Pre-chemotherapy DPYD genetic screening in England now covers a variant that occurs at elevated frequency in patients of African ancestry, closing a gap that left those patients exposed to potentially fatal chemotherapy toxicities for years after mandatory testing was introduced.

The variant in question sits within the DPYD gene, which encodes dihydropyrimidine dehydrogenase, the enzyme that breaks down fluoropyrimidine-class chemotherapy drugs. When the enzyme doesn’t work properly, drugs like 5-fluorouracil, capecitabine, and tegafur accumulate in the body rather than clear. That accumulation can produce grade 3 or grade 4 toxic events affecting the bone marrow, the bowel, and the nervous system. Deaths have occurred. The clinical stakes aren’t theoretical.

Fluoropyrimidines aren’t a niche drug class. They’re prescribed routinely across colorectal, breast, head and neck, and gastric cancers, making them among the most common chemotherapy agents in oncology practice. Without screening, a clinician prescribing a standard fluoropyrimidine dose to a patient with deficient DPD activity is operating without information that could determine whether that dose is tolerable or lethal.

The Medicines and Healthcare Products Regulatory Agency mandated DPYD genetic testing before fluoropyrimidine prescribing in October 2020. The NHS Genomic Medicine Service implemented that requirement the following month. What the rollout didn’t resolve, though, was a structural problem embedded in the test design itself: the screening panels had been built around genetic variants identified primarily in populations of European ancestry.

Roughly 3% to 5% of people with European ancestry carry a low-functioning DPD enzyme variant detectable by the original panel. That figure is reasonably well-characterized. What’s been less well-characterized is the variant burden in patients of African and Caribbean descent, partly because the research datasets underpinning pharmacogenomic guidelines have historically skewed toward European populations. Patients of African ancestry weren’t underrepresented in oncology wards; their genetic variants were underrepresented in the databases used to design clinical tools.

The expanded panel corrects one specific element of that disparity. It incorporates the variant designated c.557A>G, which appears at higher frequencies in individuals of African descent but wasn’t captured in the original NHS screening protocol. That variant is now included across all seven genetic hubs operating within the NHS Genomic Medicine Service in England.

The BMJ coverage of the expanded testing program documented the scope of the change, and the methodological rationale aligns with the Current DPYD testing guidance from the European Medicines Agency, which has long acknowledged population-specific pharmacokinetic variation as a core consideration in drug safety assessment.

Dr. Bill Newman, a consultant in genomic medicine at Manchester University NHS Foundation Trust, has been directly involved in the protocol expansion. “This change means that patients from African and Caribbean backgrounds who would previously have been missed by the test can now be identified before they receive chemotherapy,” said Dr. Newman. “It’s a significant step toward making precision medicine work for everyone, not just those whose genetic variants happen to be well-studied.”

That framing deserves some unpacking. Precision medicine’s clinical value depends entirely on the quality and representativeness of the underlying evidence base. A test calibrated on one population and applied universally doesn’t deliver precision; it delivers false reassurance for patients whose biology wasn’t modeled. The c.557A>G variant isn’t a curiosity. It’s a clinically actionable finding that was effectively invisible to the standard protocol.

This problem isn’t unique to DPYD. Pharmacogenomic research has documented similar population-stratified gaps across drug-metabolizing enzyme systems including CYP2C19, TPMT, and G6PD-related pathways. The ACC/AHA framework for cardiovascular pharmacogenomics has wrestled with comparable questions about variant representation in guideline-informing datasets. That’s worth noting here because it signals a systemic pattern, not an isolated oversight in oncology.

The 2022 period saw several international pharmacogenomics bodies begin formalizing language around population diversity in variant classification. The England update builds on that trajectory. It’s a targeted correction, not a wholesale overhaul, but targeted corrections compound over time.

From a study design standpoint, the evidence supporting c.557A>G as a clinically relevant variant derives from observational data and population-frequency analyses rather than a dedicated randomized controlled trial, which wouldn’t be ethically feasible given what’s already known about DPD deficiency and toxicity outcomes. Clinicians should understand that distinction. The mechanistic rationale is solid: reduced DPD enzyme activity from this variant produces the same downstream pharmacokinetic consequences as the European-predominant variants already covered by the panel. The 4 variants that were included in the original NHS panel each met a similar evidence threshold through comparable methodologies.

What changes practically for oncology teams? Patients of African and Caribbean ancestry presenting for fluoropyrimidine-based treatment will now be screened against a more complete genetic reference. That doesn’t mean the expanded panel captures every clinically relevant DPYD variant across all ancestry groups; it doesn’t. There are at least 25 known DPYD variants associated with reduced enzyme function, and the panel’s current scope remains selective. The c.557A>G addition is step 3 in what will likely need to be an iterative process as pharmacogenomic databases become more ancestrally diverse.

The 50 or so deaths attributed annually in England to fluoropyrimidine toxicity prior to mandatory pre-screening represent a sobering reference point. The 2020 mandate was projected to prevent a substantial fraction of those events. Whether the addition of c.557A>G will produce a statistically detectable shift in toxicity rates among patients of African ancestry specifically is a question that can’t be answered from current data but merits prospective tracking.

Manchester University NHS Foundation Trust’s involvement in developing the expanded protocol reflects a broader pattern of academic medical centers driving pharmacogenomic equity work that regulatory bodies then formalize. That collaboration structure has strengths: it brings clinical expertise and real-world patient data into the protocol design process. It also has limitations; the variants prioritized for inclusion can still reflect the research networks and patient cohorts available to the academic teams involved.

Without a more representative genomic reference database, panels will always lag behind the full spectrum of clinically relevant variation. This is why the NHS Genomic Medicine Service has invested in initiatives to diversify its genomic cohorts. The c.557A>G expansion is both a product of that effort and an argument for accelerating it.

The 80 percent figure cited in some pharmacogenomic literature represents the approximate proportion of DPYD toxicity cases potentially attributable to known variants in predominantly European-descent populations. That figure doesn’t translate cleanly to patients of African ancestry using the original 4-variant panel, which is precisely what made the gap consequential.

Pre-screening doesn’t eliminate fluoropyrimidine toxicity risk. It identifies patients who need dose modification or an alternative regimen before that risk materializes. For patients whose variant was previously undetected, that’s the difference between a manageable clinical decision and an avoidable crisis.