DPYD testing helps identify some people at increased risk of severe toxicity from fluoropyrimidine cancer medicines, including fluorouracil (5-FU) and capecitabine. It is a treatment-safety test, not a test for whether someone has cancer or whether their tumour will respond. A useful result must reach the oncology team in time to inform the treatment plan.

DPYD is the gene; DPD is the enzyme

DPYD provides instructions for dihydropyrimidine dehydrogenase, abbreviated DPD. This enzyme breaks down most administered fluorouracil. Some variants reduce its activity, increasing exposure to the medicine and the risk of serious adverse effects. Complete and partial deficiency are different clinical situations. The FDA warns about potentially fatal toxicity and the need to discuss testing and recognise unusually early or severe reactions. FDA safety communication.

A report saying “a DPYD variant was detected” is incomplete without the exact variant, its functional interpretation and the laboratory’s method. Conversely, “no tested variants detected” does not mean that every cause of reduced DPD activity has been excluded.

Genotype and phenotype tests answer related questions

ApproachWhat it measuresWhat to check
Targeted DPYD genotypingA specified set of DNA variantsExactly which variants were included; relevant variants outside that set may be missed
Broader sequence analysisMore of the gene’s sequenceCoverage and whether a rare finding has an established functional interpretation
DPD phenotypingA biochemical proxy or enzyme-activity measurementMethod, specimen handling and clinical factors affecting the measurement

EMA describes pretreatment testing using selected genetic variants or blood uracil, a substance metabolised by DPD. These approaches should not be treated as identical measurements. A biochemical result can add information, but the interpretation needs the assay’s validated context. EMA testing recommendations.

DNA variant testing and biochemical phenotyping contribute different information to an oncology treatment decision. A negative limited panel does not eliminate toxicity risk.
DNA variant testing and biochemical phenotyping contribute different information to an oncology treatment decision. A negative limited panel does not eliminate toxicity risk.

Open full-size illustration (new tab)

Read the report before looking up a dose

Pharmacogenetic guidelines connect established variants to predicted enzyme function and treatment recommendations. An activity score is a structured interpretation of the alleles; it is not a directly measured percentage of enzyme activity. The CPIC and DPWG guidance also considers subsequent titration and monitoring. A patient should not use a score or a web calculator to alter chemotherapy independently. The regimen, organ function, other medicines and treatment goals still matter. CPIC guideline; DPWG guideline.

Recommendations and product labels evolve. For example, the cited 2026 US fluorouracil label directs testing before treatment unless immediate treatment is necessary. The applicable product label and the oncology service’s current protocol should be checked, rather than relying on an older article’s wording. Fluorouracil prescribing information.

Three practical examples

Fictional example: a negative limited panel. Neda’s report lists four variants and says none was detected. Her question is whether chemotherapy can now be considered risk-free. The clinician explains that this removes those particular findings from the assessment, but does not rule out other variants or non-genetic causes of toxicity. She still receives the usual monitoring and a clear urgent-contact plan.

Fictional example: a reduced-function result arrives before treatment. Martin’s report predicts reduced DPD function. The oncology pharmacist checks the exact allele assignment and the intended regimen, documents the relevant guideline, and discusses an adjusted starting strategy or an alternative with the oncologist. The decision is recorded before the first dose; it is not left as an unread attachment in the electronic record.

Fictional example: an elevated uracil result in kidney disease. A patient with impaired kidney function has elevated plasma uracil. The team reviews whether the result accurately reflects DPD function before assigning a deficiency category. Regulators have specifically cautioned that moderate or severe renal impairment can complicate uracil-based phenotyping. 2024 healthcare-professional communication.

A practical checklist for the treatment discussion

  1. Confirm the medicine and whether the result will be available before treatment begins.
  2. Record the exact test method and variant coverage, not just “DPD tested.”
  3. Ask who will interpret the finding and enter the prescribing decision.
  4. Check whether another result conflicts with the genotype or predicted function.
  5. Agree on monitoring and which symptoms require immediate contact.
  6. Keep a copy of the report for future treatment decisions.

Severe mouth inflammation, diarrhoea, neurological changes or other unusually early or severe symptoms during treatment require prompt contact with the oncology service. A previous negative genetic result should not delay assessment. This guide explains testing; it does not provide individual dosing instructions.

For the broader framework, read genes and medicines and what a genetic test can miss. Researchers can use the paper-comparison workspace to compare the tested variants, ancestry representation, regimen and toxicity outcomes across studies.