Health Technology Review

Dihydropyrimidine Dehydrogenase Deficiency Testing for Patients Treated With 5-Fluorouracil and Capecitabine

Correction notice: This report was initially published on July 28, 2025, and subsequently revised on March 2, 2026. We had incorrectly stated that phenotype testing is conducted in New Brunswick. Phenotype testing is not conducted in New Brunswick. The text and figures have been corrected accordingly.

What Is the Issue?

• Fluoropyrimidines, such as 5-fluorouracil and capecitabine, are drugs used for the treatment of solid tumour cancers. Deficiency in the enzyme dihydropyrimidine dehydrogenase (DPD), which breaks down these drugs, can significantly increase the risk of severe toxicity and death.

• Pretreatment DPD deficiency testing, via genotyping, or phenotyping, followed by dose adjustments is recommended in several countries to reduce adverse drug reactions. However, most evidence is based on genetic variants identified in individuals from European countries such as the Czech Republic, Denmark, France, Ireland, Italy, the Netherlands, Spain, and the UK. This raises uncertainties about the transferability of the safety and effectiveness of these approaches to patients from diverse ethnic origins.

• In Canada, access to DPD deficiency testing is inconsistent and varies widely across provinces and territories.

What Did We Do?

• We conducted a national survey on the current state of DPD deficiency testing and a rapid review to identify and summarize evidence comparing the clinical and cost-effectiveness of DPD deficiency testing and test-guided dose adjustments versus usual care.

• We searched key resources, including journal citation databases, and conducted a focused internet search for relevant evidence published since 2015. One reviewer screened articles for inclusion based on predefined criteria, critically appraised the included studies, and narratively summarized the findings.

What Did We Find?

• The survey results suggest that DPYD genotyping is conducted in 5 Canadian provinces, and 2 more are set to start testing later this year. No jurisdictions reported conducting phenotype testing. The cost of testing ranged from CA$50 to CA$500 and was dependent on the testing platform and required turnaround time.

• The evidence suggests that DPYD variant carriers are at a higher risk of severe toxicities, hospitalization, and death compared to patients with the wild-type gene and that genotype-guided dose adjustments may improve these clinical outcomes in variant carriers.

• The data for the clinical utility of genotype-guided dosing is based largely on study populations from European countries, decreasing the utility of genotype testing in Canada, where the target population includes numerous ethnic origins. Further research and guideline development is required to support the validity and utility of variants more common in these groups.

• Phenotype testing provides an appealing complementary or alternative test that is independent of ethnic origin; however, evidence supporting its clinical validity and utility is minimal.

• Evidence suggests that DPYD testing with subsequent genotype-guide dose adjustments is cost-effective compared to usual care. No evidence was found on the cost-effectiveness of an extended DPYD genetic panel or of DPD phenotyping.

What Does This Mean?

• Based on the evidence identified in this report, DPYD genotyping may be clinically valid and cost-effective to improve the safety of fluoropyrimidine use in Canada for patients of European descent.

• Clinicians and decision-makers can use the evidence summarized in this review to inform decisions regarding the implementation of DPD deficiency testing.

Key Terminology

Allele: One of 2 or more alternative forms of a gene that arise by mutation and are found at the same place on a chromosome. Every individual has 2 alleles for each gene.

Clinical utility: In this Rapid Review report, this term refers to the probability that pretreatment dihydropyrimidine dehydrogenase testing followed by test-guided dose adjustments reduces fluoropyrimidine-related toxicity or mortality compared to patients who received standards doses.

Clinical validity: In this Rapid Review report, this term refers to the accuracy with which low dihydropyrimidine dehydrogenase activity (identified via either presence of a DPYD genetic variant or phenotyping tests for dihydropyrimidine dehydrogenase activity) predicts specific clinical outcomes (e.g., severe toxicity, mortality) following usual care treatment with fluorouracil or capecitabine. This is separate and distinct from analytical validity, which refers to the accuracy and reliability with which a specific test can measure or detect a specific genetic variant. Analytical validity is not considered in this report.

DPYD genotyping: A type of testing to determine whether a genetic variant (genotype) is present in an individual’s DNA, specifically their DPYD gene. Testing methods include polymerase chain reaction, Sanger sequencing, and next-generation sequencing.

DPD phenotyping: A type of testing to determine the activity level of the dihydropyrimidine dehydrogenase enzyme in an individual’s body. Testing methods include measurement of plasma uracil concentrations, dihydrouracil to uracil ratios, and activity in peripheral blood mononuclear cells.

Ethnicity: “A socially defined category or membership of people who may share a nationality, heritage, language, culture, and/or religion.”¹

Ethnic origin: “The ethnic or cultural origins of the person’s ancestors. Ancestors may have Indigenous origins, or origins that refer to different countries, or other origins that may not refer to different countries.”²

Gender: “Gender can refer to the individual and/or social experience of being a man, a woman, or neither. Social norms, expectations and roles related to gender vary across time, space, culture, and individuals.”³

Health equity: “Equity is the absence of unfair, avoidable or remediable differences among groups of people, whether those groups are defined socially, economically, demographically, or geographically or by other dimensions of inequality (e.g. sex, gender, ethnicity, disability, or sexual orientation). Health is a fundamental human right. Health equity is achieved when everyone can attain their full potential for health and well-being.”⁴

PROGRESS-Plus: An acronym used to identify characteristics that stratify health opportunities and outcomes. PROGRESS refers to place of residence, race, ethnicity, culture, language, occupation, gender, sex, religion, education, socioeconomic status, social capital. Plus refers to personal characteristics associated with discrimination (e.g., age, disability), features of relationships (e.g., smoking parents, excluded from school), and time-dependent relationships (e.g., leaving the hospital, respite care, other instances where a person may be temporarily at a disadvantage)⁵

Sex: “The classification of people as male, female, or intersex. Sex is typically assigned at birth and is based on an assessment of one’s reproductive systems, hormones, chromosomes, and other physical characteristics.”^1,3

Context

Fluoropyrimidines and Fluoropyrimidine-Associated Toxicity

Fluoropyrimidines (FPs), such as 5-fluorouracil (5-FU) and capecitabine, are widely used chemotherapy agents for treating solid tumours, including colorectal, gastric, breast, and head and neck cancers.^6,7 5-FU is typically administered through IV; whereas, capecitabine is an orally administered prodrug that is converted into 5-FU after absorption. Once inside the cell, these drugs are metabolized into active compounds that disrupt RNA and DNA synthesis, ultimately leading to cell death.⁸ While FPs have been shown to be effective in improving overall survival, approximately 30% of patients⁹ experience severe (grade ≥ 3)¹⁰ drug-related toxicities. Of note, tegafur is another FP used in the treatment of cancer; however, it will not be discussed in this report as it has not been approved for use in Canada.¹¹

5-FU is catabolized in the body by an enzyme called dihydropyrimidine dehydrogenase (DPD). As shown in Figure 1, only 1% to 5% of the original dose of 5-FU mediates the cytotoxic effects on tumour cells, while DPD catabolizes more than 80% of the drug.¹² The remaining 10% is excreted in urine. A deficiency in DPD, caused by mutations in the DPYD gene, is found in approximately 2% to 8% of people of European descent.^13,14 This deficiency markedly increases 5-FU levels in the blood, thereby increasing the risk of severe toxicity from FP treatment.

Severe toxicity is defined as adverse events graded as greater than 3 using the Common Terminology Criteria for Adverse Events (CTCAE).¹⁰ FP-related adverse events include hematological (e.g., neutropenia), gastrointestinal (e.g., nausea, diarrhea), cardiovascular (e.g., angina, myocardial infarction), and neurologic toxicities (e.g., hand-foot syndrome). Other outcomes include overall survival, FP-related hospitalization, and FP-related mortality.

DPD Deficiency Testing: DPYD Genotyping Versus DPD Phenotyping

There are 2 types of testing for DPD deficiency: DPD phenotyping and DPYD genotyping. DPYD genotyping, or genetic testing, focuses on identifying genetic variations within the DPYD gene, which encodes DPD; whereas, DPD phenotyping refers to the measurement of DPD activity.¹¹ In this report, we will use the term DPD deficiency testing when referring to both types of tests.

DPYD Genotyping

DPYD genotyping can be performed using a variety of methods, including real-time polymerase chain reaction (PCR), multiplex PCR, Sanger sequencing, and next-generation sequencing (NGS).¹⁵ PCR is a technique used to amplify specific DNA sequences using synthetic primers to target specific gene segments, followed by repeated cycles of DNA replication to generate millions of copies.¹⁶ PCR alone cannot detect novel mutations — it is limited to confirming known, prespecified variants, and is most effective for identifying common mutations. Sanger sequencing is a method used to determine the nucleotide sequence of a defined region of DNA. While more comprehensive than PCR, it is still limited to targeted regions and known areas of interest.

Figure 1: Fluoropyrimidine Metabolism

5’dFUR = 5'-deoxy-5-fluorouridine; 5’dFCR = 5’-deoxy-5-fluorocytidine; 5-FU = fluorouracil; CES = carboxylesterase; CDA = cytidine deaminase; DHFU = 5,6-dihydrofluorouracil; DPD = dihydropyrimidine dehydrogenase; dTMP = deoxythymidine monophosphate; dUMP = deoxyuridine monophosphate; FdUDP = fluorodeoxyuridine diphosphate; FdUTP = fluorouridine triphosphate; FUDP = fluorouridine diphosphate; FUTP = fluorouridine triphosphate; FUPA = fluoro-beta-ureidopropionate; TP = triphosphate; TS =thymidylate synthase.

Source: This figure was adapted from Lunenberg et al.¹² This work is licensed under the Creative Commons Attribution 4.0 International License (https://creativecommons.org/licenses/by/4.0/).

NGS is a method of genotyping capable of sequencing entire genes, exomes, or genomes. Similar to PCR or Sanger sequencing, NGS can target specific variants; however, because the targeted region is typically larger, it can detect both known and unknown variants, increasing its breadth, and efficiency. This broader scope makes NGS particularly valuable from an equity perspective, as it allows for the inclusion of diverse genetic backgrounds and rare variants that may be underrepresented in standard testing panels.

To date, more than 50 variants have been identified in the DPYD gene with varying levels of evidence regarding their impact on DPD function. The Clinical Pharmacogenetics Implementation Consortium (CPIC) primarily emphasize 4 variants that have been well-characterized and are commonly included in testing panels: DPYD*2A, DPYD*13, c.2846A>T, and c.1236G>A;¹⁷ however, a systematic review (SR) published in 2024¹⁸ identified 53 rare or novel variants associated with severe FP-related toxicity. Variants are typically categorized based on their impact on DPD enzyme activity: normal function, decreased function, or no function, and the strength of evidence supporting this claim.

DPD Phenotyping

DPD phenotyping measures the actual enzyme activity, or the metabolic products influenced by DPD function, providing a functional assessment of an individual’s ability to metabolize FPs. Three primary methods are commonly used: plasma uracil concentration, the dihydrouracil to uracil (UH₂/U) ratio, and direct measurement of DPD enzyme activity in peripheral blood mononuclear cells (PBMCs).¹⁹ Plasma uracil levels are measured using techniques like high-performance liquid chromatography or liquid chromatography-tandem mass spectrometry. Elevated uracil concentrations suggest reduced DPD activity. The UH₂/U ratio reflects the balance between uracil and its primary DPD-mediated metabolite; lower ratios indicate impaired metabolism. Measuring DPD activity in PBMCs is considered the gold standard for DPD phenotyping, though this method is technically complex and less commonly available.^17,20

Each phenotyping approach presents its own challenges, particularly related to preanalytical variables. Uracil and UH₂ levels are highly sensitive to sample handling (e.g., time to processing or improper freezing), patient’s compliance to pretest fasting, concurrent medications, and the time of day of the blood draw.²¹ PBMC-based testing, while precise, is influenced by circadian rhythms and requires rigorous laboratory infrastructure.²² Despite these limitations, phenotyping presents a valuable alternative for identifying patients at risk of FP toxicity, particularly in multiethnic populations that may not benefit from the common genetic testing panels.²³ Standardization of testing protocols and interpretation thresholds remains essential for clinical utility and therefore broader clinical implementation. A combined genotype-phenotype approach to DPD deficiency testing is also an option.^23,24

The clinical utility of DPD deficiency testing lies in the subsequent dose or treatment adjustments of FP drugs. Based on the results of the test, the initial FP dosage can be reduced, or in the case of suspected no function DPD, an alternative treatment can be chosen.^11,17,25

International and Canadian Context

In April 2020, the European Medicines Agency (EMA) recommended that patients undergo pretreatment testing for DPD deficiency — either through DPYD genotyping or phenotyping — to mitigate the risk of severe and potentially fatal toxicities.²⁶ This was followed by similar recommendations in other European countries including the UK,²⁷ Germany, Switzerland, and Belgium.²⁸ Since 2019, France is the only country where DPD testing is mandatory, primarily through phenotyping methods such as plasma uracil measurement.^29,30

FPs are widely used and effective chemotherapy drugs in Canada. Despite this, no national policies or guidelines on DPD testing exist, and current testing practices vary widely across jurisdictions. This review is important to better understand the current landscape, facilitators, and barriers to DPD testing in Canada to guide the development and implementation of standardized guidelines.

