Deficit dihydropyrimidine dehydrogenase and its association with 5-fluorouracil toxicity

Potential toxicity of 5-FU

5-fluorouracil is a cytostatic drug used, for example, in the treatment of pancreatic, breast, or colorectal cancer. One of the risk factors for severe toxicity is a deficit of the enzyme dihydropyrimidine dehydrogenase, coded by the DPYD gene. This enzyme is crucial for the catabolism of 5-FU to inactive metabolites. Therefore, the aim is to identify patients with a DPD deficit before starting therapy, individually adjust the 5-FU dose, and thus reduce the risk of severe adverse effects related to the treatment.¹

Clinical studies have identified 4 of the most common variants of the DPYD gene (found in 2–8% of the European and North American populations)² associated with an increased risk of severe 5-FU toxicity:

• c.1905+1G>A (DPYD*2A, IVS14+1G>A, rs3918290)
• c.1679T>G (DPYD*13, p.I560S, rs55886062)
• c.2846A>T (p.D949V, rs67376798)
• c.1129-5923C>G (rs75017182).

It is estimated that carriers of these alleles have a 1.6–4.4× higher risk of severe toxicity³ and a > 25× higher risk of lethal toxicity.²

Current methods for identifying patients with DPD deficit

The most widespread and practically used method is genotyping to detect the pathogenic variants of the gene mentioned above.

In research, methods such as the enzymatic DPD test in peripheral blood mononuclear cells (PBMC – peripheral blood mononuclear cells), which correlates with liver production, are utilized. Another method is determining the level of uracil and dihydrouracil (UH₂) in peripheral blood, where in the case of DPD deficit, the catabolism of uracil to UH₂ is reduced. However, clinical studies have shown that catabolite levels cannot be reliably interpreted. The breath test, in which the patient drinks an aqueous solution of 2-¹³C-uracil, and the level of ¹³CO₂ in exhaled air is measured proportionally to DPD activity, is highly expensive and thus not used in routine practice. A combination of these two methods is the oral loading test.

Pathogenic DPYD gene variants and 5-FU treatment-related mortality – meta-analysis results

A large meta-analysis² included a total of 35 studies with 13,929 patients treated with 5-FU (of which 4.1% had an identified pathogenic variant of the DPYD gene). The primary monitored parameter was the risk of death associated with cytostatic therapy. Death occurred in 14 out of 13,363 patients (0.1%) with a wild-type (non-mutated) variant of the DPYD gene (95% confidence interval [CI] 0.1–0.2) and in 13 out of 566 patients (2.3%) with the pathogenic variant of the gene (95% CI 1.3–3.9%). Thus, carriers of the pathogenic gene variant had a 25.6× higher risk of treatment-related death (95% CI 12.1–53.9; p < 0.001). After analyzing the subgroup of pathogenic variant carriers by excluding patients with the less risky gene variant c.1129-5923C>G, the mortality was 3.7%.

Individualization of 5-FU dosing based on DPYD genotyping

A Dutch prospective multicentric analysis⁴ from 2018 included 1,103 patients treated with 5-FU for malignant tumor disease from April 2015 to December 2017. All patients underwent DPYD genotyping. A pathogenic allele was identified in 85 of them (8%), while the remaining 1,018 patients (92%) were wild-type homozygotes. In carriers of the pathogenic allele, the initial 5-FU dose was reduced by 25% (variants c.2846A>T and c.1236G>A) or by 50% (DPYD*2A and c.1679T>G). The primary monitored parameter was the incidence of severe (i.e., grade ≥ 3) adverse effects (toxicity) related to treatment. This was higher in carriers of the pathogenic DPYD gene variant (n = 33; 39%) compared to patients with the wild-type allele (n = 231; 23%; p = 0.0013). The relative risk (RR) of severe toxicity in patients with the DPYD*2A variant, with dose modification based on genotyping, was 1.31 vs. 2.87 compared to the historical control cohort without dose modification. For c.1679T>G carriers, zero toxicity was achieved after dose adjustment vs. historical RR 4.30; in the c.2846A>T group, RR 2.00 vs. 3.11 was recorded, and in the c.1236G>A cohort, RR 1.69 vs. 1.72.

A recent Danish study⁵ included a total of 230 cancer patients, mostly with gastrointestinal cancer. Prior to the administration of the first 5-FU dose, DPYD genotyping was performed, and blood samples were collected for post hoc uracil level analysis. In carriers of the pathogenic DPYD gene variant, the initial 5-FU dose was reduced. In this cohort, severe adverse effects occurred in 27% of patients, compared to 24% in the control group, which was not a significant difference in incidence. However, in patients with the pathogenic DPYD gene variant and dose adjustment, the number of deaths (0 vs. 4.8% in the control cohort) and hospitalizations (0 vs. 19%) significantly decreased. In the wild-type allele group with a serum uracil level ≥ 16 ng/ml, a higher incidence of severe toxicity was observed compared to patients with a uracil level < 16 ng/ml (55 vs. 28%).

Conclusion

According to current knowledge, DPD deficiency testing is recommended before starting 5-FU therapy. Pathogenic DPYD gene variants correlate with a higher risk of severe toxicity and treatment-related death. Reducing the 5-FU dose in carriers of these pathogenic alleles significantly reduces the risk of severe toxicity, death, and hospitalization.

(mafi)

Sources:
1. Diasio R. B., Offer S. M. Testing for dihydropyrimidine dehydrogenase deficiency to individualize 5-fluorouracil therapy. Cancers (Basel) 2022; 14 (13): 3207, doi: 10.3390/cancers14133207.
2. Sharma B. B., Rai K., Blunt H. et al. Pathogenic DPYD variants and treatment-related mortality in patients receiving fluoropyrimidine chemotherapy: a systematic review and meta-analysis. Oncologist 2021; 26 (12): 1008–1016, doi: 10.1002/onco.13967.
3. Meulendijks D., Henricks L. M., Sonke G. S. et al. Clinical relevance of DPYD variants c.1679T>G, c.1236G>A/HapB3, and c.1601G>A as predictors of severe fluoropyrimidine-associated toxicity: a systematic review and meta-analysis of individual patient data. Lancet Oncol 2015; 16 (16): 1639–1650, doi: 10.1016/S1470-2045(15)00286-7.
4. Henricks L. M., Lunenburg C. A. T. C., de Man F. M. et al. DPYD genotype-guided dose individualisation of fluoropyrimidine therapy in patients with cancer: a prospective safety analysis. Lancet Oncol 2018; 19 (11): 1459–1467, doi: 10.1016/S1470-2045(18)30686-7.
5. Paulsen N. H., Pfeiffer P., Ewertz M. et al. Implementation and clinical benefit of DPYD genotyping in a Danish cancer population. ESMO Open 2023; 8 (1): 100782, doi: 10.1016/j.esmoop.2023.100782.