. A genome-wide association study (GWAS) with severe fluoropyrimidine-related toxicity was performed to identify variants outside DPYD. Association with severe toxicity was assessed using matched-pair analyses for the exon sequencing and logistic, Cox, and ordinal regression analyses for GWAS.

Results

Twenty-four non-synonymous, frameshift, and splice site DPYD variants were detected in ten of 986 patients. Seven of these variants (c.1670C > T, c.1913 T > C, c.1925 T > C, c.506delC, c.731A > C, c.1740 + 1G > T, c.763 − 2A > G) were predicted to be deleterious. The carriers of either of these variants showed a trend towards a 2.14-fold (95% CI, 0.41–11.3, P = 0.388) increased risk of severe toxicity compared to matched controls (N = 30). After GWAS of 942 patients, no individual single nucleotide polymorphisms achieved genome-wide significance (P ≤ 5 × 10−8), however, five variants were suggestive of association (P < 5 × 10−6) with severe toxicity.

Conclusions

Results from DPYD exon sequencing and GWAS analysis did not identify additional genetic variants associated with severe toxicity, which suggests that testing for single markers at a population level currently has limited clinical value. Identifying additional variants on an individual level is still promising to explain fluoropyrimidine-related severe toxicity. In addition, studies with larger samples sizes, in more diverse cohorts are needed to identify potential clinically relevant genetic variants related to severe fluoropyrimidine toxicity.

Background

Fluoropyrimidines, including 5-fluorouracil (5-FU) and capecitabine, represent the backbone of chemotherapeutic regimens used in the treatment of solid tumours, such as gastroesophageal, colorectal, and breast cancer. Depending on the treatment regimen administered, severe fluoropyrimidine-induced toxicity affects approximately 35% of recipients and can be lethal in up to 1% of the patients [1]. Common fluoropyrimidine-induced adverse events include diarrhoea, mucositis, hand-foot syndrome, and myelosuppression. An increased risk for the development of fluoropyrimidine-induced toxicity exists in patients with a deficiency of dihydropyrimidine dehydrogenase (DPD), an enzyme that is encoded by the DPYD gene and responsible for catalysing 5-FU degradation into inactive metabolites [2]. Both DPD activity and genetic variants in DPYD have been widely investigated and partially explain severe fluoropyrimidine-induced toxicity. Previous studies and meta-analyses have shown a strong association between four DPYD variants (c.1905 + 1G > A/rs3918290, c.1236G > A/rs56038477, c.2846A > T/rs67376798, and c.1679 T > G/rs55886062) and severe fluoropyrimidine-induced toxicity [3, 4]. Recently, we showed that patients’ safety indeed improved following fluoropyrimidine dose individualization based on DPYD-genotyping of the four DPYD variants mentioned above [5]. Consistent with these findings, the European Medicine Agency (EMA) recently recommended that all patients scheduled for fluoropyrimidine treatment should be tested for DPD deficiency before the start of treatment [6].

Despite the recognition of the importance of the abovementioned four variants in reducing toxicity, approximately 25% of DPYD wild-type patients still experienced severe fluoropyrimidine-induced toxicity [5]. This suggests that additional factors, including other DPYD genetic variants and/or variants affecting other genes involved in fluoropyrimidine metabolism, may contribute to toxicity. Indeed, low-frequency germline variants (minor allele frequencies (MAFs) < 1%) may explain approximately 30–40% of inter-individual functional variability in pharmacogenes [7]. However, the effect of these low-frequency variants in DPYD has not been assessed comprehensively in fluoropyrimidine-treated patient populations.

In the present study, we sought to identify potential biomarkers of severe fluoropyrimidine toxicity risk in a patient population that did not carry any of the four well-characterized risk alleles in DPYD. To accomplish this goal, we used complementary approaches for genotyping that included both targeted sequencing of the exon-coding region for DPYD and genome-wide association study (GWAS) in cancer patients treated with fluoropyrimidines.

Methods

Patients

Clinical data including baseline characteristics and toxicity data were derived from patients included in the Alpe-DPD study (clinicaltrial.gov identifier NCT02324452) [5]. The design, study population, and results of the Alpe-DPD study have been previously published [5]. In brief, adult patients (≥ 18 years) with cancer who were intended to start treatment with fluoropyrimidine-based therapy in 17 hospitals in the Netherlands were included. Patients with all tumour types for which fluoropyrimidine treatment was indicated were eligible. Prospective genotyping for DPYD*2A, c.2846A > T, c.1679 T > G, and c.1236G > A was performed. Heterozygous DPYD variant carriers received an initial dose reduction of 50% (DPYD*2A and c.1679 T > G) or 25% (c.1236G > A and c.2846A > T). DPYD wildtype patients were treated according to standard-of-care. The primary endpoint of the study was the frequency of severe fluoropyrimidine-related toxicity across the entire treatment duration. All patients (N = 1181 of which 1103 were evaluable) signed informed consent before inclusion in the study, which included approval for the use of clinical data and remaining material (whole blood samples taken before the start of the chemotherapy in a 4-ml EDTA tube) to perform DPYD sequencing and GWAS. In- and exclusion criteria can be found in the Additional file 1. The baseline characteristics of the cohort can be found in Table 1. Toxicity was graded according to the National Cancer Institute common terminology criteria for adverse events (CTC-AE; version 4.03) and severe toxicity was defined as CTC-AE grade ≥ 3 [8]. Only the highest graded adverse events classified as possible, probable, or definite related to fluoropyrimidines were included in the analyses [5].