Purpose and Objectives

This report was undertaken to support decision-makers across jurisdictions in Canada by providing a timely update on the status of DPD testing and funding throughout the country, and to summarize relevant evidence on its clinical validity, clinical utility, and cost-effectiveness.

This report also aligns with the CDA-AMC broader efforts to develop a standardized framework to inform decisions about the adoption and implementation of molecular (genetic and genomic biomarker testing) across jurisdictions in Canada.

This report has 3 primary objectives:

• to provide a summary on the status and availability of DPD testing in jurisdictions across Canada

• to identify and summarize available evidence-based guidelines for DPD testing

• to identify, summarize, and critically appraise evidence on the clinical validity, clinical utility, and cost-effectiveness of DPD testing and testing-guided dose adjustments for patients being treated with 5-FU or capecitabine.

Research Questions

• What is the status and availability of genotype and phenotype testing for dihydropyrimidine dehydrogenase deficiency in patients being treated with 5-FU or capecitabine across Canada?

• What are the available evidence-based guidelines on genotype and phenotype testing for dihydropyrimidine dehydrogenase deficiency in patients being treated with 5-FU or capecitabine?

• What is the clinical validity of genotype and phenotype testing for dihydropyrimidine dehydrogenase deficiency in patients being treated with 5-FU or capecitabine?

• What is the clinical utility of genotype and phenotype testing for dihydropyrimidine dehydrogenase deficiency in patients being treated with 5-FU or capecitabine?

• What are the cost implications of genotype and phenotype testing for dihydropyrimidine dehydrogenase deficiency in patients being treated with 5-FU or capecitabine?

Approach

We conducted a survey to inform the status and availability of DPD testing in jurisdictions across Canada, and a Rapid Review of the literature to identify:

• evidence-based guidelines for genotype or phenotype testing and associated dose adjustments

• SRs and primary studies to evaluate the clinical validity of DPD testing, including both DPYD genotyping and DPD phenotyping

• SRs and primary studies to evaluate the clinical utility of DPD testing-guided dose adjustments for patients treated with 5-FU and capecitabine

• economic and cost considerations associated with DPD testing for patients treated with 5-FU and capecitabine.

We surveyed representatives from each province and territory responsible for assessment, implementation, or funding decisions regarding DPD testing in March 2025 and April 2025. The survey was hosted on SurveyMonkey³¹ and was sent via email, with a reminder email 1 week following the original email. Questions focused on the availability and accessibility of testing, testing methods, and funding details.

An information specialist conducted a customized literature search, balancing comprehensiveness with relevancy, of multiple sources and grey literature on March 4, 2025. One reviewer screened citations and selected studies based on the inclusion criteria presented in and critically appraised included publications using established critical appraisal tools. Appendix 1 presents a detailed description of methods and selection of included studies.

Summary of Survey Findings

Status and Availability of DPD Testing

We received 14 responses regarding the availability of DPD testing from 13 of 13 Canadian jurisdictions — Newfoundland and Labrador, Prince Edward Island, Nova Scotia, New Brunswick, Quebec, Ontario, Manitoba, Saskatchewan, Alberta, British Columbia, Yukon, Northwest Territories, and Nunavut. Respondents provided replies directly via email (for the 3 territories) or via the online survey (13 from Canadian provinces). Details about the availability of DPD testing in Ontario were obtained from public resources. Respondents reported holding positions in medical oncology, genetics, pathology, and pharmacy. A narrative overview of the findings related to the status and availability of DPD testing in Canada is presented in the following section. A list of participating organizations and detailed findings by jurisdiction are available in Appendix 2.

Jurisdictions with DPD Testing

Availability

Five jurisdictions – New Brunswick, Quebec, Ontario, Saskatchewan, and British Columbia - reported the availability of publicly funded DPD testing. One jurisdiction (Alberta) reports that at the time of survey completion, DPD testing is recommended and available with a private payer option only; however, an application for public funding is under way with an estimated availability date of July 2025. Another jurisdiction (Manitoba) reported that while testing was not available at the time of survey completion implementation is under way, with testing set to begin in June 2025. Respondents indicated that testing practices are guided by international and jurisdictional guidelines, safety announcements from regulatory agencies, and established treatment algorithms.

All 5 provinces report conducting genotype testing for the 4 most common tier 1³² genetic variants:

• DPYD*2A

• DPYD*13

• DPYD c.2846A>T

• DPYD c.1236G>A.

DPYD c.1236G>A is listed in ON as an optional variant to test. Three provinces reported testing additional variants: HapB3 (c.1129 to 5923C>G and c.1236G>A) in Ontario and Saskatchewan, and c557A>G and c2279C>T in British Columbia. The latter 2 are considered tier 2 variants, which are emerging variants with less supporting evidence.³² One province (New Brunswick) noted plans to expand testing to all recognized tier 1 and tier 2³² variants.

All provinces reported that patient demographic characteristics did not influence the genetic variant tested. No jurisdictions reported conducting phenotype testing. Figure 2 presents the availability of DPD testing across Canada, based on findings from the survey responses.

Figure 2: Availability of DPD Testing Across Canada

NL = Newfoundland and Labrador; PEI = Prince Edward Island; NS = Nova Scotia; NB = New Brunswick; QC = Quebec; ON = Ontario MB = Manitoba; SK = Saskatchewan; AB = Alberta; BC = British Columbia; YT = Yukon; NT = Northwest Territories; NU = Nunavut.

Notes: For Manitoba, the implementation of DPD testing is under way and set to start as soon as June 2025. For Alberta, DPD testing is recommended with a private payer option currently available — publicly funded testing set to start as soon as July 2025.

Survey responses for DPD testing availability were available for 10 provinces and the 3 territories.

Data were derived from the following survey questions: “Do you currently test for dihydropyrimidine dehydrogenase (DPD) deficiency (either via genotype or phenotype testing) in people being treated with fluorouracil and capecitabine in your province or territory?” and “What type of testing do you use?”

Accessibility

In most jurisdictions the same genotype test is used for all patients. While this suggests equality in access to the test, it also suggests minimal consideration for ethnic origin, that might necessitate testing for different variants. A response from Saskatchewan specified test type may vary based on availability, while a response from New Brunswick noted that patient knowledge or preference can influence testing decisions.

All the provinces reported that DPD testing is generally accessible across the province, although barriers to widespread accessibility were identified as:

• lack of clinician and/or patient awareness

• location of testing centres

• stigma or prejudice based on ethnicity, sex, and/or gender

• turnaround time delays.

Testing Practices

All provinces except 1 reported conducting DPD testing before treatment initiation. A standard order form was used to initiate DPD testing in all but 1 province (Ontario). A response from Saskatchewan indicated that while testing is usually conducted before treatment, it is occasionally conducted after some patients have experienced severe toxicity to FP therapy. Factors that influence when patients are tested were identified as institutional or jurisdictional guidelines, family history of DPD deficiency, and clinician knowledge.

The turnaround times for tests ranged from 5 to 10 days, with most provinces reporting that results were ready a week after testing. For respondents that provided turnaround times for emergency conditions, results were made available within 24 hours to 48 hours.

Resource Implications

DPD testing is publicly reimbursed in 5 provinces. A respondent from British Columbia indicated that all components of the test are publicly funded. Cost estimates per test provided by 5 respondents ranged from CA$50 to CA$500.

Implementation Barriers and Facilitators

Respondents identified several facilitators to the implementation of DPD testing, including:

• number of testing centres (2)

• location of testing centres (2)

• clinician and patient awareness (3)

• clinician and patient educational resources (1)

• reimbursement of test costs (4)

• well-established guidelines (3)

• established billing infrastructure (1)

• electronic medical record system workflow integration (2).

The most reported facilitators of widespread DPD testing were reimbursement of test costs (4 provinces), the presence of well-established guidelines (3 provinces), as well as clinician and patient awareness (3 provinces). Only 1 province identified established billing infrastructure as a facilitator.

Reported barriers to implementation included:

• lack of clinician/ patient awareness of testing (1)

• time constraints related to test turnaround times and treatment schedules (2)

• location of testing sites (2)

• lack of guidelines (1)

• limited lab resources, including lack of technologists and validation plan or platform (1).

The most commonly reported barriers were time constraints and location of testing sites.

Jurisdictions Without DPD Testing

Publicly funded DPD testing was reported unavailable in 5 provinces (Newfoundland and Labrador, Prince Edward Island, Nova Scotia, Manitoba, and Alberta) and the 3 territories. A respondent from Nunavut reported that all oncology patients in the territories are referred to neighbouring provinces where patient assessments (including DPYD testing) and treatments are conducted.

A response from Manitoba reported that while testing was not available at the time of survey completion implementation is under way, with testing set to begin in June 2025. A response from Alberta noted that DPD testing is under consideration for 6 of the most common genetic variants (DPYD*2A, DPYD*13, c.2846A>T, c.1129 to 5923C>G, c557A > G, c2279C > T), with an application submitted to support public funding. In previous years, DPD testing was not considered as there was no demand from oncologists. The response from Newfoundland and Labrador reported that testing is not available in the province, but if it is requested by a clinician (rarely, and typically in response to greater than anticipated toxicities to FP treatment) tests are sent to the Mayo Clinic. These tests are not funded. Similarly, Prince Edward Island indicated that testing is not available but can be accessed in a neighbouring province.

The barriers to implementation of DPD testing were identified as:

• cost of testing (3)

• lack of guidelines (3)

• limited ability to interpret test results (2)

• time constraints (i.e., test turnaround times and treatment start date) (2)

• electronic medical record workflow and integration (e.g., automatic alerts) (3)

• lack of guidance on test selection (i.e., limited targeted panel of common variants in a population of white people, versus broader panel, versus whole gene sequencing) (1)

• lack of billing infrastructure (2)

• lack of infrastructure and human resources to implement testing (2)

• limited testing availability and capacity volumes (1)

• lack of availability of best test (1)

• perceived undertreating or lack of clinical utility (2)

• lack of clinician and/or patient awareness of testing (1)

• location of testing centres (1).

The most frequently reported barrier was the lack of guidelines (3 responses). The response from Prince Edward Island emphasized a lack of clarity around the clinical utility of testing and the need for national standardization of DPD testing practices. The response from Nova Scotia specified that the potential for undertreating patients was a barrier to implementation.

Summary of Literature Review Evidence

We included 16 publications that met the inclusion criteria of this report. These comprised 1 health technology assessment (HTA), 5 SRs, 5 nonrandomized studies, 1 economic evaluation, and 4 evidence-based guidelines. Appendix 1 presents the PRISMA³³ flow chart of the study selection. A summary of the study characteristics can be found in Appendix 3. Additional references of potential interest that did not meet the inclusion criteria of this review are provided in Appendix 7.

Quality of Evidence

The included publications were critically appraised using established tools. The overall quality of guidelines was high, with 2 guidelines providing a clear link between evidence and recommendations, while 1 provided recommendations based on expert consensus. The quality of evidence assessing clinical validity and clinical utility was variable. While several studies demonstrated methodological strengths, such as a standardized means for measuring outcomes and consistent reporting, most were limited by small sample sizes, historical control groups, and numerous confounding variables. Heterogeneity in study populations and testing methods further limited comparability across studies, so conclusions should be interpreted with caution. Further large observational comparative studies should be conducted to support the findings. The cost studies were appraised based on their alignment with the decision problem, relevance to the setting, and fit for purpose of its main input parameters. A summary of the critical appraisal of the included publications, along with details regarding their strengths and limitations can be found in Appendix 4.

Summary of Findings

Appendix 5 presents additional details regarding the main study findings.

Guidelines Regarding Genotype or Phenotype Testing for DPD Deficiency

Three evidence-based guidelines³² and 1 consensus-based guideline³² were included in this review. The consensus document³² was published in 2024 by the Association for Molecular Pathology (AMP) which is the Pharmacogenomics Working Group of the Clinical Practice Committee. The 3 evidence-based guidelines^11,12,17 were produced by Ontario Health-Cancer Care Ontario (OH-CCO) published in 2023, the Dutch Pharmacogenetics Working Group (DPWG) of The Royal Dutch Pharmacists Association published in 2019, and the CPIC published in 2017, with an update posted on the webpage in 2024. One guideline was developed in Canada,¹¹ 1 in the Netherlands,¹² and 2 by groups based in the US.^17,32

AMP³² provided a minimum set of variants that should be included in clinical genotyping assays and classified them into 2 groups (tier I and tier II). Tier I recommended variants are those that meet the following requirements:

• have an effect on the function of the DPD protein

• are represented commonly in at least 1 population or ancestral group

• have publicly available reference materials

• clinical laboratories can feasibly analyze with standard molecular testing methods.

Tier II variants meet at least 1 of the tier I criteria (but not all) and can be reclassified as tier I if more information becomes available. Although AMP identified variants for each tier, they urge laboratories to consider genetic variations represented in their population to correctly identify patients who may be at risk of developing severe FP toxicity. Specific gene variants identified by AMP for the 2 tiers are presented in Table 1.

