Impact of pharmacogenomic DPYD variant guided dosing on toxicity in patients receiving fluoropyrimidines for gastrointestinal cancers in a high-volume tertiary centre

Abstract

Background Dihydropyrimidine dehydrogenase (DPD) is a key enzyme in the metabolism of fluoropyrimidines. Variations in the encoding DPYD gene are associated with severe fluoropyrimidine toxicity and up-front dose reductions are recommended. We conducted a retrospective study to evaluate the impact of implementing DPYD variant testing for patients with gastrointestinal cancers in routine clinical practice in a high volume cancer centre in London, United Kingdom. Methods Patients receiving fluoropyrimidine chemotherapy for gastrointestinal cancer prior to, and following the implementation of DPYD testing were identified retrospectively. After November 2018, patients were tested for DPYD variants c.1905+1G>A (DPYD*2A), c.2846A>T (DPYD rs67376798), c.1679T>G (DPYD*13), c.1236G>A (DPYD rs56038477), c.1601G>A (DPYD*4) prior to commencing fluoropyrimidines alone or in combination with other cytotoxics and/or radiotherapy. Patients with a DPYD heterozygous variant received an initial dose reduction of 25–50%. Toxicity by CTCAE v4.03 criteria was compared between DPYD heterozygous variant and wild type carriers. Results Between 1st December 2018 and 31st July 2019, 370 patients who were fluoropyrimidine naïve underwent a DPYD genotyping test prior to receiving a capecitabine (n = 236, 63.8%) or 5FU (n = 134, 36.2%) containing chemotherapy regimen. 33 patients (8.8%) were heterozygous DPYD variant carriers and 337 (91.2%) were wild type. The most prevalent variants were c.1601G > A (n = 16) and c.1236G > A (n = 9). Mean relative dose intensity for the first dose was 54.2% (range 37.5–75%) for DPYD heterozygous carriers and 93.2% (42.9–100%) for DPYD wild type carriers. Overall grade 3 or worse toxicity was similar in DPYD variant carriers (4/33, 12.1%) as compared to wild-type carriers (89/337, 25.7%; P = 0.0924). Conclusions Our study demonstrates successful routine DPYD mutation testing prior to the initiation of fluoropyrimidine chemotherapy with high uptake. In patients with DPYD heterozygous variants with pre-emptive dose reductions, high incidence of severe toxicity was not observed. Our data supports routine DPYD genotype testing prior to commencement of fluoropyrimidine chemotherapy.

Implementation of upfront DPYD genotyping with a low-cost and high-throughput assay to guide fluoropyrimidine treatment in cancer patients

“…This is in accordance with previous studies in which the combined frequency of the four variants explained about 12-30% of fluoropyrimidine toxicity [4,20,34]. Regarding the pretreatment group, 2.5% of the patients presented DPYD gene variants, a value somewhat inferior to that described in other European populations, in which the four variants combined were detected in about 5-10% of the patients in a pretherapeutic screening setting [20,29,31,32]. The frequency of DPYD variants can differ among populations, as previously described for the c.1905 + 1G>A variant, for which the incidence among European populations ranges from 0.7-3.5% [4,10,17,20,26,29,[31][32][33][35][36][37].…”

“…Regarding the pretreatment group, 2.5% of the patients presented DPYD gene variants, a value somewhat inferior to that described in other European populations, in which the four variants combined were detected in about 5-10% of the patients in a pretherapeutic screening setting [20,29,31,32]. The frequency of DPYD variants can differ among populations, as previously described for the c.1905 + 1G>A variant, for which the incidence among European populations ranges from 0.7-3.5% [4,10,17,20,26,29,[31][32][33][35][36][37]. Likewise, the frequency of the c.2846A>T and c.1129-5923C>G variants also varies between populations, ranging from 0.4-3% to 2.4-6.2% in European populations, respectively [4,10,17,20,29,[31][32][33].…”

“…The frequency of DPYD variants can differ among populations, as previously described for the c.1905 + 1G>A variant, for which the incidence among European populations ranges from 0.7-3.5% [4,10,17,20,26,29,[31][32][33][35][36][37]. Likewise, the frequency of the c.2846A>T and c.1129-5923C>G variants also varies between populations, ranging from 0.4-3% to 2.4-6.2% in European populations, respectively [4,10,17,20,29,[31][32][33]. The c.1679T>G variant is rarer, being reported in frequencies of 0.1 to less than 1% [4,10,17,20,29,[31][32][33].…”

“…Several studies demonstrated the clinical validity of genotyping the four most common and well-established variants associated with fluoropyrimidine toxicity, namely the c.1129-5923C>G, the c.1679T>G, the c.1905 + 1G>A, and the c.2846A>T DPYD variants [4,16,20,29,[31][32][33]. However, these four DPYD variants correlate differently with DPD enzyme activity; for instance, the c.1129-5923C>G variant appears to present a milder impact on DPD activity, and the available data are unclear regarding the appropriate dose reduction for carriers [15,17,18,20].…”

“…Several authors suggested that upfront genotyping of DPYD variants followed by individualized dose adjustment would improve the safety of fluoropyrimidine therapy for patients and reduce overall treatment costs [17,20,[24][25][26][27][28][29]. On 13 March 2020, the European Medicines Agency's (EMA's) Pharmacovigilance Risk Assessment Committee has recommended that patients receiving fluorouracil given by injection or infusion and the related medicines capecitabine and tegafur should be tested for the lack of DPD enzyme, either by measuring the level of uracil in the blood or DPYD gene variant analysis, before starting treatment.…”