Table 1 Patient characteristics
Full size table

DPYD sequencing

Genotyping

Targeted DNA sequencing was performed for specimens with adequate DNA (N = 1103) extracted from whole blood samples (4 ml in an EDTA tube). Sequencing libraries were generated using Access Array chemistry (Fluidigm, South San Francisco, CA) as previously described, with modifications [9]. Custom primer panels were designed to cover all 23 exons of the DPYD gene and the intronic region containing rs75017182, the causal single nucleotide polymorphisms (SNP) in perfect linkage with c.1236G > A/HapB3. Target amplification and sample indexing were performed using Juno Targeted Sequencing LP 192.24 Integrated Fluidic Circuits (IFCs) on a Juno instrument (Fluidigm). Indexed sequencing libraries from 2 IFCs were pooled, and paired-end sequencing was performed using an Illumina HiSeq 4000 in the Mayo Clinic Cancer Center Genome Analysis Core. Patient sequence data were demultiplexed using barcode sequences added during library preparation. Adapter and region-specific primer sequences were pruned, and reads were aligned to targeted regions of the hg38 human reference genome using BWA-MEM. Variants were identified using GATK HaplotypeCaller. A QUAL score of ≥ 500 across the population of samples tested was used as a threshold for variant inclusion in subsequent analyses. The presence of toxicity-associated variants (DPYD*2A, c.1236G > A, c.2846A > T, or c.1679 T > G) was confirmed using previous genotyping data [5]. The genotypes for additional rare variants with allele frequencies less than 1% in the study population were confirmed in carriers by Sanger sequencing of the relevant exon at the Mayo Clinic Cancer Center Genomics Analysis Core using methods that have been previously described [9].

Variant classification

In this study, different in vitro and in silico approaches were used to assess the potential effect of identified DPYD variants (Fig. 1). Missense variants were evaluated using a previously published in vitro expression system in HEK293T/c17 cells. If available, results for variants were reused, otherwise, novel variants were expressed in the in vitro system [10, 11]. Detailed primer sequences used to generate the expression plasmids for selected variants are reported in the Additional file 1: Table S1. The DPYD-Varifier, a DPYD-specific in silico prediction tool applied for eligible variants [12]. Results of the in vitro assay are used as the final decision of the variant function if the DPYD-Varifier has an inconsistent prediction [12]. Frameshift variants were considered deleterious based on previous findings [11]. The potential impact of splice site variants was predicted using MMsplice, a modelling-based tool to predict genetic variation effects on splicing [13].

Fig. 1
figure 1

Study design. Severe fluoropyrimidine-related toxicity was defined as CTC-AE grade ≥ 3. *The four known DPYD variants are DPYD*2A, c.1236G > A, c.2846A > T, and c.1679 T > G

Full size image

Statistical analysis

After DPYD sequencing, all patients who carried at least one predicted deleterious variant were matched with wild-type controls identified from the Alpe-DPD study participants to compare toxicity. To perform matching, we first studied associations of factors with toxicity one by one. Automatic matching (in a 1:3 ratio) was performed based on the three parameters that were most strongly associated with toxicity: treatment regimen, tumour type, and disease stage. If more than three eligible wild-type controls that fulfilled all matching criteria were available, these matches were selected at random from these eligible controls. Fisher’s exact test (conducted in SPSS) was conducted to compare the incidence of severe toxicity between deleterious variant carriers and their matched controls as the variant frequency was low. Due to the small sample size, we did not include a random effect for the matching cluster. Matching can improve the stability of statistical models as the matching variables do not have to be included in the model. For comparison, we have also analysed the non-matched sample with a logistic regression including the matching variables as covariates (conducted in R). Statistical analyses were conducted using SPSS version 25.0 and R version 4.2.3.