Table 1: Tier Classifications of Gene Variants to Test

Three guidelines^11,12,17 provided genotype-based dosing recommendations, classifying patients by their Gene Activity Scores (GAS). This is a standardized system used to quantify DPD phenotype (DPD activity) based on DPYD genotype.³⁴ Because each individual carries 2 copies of a gene (i.e., allele), individuals are assigned a score ranging from 0 to 2, based on the combination of variants that they carry. Carriers of 2 no function or 1 no function and 1 decreased function variant are considered poor metabolizers and given a score of 0.0 or 0.5; carriers of 1 no function or decreased function variant are intermediate metabolizers and are given a score of 1.0 or 1.5; and carriers with 2 fully functional variants are considered normal metabolizers, with a score of 2.

A summary of the recommendations provided by the 3 DPYD guideline documents^11,12,17 can be found in Table 2.

Patients With Complete DPD Deficiency (GAS 0 or 0.5)

• DPWG and CPIC recommend avoiding 5-FU or capecitabine.

• In cases where avoiding FP therapy is not possible, the groups recommend starting therapy at a reduced dose and monitoring effects at earliest time points to minimize toxicity.

Patients With Decreased DPD Activity (GAS 1 or 1.5)

• DPWG recommends starting therapy at 50% of the standard dose or to avoid 5-FU and capecitabine.

• CPIC recommends:

  • reducing the starting dose by 50% for patients with GAS of 1

  • reducing the starting dose by 25% to 50% for patients with a GAS of 1.5.

Patients With Fully Functional DPD Activity (GAS 2)

• DPWG and CPIC recommend using the standard dose recommended on the label for 5-FU and capecitabine.

Patients Who Have an Unknown Genotype

• DPWG and CPIC recommend a phenotyping test to determine DPD activity.

• CPIC states that if a phenotyping test is unavailable, a dose reduction of 75% is recommended.

The OH-CCO guideline adapted its recommendations from the CPIC guideline and supplementary materials. However, the group provided recommendations for DPYD testing:

• Patients with planned FP-based therapies should be informed about DPD deficiency, risks associated with reduced activity, and available tests to determine functionality.

• They recommend that DPYD genotype tests should be involved in the planning of FP-based therapies and screening for clinically relevant DPYD should happen before the start of treatment.

In addition, the OH-CCO guideline¹¹ provided recommendations for implementing DPYD testing. These include identifying patients who are candidates for FP-based therapy early so testing can be conducted at the earliest convenience and recommending that DPYD testing be a standard part of the prechemotherapy check process with results of the test informing an initial treatment plan.

Clinical Validity of Genotype or Phenotype Testing for DPD Deficiency

Evidence regarding the clinical validity of genotype and phenotype testing for DPD deficiency was available from 1 HTA,¹¹ 4 SRs,^35-38 and 1 primary study.⁹ Two of these SRs^35,36 included a meta-analysis of results. The HTA, published in 2021, was conducted by Ontario Health to assess the validity, utility, and cost-effectiveness of DPYD genotyping. The SRs were all published between 2022 and 2024 and aimed to assess the risk of severe toxicity, hospitalization, reduced survival, and death in adult cancer patients (normal versus reduced DPD activity) administered a standard dose of FPs. The primary study, published in 2023, aimed to establish an association between reduced DPD activity determined via PBMCs and adverse events of FP therapy. The 1 HTA and 4 SRs^35-38 included data from a total of 90 unique primary clinical studies; however, there was considerable overlap among the included primary studies. As a result, the pooled effect estimates and narrative summaries from separate reviews are based on some of the same data (refer to Appendix 6 regarding overlap).

Table 2: Comparison of Genotype-Guided Dosing Recommendations

CPIC = Clinical Pharmacogenetics Implementation Consortium; DPD = dihydropyrimidine dehydrogenase; DPWG = Dutch Pharmacogenetics Working Group; FU = fluorouracil; GAS = gene activity score.

Severe (Grade ≥ 3) FP-Related Toxicity

In the SR conducted as part of the HTA,¹¹ pooled results of 7 studies, that included patients with colorectal, breast, gastrointestinal, esophageal, or head and neck cancer, indicated a higher risk of overall toxicity in DPYD variant carriers (any of the 4 main variants) compared to patients with the wild-type gene (risk ratio [RR] = 2.63; 95% confidence interval [CI], 2.15 to 3.96). Similarly, the pooled risk ratio of neutropenia, from 4 included studies was 4.42 (95% CI, 1.59 to 9.18). In 1 included study, 0 of 34 variant carriers experienced hand-foot syndrome compared to 24 of 771 patients with the wild-type gene.

One meta-analysis³⁶ found an increase in overall toxicity, hematological toxicity, neutropenia, and diarrhea in DPYD*2A variant carriers compared to patients with wild-type DPYD with odds ratios [ORs] of 1.73, 2.37, 1.87, and 1.43, respectively. One meta-analysis³⁶ found an increase in overall toxicity, hematological toxicity, neutropenia, and diarrhea in DPYD*2A variant carriers compared to patients with wild-type DPYD, with ORs of 1.73, 2.37, 1.87, and 1.43, respectively. No significant difference was found between DPYD*2A variant carriers and patients with DPYD for gastrointestinal toxicity (OR = 1.22; 95% CI, 0.93 to 1.61).

One SR³⁷ evaluated toxicity in patients with colorectal cancer who were treated with capecitabine specifically and found 3 studies that reported an increased risk of toxicity in patients who carried the c.1601G>A variant. Two of these studies also found a significant association between the presence of the c.85T>C variant and severe adverse events. Findings on the DPYD variant c.496A > G were inconsistent: 1 study reported a significant association with capecitabine toxicity, while 2 others did not.

One SR³⁸ and 1 nonrandomized study (NRS)⁹ reported on severe toxicities and DPD deficiency in patients with various types of cancer, as measured via phenotype methods. Doornhof et al.⁹ measured PBMCs, while Paulsen et al³⁸ included studies that measured plasma uracil concentrations and U/UH₂ or UH₂/U ratios. Paulsen et al³⁸ narratively summarized 7 observational studies and found that the data regarding the correlation between uracil concentration or the UH₂/U ratio and severe FP-related toxicity is insufficient to draw any reliable conclusions. They suggest the need for adequately powered prospective clinical trials to properly validate the current uracil concentration threshold value proposed by the EMA.²⁶

The 1 included NRS⁹ found statistically significant associations between DPD deficiency, as measured in PBMCs, and overall, hematological, and gastrointestinal toxicities. A multivariable logistic regression adjusting for age, sex, FP dosage, chemotherapy regimen, kidney and liver function, found that gastrointestinal, cardiovascular and neurologic adverse events were significantly higher in patients with DPD deficiency.

FP-Related Mortality

One meta-analysis³⁵ found that the presence of DPYD variants (DPYD*2A, DPYD*13, c.2846A>T, or c.1236G>A) was significantly associated with treatment-related mortality, compared to DPYD patients with wild-type, with an OR of 34.86 (95% CI, 13.96 to 87.05). In pooled results across 13 studies, 13 out of 322 variant carriers died from FP-related toxicity, compared to 14 out of 6,952 patients with the wild-type gene. The authors found that the DPYD*2A variant was the most prevalent among fatalities, followed by DPYD*13 and c.1129 to 5923C>G and c.1236G>A (HapB3).

One SR¹¹ narratively summarized 9 observational studies and found that in heterozygous DPYD carriers mortality ranged from 0.0% to 100%, and 0.0% to 2.0% in patients with the wild-type gene. Two studies that included only DPYD*2A carriers found a higher mortality risk in carriers compared to the wild-type gene (RR = 50.00 and 95% CI, 6.21 to 74.53; RR = 52.63 and 95% CI, 10.40 to 120.90).

FP-Related Hospitalization

One SR¹¹ narratively summarized 5 observational studies reporting rates of hospitalization in DPYD variant carriers versus patients with the wild-type gene. Three studies found a higher risk of hospitalization in DPYD variant carriers compared to patients with the wild-type gene (RR = 2.26 and 95% CI, 0.69 to 5.14; RR = 4.46 and 95% CI, 3.26 to 5.29; RR = 58.82 and 95% CI, 15.19 to 168.60). The risk ratios of the other 2 studies could not be calculated because they reported frequencies of hospitalizations in variants carriers, but not patients with the wild-type gene.

Sensitivity, Specificity, Positive Predictive Value, Negative Predictive Value

The authors of the included HTA¹¹ calculated the sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of DPYD genotyping (3 to 4 variants) to detect severe toxicity for 9 observational studies based on data reported within the studies. They found that DPYD genotyping had a high specificity for severe toxicity (median of 98.6%), but low sensitivity (median of 8.1%), as many patients with the wild-type gene also experienced severe toxicity. In other words, the test is good at identifying individuals who do not experience severe toxicity (low false positives) but is not as good at identifying those who do experience severe toxicity (high false negatives). The authors noted that this may be due to the fact that other factors can contribute to severe toxicity including other unmeasured DPYD variants and baseline or treatment characteristics (e.g., age, sex, kidney function, cancer type, FP dosage). Previous studies not included in this report, have reported that approximately 30% to 80% of toxicities could be attributable to DPD deficiency.³⁹ The median PPV and NPV were 61.1% and 84.5%, respectively.

Clinical Utility of Genotype or Phenotype Testing for DPD Deficiency

Evidence regarding the clinical utility of genotype and phenotype testing for DPD deficiency was available from 1 HTA,¹¹ 2 SRs,^20,38 and 4 NRSs.^40-43 One of the SRs²⁰ included a meta-analysis of results. The HTA¹¹ and 2 SRs,^20,38 included data from a total of 67 primary clinical studies, 27 of which were relevant to this research question. There was considerable overlap among the included primary studies of the included HTA and SRs and as a result, the pooled effect estimates and narrative summaries from separate reviews are based on some of the same data (refer to Appendix 6 regarding overlap).

Severe (Grade ≥ 3) FP-Related Toxicity

DPYD-Guided Dose in Variant Carriers Versus Usual Care in Variant Carriers

One study summarized in 2 SRs^38,44 directly compared DPYD-guided dose adjustments in variant carriers to usual care in variant carriers. The severe toxicity rates (grade ≥ 3 assessed by the CTCAE¹⁰) were comparable between the 2 groups (5 out of 22 [22%] and 8 out of 34 [21%], respectively); however, the authors of 1 SR noted that imbalances in the distribution of DPYD variants between groups as well as imprecision in study results reduced the strength of these findings. The NRS by Paulsen et al.⁴² found that variant carriers who received reduced doses experienced less severe toxicity compared to those with standard doses (23% of patients and 29% of patients, respectively).

DPYD-Guided Dose Versus Usual Care (No Distinction Between Variant Carriers and Wild-Type Carriers)

One SR with meta-analysis found a statistically significant decrease in both overall toxicity and diarrhea in patients who received DPYD-guided dosing versus those who received usual care,²⁰ suggesting that pharmacogenetic-guided dosing was associated with improved patient outcomes. The NRS by Paulsen et al.⁴² found an increased risk of overall grade 3 or higher toxicity in all patients of the DPYD-guided group versus a historical control group treated with standard doses (RR = 1.20; 95% CI, 0.92 to 1.57).

DPYD-Guided Dose in Variant Carriers Versus Usual Care in the Wild-Type Gene

Findings regarding risk of severe toxicity in DPYD-guided dose reductions in variant carriers versus usual care in patients with the wild-type gene were limited and contradictory. The HTA⁴⁴ authors wrote that due to the design of the included studies, they were unable to determine whether reducing the treatment dose results in a risk of severe toxicity that is comparable or lower than that observed in patients with the wild-type gene receiving a standard dose. Paulsen et al³⁸ summarized 3 studies; 2 of which found higher toxicity rates in variant carriers receiving reduced doses, and 1 of which found higher toxicity rates in patients with the wild-type gene receiving standard doses.

Phenotype-Guided Dose in Patients With Reduced DPD Activity Versus Usual Care in Patients With Normal DPD Activity

UH₂/U ratio: One included NRS found that the incidence of severe toxicity was 11% in patients who received a reduced dose based on their UH₂/U ratio and 13% in patients who received a standard dose.³⁸

Plasma Uracil: One study reported similar rates of severe toxicity between patients with reduced DPD activity (U = 16 ng/mL to 150 ng/mL) who received a reduced dose and those with normal DPD activity (U < 16 ng/mL) who received a standard dose (12 out of 27 [44%] versus 43 out of 92 [46%]).⁴⁰ However, 26 (28%) of the patients with normal DPD activity received an initial FP dose reduction based on factors other than DPD activity (i.e., fragile baseline condition).

Combined genotype and phenotype dosing: One study reported that 23% (14 out of 60) patients who received reduced doses, based off of a combination of genotype and PBMC levels, experienced severe grade greater than or equal to 3 toxicity, compared to 30% (50 out of 168) of patients with the wild-type gene, and normal PBMC levels.⁴³

FP-Related Mortality

FP-related mortality was reported by 1 HTA⁴⁴ and 1 NRS.⁴² The authors of the HTA⁴⁴ noted only 1 FP-related death in a DPYD variant carrier (n = 1,103) in the included studies, which occurred “after the patient was wrongly prescribed a standard fluoropyrimidine dose for two cycles” (p. 52).⁴⁴ Mortality rates in patients with the wild-type gene across 3 included studies ranged from 0.1% to 0.7%.

