Patient-reported symptom assessments and premature chemotherapy discontinuation in patients with gastrointestinal cancers.

383 Background: Premature chemotherapy (CTX) discontinuation in patients with gastrointestinal (GI) cancers may compromise treatment effectiveness and worsen clinical outcomes. Patient-reported symptom data can help to reveal problematic symptoms so treatment plans can be adjusted to preserve patient outcomes. Using routinely collected patient-reported data, we examined the association between levels of patient engagement in symptom assessments and premature CTX discontinuation. Methods: We included adult patients with GI cancers receiving CTX at a single health system between 2019 and 2024. The Patient-Reported Outcome version of the Common Terminology Criteria for Adverse Events was used to evaluate the 12 common CTX-related symptoms. We identified patients who completed any assessments within 15 days of CTX initiation, at one, two, or three months post-initiation. The primary outcome was discontinuation due to toxicity within three months of CTX initiation. Using the Fine-Gray model to account for the competing risks of death and CTX changes due to disease progression, we separately predicted discontinuation based on the number of assessments completed (1-4) as well as whether at least two were completed (yes vs. no). Results: In 1,673 patients, 80% were non-Hispanic White, 59% male, and 52% ≥65 years old. Colorectal (36%) and pancreatic (29%) cancers are common. The most frequently administered CTX drugs were 5-fluorouracil (64%), oxaliplatin (55%), and irinotecan (33%). The average completed assessments was 2(SD = 1.1). About 58% completed ≥2 assessments, of whom 86% had one assessment at baseline. After adjusting for covariates, a higher number of completed assessments was associated with a reduced risk of discontinuation (SHR: 0.68 [95%CI: 0.51 – 0.90], Table). Compared to those with one assessment, patients with at least two had a lower risk of discontinuation (SHR: 0.56 [95%CI: 0.31 – 1.00], Table). Conclusions: Greater level of engagement in symptom assessments was associated with a lower risk of premature CTX discontinuation due to toxicity. Assessing symptoms at CTX initiation and at least one follow-up time point may be a feasible and effective strategy to prevent premature CTX discontinuation. Fine-Gray model of three-month non-tolerance-related discontinuation predicted by the number of symptom assessments in gastrointestinal patients receiving chemotherapy, 2019-2024 (N = 1673). SHR (95% CI) Continuous variable Numbers of symptom assessments 0.68 (0.51 – 0.90) Categorical variable Only one symptom assessment Reference At least two symptom assessments 0.56 (0.31 – 1.00) SHR: subdistribution hazard ratios. Age, race, language, marital status, level of education, employment status, insurance type, cancer site, concurrent target therapy, concurrent immunotherapy, concurrent radiation therapy, duration after diagnosis, and comorbidity were controlled as covariates in the analyses.