Genome-wide association study

Genotyping was performed at the Human Genotyping Facility of the Erasmus University Medical Center, using the Infinium Global Screening Array (GSA) v1.0 [14]. The array contains 692,842 SNPs and includes rare variants with allele frequencies < 1%. A minor allele frequency (MAF) threshold of 0.5% was used for the primary analysis. 1000 Genomes reference phase 3 GRCh37.p13 was used to impute the data [15, 16] Quality control (QC) checks can be found in the Additional file 1: Text “GWAS—Quality control”. Genetic variants were tested for an association with the onset of severe fluoropyrimidine-induced toxicity. The primary outcome was severe (grade ≥ 3) fluoropyrimidine-induced toxicity, compared to grade ≤ 1 fluoropyrimidine-induced toxicity. Grade 2 toxicity was excluded from the primary analysis to maximize the contrast between toxicities (Fig. 1). Additionally, as a sensitivity analysis, severe fluoropyrimidine-induced toxicity was also compared to grade ≤ 2 fluoropyrimidine-induced toxicity. Gender, age, baseline body surface area (BSA), and treatment regimen (grouped as previously published [5]) were used as pre-specified covariates. Statistical analyses were performed in R statistics version 4.3.2 [17]. Base packages stats, survival, and MASS were used to evaluate logistic, Cox, and ordinal regression analyses, respectively. A p value threshold of ≤ 5 × 10−8 was used for determining significance at the genome-wide level. Post-association QC was performed by visual inspection of Quantile–Quantile (QQ) plots of p values of association tests and computation of the inflation factor. Online databases (Linkage-Disequilibrium tools, hapreg, and genome browser) were used to explore possible biological mechanisms of genome-wide associated or suggestive novel SNPs [18,19,20].

Power analysis

We performed power calculations based on the marginal event rate of 34% for the primary outcome. We based the calculations on 559 individuals (or 1118 alleles). Based on several allele frequencies, we calculated detectable effect size for a power of 80% and an alpha level of 5e − 8. For allele frequencies of 10%, 25%, and 50%, detectable ORs of risk alleles are 3.8, 2.4, and 2.2, respectively (Additional file 1: Table S2).

Results

Cohort

Patient characteristics are shown in Table 1. In total, 1181 were included in the Alpe-DPD study, of which 1103 were evaluable (Fig. 2). Of these, 85 DPYD variant carriers (DPYD*2A, c.1236G > A/HapB3, c.2846A > T, or c.1679 T > G) were treated with a reduced dose and consequently excluded from analyses, resulting in 1018 patients being evaluable for DPYD sequencing and GWAS analysis. Whole exon sequencing failures and GWAS quality control checks led to the exclusion of 32 and 74 patients, resulting in 986 and 942 patients being included in the DPYD sequencing analysis and GWAS, respectively. As stated in the methods section, patients with grade 2 toxicity were disregarded in the GWAS analysis, leading to 599 patients in the GWAS cohort.

Fig. 2
figure 2

Flowchart of patient inclusion. Patients who experienced grade 2 toxicity were excluded from the GWAS analyses to maximize the contrast between severe and non-severe toxicity. Abbreviations: QC, quality control; DPYD, gene encoding dihydropyrimidine dehydrogenase; NT, not tested, NA, prediction not available

Full size image

DPYD sequencing and variant function prediction

A total of 24 non-synonymous, frameshift, and splice site variants were detected in 986 individuals (excluding patients carrying any of the four known variants (DPYD*2A, c.1236G > A, c.2846A > T, and c.1679 T > G) tested for in the Alpe-DPD study. Of these variants, 20 were in exons and four were in introns (Table 2). The frequencies and results of the functional assessment with the in vitro assay and the MMsplice are described in Table 2. In total, seven rare deleterious variants were identified, which were carried by 10 individuals. Five variants in the coding region (c.1670C > T, c.1913 T > C, c.1925 T > C, c.506delC, and c.731A > C) and two variants in the flanking splice region (c.1740 + 1G > T and c.763 − 2A > G) were predicted to be deleterious. Of these seven variants, only c.1670C > T and c.763 − 2A > G have been reported previously. The remaining seventeen non-synonymous variants were predicted benign, of which 3 have not yet been previously noted in dbSNP. None of the 24 variants were classified as decreased function or no function allele according to the CPIC guideline.

Table 2 Frequencies and results of functional assessment of all variants
Full size table

Out of the patients who carried predicted deleterious variants, 3 of 10 (30%) patients developed severe toxicity. None of the 10 deleterious variants was related to a statistically significant increased risk of several toxicities compared with matched controls who did not carry any deleterious DPYD variant, of which, 16.7% (5 out of 30 patients) experienced severe toxicities (OR 2.143, P < 0.388; Table 3). In addition, all grade toxicity was comparable between carriers and matched non-carriers (OR 2.25, p < 0.656; Table 3). The patient characteristics of ten carriers and their matched control are shown in the Additional file 2: Table S3. In a post hoc exploratory analysis also no statistically significant difference in severe toxicity was found between patients carrying a predicted deleterious variant (n = 10) and 976 non-carriers (logistic regression: OR 1.49 CI 0.38 to 5.87, P = 0.57, Additional file 2: Table S4).