The NRS⁴² found a decreased risk of FP-related death in variant carriers receiving DPYD-guided doses (0 out of 22) versus variant carriers receiving usual care (2 out of 42) (RR = 0.37; 95% CI, 0.02 to 7.46).

FP-Related Hospitalization

Paulsen et al⁴² found a decreased risk of FP-related hospitalization in variant carriers receiving DPYD-guided doses (0/22) versus variant carriers receiving usual care (8 out of 42) (RR = 0.11; 95% CI, 0.01 to 1.82). The authors of the included HTA⁴⁴ note that the point estimates of 4 studies indicated a higher risk of treatment-related hospitalization in variant carriers compared to the wild-type gene, however the confidence intervals also included the possibility of lower risk in DPYD carriers.

In a meta-analysis (of 4 studies) conducted by Glewis et al.,²⁰ it was found that variant carriers had statistically significant more hospitalizations compared to patients with wild-type. However, the authors noted that this could be a result of variation in dose reductions as well as noncompliance to the recommended dose reductions in the included studies. For example, in 1 study, 4 DPYD variant carriers received standard FP doses, leading to fatal toxicity in all.

Disease Response

One SR reported no statistically significant difference between variant carriers who received dose adjustments and patients with the wild-type gene who received usual care in terms of complete and partial disease response, based on a meta-analysis of 3 studies (RR = 1.31; 95% CI, 0.93 to 1.85; P = 0.12; I2 = 0%). Similarly, there was no statistically significant difference in stable disease between the same groups, based on meta-analysis of 2 studies (RR = 1.27; 95% CI, 0.66 to 2.44; P = 0.47; I2 = 0%).

Cost Implications of Genotype or Phenotype Testing for DPD Deficiency

Evidence regarding the cost-effectiveness DPYD genotyping was available from 1 primary economic evaluation performed as part of the included Ontario Health HTA.⁴⁴ The findings from the economic evaluation that was conducted from the perspective of a health care system in Canada suggest that universal pretreatment DPYD genotyping (of the following 4 variants: DPYD*2A, DPYD*13, c.2846A>T, c.1236G>A) for all patients undergoing treatment with FPs, followed by genotype-guided dose adjustments, is likely cost-effective compared to usual care, at the willingness-to-pay thresholds of $50,000 and $100,000 per quality-adjusted life-year (QALY) gained (91% and 96% probability of being cost-effective, respectively). DPYD genotyping remained cost-saving and slightly more effective (resulting in greater QALYs) compared to usual care in all scenarios modelled in sensitivity analyses.

We identified 9 other economic evaluations^45-53 and 1 budget impact analysis, included in the Ontario Health HTA⁴⁴ that were excluded from evaluation in this report as they were conducted outside of Canada and therefore have limited transferability to our setting of interest. However, it is worth noting that DPYD genotyping was found cost-effective compared to usual care across all studies from different settings. A brief summary of the results of these studies can be found in Appendix 5.

Limitations

Risk of Bias of Included Studies

The authors of the SRs included in this report noted some methodological limitations of the included primary studies affecting the validity and reliability of their results. For example, the comparability of cohorts was limited by numerous confounders in this population including age, sex, body surface area, ethnicity, comorbidities, cancer type, treatment regimen, DPYD variant, and kidney and liver function. Because of a lack of clinical equipoise, randomized controlled trials were extremely limited and occurred at the earlier end of the date range. Therefore, the methodological and analytical limitations inherent to both retrospective and prospective cohort studies (e.g., selection bias, missing or incomplete records) were present in the evidence.

Some studies noted discrepancies in DPYD testing assays among laboratories⁴³ and significant variability in uracil measurements due to factors such as participants’ food intake and circadian rhythms, and other preanalytical conditions.³⁸ While in most clinical utility studies, test-guided dose adjustments were made using accepted guidelines,^12,17 exact doses were also dependent on other patient characteristics and clinician preference. Thus, even in the intervention group, doses varied widely.

Evidence Gaps

Evidence regarding the clinical validity and utility of DPYD variants beyond the 4 identified by CPIC,¹⁷ as clinically relevant, was limited to 1 guideline³² and 2 SRs.^36,37 Of note, we did identify 2 large SRs^18,54 in our literature search that assessed evidence on rare or novel DPYD variants; however, they were excluded from the present report as they did not satisfy other search criteria (i.e., they included primarily case reports and case-control studies with limited sample sizes, lacked relevant comparators, and did not consistently report on our outcomes of interest).

Evidence regarding the clinical utility of phenotype-guided dosing was limited to 2 SRs^20,38 and 2 NRS,^40,41 with 2 included testing methods: plasma uracil concentrations and UH₂/U ratio. Of note, our literature search did identify 1 consensus guideline focused on phenotype-guided dose recommendations;²⁹ however, it was excluded from this report as the recommendations were based on expert consensus rather than a formal SR of evidence.

The assessment of the cost-effectiveness of DPYD genotyping and subsequent dose adjustments versus usual care in a public health care payer setting in Canada was limited to 1 economic evaluation.⁴⁴ This study only included genotyping for the 4 primary variants. We also found no evidence assessing the cost-effectiveness of DPD phenotyping for reducing FP-related toxicity or mortality. Consequently, no conclusions can be drawn regarding the cost-effectiveness of either extended DPYD genetic testing (via NGS or a larger panel of targeted variants) or phenotype testing.

Finally, none of the included studies reported quality of life outcomes specific to DPD testing or test-guided dosing, so the impact of these interventions on patient-reported outcomes is unknown.

Generalizability

We used PROGRESS-Plus criteria^55,56 to guide data extraction and to provide insights into whether the clinical studies conducted to date included diverse patient populations who could be representative of those in Canada. However, the literature we reviewed for this report provided limited information on participant characteristics, often only reporting a few factors such as age, sex, or ethnicity. In cases where participant sex or gender were reported, the authors did not provide any information on how they were defined or measured. Similarly, ethnicity was poorly reported across studies, with some authors reporting patient-reported ethnic groups and others simply reporting the country in which the study was completed. Where it was reported, most studies included participants from European countries, such as the Czech Republic, Denmark, France, Ireland, Italy, the Netherlands, Spain, and the UK. This may limit the generalizability of the evidence to settings in Canada, where patients belong to numerous ethnic groups, including those of African, East Asian, Latin American, Middle Eastern, or South Asian descent.

Relatedly, most studies assessed the 4 DPYD variants most prevalent in patients who identified as being of European ancestry, with limited evidence on the validity and utility of other variants. As such, the generalizability of results to diverse population groups is limited.

Some of the phenotype tests included in this report require specific analytical conditions that could be difficult to achieve if sample collection occurs at a location at a great distance from the lab conducting the analysis. Given the geography of Canada and many rural and remote residents, this may not always be achievable.

Conclusions and Implications for Decision- or Policy-Making

This review includes 1 HTA (which included 1 SR and 1 de novo economic evaluation), 4 guideline documents,^11,12,17,32 5 SRs^20,35-38 (2 with meta-analyses^20,36), and 5 NRS^9,40-43 regarding the clinical validity, utility, or cost-effectiveness of DPD deficiency testing and test-guided dose adjustments for detecting and preventing severe toxicities, mortality, and hospitalizations.

Summary of Evidence

The survey conducted as part of this report suggests that DPYD genotype testing is conducted in 5 jurisdictions in Canada. Two additional provinces reported that they are set to begin genotype testing later in 2025. No jurisdictions reported conducting phenotype testing. All jurisdictions publicly reimburse the test costs, which ranged from CA$50 to CA$500, depending on testing platform (PCR versus NGS), with NGS costing more. The most frequently reported facilitators in those P/Ts were reimbursement of test costs, the presence of well-established guidelines, and clinician and patient awareness. The most frequently reported barriers in jurisdictions without DPD testing was the lack of well-established guidelines.

The evidence summarized in this report indicates that individuals carrying a DPYD variant are likely to be at higher risk of severe toxicity, treatment-related hospitalization, and mortality, compared to patients with the wild-type gene. The evidence also suggests that genotype- or phenotype-guided dosing may reduce severe toxicities and mortality in patients being treated with FPs, without reducing treatment effectiveness (disease response); however, evidence on phenotype-guided dosing is extremely limited. The included studies suggest that FP-related hospitalization may be higher in DPYD-guided variant carriers compared to patients with the wild type; however, inconsistency in dose reduction compliance may have affected the results. Large observational comparative studies should be conducted to support these findings. These conclusions may only be applicable to persons with European ancestry, as there is insufficient evidence regarding validity and utility in other groups.

One economic evaluation conducted from a health care perspective in Canada reported that DPYD genetic testing for the 4 primary variants (based on data from study populations primarily with European ancestry) with subsequent guided dose adjustments was cost-effective compared to usual care, at the willingness-to-pay thresholds of $50,000 and $100,000 per QALY gained. We did not identify any relevant studies that evaluated the cost-effectiveness of whole genome sequencing or phenotype testing.

Considerations for Future Research

While current evidence on DPYD genetic testing supports the clinical validity and utility of a small number of well-characterized variants, questions remain about the relevance of additional, less-studied variants. Future research could focus on assessing the clinical validity for predicting FP-related toxicity, as well as the utility of dose adjustments for these variants. Such studies would provide a more comprehensive understanding of the genetic contributors to DPD deficiency and inform whether expanded testing panels could enhance predictive accuracy and clinical benefit. Further, evidence-based dosing guidelines are also largely limited to 4 DPYD variants. As evidence increases regarding other potentially actionable variants, there is a need for the development and validation of dosing recommendations for these variants. These guidelines should be based on robust clinical outcome data and supported by consensus among pharmacogenomic experts.

Evidence on the clinical utility of phenotype testing for DPD deficiency remains limited. As more data becomes available, future economic evaluations should assess the cost-effectiveness of phenotype testing, both alone and in combination with genotype testing. These analyses would help clarify the value of phenotype-guided dosing strategies and help guide health resource allocation decisions. Similarly, the cost-effectiveness of NGS testing could be assessed.

Finally, given Canada’s ethnically diverse population and the equity implications of pharmacogenomic testing, future studies should include detailed reporting of equity-relevant population characteristics including residence, race, ethnicity, culture, language, occupation, gender, sex, religion, education, socioeconomic status, and social capital. This information will be essential to understanding whether findings are generalizable to settings in Canada and whether expanded testing strategies mitigate or inadvertently reinforce existing disparities in cancer treatment outcomes.

Equity Considerations

A lack of diversity in genotype research and DPYD variant testing has implications for diverse populations such as Canada’s. As described throughout this report, evidence is dominated by 4 DPYD variants — DPYD*2A, DPYD*13, c.2846A>T, and c.1236G>A — identified predominantly in individuals of European ancestry. While these variants have been shown to be clinically valid, relying solely on them may result in missed diagnoses of DPD deficiency in patients from other ethnic origins, who may carry other, less-studied variants. This limits the effectiveness and inclusiveness of genotype-guided dosing strategies and raises concerns about equitable access to safe and effective cancer treatment. Testing panels that include a broader range of variants and DPD phenotype testing, which directly measures enzyme activity regardless of genotype, are compelling options that might allow for a more inclusive and accurate risk assessment, ensuring that the benefits of personalized medicine extend to all patients.

Further, laboratories that conduct DPD testing are typically located in urban centres, thereby limiting equitable access to patients in rural, remote, or underserved communities. This may delay or prevent testing.

Implications for Clinical Practice and Policy-Making

In Canada, implementation of DPD deficiency testing is still in its early stages. In our survey we identified 5 provinces currently offering DPYD genotype testing. While the findings of this report suggest that DPYD genetic testing with subsequent dose adjustments may improve clinical outcomes and be cost-effective compared to usual care, there are equity concerns in applying this evidence within the context in Canada. Much of the existing research has focused on 4 DPYD variants identified by the CPIC as clinically relevant in populations of European descent. Consequently, there is a risk of disproportionate harms and benefits for groups not of European ancestry, as well as uncertainty around the interpretation of results in these groups. To address these gaps, decision-makers may wish to work with laboratories to consider the feasibility and cost of an extended panel of variants and develop a panel relevant to the patient population present in their jurisdiction or explore whole genome sequencing via NGS. In the meantime, information could be shared with clinicians regarding the current limitations of DPYD genetic testing, and informed consent conversations with patients could outline the insufficiency of evidence to determine the applicability of the test for individuals not of European ancestry.

Phenotype testing may offer a more inclusive alternative, as it is not limited by ancestry. However, current evidence supporting its clinical utility remains limited. Furthermore, we did not find any evidence-based guidelines for phenotype-guided dosing. A combination of genotype and phenotype testing has shown promise in some contexts and may help address clinician concerns about underdosing. Decision-makers may wish to monitor the evolving literature and consider the characteristics of their patient population, to determine which testing strategies to implement to support safe and equitable chemotherapy dosing.

