Implementing DPYD/UGT1A1testing to prevent treatment related adverse events (TRAE) in patients with gastrointestinal (GI) cancer: Results of a pragmatic implementation trial.

Abstract

10599 Background: Fluoropyrimidines (FP), 5-FU and capecitabine, are standard treatments for many GI cancers. Irinotecan may be used with FPs in both the curative and palliative settings. 30-40% of patients experience severe TRAEs which can lead to treatment interruption and hospitalization. TRAE risk is higher in patients with pharmacogenetic variants in DPYD and/or UGT1A1. Methods: This is a non-randomized, open-label implementation study at 3 cancer care sites within an academic health system to determine feasibility of integrating DPYD/UGT1A1 testing into clinical care (IMPACT-GI; NCT#04736472). Test ordering and discrete variant resulting were available within the Electronic Health Record. The lab detected 5 DPYD variants with reduced or no function (c.1905+1G>A, c.2846A>T, c.1679T>G, c.1236G>A, c.557A>G) and 2 UGT1A1 (*6, *28) alleles. Patients with a GI cancer initiating treatment with 5-FU, capecitabine, or irinotecan underwent testing along with preemptive chemotherapy dose modifications. Carriers of 1 DPYD variant (i.e., intermediate metabolizers [IM]) and UGT1A1 homozygous carriers (i.e., poor metabolizers [PM]) were recommended a 50% and 30% dose reduction, respectively. The primary endpoints were feasibility (proportion of results available prior to cycle 1) and fidelity (genotype concordant dose modifications). Secondary endpoints included severe TRAE (requiring emergency room visit or hospitalization) comparing variant carriers vs. wildtype over the first 6 cycles. Results: From 3/2021- 12/2022, 224 patients received testing (54% male, 74% White, 18% Black, mean age 62 ±17 years. The most common diagnoses were 47% colorectal and 30% pancreatic cancers. The median lab turnaround time was 10 (IQR 9-13) calendar days with 57.2% of results available prior to cycle 1. Results available prior to cycle 1 varied by distance to lab (64.7% near sites vs. 32.7% far site, p<0.001). We identified 7 (3%) DPYD intermediate metabolizers (IM) and 33 (15%) UGT1A1 poor metabolizers (PM). Four (2%) patients carried a variant in both DPYD and UGT1A1. Chemotherapies were preemptively dose reduced for 6 of 7 DPYD carriers receiving FP and 5 of 8 UGT1A1 PMs receiving irinotecan. 2 (28%) DPYD carriers vs. 71 (33%) wildtype carriers had a severe TRAE. 4 (50%) UGT1A1 PMs vs. 11 (36%) UGT1A1 wildtype carriers experienced a severe TRAE. Conclusions: DPYD/UGT1A1 testing is feasible with oncology providers following the genotype guided dose reductions when results were available prior to cycle 1. Pretreatment DPYD/UGT1A1 testing in conjunction with preemptive dose reductions resulted a similar proportion of severe TRAE as compared with wild-type patients. Further workflow modifications are required to decrease test turnaround time to ensure results are available prior to chemotherapy initiation. Clinical trial information: NCT04736472 .