Prospective DPYD testing in colorectal cancer patients in a realworld UK population

Abstract

Background: Polymorphisms within the DPYD gene are present in 7% of Europeans but account for 23% of life-threatening toxicity from fluoropyrimidine (FP) chemotherapy. Four DPYD variants have an adjusted relative risk for toxicity of 1.59 - 4.4. Upfront genotyping is safe and cost effective but not mandated by ESMO guidelines. To reduce the risk of life-threatening toxicity we implemented prospective DPYD testing as standard practice. Methods: Consecutive colorectal cancer (CRC) patients in a UK cancer centre due to receive FP chemotherapy were genotyped by real time PCR for known clinically relevant DPYD mutations: c.1905+G>A 2∗, c.2846A>T, c.1679T>G and c.1605 G>A and from March 2017, c.1236G>A/HapB. We followed published recommendations for dose reduction or alternative drug. Demographics, dose, toxicity and survival data were collected. Results: Between 1/1/16-31/12/2017, 230 patients were tested. 72% had capecitabine, 24% 5-fluorouracil, and 4% raltitrexed combinations. After dose reduction or alternative therapy, grade 3/4 diarrhoea was similar in wildtype and mutations (10 vs 13%) and any toxicity admissions were not significantly different (p=0.284). There were no treatment deaths. Conclusions: To our knowledge, we are the only UK centre to implement prospective DPYD testing in routine clinical practice for CRC patients. In published data of unselected CRC patients the G3/4 GI toxicity is 15%, but if 2∗ variant is present this increases to 73%. In the latter population, genotype guided dosing reduces the risk to 28%. Our rate for all variants was 10% but limited by small numbers. Pharmacokinetics in another study showed adequate 5FU exposure with a 50% dose reduction, alleviating underdosing concerns. The growing evidence supports prospective DPYD testing. We have shown it is practical and may mitigate serious toxicities.