Unrecognized mutations in DPYD* 2 A wild-type rectal cancer patients receiving postoperative 5-FU-based chemotherapy - do they have a clinical impact?

Purpose: The impact of the unrecognized mutational dihydropyrimidine-dehydrogenase-gene-(DPYD)-status on high-grade CTC-AE-grades ≥ 3 (NCI-Common Terminology Criteria for Adverse Events, vs. 3.0) was assessed in patients with upper rectal cancer (inferior tumor margin ≥ 12 cm above the anal verge) treated with upfront surgery and 5-Fluorouracil (5-FU) based adjuvant chemotherapy (CTx).

Methods: 75 participants of the GAST-05-phase-IIb-trial (ISRCTN35198481) were tested in this single center analysis for DPYD*2A-wildtype (WT) at staging. After surgery, 43 patients (stages II and III, according to the current 8th TNM/UICC-classification, 2017) received FOLFOX-CTx and entered follow-up (median: 101 months). According to recent recommendations of the European Medicines Agency (EMA) and national guidelines, post-hoc genotyping for DPYD*2A (c.1905 + 1G > A; IVS14 + 1G > A; rs3918290), DPYD*13 (c.1679T > G; rs55886062), polymorphism c.2846 A > T (rs67376798) and Haplotype B3 (HapB3) (c.1236G > A; c.1129-5923 C > G) was performed using cryopreserved blood samples and standardized PCR-techniques.

Results: Five patients were found to have a heterozygous (het_) DPYD-HapB3-status. Across all patients, the adherence to CTx-cycles 1 to 4 was 100%, 97.7%, 95.3%, and 93.0%, respectively. Grade ≥ 3 CTC-AEs were observed in 0.9% of both het_HapB3- and WT-patients. The mean administered dose of 5-FU was 68.8% of the target in DPYD-HapB3 carriers, compared to 92.6% in 38 WT patients. Logistic regression analysis revealed that 5-FU dose reductions were significantly associated with DPYD-HapB3 carrier status (odds ratio [OR] 12.55, p = 0.044) and male sex (OR 0.23, p = 0.049). During follow-up het_HapB3-patients had a recurrence rate of 60.0%, compared to 13,6% for WT-patients. The disease-free survival (DFS) for het_HapB3-patients was significantly reduced vs. WT (p = 0.010). Multivariable analysis showed that het_HapB3-patients had an increased risk for reduced DFS (HR 3.774; p = 0.057). Interestingly, 5-FU dose reductions per se were not significantly associated with limited DFS in the total population.

Conclusion: DPYD genotyping revealed a het_HapB3 variant in 11.6% of DPYD*2A-WT patients treated with FOLFOX. While not linked to increased toxicity, HapB3 status was associated with reduced DFS, suggesting an impact on treatment efficacy. These results support DPYD genotyping and highlight the need for adequate 5-FU plasma level assessment followed by subtile dose escalation (therapeutic drug monitoring) to personalize 5-FU dosing more precisely, safely and most effective.