Association Between DPYD Gene Variants and Severe Fluoropyrimidine-Related Toxicity in Asian and African Populations: A Systematic Review.

DPYD

DPYD (dihydropyrimidine dehydrogenase) variants associated with fluoropyrimidine toxicity, with recurrent variants specific to Asian and African populations

This systematic review, conducted according to PRISMA 2020, assesses the association between DPYD variants and severe fluoropyrimidine toxicity in Asian and African populations, where the predictive value of variants prioritized by predominantly European guidelines remains unclear. PubMed, Embase, Scopus, the Cochrane Library and Google Scholar were searched; 23 studies published between 2007 and 2025 across 14 countries were included, six at low risk of bias, nine at moderate and eight at high risk. Recurrent variants were c.1627A>G (DPYD*5), c.85T>C (DPYD*9A), c.1905+1G>A (DPYD*2A), c.496A>G, c.1896T>C and c.2194G>A (DPYD*6); associations were most consistent for c.85T>C (unadjusted odds ratios of 2.3 for hand-foot syndrome and 2.7 for diarrhea in South India, adjusted odds ratio of 4.35 for mucositis in Tunisia, all for any-grade toxicity) and for c.1905+1G>A in Bangladesh (adjusted odds ratios of 6.47 for grade 3-4 neutropenia and 5.76 for diarrhea). Only one study corrected for multiple testing, and African data, confined to Tunisia, Algeria, Sudan and Zimbabwe, are too sparse for strong conclusions. In Zimbabwe, no carrier of the four guideline variants was found among 150 genotyped patients, yet 36% of the 50 who received a fluoropyrimidine had severe adverse events.

Synthesis written by Geno'X. For the full original abstract, please refer to the source publication.

The Zimbabwean finding sums up the issue: genotyping restricted to the four guideline DPYD variants can come back negative in patients who nonetheless develop severe toxicity. The authors are rightly cautious — heterogeneous regimens, grading and platforms, a single correction for multiple testing — and no population-specific variant is actionable as it stands. For a laboratory, this argues for reading the whole DPYD gene, already captured by an exome or a genome, rather than four positions in patients of non-European ancestry.

Analysis by Dr Thibaut Benquey