Characterization of DPYD alterations and treatment-related toxicity in a Chilean cohort of cancer patients.

Abstract

850 Background: Fluoropyrimidines are widely used in the treatment of solid tumors. Severe treatment-related toxicities occur in 10–30% of patients, with dihydropyrimidine dehydrogenase (DPYD) gene variants recognized as major pharmacogenetic determinants. Although the prevalence of clinically relevant DPYD alterations is well documented in European and North American cohorts, data from Latin America remain scarce. We investigated the prevalence of DPYD variants in a Chilean cohort of cancer patients and evaluated their correlation with chemotherapy tolerability. Methods: We retrospectively analyzed 290 Chilean cancer patients who underwent next-generation sequencing (NGS) with the Oncomine Comprehensive Assay Plus for predictive purposes. Among these, 199 individuals with at least one DPYD alteration were included. Because the assay reports somatic variants, DPYD alterations were considered germline if the allelic frequency (AF) was ≥40%, a threshold associated with ~94% probability of being germline. Clinical data, including toxicity outcomes and therapy discontinuation, were obtained from chart review. Statistical associations were assessed using Chi-square or Fisher’s exact test, and logistic regression was applied to evaluate the effect of the number of variants on outcomes. Results: Overall, 167 of 199 patients (84.0%) harbored at least one DPYD variant with AF ≥40%. The most prevalent variants were c.1627A>G (32.4% of 290), c.496A>G (12.1%), c.1896T>C (9.7%), and c.2194G>A (8.3%). Among the 65 patients treated with fluoropyrimidines, 56 carried ≥1 variant with AF ≥40%. Across this subset, grade ≥3 toxicities occurred in 8.0%, early dose adjustments within four cycles in 8.5%, and treatment discontinuations in 4.5%. No single variant or group of variants showed statistically significant associations with these outcomes (all p > 0.05). Logistic regression suggested a non-significant trend towards increased risk of severe toxicity with multiple variants (OR per variant 1.44, 95% CI 0.59–3.52, p=0.42). Notably, one patient harbored DPYD 9B in heterozygosis and received fluorouracil without major toxicity, while another carried DPYD 2A but was not treated with fluoropyrimidines. No patients presented with DPYD 13, DPYD p.Y186C, or the HapB3 collection of SNPs, yielding an estimated frequency of 0.68% for these high-risk variants. Conclusions: This study represents the first characterization of DPYD variants in a Chilean cancer cohort. While the prevalence of classical “drug response” variants was high, no significant correlation with fluoropyrimidine-related toxicities was observed. These findings highlight the importance of larger Latin American cohorts to refine regional guidelines for DPYD testing and dose optimization, while contributing novel insights into the pharmacogenetic landscape of DPYD in Chile.