The findings of our survey identified clinician and patient awareness as a facilitator to DPD testing implementation. Jurisdictions may wish to consider a knowledge dissemination plan to increase awareness and access should they introduce testing. Similarly, no clinician or patient-reported outcomes were reported in the included studies. Those who intend to implement DPD testing as a part of routine clinical care may want to consider conducting ongoing monitoring to determine whether any clinician or patient-specific challenges arise. All provinces that reported conducting DPD testing offered public reimbursement, however we did not assess from where jurisdictions allotted funds for this testing (e.g., cancer budget, lab budget, other). Cost of testing will be dependent on the testing method (e.g., PCR, Sanger, NGS) and the expected number of tests, which will be specific to the population and characteristics of each individual province and territory.

A recent Environmental Scan conducted by the CDA-AMC⁵⁷ described current assessment frameworks, processes, and guiding principles used to inform the evaluation and implementation of genetic and genomic biomarker testing in cancer care across jurisdictions in Canada. The goal of this report was to support the development of a standardized decision-making framework, in response to the rapid emergence and adoption of precision medicine technologies within Canada. Drawing from a literature review and consultations, the report outlined 3 key categories of assessment criteria when considering the implementation of a genetic biomarker test: evidentiary, implementation, and decision-making. Within these categories, specific criteria included the evaluation of the clinical validity and utility of tests, economic considerations such as cost-effectiveness, barriers and facilitators to implementation, and identification of evidence gaps and future research priorities. All these criteria have been applied in the current report, which may serve as a pilot evidence review when assessing a potential new genetic test. In addition, the survey findings presented here underscore the varied landscape of implementation of genetic and genomic testing, further reinforcing the need for a standardized framework across Canada.

The limitations of the included literature, such as the lack of evidence directly comparing DPYD variant carriers with reduced dose to DPYD variant carriers with standard doses, the variable quality of primary studies included in identified SRs, the lack of randomized trials to minimize confounding variables, and concerns regarding the generalizability of findings to settings in Canada should be considered when interpreting the conclusions of this report.

References

1.Rainbow Health Ontario. Glossary. Accessed April 24, 2025, https://www.rainbowhealthontario.ca/news-publications/glossary/

2.Statistics Canada. Ethnic or cultural origin. Accessed May 20, 2025, https://www12.statcan.gc.ca/census-recensement/2021/ref/dict/az/definition-eng.cfm?ID=pop030

3.The 519. The 519 glossary of terms. Accessed April 24, 2025. https://www.the519.org/education-training/glossary/

4.World Health Organization. Health Equity. Accessed April 24, 2025. https://www.who.int/health-topics/health-equity#tab=tab_1

5.Cochrane Methods Equity. PROGRESS-Plus. Cochrane. Accessed May 12, 2025, https://methods.cochrane.org/equity/projects/evidence-equity/progress-plus

6.JAMP Pharma Corporation. Capecitabine Product Monograph. JAMP Pharma Corporation; 2022. Accessed April 20, 2025. https://pdf.hres.ca/dpd_pm/00066694.PDF

7.Sandoz Canada Inc. Fluorouracil Injection Product Monograph. Sandoz Canada Inc; 2024. Accessed April 20, 2025. https://pdf.hres.ca/dpd_pm/00016156.PDF

8.Longley DB, Harkin DP, Johnston PG. 5-fluorouracil: mechanisms of action and clinical strategies. Nat Rev Cancer. 2003;3(5):330-8. doi:10.1038/nrc1074 PubMed

9.Doornhof KR, van der Linden PD, Boeke GM, Willemsen A, Daskapan A. Dihydropyrimidine dehydrogenase phenotype in peripheral blood mononuclear cells is related to adverse events of fluoropyrimidine-therapy. Eur J Clin Pharmacol. 2023;79(4):493-501. doi:10.1007/s00228-023-03466-8 PubMed

10.US Department of Health and Human Services. Common Terminology Criteria for Adverse Events (CTCAE) Version 5.0. National Institutes of Health, National Cancer Institute. Updated November 27, 2017. Accessed May 12, 2025, https://ctep.cancer.gov/protocoldevelopment/electronic_applications/ctc.htm

11.Salama S, Gallo-Hershberg D, Forbes L. Fluoropyrimidine Treatment in Patients with Dihydropyrimidine Dehydrogenase (DPD) Deficiency, Guidance for Clinicians. Ontario Health, Cancer Care Ontario; 2023. https://www.cancercareontario.ca/en/guidelines-advice/types-of-cancer/73951

12.Lunenburg CATC, van der Wouden CH, Nijenhuis M, et al. Dutch Pharmacogenetics Working Group (DPWG) guideline for the gene-drug interaction of DPYD and fluoropyrimidines. Eur J Hum Genet. 2020;28(4):508-517. doi:10.1038/s41431-019-0540-0 PubMed

13.Henricks LM, Opdam FL, Beijnen JH, Cats A, Schellens JHM. DPYD genotype-guided dose individualization to improve patient safety of fluoropyrimidine therapy: call for a drug label update. Ann Oncol. 2017;28(12):2915-2922. doi:10.1093/annonc/mdx411 PubMed

14.Garcia-Alfonso P, Saiz-Rodriguez M, Mondejar R, et al. Consensus of experts from the Spanish Pharmacogenetics and Pharmacogenomics Society and the Spanish Society of Medical Oncology for the genotyping of DPYD in cancer patients who are candidates for treatment with fluoropyrimidines. Clinical & Translational Oncology: Official Publication of the Federation of Spanish Oncology Societies & of the National Cancer Institute of Mexico. 2022;24(3):483-494. doi:10.1007/s12094-021-02708-4 PubMed

15.ThermoFisher Scientific. What is Genotyping? ThermoFisher Scientific. Accessed April 15, 2025, 2025. https://www.thermofisher.com/ca/en/home/life-science/pcr/real-time-pcr/real-time-pcr-learning-center/genotyping-analysis-real-time-pcr-information/what-is-genotyping.html

16.Smith M. Polymerase Chain Reaction (PCR). National Human Genome Research Institute. Accessed April 15, 2025, 2025. https://www.genome.gov/genetics-glossary/Polymerase-Chain-Reaction-PCR

17.Amstutz U, Henricks LM, Offer SM, et al. Clinical Pharmacogenetics Implementation Consortium (CPIC) Guideline for Dihydropyrimidine Dehydrogenase Genotype and Fluoropyrimidine Dosing: 2017 Update. Clin Pharmacol Ther. 2018;103(2):210-216. doi:10.1002/cpt.911 PubMed

18.Chan TH, Zhang JE, Pirmohamed M. DPYD genetic polymorphisms in non-European patients with severe fluoropyrimidine-related toxicity: a systematic review. Systematic Review. Br J Cancer. Aug 2024;131(3):498-514. doi:10.1038/s41416-024-02754-z PubMed

19.Deac AL, Burz CC, Bocse HF, Bocsan IC, Buzoianu AD. A review on the importance of genotyping and phenotyping in fluoropyrimidine treatment. Med. 2020;93(3):223-230. doi:10.15386/mpr-1564 PubMed

20.Glewis S, Alexander M, Khabib MNH, et al. A systematic review and meta-analysis of toxicity and treatment outcomes with pharmacogenetic-guided dosing compared to standard of care BSA-based fluoropyrimidine dosing. Br J Cancer. 2022;127(1):126-136. doi:10.1038/s41416-022-01779-6 PubMed

21.Knikman JE, Rosing H, Guchelaar HJ, Cats A, Beijnen JH. Assay performance and stability of uracil and dihydrouracil in clinical practice. Cancer Chemother Pharmacol. 2023;91(3):257-266. doi:10.1007/s00280-023-04518-7 PubMed

22.Jacobs BA, Deenen MJ, Pluim D, et al. Pronounced between-subject and circadian variability in thymidylate synthase and dihydropyrimidine dehydrogenase enzyme activity in human volunteers. Br J Clin Pharmacol. 2016;82(3):706-716. doi:10.1111/bcp.13007 PubMed

23.De Metz C, Hennart B, Aymes E, et al. Complete DPYD genotyping combined with dihydropyrimidine dehydrogenase phenotyping to prevent fluoropyrimidine toxicity: A retrospective study. Cancer Med. Mar 2024;13(6):e7066. doi:10.1002/cam4.7066 PubMed

24.Hodroj K, Barthelemy D, Lega JC, et al. Issues and limitations of available biomarkers for fluoropyrimidine-based chemotherapy toxicity, a narrative review of the literature. ESMO open. 2021;6(3):100125. doi:10.1016/j.esmoop.2021.100125 PubMed

25.Henricks LM, van Merendonk LN, Meulendijks D, et al. Effectiveness and safety of reduced-dose fluoropyrimidine therapy in patients carrying the DPYD 2A variant: A matched pair analysis. Int J Cancer. 2019;144(9):2347-2354. doi:10.1002/ijc.32022 PubMed

26.European Medicines Agency. EMA Recommendations on DPD Testing Prior to Treatment with Fluorouracil, Capecitabine, Tegafur and Flucytosine. European Medicines Agency. Accessed May 12, 2025, https://www.ema.europa.eu/en/news/ema-recommendations-dpd-testing-prior-treatment-fluorouracil-capecitabine-tegafur-and-flucytosine

27.Medicines and Healthcare Products Regulatory Agency. 5-Fluorouracil (intravenous), Capecitabine, Tegafur: DPD Testing Recommended Before Initiation to Identify Patients at Increased Risk of Severe and Fatal Toxicity. gov.uk. Accessed April 22, 2025, https://www.gov.uk/drug-safety-update/5-fluorouracil-intravenous-capecitabine-tegafur-dpd-testing-recommended-before-initiation-to-identify-patients-at-increased-risk-of-severe-and-fatal-toxicity

28.Casneuf V, Borbath I, Van den Eynde M, et al. Joint Belgian recommendation on screening for DPD-deficiency in patients treated with 5-FU, capecitabine (and tegafur). Acta Clin Belg. 2022;77(2):346-352. doi:10.1080/17843286.2020.1870855 PubMed

29.Loriot MA, Ciccolini J, Thomas F, et al. Dihydropyrimidine dehydrogenase (DPD) deficiency screening and securing of fluoropyrimidine-based chemotherapies: Update and recommendations of the French GPCO-Unicancer and RNPGx networks. Depistage du deficit en dihydropyrimidine deshydrogenase (DPD) et securisation des chimiotherapies a base de fluoropyrimidines: mise au point et recommandations nationales du GPCO-Unicancer et du RNPGx. Bull Cancer. 2018;105(4):397-407. doi:10.1016/j.bulcan.2018.02.001 PubMed

30.Loriot MA, Masskouri F, Carni P, et al. Dihydropyrimidine dehydrogenase deficiency screening for management of patients receiving a fluoropyrimidine: Results of two national practice surveys addressed to clinicians and biologists. Interets et limites de la recherche du deficit en dihydropyrimidine deshydrogenase dans le suivi des patients traites par fluoropyrimidines: resultats de deux enquetes nationales de pratiques realisees aupres des medecins et des biologistes. Bull Cancer. 2019;106(9):759-775. doi:10.1016/j.bulcan.2019.04.013 PubMed

31.SurveyMonkey Inc. SurveyMonkey. Accessed March 12, 2025, https://www.surveymonkey.com

32.Pratt VM, Cavallari LH, Fulmer ML, et al. DPYD Genotyping Recommendations: A Joint Consensus Recommendation of the Association for Molecular Pathology, American College of Medical Genetics and Genomics, Clinical Pharmacogenetics Implementation Consortium, College of American Pathologists, Dutch Pharmacogenetics Working Group of the Royal Dutch Pharmacists Association, European Society for Pharmacogenomics and Personalized Therapy, Pharmacogenomics Knowledgebase, and Pharmacogene Variation Consortium. Review. J Mol Diagn. Oct 2024;26(10):851-863. doi:10.1016/j.jmoldx.2024.05.015 PubMed

33.Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. J Clin Epidemiol. 2009;62(10):e1-e34. doi:10.1016/j.jclinepi.2009.06.006 PubMed

34.Henricks LM, Lunenburg CATC, Meulendijks D, et al. Translating DPYD genotype into DPD phenotype: Using the DPYD gene activity score. Pharmacogenomics. 2015;16(11):1277-1286. doi:10.2217/PGS.15.70 PubMed

35.de Moraes FCA, de Almeida Barbosa AB, Sano VKT, Kelly FA, Burbano RMR. Pharmacogenetics of DPYD and treatment-related mortality on fluoropyrimidine chemotherapy for cancer patients: a meta-analysis and trial sequential analysis. BMC Cancer. 2024;24(1):1210. doi:10.1186/s12885-024-12981-5 PubMed

36.Kim W, Cho YA, Kim DC, Lee KE. Elevated Risk of Fluoropyrimidine-Associated Toxicity in European Patients with DPYD Genetic Polymorphism: A Systematic Review and Meta-Analysis. Journal of Personalized Medicine. 2022;12(2):225. doi:10.3390/jpm12020225 PubMed

37.Cura Y, Perez-Ramirez C, Sanchez-Martin A, et al. Influence of Single-Nucleotide Polymorphisms on Clinical Outcomes of Capecitabine-Based Chemotherapy in Colorectal Cancer Patients: A Systematic Review. Review. Cancers (Basel). 2023;15(6):1821. doi:10.3390/cancers15061821 PubMed