Table 3 Matched pair analysis of novel deleterious variants
Full size table

Genome-wide association analysis

GWAS was assessed for severe (grade ≥ 3) toxicity and was compared to grade 0 or 1 toxicity in 599 patients (excluding 343 patients with grade 2 toxicity, Fig. 2). The number of patients varied per SNP due to genotype missingness, which was limited to up to 3% as per QC. An association test for severe fluoropyrimidine-induced toxicity (CTC-AE grades 3–5) was performed for a total of 4,650,899 markers. Gender, age, baseline BSA, and treatment type were included as covariates. The corresponding Manhattan and QQ plots are shown in the Additional file 2: Fig. S1 and Fig. S2. The inflation factor is 1.04. While none of the individual SNPs achieved genome-wide significance as per the pre-specified definition (p ≤ 5 × 10−8), five variants (rs17114875, rs367239, rs77579689, rs114105116, and rs12622722) showed a suggestive association with severe toxicity, with p values between 5 × 10−8 and 5 × 10−6. The closest annotated genes to rs171114875 are PRKD1 and MIR548AI. The closest annotated gene to rs77579689 is KHDRBS3. The closest annotated genes to rs367239 are VENTXP7 and ZNF385D. Additionally, rs367239 is in linkage disequilibrium with rs1396004 and rs341838 which are both SNPs located in VENTXP7. The other two suggestive variants are listed as intronic variants of the non-coding RNA gene LOC101927414 (rs114105116) and protein-coding gene COL6A3 (rs12622722). The 30 most significantly associated markers are shown in Table 4. None of these SNPs have previously been reported in publications or the ClinVar database of the National Center for Biotechnology Information (NCBI) [15].

Table 4 Thirty genetic variants with the lowest p values
Full size table

Discussion

While applying prospective DPYD genotyping to clinical practice has successfully reduced the incidence of severe toxicity, a substantial number of patients treated with fluoropyrimidines still experience severe treatment-related toxicity [5]. We hypothesized that in addition to the four established DPYD variants, other genetic variations in- and outside DPYD might be associated with the onset of severe fluoropyrimidine-related toxicity. Therefore, we performed comprehensive genetic analyses including whole exon sequencing of DPYD and a GWAS analysis in a large well-characterized cohort derived from a prospective clinical study consisting of 1103 mostly Caucasian patients (95%) treated with fluoropyrimidine-based chemotherapy [5]. Within DPYD, we detected 24 non-synonymous and splice site variants, of which 7 allele variants that were carried in 10 patients were predicted to be deleterious. In the matched-pair analysis, the carriers of these deleterious variants showed a statistically non-significant twofold higher risk of severe toxicity. These findings imply that patients with rare deleterious variants may be at increased risk of severe fluoropyrimidine-related toxicity.

Out of the 24 detected variants, 5 deleterious variants are novel and would have been missed with a pre-designed panel test, highlighting the potential of the combination of next-generation sequencing (NGS) with available functionality assessment tools in detecting deleterious variants and preventing life-threatening toxicity. Yet, despite analysis of a large cohort of over 1000 patients, the number of novel deleterious DPYD variants remains low. Moreover, it is challenging to connect these unique variants to clinical decisions or upfront dose reductions because of the risk of undertreatment, limiting clinical application. By contrast, a study consisting of 120 patients developing grade 3–5 fluoropyrimidine-related toxicity and 104 matched controls identified a significant increased risk of patients carrying at least one rare missense DPYD variant [21]. Therefore, additional studies on implementing these approaches are needed, especially in understudied populations, which are more likely to carry other deleterious DPYD variants in addition to the four commonly tested ones [9]. However, in our cohort, even after accounting for the additional deleterious variants in DPYD, unexplained severe fluoropyrimidine-induced toxicity remained. Potentially, this remaining toxicity is the result of genetic variation outside DPYD. Several GWAS studies have been performed in patients and cell lines in attempts to identify novel risk variants [22,23,24]. These previous studies failed to identify associations that reached genome-wide significance, possibly due to limitations including small sample size and focus on specific toxicities such as neutropenia or leucopenia [22,23,24]. Similarly, no variants in our GWAS reached genome-wide significance despite the comparatively large sample size and broader definition of fluoropyrimidine-associated toxicity, suggesting that non-genetic variables and/or more complex interactions between genetic components, with each exerting a small effect size, contribute to the occurrence of severe fluoropyrimidine induced toxicity. Polygenic risk models are an attractive approach to address this issue; however, such analyses require far more patients than are available in our study.

Although no genome-wide significant SNPs were identified, we did identify five variants suggestive of association with severe fluoropyrimidine-induced toxicity that might provide insight into possible alternative mechanisms that contribute to fluoropyrimidine toxicity. To the best of our knowledge, these SNPs have not been previously described in relation to fluoropyrimidines [25, 26].

One trade-off of not considering patients who experienced grade 2 toxicity is that not all patients of the cohort are included in the association analysis. Therefore, we conducted a sensitivity analysis by including the patients with grade 2 fluoropyrimidine-related toxicity (grade 0–2 vs. grade 3–5), thereby increasing the number of patients while reducing the contrast between toxicities. Yet, this did not result in a different outcome (Additional file 2: Fig. S3 and Table S5). Furthermore, as toxicities can differ between capecitabine and 5-FU, we repeated the GWAS with patients receiving capecitabine as this was the majority of patients (494 (82%) of 599). This analysis did not result in a different outcome.

