Rare genetic variants in DPYP increase toxicity and screening for them prevents serious complications by upfront reduction in 5FU dose; however, most patients with severe toxicities do not have a rare mutation. We have previously shown that 2 common DPYD variants were associated with toxicity in patients with advanced colorectal cancer treated on COIN & COIN-B (abstract 3509, ASCO 2013): Cys29Arg [rs1801265] (Minor Allele Frequency (MAF) 0.21) and Val732Ile [rs1801160] (MAF 0.04). We have now genotyped 4 rare variants using the same cohort.
Blood samples were available from 2183 patients treated with first line oxaliplatin-5FU ± cetuximab. We assayed IVS14 + 1G>A [rs3918290], Asp949Val [rs67376798], Lys259Glu [rs45589337] and Ser534Asn [rs1801158] using KASPar. Primary endpoint was dose reduction or delay in chemotherapy in the first 12 weeks of treatment due to any toxicity except neuropathy. Secondary endpoints were grade ≥2 versus grade <2 for neutropenia, lethargy, Nausea & Vomiting (N&V), diarrhoea, stomatitis, Hand-Foot Syndrome (HFS) and infection.
Two rare variants were associated with toxicity (OR (95% CI)): Asp949Val with neutropenia 3.2 (1.2-8.2) P = 0.019, N&V 3.4 (1.5-7.3) P = 0.002, diarrhoea 4.6 (2.1-10.1) P < 0.001 and infection 5.5 (1.3-24.2) P = 0.024; IVS14 + 1G>A with lethargy 5.3 (1.9-14.9) P = 0.002, diarrhoea 4.4 (1.7-11.0) P = 0.002, stomatitis 4.6 (1.7-12.6) P = 0.003, HFS 3.8 (1.2- 11.8) P = 0.021 and infection 19.2 (5.0-73.8) P < 0.001. MAF was 0.007 and 0.005, respectively. The effect on toxicity for our 2 common variants was not as marked (OR (95% CI)): Cys29Arg 0.8 (0.7-1.0) P = 0.008 (protective) and Val732Ile 1.6 (1.1-2.1) P = 0.006 for the primary endpoint.
We have validated 2 mutations, Asp949Val and IVS14 + 1G>A, as predictors for 5FU toxicity in a large cohort of patients and recommend they should be screened for. Our data suggest that common DPYD variants are also associated with toxicity but not to the same level seen with rare ones. While the presence of a single common variant is not an indication for dose modification, the presence of multiple variants in a patient might be. Further work is needed to establish what combinations of common DPYD variants would necessitate 5FU dose alteration.
Clinical trial identification
Legal entity responsible for the study
Cardiff University and MRC CTU.
This work was supported by The Bobby Moore Fund from CRUK, Cancer Research Wales, Tenovus, the Wales Gene Park, and an unrestricted research grant from Merck Serono.
T.S. Maughan, J.P. Cheadle: Part funded by an unrestricted research grant from Merck Serono. All other authors have declared no conflicts of interest.