38.Paulsen NH, Vojdeman F, Andersen SE, et al. DPYD genotyping and dihydropyrimidine dehydrogenase (DPD) phenotyping in clinical oncology. A clinically focused minireview. Review. Basic Clin Pharmacol Toxicol. 2022;131(5):325-346. doi:10.1111/bcpt.13782 PubMed

39.van Kuilenburg ABP, Meinsma R, van Gennip AH. Pyrimidine Degradation Defects and Severe 5-Fluorouracil Toxicity. Nucleosides, nucleotides & nucleic acids. 2004;23(8-9):1371-1375. doi:10.1081/NCN-200027624 PubMed

40.Tejedor-Tejada E, Rubio Calvo D, Garcia Andreo A. Determination of plasma uracil as a screening for dihydropyrimidine dehydrogenase deficiency: clinical application in oncological treatments. Observational Study. Eur. Feb 22 2024;31(2):124-126. doi:10.1136/ejhpharm-2021-003210 PubMed

41.Laures N, Konecki C, Brugel M, et al. Impact of Guidelines Regarding Dihydropyrimidine Dehydrogenase (DPD) Deficiency Screening Using Uracil-Based Phenotyping on the Reduction of Severe Side Effect of 5-Fluorouracil-Based Chemotherapy: A Propension Score Analysis. Pharmaceutics. 2022;14(10):06. doi:10.3390/pharmaceutics14102119

42.Paulsen NH, Pfeiffer P, Ewertz M, et al. Implementation and clinical benefit of DPYD genotyping in a Danish cancer population. ESMO open. 2023;8(1):100782. doi:10.1016/j.esmoop.2023.100782 PubMed

43.Ockeloen CW, Raaijmakers A, Hijmans-van der Vegt M, et al. Potential added value of combined DPYD/DPD genotyping and phenotyping to prevent severe toxicity in patients with a DPYD variant and decreased dihydropyrimidine dehydrogenase enzyme activity. J Oncol Pharm Pract. 2023;29(1):5-13. doi:10.1177/10781552211049144 PubMed

44.Ontario Health (Quality). DPYD Genotyping in Patients Who Have Planned Cancer Treatment With Fluoropyrimidines: A Health Technology Assessment. 2021:1-186. https://www.hqontario.ca/Portals/0/Documents/evidence/reports/hta-dpyd-genotyping-in-patients-who-have-planned-cancer-treatment-with-fluoropyrimidines.pdf

45.Koleva-Kolarova R, Vellekoop H, Huygens S, et al. Cost-effectiveness of extended DPYD testing before fluoropyrimidine chemotherapy in metastatic breast cancer in the UK. Per Med. 2023;20(4):339-355. doi:10.2217/pme-2022-0099 PubMed

46.Fragoulakis V, Roncato R, Bignucolo A, et al. Cost-utility analysis and cross-country comparison of pharmacogenomics-guided treatment in colorectal cancer patients participating in the U-PGx PREPARE study. Review. Pharmacol Res. 2023;197:106949. doi:10.1016/j.phrs.2023.106949 PubMed

47.Fragoulakis V, Roncato R, Fratte CD, et al. Estimating the Effectiveness of DPYD Genotyping in Italian Individuals Suffering from Cancer Based on the Cost of Chemotherapy-Induced Toxicity. Am J Hum Genet. 2019;104(6):1158-1168. doi:10.1016/j.ajhg.2019.04.017 PubMed

48.Brooks GA, Tapp S, Daly AT, Busam JA, Tosteson ANA. Cost-effectiveness of DPYD Genotyping Prior to Fluoropyrimidine-based Adjuvant Chemotherapy for Colon Cancer. Clin Colorectal Cancer. 2022;21(3):e189-e195. doi:10.1016/j.clcc.2022.05.001 PubMed

49.Deenen MJ, Meulendijks D, Cats A, et al. Upfront Genotyping of DPYD 2A to Individualize Fluoropyrimidine Therapy: A Safety and Cost Analysis. J Clin Oncol. 2016;34(3):227-234. doi:10.1200/JCO.2015.63.1325 PubMed

50.Henricks LM, Lunenburg CATC, de Man FM, et al. A cost analysis of upfront DPYD genotype-guided dose individualisation in fluoropyrimidine-based anticancer therapy. Eur J Cancer. 2019;107:60-67. doi:10.1016/j.ejca.2018.11.010 PubMed

51.Murphy C, Byrne S, Ahmed G, et al. Cost Implications of Reactive Versus Prospective Testing for Dihydropyrimidine Dehydrogenase Deficiency in Patients With Colorectal Cancer: A Single-Institution Experience. Dose-Response. 2018;16(4)doi:10.1177/1559325818803042 PubMed

52.Cortejoso L, Garcia-Gonzalez X, Garcia MI, Garcia-Alfonso P, Sanjurjo M, Lopez-Fernandez LA. Cost-effectiveness of screening for DPYD polymorphisms to prevent neutropenia in cancer patients treated with fluoropyrimidines. Pharmacogenomics. 2016;17(9):979-984. doi:10.2217/pgs-2016-0006 PubMed

53.Abushanab D, Mohamed S, Abdel-Latif R, et al. Dihydropyrimidine Dehydrogenase Deficiency (DPYD) Genotyping-Guided Fluoropyrimidine-Based Adjuvant Chemotherapy for Breast Cancer. A Cost-Effectiveness Analysis. Clin Drug Investig. 2025;45(3):151-163. doi:10.1007/s40261-024-01413-8 PubMed

54.De Mattia E, Milan N, Assaraf YG, Toffoli G, Cecchin E. Clinical Implementation of Rare and Novel DPYD Variants for Personalizing Fluoropyrimidine Treatment: Challenges and Opportunities. Review. International Journal of Biological Sciences [Electronic Resource]. 2024;20(10):3742-3759. doi:10.7150/ijbs.97686 PubMed

55.O'Neill J, Tabish H, Welch V, et al. Applying an equity lens to interventions: using PROGRESS ensures consideration of socially stratifying factors to illuminate inequities in health. J Clin Epidemiol. 2014;67(1):56-64. doi:10.1016/j.jclinepi.2013.08.005 PubMed

56.Oliver S KJ, Caird J, Lorenc T, Oliver K, Harden A, Thomas J, Greaves A, Oakley A.. Health Promotion, Inequalities and Young People's Health: A Systematic Review of Research. EPPI-Centre, Social Science Research Unit, Institute of Education, University of London; 2008. Accessed April 20, 2025. https://eppi.ioe.ac.uk/cms/Portals/0/PDF%20reviews%20and%20summaries/Inequalities%20Young%20People%20R2008Oliver.pdf

57.Canada’s Drug Agency (CDA-AMC). Environmental Scan of Genetic and Genomic Biomarker Testing Assessment Frameworks, Processes, and Inventories in Cancer Care. 2025. Accessed April 25, 2025. https://www.cda-amc.ca/sites/default/files/hta-he/CP0037-000-GG-Test-Frameworks-Report_revised.pdf

58.Shea BJ, Reeves BC, Wells G, et al. AMSTAR 2: a critical appraisal tool for systematic reviews that include randomised or non-randomised studies of healthcare interventions, or both. BMJ. 2017;358:j4008. doi:10.1136/bmj.j4008 PubMed

59.Downs SH, Black N. The feasibility of creating a checklist for the assessment of the methodological quality both of randomised and non-randomised studies of health care interventions. J Epidemiol Community Health. 1998;52(6):377-384. doi:10.1136/jech.52.6.377 PubMed

60.Agree Next Steps Consortium. The AGREE II Instrument. AGREE Enterprise; 2017. Accessed April 15, 2025. https://www.agreetrust.org/wp-content/uploads/2017/12/AGREE-II-Users-Manual-and-23-item-Instrument-2009-Update-2017.pdf

61.Swen JJ, Nijenhuis M, Boer A, et al. Pharmacogenetics: From Bench to Byte— An Update of Guidelines. Clin Pharmacol Ther. 2011;89(5):662-673. doi:10.1038/clpt.2011.34 PubMed

62.Valdes R, Jr., Payne DA, Linder MW. Laboratory Analysis and Application of Pharmacogenetics to Clinical Practice. National Academy of Clinical Biochemistry; 2010. Accessed April 22, 2025. https://myadlm.org/science-and-research/practice-guidelines/pharmacogenetics

63.Husereau D, Drummond M, Augustovski F, et al. Consolidated Health Economic Evaluation Reporting Standards 2022 (CHEERS 2022) Statement: Updated Reporting Guidance for Health Economic Evaluations. BMJ. 2022;376:e067975. doi:10.1136/bmj-2021-067975 PubMed

Appendix 1: Detailed Methods and Selection of Included Studies

Please note that this appendix has not been copy-edited.

Literature Search Methods

An information specialist conducted a literature search on key resources including MEDLINE and Embase via OVID, the Cochrane Database of Systematic Reviews, the International HTA Database, the websites of health technology assessment agencies in Canada and major international HTA agencies, as well as a focused internet search. The search approach was customized to retrieve a limited set of results, balancing comprehensiveness with relevance. The search strategy comprised both controlled vocabulary, such as the National Library of Medicine’s MeSH (Medical Subject Headings), and keywords. Search concepts were developed based on the elements of the research questions and selection criteria. The main search concepts were dihydropyrimidine dehydrogenase deficiency and testing. Comments, newspaper articles, editorials, conference abstracts, and letters were excluded. Retrieval was limited to the human population. The search was completed on March 4, 2025 and limited to English-language documents published since January 1, 2015.

Selection Criteria and Methods

One reviewer screened citations and selected studies. In the first level of screening, titles and abstracts were reviewed and potentially relevant articles were retrieved and assessed for inclusion. The final selection of full-text articles was based on the inclusion criteria presented in Table 3.

Table 3: Selection Criteria

ICER = incremental cost-effectiveness ratio; QALY = quality-adjusted life-year; vs. = versus.

Exclusion Criteria

Articles were excluded if they did not meet the selection criteria outlined in Table 3, they were duplicate publications, were published before 2015, or were not available in English. Systematic reviews in which all relevant studies were captured in other more recent or more comprehensive systematic reviews were excluded. Primary studies retrieved by the search were excluded if they were captured in 1 or more included systematic reviews. Guidelines with unclear methodology or without a formal evidence review were also excluded.

Critical Appraisal of Individual Studies

The included publications were critically appraised by 1 reviewer using the following tools as a guide: A MeaSurement Tool to Assess systematic Reviews 2 (AMSTAR 2)⁵⁸ for systematic reviews, the Downs and Black checklist⁵⁹ for randomized and nonrandomized studies, and the Appraisal of Guidelines for Research and Evaluation (AGREE) II instrument⁶⁰ for guidelines. Summary scores were not calculated for the included studies; rather, the strengths and limitations of each included publication were described narratively.

Survey

The invitation to participate in our voluntary survey was sent via email to the representative believed to be responsible for decisions regarding DPD testing and funding in 10 of the 13 Canadian jurisdictions in March and April 2025. Consent was obtained from participants on the first page of the survey, following provision of information regarding the purpose of the survey, time involved, and confidentiality. The survey was hosted by SurveyMonkey, with all appropriate licensing, but respondents also had the option to respond to survey questions directly via email.

The survey consisted of 34 questions more than 13 pages: 1 consent question, 6 demographic questions, 16 questions regarding the status of DPD testing in their jurisdiction, 7 questions assessing barriers and facilitators to implementation, 3 questions on cost and reimbursement, and 1 question requesting any additional information not otherwise captured. The respondent for Nunavut provided information regarding the Northwest Territories and the Yukon. Details regarding the availability of DPD testing for Ontario were pulled from public resources.

Figure 3: Selection of Included Studies

Appendix 2: Survey Findings

Please note that this appendix has not been copy-edited.

Table 4: Information on Survey Respondents and Contacts

QEII = Queen Elizabeth II.

^aContact provided information on other territories – Northwest Territories and Yukon.

Table 5: Summary of Status and Availability of DPD Testing Across Canada

EMR = electronic medical records; NA = not applicable; NR = not reported; TAT = turnaround time; NAAT = nucleic acid amplification test.

^aSurvey response was pulled from public sources provided by contact.

Table 6: Implementation of DPD Testing across Canada

AMP = Association for Molecular Pathology; CCO = Cancer Care Ontario; CPIC = Clinical Pharmacogenetics Implementation Consortium; DPD = dihydropyrimidine dehydrogenase; EMA = European Medicine Agency; EMR = electronic medical records; FDA = NA = not applicable; NR = not reported OH = Ontario Health; TAT = turnaround time.

^aSurvey response was pulled from public sources provided by contact.

Appendix 3: Characteristics of Included Publications

Please note that this appendix has not been copy-edited.

Summary of Study Characteristics

Summaries of study characteristics are organized by research question. Additional details regarding the characteristics of included publications are provided in Table 7 to Table 10.