Our results indicate that DPYD exonic variants, especially predicted deleterious variants, as well as the five GWAS variants that were found to be suggestive of association with severe fluoropyrimidine-induced toxicity, are candidate SNPs that are valuable for further study. However, a substantial part of the observed fluoropyrimidine-related toxicity remains unexplained and other explanations such as the contribution of variants in non-coding regions and rare variants outside DPYD should also be considered. Recently, the association between rare variants in DPYS, a gene involved in the catabolic pathway with DPYD, and fluoropyrimidine-related severe toxicity was reported, which suggested a fourfold increased risk of cumulative severe toxicity [25]. Furthermore, while exon sequencing had a limited contribution to explaining the remaining severe fluoropyrimidine-toxicity in our patient population consisting of mostly Caucasians (95%), other genetic variants may be present in more ethnically diverse study populations as it is well-known that these facilitate the identification of genetic risk factors [26]. Additional studies in populations with greater ancestral diversity are therefore needed. Our analyses applied a toxicity definition of overall ≥ grade 3 toxicity during the entire treatment duration as used previously [25]. However, the effect of genetic contributors to toxicity might be more obvious in the earlier cycles of treatment. To further address this, we performed a sensitivity analysis with overall ≥ grade 3 toxicity during the first two treatment cycles as the endpoint for both the GWAS and the matched pair analyses. These analyses did not detect any new signals (Additional file 2: Fig. S4 and Table S6, S7, S8).

In conclusion, our results from DPYD exon sequencing and GWAS analysis suggest that at a population level it is not likely that, besides the four established DPYD variants, genetic variants either inside or outside DPYD have a clinically relevant contribution to severe fluoropyrimidine-induced toxicity in patients treated with fluoropyrimidines. However, at patient level, it cannot be excluded—based upon this study—that a rare variant is causing toxicity. Therefore, larger studies, in more diverse populations are needed to identify these additional variants.

Conclusions

Results from DPYD exon sequencing and GWAS analysis did not identify additional genetic variants associated with severe fluoropyrimidine-related toxicity, which suggests that testing for single markers at a population level currently has limited clinical value. Identifying additional variants on an individual level is still promising to explain fluoropyrimidine-related severe toxicity. In addition, studies with larger sample sizes, in more diverse cohorts are needed to identify potential clinically relevant genetic variants related to severe fluoropyrimidine toxicity.

Availability of data and materials

Sequence and GWAS data that support the findings of this study are available in our European Genome and Phenotype Archive (EGA) account EGA-box-1314 accession code EGAS00001007855.

References

  1. Hoff PM, Ansari R, Batist G, et al. Comparison of oral capecitabine versus intravenous fluorouracil plus leucovorin as first-line treatment in 605 patients with metastatic colorectal cancer: results of a randomized phase III study. J Clin Oncol. 2017;19(8):2282–92.

    Article  Google Scholar 

  2. Van Kuilenburg ABP. Dihydropyrimidine dehydrogenase and the efficacy and toxicity of 5-fluorouracil. Eur J Cancer. 2004;40(7):939–50.

    Article  PubMed  Google Scholar 

  3. Terrazino S, Cargnin S, Del Re M, Danesi R, Canonico PL, Genazzani AA. DPYD IVS14 + 1G > A and 2846A > T genotyping for the prediction of severe fluoropyrimidine-related toxicity : a meta-analysis. Pharmacogenomics. 2013;14(11):1255–72.

    Article  Google Scholar 

  4. Meulendijks D, Henricks LM, Sonke GS, et al. Clinical relevance of DPYD variants c1679T>G, c.1236G>A/HapB3, and c.1601G>A as predictors of severe fluoropyrimidine-associated toxicity: A systematic review and meta-analysis of individual patient data. Lancet Oncol. 2015;16(16):1639–50.

    Article  CAS  PubMed  Google Scholar 

  5. Henricks LM, Lunenburg CATC, de Man FM, et al. DPYD genotype-guided dose individualisation of fluoropyrimidine therapy in patients with cancer: a prospective safety analysis. Lancet Oncol. 2018;19(11):1459–67.

    Article  CAS  PubMed  Google Scholar 

  6. European Medicines Agency. EMA recommendations on DPD testing prior to treatment with fluorouracil, capecitabine, tegafur and flucytosine. EMA recommendations on DPD testing prior to treatment with fluorouracil, capecitabine, tegafur and flucytosine 2020;31(April):3. Available from: https://www.ema.europa.eu/en/news/ema-recommendations-dpd-testing-prior-treatment-fluorouracil-capecitabine-tegafur-flucytosine

  7. Kozyra M, Ingelman-Sundberg M, Lauschke VM. Rare genetic variants in cellular transporters, metabolic enzymes, and nuclear receptors can be important determinants of interindividual differences in drug response. Genet Med. 2017;19(1):20–9.

    Article  CAS  PubMed  Google Scholar 

  8. NCI. National Cancer Institute: Common Terminology Criteria for Adverse Events v4.03. 2017;https://evs.nci.nih.gov/ftp1/CTCAE/CTCAE_4.03_2010.

  9. Elraiyah T, Jerde CR, Shrestha S, et al. Novel Deleterious Dihydropyrimidine Dehydrogenase variants may contribute to 5-Fluorouracil sensitivity in an East African Population. Clin Pharmacol Ther. 2017;101(3):382–90.

    Article  CAS  PubMed  Google Scholar 

  10. Offer SM, Wegner NJ, Fossum C, Wang K, Diasio RB. Phenotypic profiling of DPYD variations relevant to 5-fluorouracil sensitivity using real-time cellular analysis and in vitro measurement of enzyme activity. Cancer Res. 2013;73(6):1958–68.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  11. Offer SM, Fossum CC, Wegner NJ, Stuflesser AJ, Butterfield GL, Diasio RB. Comparative functional analysis of dpyd variants of potential clinical relevance to dihydropyrimidine dehydrogenase activity. Cancer Res. 2014;74(9):2545–54.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  12. Shrestha S, Zhang C, Jerde CR, et al. Gene-specific variant classifier (DPYD-Varifier) to identify Deleterious Alleles of Dihydropyrimidine Dehydrogenase. Clin Pharmacol Ther. 2018;104(4):709–18.

    Article  CAS  PubMed  Google Scholar 

  13. Cheng J, Nguyen TYD, Cygan KJ, et al. MMSplice: Modular modeling improves the predictions of genetic variant effects on splicing. Genome Biol. 2019;20(1):48.

    Article  PubMed  PubMed Central  Google Scholar 

  14. Illumina. Infinium Global Screening Array-24 v1 . 0. 2017;http://glimdna.org/assets/2017-infinium-global-scr.

  15. NCBI. National Center for Biotechnology Information. ClinVar database. 2021. :https://www.ncbi.nlm.nih.gov/clinvar/. Accessed: 05 August

  16. NCBI. National Center for Biotechnology Information. SNP database (dbSNP). :https://www.ncbi.nlm.nih.gov/snp/. Accessed: 05 August 2021

  17. Team. RCore. R: A Language and Evironment for Statistical Computing. 2018;https://www.R-project.org.

  18. University of California, Genomics Institute. UCSC Genome Browser. Retrieved from https://genome.ucsc.edu/ on January 2023.

  19. National Institutes of Health, National Cancer Institute. LDproxy Tool. Retrieved from https://ldlink.nih.gov/?tab=ldproxy on January 2023.

  20. Massachusetts Institute of Technology. Haploreg v4.1. Retrieved from https://pubs.broadinstitute.org/mammals/haploreg/haploreg.php on January 2023.

  21. De Mattia E, Silvestri M, Polesel J, et al. Rare genetic variant burden in DPYD predicts severe fluoropyrimidine-related toxicity risk. Biomed Pharmacother. 2022;154:13644.

    Article  Google Scholar 

  22. Low SK, Chung S, Takahashi A, et al. Genome-wide association study of chemotherapeutic agent-induced severe neutropenia/leucopenia for patients in Biobank Japan. Cancer Sci. 2013;104(8):1074–82.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  23. O’Donnell PH, Stark AL, Gamazon ER, et al. Identification of novel germline polymorphisms governing capecitabine sensitivity. Cancer. 2012;118(16):4063–73.

    Article  PubMed  Google Scholar 

  24. Van Huis-Tanja LH, Ewing E, Van Der Straaten RJHM, et al. Clinical validation study of genetic markers for capecitabine efficacy in metastatic colorectal cancer patients. Pharmacogenet Genomics. 2015;25(6):279–88.

    Article  PubMed  Google Scholar 

  25. De Mattia E, Polesel J, Silvestri M, et al. The burden of rare variants in DPYS gene is a novel predictor of the risk of developing severe fluoropyrimidine-related toxicity. Hum Genomics. 2023;17(1):99.

    Article  PubMed  PubMed Central  Google Scholar 

  26. Sirugo G, Williams SM, Tishkoff SA. The missing diversity in human genetic studies. Cell. 2019;177(1):26–31. https://doi.org/10.1016/j.cell.2019.02.048.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

Download references

Acknowledgements

The authors thank the Human Genotyping Facility of the Erasmus Medical Center for the execution of the genotyping and all participating centres of the NCT02324452 study for their help in including patients in the study.

Funding

This work was supported by a grant from Alpe-d’HuZes/Dutch Cancer Society NKI2013-6249. This study was supported by the ZonMw, The Netherlands Organization for Health Research and Development (project no. 848016007). Carin Lunenburg was supported by an unrestricted grant from Roche Pharmaceuticals. Linda Henricks and Carin Lunenburg were sponsored by the Dutch Cancer Society. There was no involvement from any of the funding sources in the study design, data collection, analysis, or interpretation of the data. Qinglian Zhai was supported by a Chinese Scholarship Council grant.