Included Studies for Question 2: Guidelines on Genotype or Phenotype Testing for DPD Deficiency

Three evidence-based guidelines^11,12,17 and 1 consensus-based guideline³² were included in this review. The consensus document³² was published in 2024 by the AMP which is the Pharmacogenomics Working Group of the Clinical Practice Committee. The 3 evidence-based guidelines^11,12,17 were produced by OH-CCO published in 2023, the Dutch Pharmacogenetics Working Group (DPWG) of The Royal Dutch Pharmacists Association published in 2019, and the CPIC published in 2017, and an update posted on the webpage in 2024. One guideline was developed in Canada,¹¹ 1 in the Netherlands,¹² and 2 by groups based in the US.^17,32 One guideline³² was developed through joint consensus among subject matter experts from multiple professional organizations including the CDC, CPIC, and DPWG. Three guidelines^11,12,17 conducted a systematic literature review to create the evidence base for recommendations. Recommendations were developed by evidence and expert consensus.^11,17 The OH guideline¹¹ was additionally reviewed by external experts and peer consultations. The DPWG¹² used a 5-point rating scale (0 to 4) created by Swen et al.⁶¹ to assess the strength of evidence, while CPIC¹⁷ used a scale modified from Valdes et al.⁶² (weak, moderate, high). The CPIC group determined the strength of recommendations, while OH and DPWG did not report a formal method for rating the strength of recommendations. The consensus-based guideline by the AMP was a technical testing standard intended to provide laboratory guidance for pharmacogenomic testing of DPYD variants. Three guidelines focused on patients who are candidates for systemic treatment with FPs (5-FU or capecitabine) to provide genotype-guided dosing recommendations to minimize the risk of severe toxicity. The DPWG also considered recommendations for cutaneous administration of 5-FU or capecitabine. The OH group additionally provided recommendations on DPYD testing procedures. Intended users were clinicians,^11,17 physicians,^11,12 pharmacists,¹² or laboratory technicians³² involved in the care of patients with cancer.

Included Studies for Question 3: Clinical Validity of Genotype or Phenotype Testing for DPD Deficiency

We identified 1 HTA,⁴⁴ 4 SRs,^35-38,44 and 1 primary study⁹ to address this research question. Two of the SRs^35,36 included a meta-analysis of results. The HTA⁴⁴ and SRs^35-38,44 included data from a total of 90 unique primary clinical studies; however, there was considerable overlap among the included primary studies. As a result, the pooled effect estimates and narrative summaries from separate reviews are based on some of the same data, although not all reviews reported the same outcomes. A citation matrix illustrating the degree of primary study overlap is presented in Appendix 6.

The authors of the SR conducted as part of the HTA by Ontario Health⁴⁴ searched for systematic reviews and HTAs published from database inception to February 2020, to use as a source of primary studies published until their literature search dates. They then performed a search for primary clinical studies published from January 2018 (earliest search end date for included SRs and HTAs) to February 2020. In total, the authors included 29 primary clinical studies (25 relevant to this research question; clinical validity).

In the 2 included SRs with meta-analyses^35,36 study design was limited to RCTs or cohort studies. The SR by de Moraes et al³⁵ included 9 RCTs, 11 retrospective cohort, and 16 prospective cohort studies published up to June 18, 2024, while the SR by Kim et al³⁶ included 11 RCTs published up to October 18, 2021.

The SR by Cura et al³⁷ searched for studies of any design published up to December 30, 2022 that evaluated association between mutations in genes involved in the breakdown of capecitabine with toxicity or treatment effectiveness. They identified 12 studies in total, 6 of which were relevant to the present review. Finally, Paulsen et al³⁸ conducted 2 separate literature searches, both for any human clinical trials published up to June 10, 2021. The first search for measure of participants DPYD genotype, and the second for measurement of participants’ DPD phenotype in plasma (uracil and/or dihydrouracil). They found 10 genotyping studies and 7 phenotyping studies relevant to this research question.

The primary study included⁹ in this question was conducted in the Netherlands and published in 2023.

The HTA,⁴⁴ 2 of the SRs^36,37,44 and the NRS⁹ provided information on the age, sex, and ethnicity of participants from the included primary studies; however, the authors did not report how sex was defined or measured. In most cases, ethnicity was poorly reported within the primary studies. None of the included SRs provided participant information for other PROGRESS-Plus criteria,^55,56 such as place of residence, race, culture, language, occupation, religion, education, socioeconomic status, or social capital.

Four of the SRs^35-37,44 focused solely on DPYD genotyping to test DPD activity, while 1 SR³⁸ included 12 studies focused on DPYD genotyping and 9 studies on DPD phenotyping (via plasma U concentrations and UH₂/U ratios). The NRS⁹ evaluated DPD activity measured in PBMCs.

Outcomes assessed across the HTA, 4 SRs^35-38,44 and 1 NRS⁹ to address research question 3 included:

• severe (grade ≥ 3) toxicity (overall and by category: cardiovascular, gastrointestinal, neurologic, and hematological)

• FP-related mortality

• FP-related hospitalization

• sensitivity, specificity, PPV, and NPV of DPYD genotyping to predict severe FP-related toxicity.

Included Studies for Question 4: Clinical Utility of Genotype or Phenotype Testing for DPD Deficiency

We identified 1 HTA,⁴⁴ 2 SRs^20,38 and 4 primary studies^40-43 to address this research question. One of the SRs²⁰ included a meta-analysis of results. The HTA⁴⁴ and SRs^20,38 included data from a total of 67 primary clinical studies, however only 27 were relevant to the present question. There was considerable overlap among the included primary studies and as a result, the pooled effect estimates and narrative summaries from separate reviews are based on some of the same data, although not all reviews reported the same outcomes. A citation matrix illustrating the degree of primary study overlap is presented in Appendix 6.

The authors of the HTA by Ontario Health⁴⁴ published in 2021, searched for SRs and HTAs published from database inception to February 2020, to use as a source of primary studies published until their literature search dates. They then performed a search for primary clinical studies published from January 2018 (earliest search end date for included SRs and HTAs) to February 2020. In total, the authors included 29 primary clinical studies (6 relevant to this relevant to this research question).

In the 1 included SR with a meta-analysis,²⁰ published in 2022, included study designs was limited to RCTs or cohort studies. It included 8 prospective studies, 6 retrospective studies, and 3 combined retrospective and prospective studies published up to December 7, 2020. The SR by Paulsen et al.,³⁸ published in 2022, conducted 2 separate literature searches, both including any human clinical trials published up to June 10, 2021. The first search looked for studies assessing participants’ DPYD genotype, and the second for studies assessing participants’ DPD phenotype (as measured by plasma uracil and/or dihydrouracil concentrations). The authors included 21 primary studies in total, 5 of which were relevant to this research question.

All 4 included NRSs^40-43 were published between 2022 and 2024, and were conducted in hospital settings in Spain,⁴⁰ the Netherlands,⁴³ Denmark,⁴² and France.⁴¹ Two are retrospective cohort studies,^40,43 1 retrospective, before-and-after study with a propensity score analysis,⁴¹ and 1 prospective cohort, before-and-after study with a historical control group.⁴²

All included studies for this research question^20,38,40-44 provided information on the age and sex of participants; however, the authors did not report how sex was defined or measured. The HTA⁴⁴ and 1 SR}²⁰ reported on the ethnicity of study participants. None of the included studies provided participant information for other PROGRESS-Plus criteria,^55,56 such as place of residence, race, culture, language, occupation, religion, education, socioeconomic status, or social capital.

The HTA,⁴⁴ 1 SR,²⁰ and 1 NRS⁴² assessed genotype-guided dosing only. 1 SR³⁸ included studies assessing either genotype or phenotype-based dosing. Two of the included NRSs^40,41 evaluated phenotype-guided dosing based on plasma uracil concentrations, and 1 NRS⁴³ assessed a combined genotype and phenotype dosing method. In that study, DPD activity levels were measured in PBMCs.

Outcomes assessed across the 1 HTA, 2 SRs and 4 NRS to address research question 4 included:

• severe (grade ≥ 3) toxicity (overall toxicity and specific categories or types: cardiovascular, gastrointestinal, neurologic, and hematological), assessed using the CTCAE¹⁰

• FP-related mortality

• FP-related hospitalization

• disease response.

Included Studies for Question 5: Cost Implications of Genotype or Phenotype Testing for DPD Deficiency

We identified 1 economic evaluation, conducted as part of an Ontario HTA,⁴⁴ to address this research question. The study assessed the cost-effectiveness of DPYD genotyping followed by genotype-guided dosing versus usual care. We found no relevant evidence regarding the cost-effectiveness of phenotype testing for DPD deficiency versus usual care; therefore, no summary can be provided. Of note, we found 9 additional economic evaluations^45-53 and 1 budget impact analysis⁴⁴ assessing DPYD genotyping, which were excluded from this report due to their settings (i.e., they were not conducted from a Canadian health care perspective). A brief summary of the results of these excluded economic evaluations can be found in Appendix 5.

The included economic evaluation⁴⁴ conducted a probabilistic cost-utility analysis and a cost-effectiveness analysis, using a decision-tree model with a 6-month time horizon, from the perspective of the Ontario Ministry of Health, in Ontario, Canada. Outcomes for the probabilistic cost-utility analysis were costs and QALYs, and for the cost-effectiveness analysis were the proportion of patients with severe FP-related toxicities and the number of severe toxicities.

Clinical model inputs (e.g., patient characteristics, DPYD variant prevalence, probabilities of severe toxicity) were drawn from various sources of published literature, pooled prevalences from the meta-analysis conducted as part of the HTA⁴⁴ to inform clinical validity, clinical expert opinion, various sources of published literature and assumptions where required. Cost inputs were drawn from the Canadian Institute of Health Information (CIHI), the Ontario Health Insurance Program (OHIP), Ontario Drug Benefit (ODB) program, the pan-Canadian Oncology Drug Review (pCODR), laboratory expert opinion, and various sources of published literature. Costs were inflated to 2020 CA$.

The patient population was based on the characteristics of patients who received FPs in Ontario from 2014 to 2019, including different types of cancer (e.g., colorectal, breast, gastrointestinal, other), and receiving either 5-FU, capecitabine, or a combination regimen. Age and sex of the patient population was reported however, the authors did not report how sex was defined or measured. The evaluation did not provide information for other PROGRESS-Plus criteria,^55,56 such as place of residence, race, ethnicity, culture, language, occupation, religion, education, socioeconomic status, or social capital. The intervention in this study was pretreatment DPYD genotyping for the 4 primary variants (DPYD*2A, DPYD*13, c.2846A>T, c.1236G>A) in all patients with planned FP treatment, followed by genotype-guided dose adjustments made according to the 2017 CPIC guidelines,¹⁷ compared to no testing and standard doses.

Table 7: Characteristics of Included Guidelines

5- FU = 5-fluorouracil; AMP = Association for Molecular Pathology; CCO = Cancer Care Ontario; CPIC = Clinical Pharmacogenetics Implementation Consortium; DPWG = Dutch Pharmacogenetics Working Group; OH = Ontario Health; NR = not reported.

^aEvidence was given a clinical impact score on a 7-point scale ranging from AA^{# (}positive effect) to F (highest negative effect).

Table 8: Characteristics of Included HTA and Systematic Reviews

CI = confidence interval; GI = gastrointestinal; CTCAE = Common Terminology Criteria for Adverse Events; NA = not applicable; NR = not reported.; OR = odds ratio.

Table 9: Characteristics of Included Primary Clinical Studies

NA = not applicable; NR = not reported.; IQR = interquartile range; BSA = body surface area; DPD = dihydropyrimidine dehydrogenase; sd = standard deviation; PMBCs = peripheral blood mononuclear cells; CPIC = clinical pharmacogenetics implementation consortium; AE = adverse event, CTCAE = common terminology criteria for adverse events.

^aThe main PROGRESS-Plus criteria include place of residence, race, ethnicity, culture, language, occupation, gender, sex, religion, education, socioeconomic status, and social capital, personal characteristics associated with discrimination (e.g., age, disability), features of relationships, and time-dependent relationships.^55,56

Table 10: Characteristics of Included Economic Evaluation

NR = not reported. CPIC = Clinical Pharmacogenetics Implementation Consortium; NHS = National Health Service; QALY = quality-adjusted life-years.

Appendix 4: Critical Appraisal of Included Publications

Please note that this appendix has not been copy-edited.

Summary of Critical Appraisal

Critical appraisal summaries are organized by study design. Table 11 to Table 15 present additional details regarding the strengths and limitations of the included publications.

Evidence-Based Guidelines

The overall scope and purpose were clearly outlined in all guidelines,^11,12,17,32 including their objectives and target population, although the health questions being addressed in the guidelines were not explicitly stated or clearly defined in any of the documents. All guidelines reported details regarding their guideline development group and target users.^11,12,17,32 Two guidelines specified that a multidisciplinary group of professionals was involved in formulating the recommendations.^12,17

A systematic literature review was used by 3 guideline development groups^11,12,17 however, only 2 groups^12,17 provided details about the criteria, search strategies, and strengths and limitations of evidence. Systematic review details of the OH-CCO guideline¹¹ were provided in a HTA published by Ontario Health.⁴⁴ One guideline group relied on consensus between subject matter experts but details on how the evidence base was created were scarce.³² Two guidelines described the methods used to formulate recommendations, expert consultation and consensus, though neither described how expert consensus was achieved.^11,17 A clear explicit link between the recommendations and evidence was provided in 2 guidelines with a table showing the rationale and strength of evidence.^11,17

Only 1 guideline included details of an external review process with the names and occupation of each external reviewer provided.¹¹ None of the groups provided details about procedures for updating the guideline, although 1 guideline was updated.¹⁷ Recommendations in each of the guidelines were specific, clear, and unambiguous with key recommendations easy to identify.