Author information

Author notes

  1. Jonathan E. Knikman and Qinglian Zhai contributed equally to this work.

Authors and Affiliations

  1. Division of Pharmacology, The Netherlands Cancer Institute, Amsterdam, The Netherlands

    Jonathan E. Knikman

  2. Department of Clinical Pharmacy and Toxicology, Leiden University Medical Center, Leiden, The Netherlands

    Qinglian Zhai, Stefan Böhringer, Maaike van der Lee, Henk-Jan Guchelaar & Jesse J. Swen

  3. Department of Medical Oncology, Leiden University Medical Center, Leiden, The Netherlands

    Carin A. T. C. Lunenburg, Johanna E. A. Portielje & Hans Gelderblom

  4. Department of Clinical Chemistry and Laboratory Medicine, Leiden University Medical Center, Leiden, The Netherlands

    Linda M. Henricks

  5. Department of Biomedical Data Sciences, Leiden University Medical Center, Leiden, The Netherlands

    Stefan Böhringer

  6. Department of Medical Oncology, Erasmus MC Cancer Institute, Erasmus University Medical Center, Rotterdam, The Netherlands

    Femke M. de Man & Ron H. J. Mathijssen

  7. Department of Pathology, Carver College of Medicine, University of Iowa, Iowa City, IA, USA

    Steven M. Offer

  8. Holden Comprehensive Cancer Center, University of Iowa, Iowa City, IA, USA

    Steven M. Offer

  9. Mayo Clinic Graduate School of Biomedical Sciences, Mayo Clinic, Rochester, MN, USA

    Steven M. Offer

  10. Division of Pulmonary and Critical Care Medicine, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA, USA

    Shikshya Shrestha

  11. Department of Medical Oncology, Catharina Hospital, Eindhoven, The Netherlands

    Geert-Jan Creemers

  12. Department of Internal Medicine, Hospital Gelderse Vallei, Ede, The Netherlands

    Arnold Baars

  13. Department of Internal Medicine, Reinier de Graaf Hospital, Delft, The Netherlands

    Vincent O. Dezentjé

  14. Division of Medical Oncology, The Netherlands Cancer Institute, Amsterdam, The Netherlands

    Vincent O. Dezentjé

  15. Department of Internal Medicine, Deventer Hospital, Deventer, The Netherlands

    Alexander L. T. Imholz

  16. Department of Internal Medicine, Haaglanden Medical Center, The Hague, The Netherlands

    Frank J. F. Jeurissen

  17. Department of Internal Medicine, Maastricht University Medical Center, Maastricht, The Netherlands

    Rob L. H. Jansen

  18. Department of Internal Medicine, Franciscus Gasthuis en Vlietland, Rotterdam, The Netherlands

    Paul Hamberg

  19. Department of Internal Medicine, Bravis Hospital, Roosendaal, The Netherlands

    Helga J. Droogendijk

  20. Department of Medical Oncology, University Medical Center Utrecht, Utrecht, The Netherlands

    Miriam Koopman

  21. Department of Internal Medicine, Wilhelmina Hospital Assen, Assen, The Netherlands

    Peter Nieboer

  22. Department of Internal Medicine, Laurentius Hospital, Roermond, The Netherlands

    Marlène H. W. van de Poel

  23. Department of Internal Medicine, Canisius-Wilhelmina Hospital, Nijmegen, The Netherlands

    Caroline M. P. W. Mandigers

  24. Department of Clinical Chemistry, Erasmus University Medical Center, Rotterdam, Netherlands

    Ron H. N. van Schaik

  25. Department of Pharmaceutical Sciences, Utrecht University, Utrecht, The Netherlands

    Jan H. M. Schellens

  26. Department of Gastroenterology and Hepatology, Division of Medical Oncology, The Netherlands Cancer Institute, Amsterdam, The Netherlands

    Annemieke Cats

Authors
  1. Jonathan E. Knikman
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Qinglian Zhai
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Carin A. T. C. Lunenburg
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Linda M. Henricks
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Stefan Böhringer
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Maaike van der Lee
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. Femke M. de Man
    View author publications

    You can also search for this author in PubMed Google Scholar

  8. Steven M. Offer
    View author publications

    You can also search for this author in PubMed Google Scholar

  9. Shikshya Shrestha
    View author publications

    You can also search for this author in PubMed Google Scholar

  10. Geert-Jan Creemers
    View author publications

    You can also search for this author in PubMed Google Scholar

  11. Arnold Baars
    View author publications

    You can also search for this author in PubMed Google Scholar

  12. Vincent O. Dezentjé
    View author publications

    You can also search for this author in PubMed Google Scholar

  13. Alexander L. T. Imholz
    View author publications

    You can also search for this author in PubMed Google Scholar

  14. Frank J. F. Jeurissen
    View author publications

    You can also search for this author in PubMed Google Scholar

  15. Johanna E. A. Portielje
    View author publications

    You can also search for this author in PubMed Google Scholar

  16. Rob L. H. Jansen
    View author publications

    You can also search for this author in PubMed Google Scholar

  17. Paul Hamberg
    View author publications

    You can also search for this author in PubMed Google Scholar

  18. Helga J. Droogendijk
    View author publications

    You can also search for this author in PubMed Google Scholar

  19. Miriam Koopman
    View author publications

    You can also search for this author in PubMed Google Scholar

  20. Peter Nieboer
    View author publications

    You can also search for this author in PubMed Google Scholar

  21. Marlène H. W. van de Poel
    View author publications

    You can also search for this author in PubMed Google Scholar

  22. Caroline M. P. W. Mandigers
    View author publications

    You can also search for this author in PubMed Google Scholar

  23. Ron H. N. van Schaik
    View author publications

    You can also search for this author in PubMed Google Scholar

  24. Hans Gelderblom
    View author publications

    You can also search for this author in PubMed Google Scholar

  25. Ron H. J. Mathijssen
    View author publications

    You can also search for this author in PubMed Google Scholar

  26. Jan H. M. Schellens
    View author publications

    You can also search for this author in PubMed Google Scholar

  27. Annemieke Cats
    View author publications

    You can also search for this author in PubMed Google Scholar

  28. Henk-Jan Guchelaar
    View author publications

    You can also search for this author in PubMed Google Scholar

  29. Jesse J. Swen
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

HJG, JJS, JK, CL, and LH conceptualized and designed the study. JK, QZ, CL, LH, SB, MvdL, JJS, and HJG developed the methodology. JK, QZ, CL, SB, FM, SO, and SS performed the analyses. GC, AB,VD, AI, FJ, JP, RJ, PH, HD, MK, PN, MvdP, CM, RS, HG, RM, JS, and AC provided study material of patients and collected data. JK, QZ, CL, JJS, and HJG drafted the manuscript. All authors read and approved the final manuscript.

Corresponding author

Correspondence to Henk-Jan Guchelaar.

Ethics declarations

Ethics approval and consent to participate

This study was approved by the institutional review board of the Netherlands Cancer Institute (Amsterdam, The Netherlands) and approval from the board of directors of each individual hospital was obtained for all participating centers and conformed to the principles of the Helsinki Declaration. No additional informed consent was needed as patients from the Alpe-DPD study which was approved by the institutional review board of The Netherlands Cancer Institute (Amsterdam, Netherlands) had already consented to use of their data for future studies, and data from the patients from routine clinical care were anonymized.

Consent for publication

Not applicable.

Competing interests

The authors declare no conflict of interest in relationship with this work. Carin Lunenburg is currently employed by Bristol-Myers Squibb, not related to the contents of this manuscript. She was supported by an unrestricted grant from Roche Pharmaceuticals. Jan Schellens is an (part time) employee, stock- and patent holder of Modra Pharmaceuticals, a spin out company developing oral taxane formulations; Jan Schellens is also a part time employee of Byondis bv and received consultancy fees from Debiopharm, all not related to the contents of the manuscript. Steven Offer has received consultancy fees from Processa Pharmaceuticals, Inc. for activities not related to the contents of this manuscript. The remaining authors declare that they have no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

13073_2024_1354_MOESM1_ESM.docx

Additional file 1. Supplementary methods. Supplementary information regarding the inclusion and exclusion criteria of the Alpe-DPD study and the quality control of the GWAS. Table S1: In vitro assay - The primer sequences used to perform site-directed mutagenesis on the expression plasmids. Table S2. The power analysis of the GWAS.

13073_2024_1354_MOESM2_ESM.docx

Additional file 2. Supplementary results. Table S3. The characteristics of patients included in the matched-pair analysis. Table S4. The severe toxicity between the predicted deleterious variants carriers and non-carriers compared by the logistic regression. Fig S1: Manhattan plot for association with severe fluoropyrimidine-induced toxicity. Fig S2: QQ-plot of p-values. Fig S3: Manhattan plot for association with severe fluoropyrimidine-induced toxicity (Sensitivity analysis). Table S5: Thirty genetic variants with the lowest p values (Sensitivity analysis). Table S6. Severe toxicities in the first two cycles and the entire treatment duration. Table S7. The association between novel DPYD deleterious variants and ≥ grade 3 toxicity in the early two cycles of treatment. Fig S4: Manhattan plot for association with severe fluoropyrimidine-induced toxicity (Sensitivity analysis). Table S8: Table of the top 30 associated variants related to severe toxicity in the early two cycles.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Knikman, J.E., Zhai, Q., Lunenburg, C.A.T.C. et al. Discovering novel germline genetic variants linked to severe fluoropyrimidine-related toxicity in- and outside DPYD. Genome Med 16, 101 (2024). https://doi.org/10.1186/s13073-024-01354-z

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s13073-024-01354-z

Keywords

Genome Medicine

ISSN: 1756-994X

Contact us