Two guidelines^11,17 described the facilitators and barriers of applying the recommendations, as well as implementation tools and workflow that intended users could apply. One guideline provided guidelines for implementing DPYD testing.¹¹ None of the guidelines reported potential resource implications for applying the recommendations. Two guidelines^12,17 provided a funding statement but only 1¹² specified that the funding bodies were not involved in the development of guideline. Conflicts of interest were declared in all 4 documents,^11,12,17,32 but one¹¹ did not include a formal statement.

Systematic Reviews

The authors of all 6 SRs,^20,35-38,44 including the SR from the HTA,⁴⁴ clearly defined their objectives and eligibility criteria, conducted comprehensive literature searches across multiple databases, and provided details on key search terms and search dates. They also included flow charts illustrating the study selection along with their reasons for excluding studies. These methodological strengths increase the reproducibility of the SRs. The review methods for 3 of the SRs^20,35,44 were established before conducting the reviews (i.e., they were documented in registered protocols), reducing the risk of reporting bias. In 5 SRs,^20,35-38 study selection was performed by either 2 or 3 authors independently.

The quality of the included primary studies was assessed using transparent and satisfactory techniques in 5 SRs,^20,35-37,44 and publication bias was assessed by the authors of 4 SRs^20,35,36,44 using funnel plots and/or Egger’s and Begg’s regression tests. In all cases, the authors suggested that the risk for publication bias was low. Three of the SRs^20,35,44 reported the characteristics of included studies in sufficient detail (e.g., study design, number of participants, study location) and used appropriate methods for the statistical combination of results and assessing statistical heterogeneity (e.g., the I2 statistic).

As for methodological limitations, the authors of 5 SRs^20,35-38 did not report conducting a grey literature search, increasing the risk of missing relevant studies that are not published commercially and that may be inaccessible via bibliographic databases (i.e., nonindexed studies). Three SRs^36-38 did not report that methods were established a priori. None of the SRs reported the sources of funding for the included primary studies, potentially influencing conflicts of interest within the studies. All the SRs,^20,35-38,44 limited included studies to those published in English or did not specify which languages were eligible for inclusion, potentially introducing language bias and omitting relevant data from non-English studies. In all 6 SRs^20,35-38,44 it was unclear if data extraction and critical appraisal were conducted by a single reviewer or multiple reviewers, creating a risk for inaccuracies in these processes.

The generalizability of findings from all 6 SRs^20,35-38,44 to settings in Canada was unclear because of limited reporting on the characteristics of primary study participants (e.g., across PROGRESS-Plus criteria^55,56). Finally, the authors of 1 SR¹¹ did not state their potential conflicts of interest, and the sources of funding for 3 SRs^35,36,44 was unclear.

Nonrandomized Cohort Studies

The authors of all 5 nonrandomized cohort studies (NRS)^9,40-43 provided clear descriptions of their study objectives, outcomes, interventions, comparators, participant eligibility criteria, and main findings. Principal confounders were listed and compared between groups in 4 of the studies.^9,41-43 Additional methodological strengths were that the measurement of outcomes, including FP-related toxicity, were standardized using the CTCAE.¹⁰ However, 1 study⁴⁰ only reported on the CTCAE grade of AE, and not on the specific type of AE. The authors of 4 NRSs^9,41-43 reported estimates of uncertainty (e.g., confidence intervals) and P values. All 5 studies^9,40-43 recruited patients treated in large hospitals, with care providers and treatment pathways representative of the target population. For 1 NRS,⁴² the proportion of patients who were asked to participate in the study and those who agreed was not stated, potentially introducing selection bias and differences between the study sample and source population. In 1 NRS,⁴³ the sampling method was not well described so we were unable to determine whether the study sample was representative of the source population. The authors of all 5 NRSs^9,40-43 declared that they had no potential conflicts of interest related to this work and reported their sources of funding, which were unrelated to industry and considered unlikely to have influenced the study’s findings.

Several factors affected the internal and external validity of the included NRSs. In all 5 NRSs,^9,40-43 subjects were not blinded to the intervention they received, potentially introducing performance or detection bias. In none of the studies was there any indication that the outcome assessors (i.e., those grading the AEs experienced by participants) were blinded to the intervention, which could lead to bias in their severity grading. However, all studies reported standardized grading using a validated tool¹⁰ which could help minimize this bias. Additionally, no mention was made in any of the studies regarding missing data or incomplete records. Missing confounders can introduce selection bias and missing outcomes could result in attrition bias or systematic bias if the missing data patterns are related to the incidence of AEs.

In 2 studies^41,42 different intervention groups were recruited over different periods of time (pre- and post- implementation of standardized DPD testing at the study centres), potentially introducing other variables (e.g., referral patterns, staff changes, treatment guidelines) that might have confounded the incidence of severe toxicities. Due to the nature of the intervention in the included studies (FP dose reductions based on DPD activity), compliance with the intervention was not always reliable. In 1 NRS,⁹ some patients were missing phenotype data and therefore received doses based on other factors. In another study⁴³ some low DPD activity patients did not receive the recommended dose reduction because testing was performed after the start of treatment.

Although the authors reported some relevant baseline participant characteristics (e.g., age, sex, body surface area, comorbidities), many important characteristics that stratify health opportunities and outcomes were not reported, such as race, ethnicity, culture, language, place of residence, socioeconomic status, and other PROGRESS-Plus criteria.^55,56 As a result, it remains unclear whether the findings of these NRSs^9,40-43 can be generalized to settings in Canada. Further, the 2 studies that involved DPYD genotyping^42,43 only evaluated the 4 variants most relevant in populations of European descent, limiting the generalizability of results to more ethnically diverse populations that might carry other DPYD variants affecting DPD activity.

Economic Evaluations

The authors of the economic evaluation⁴⁴ clearly stated their research question, objectives, the economic importance of the research question, the interventions compared, and rationale for conducting the analysis from the perspective of the Ontario Ministry of Health using a 6-month horizon. They provided detailed information on the sources of the effectiveness estimates, utility values, and treatment costs. The authors recorded the currency and price data used and the methods for adjusting prices for inflation, described their approach to sensitivity analyses, reported incremental analyses, provided an answer to the study question, and summarized the findings with conclusions accompanied by appropriate caveats.

The primary strength of the economic evaluation⁴⁴ is its concordance with the decision problem of interest to this report. Given that it was conducted from a health care perspective in Canada, it is more generalizable to other settings in Canada. The model-based evaluation allowed for the comparison of broad patient populations, including numerous cancer types, and different chemotherapy regimens (5-FU, capecitabine, and alternative treatments for DPYD poor metabolizers). Similarly, the choice of a decision-tree model was justified due to the acute nature of chemotherapy treatment and its associated adverse events, which typically resolve within a few months.⁶² The intervention (DPYD genotyping followed by genotype-guided dosing) is conducted already in some provinces across Canada and so is of direct interest. The dose adjustments were guided by an evidence-based guideline.¹⁷

The primary limitation of the economic evaluation is the uncertain generalizability of results among broad ethnic groups. The studies from which the clinical input data on genotype-guided dosing utility were sourced, generally did not report any PROGRESS-Plus criteria,^55,56 except age and sex. Further, the evaluation focused on 4 primary variants that are much less common in other ethnic groups other than those of European descent. Therefore, DPYD variant prevalences and probability data for severe toxicities, hospitalizations, and mortality may have limited generalizability beyond the ethnic groups represented in the studies; a critical limitation given the diverse population of Canada. Conclusions about cost-effectiveness may be different if the intervention included testing for a broader panel of variants that better represented this population. Similarly, health-utility inputs were not drawn from sources in Canada and therefore may not accurately reflect preferences across Canada.

Finally, the study⁴⁴ only looked at the impact of DPYD genotyping, without any analysis of phenotype testing for DPD activity. Therefore, no evidence is available regarding the cost-effectiveness of phenotype testing.

Table 11: Strengths and Limitations of Guidelines Using AGREE II Tool⁶⁰

AGREE II = Appraisal of Guidelines for Research and Evaluation II; AMP = Association for Molecular Pathology; CCO = Cancer Care Ontario; CPIC = Clinical Pharmacogenetics Implementation Consortium; DPWG = Dutch Pharmacogenetics Working Group; HTA = health technology assessment; NA = not applicable; NR = not reported; OH = Ontario Health.

Note: Categories Yes, Partial Yes, and No were determined using the Likert Scale on the AGREE-II tool. Yes (Scores 5 to 7); Partial Yes (Scores 3 to 4); No (Scores 1 to 2).

Table 12: Strengths and Limitations of HTAs and SRs Using AMSTAR 2⁵⁸

AMSTAR 2 = A MeaSurement Tool to Assess Systematic Reviews 2; NA = not applicable; NR = not reported; ROBIS = Risk of Bias in Systematic Reviews; PROSPERO = international prospective register of systematic reviews.

Table 13: Strengths and Limitations of Clinical Studies Using the Downs and Black Checklist⁵⁹

NA = not applicable; NR = not reported.; BSA = body surface area; CTCAE = Common Terminology Criteria for Adverse Events.

Table 14: Strengths and Limitations of Economic Evaluations

CA$ = Canadian dollars; CHEERS = Consolidated Health Economic Evaluation Reporting Standard; DPD = dihydropyrimidine dehydrogenase.

Appendix 5: Literature Review Findings

Please note that this appendix has not been copy-edited.

Table 15: Summary of DPYD Testing Recommendations From Ontario Health, Cancer Care Ontario

DPD = dihydropyrimidine dehydrogenase; EE = economic evaluation; HTA = health technology assessment; MA = meta-analysis; NR = not reported; NRS = nonrandomized study; SR = systematic review.

^aGenotype-guided dosing recommendations for patients are adapted from the 2017 CPIC Guidelines and Supplementary Tables¹⁷ which are subject to updates and modifications. Refer to Table 16.

Table 16: Summary of Genotype-Guided Dosing Recommendations in Included Guidelines

CPIC = Clinical Pharmacogenetics Implementation Consortium; DPD = dihydropyrimidine dehydrogenase; DPWG = Dutch Pharmacogenetics Working Group; FU = fluorouracil; GAS = gene activity score; MA = meta-analysis; NRS = nonrandomized study; NR = not reported; SR = systematic review; TDM = therapeutic data monitoring.

^aPatients are at risk of severe or even fatal drug toxicity when treated with FP drugs.

^bDPD enzyme activity cannot be predicted correctly, an additional phenotyping test is required to determine the DPD enzyme activity.

Table 17: Clinical Validity Findings by Outcome — Severe (Grade ≥ 3) Toxicity

CI = confidence interval; EMA = European Medicines Agency; HFS = hand and foot syndrome; NA = not applicable; NR = not reported; OR = odds ratio; RR = risk ratio; U = uracil; U/UH₂ = uracil to dihydrouracil.

^aAlso known as palmar-plantar erythrodysesthesia.

Table 18: Clinical Validity Findings by Outcome — Sensitivity, Specificity, PPV, NPV of DPYD Genotyping (3 to 4 variants) to Detect Severe Toxicity

CI = confidence interval; NA = not applicable; NR = not reported; NPV = negative predictive value; PPV = positive predictive value.

Note: Severe FP-related toxicity was used as the reference standard to calculate sensitivity, specificity, PPV, and NPV (i.e., if toxicity occurred in a DPYD variant carrier, this was considered a true positive; if toxicity occurred in a patient with the wild-type gene, this was considered a false-negative).

^a1 study calculated and reported sensitivity, specificity, PPV, and NPV values. For the remaining 8 studies, outcomes were calculated by the Ontario Health (2021)⁴⁴ authors for each included study based on data presented within the study. Results are summarized here using median and min-max.

Table 19: Clinical Validity Findings by Outcome — Other Patient-Related Outcomes

CI = confidence interval; NR = not reported; OR = odds ratio.

Table 20: Clinical Utility Findings by Outcome — Severe (Grade ≥ 3) Toxicity

NA = not applicable; NR = not reported.; PGD = pharmacogenetic-guided dosing; RR = relative risk.

^aThese results refer to the incidence of toxicity of any grade (1 to 4).

^bTwenty-six (28%) of these patients received an initial FP dose reduction based on factors other than DPD activity (i.e., fragile baseline condition).

Table 21: Clinical Utility Findings by Outcome — Other Patient-Related Outcomes

NA = not applicable; NR = not reported; RR = risk ratio, CI = confidence interval.

Table 22: Summary of Findings of Included Economic Evaluations

CrI = Credible interval; ICER = incremental cost-effectiveness ratio; QALY = quality-adjusted life-year.

Table 23: Summary of Excluded Economic Evaluations

CA$ = Canadian Dollar; £ = euro; £= British Pound; GLM = generalized linear model; NR = not reported; US$ = US Dollar; QAR = Qatari Riyal; QALY = quality-adjusted life-year; SOC = standard of care.

Appendix 6: Overlap Between Included Systematic Reviews

Please note that this appendix has not been copy-edited.

Table 24: Overlap in Relevant Primary Studies Between Included Systematic Reviews