Outcomes of young-onset colorectal cancer vs late-onset colorectal cancer patients on phase 1 matched and non-matched therapies.

3549 Background: Systemic therapy recommendations for young-onset colorectal cancer (YOCRC), CRC diagnosed at < 50 years old, are similar for late-onset CRC (LOCRC) despite possible differences in biologic behavior. This study aims to compare outcomes among YOCRC and LOCRC patients on Phase 1 matched and non-matched therapies. Methods: This was a single-institution retrospective analysis of patients with CRC who received treatment on a Phase 1 clinical trial. Only the first Phase 1 therapy for each patient was included in analysis. Matched therapy was defined as therapy targeting genomic alterations or their signaling pathways. Distributions of progression-free survival (PFS) were estimated by the Kaplan-Meier method. Log-rank test was performed to test the difference in survival between groups. A propensity score matched analysis was created using a multivariate logistic regression model. Covariates in the model included: gender, race, lung metastasis, liver metastasis and tumor sidedness. Results: 577 patients were included in analysis (Table 1). 252 patients had YOCRC (43.7%) and 325 had LOCRC (56.3%). 100 YOCRC patients (39.7%) and 90 LOCRC patients (27.7%) received matched therapies. Before propensity score matching YOCRC patients on matched therapy had higher odds of achieving a response compared to LOCRC patients on matched therapy (complete response/partial response) (10.5% vs 3.4%, odds ratio (OR): 3.294 (95% confidence interval (CI)): 0.876, 12.390), p=0.0777). After propensity score matching YOCRC patients on matched therapy had higher odds of achieving a response compared to LOCRC patients on matched therapy (13.3% vs 3.9%, OR was not estimable, p=0.0082). No significant differences in overall response rate were detected between YOCRC and LOCRC patients on non-matched therapy before or after propensity score matching (5.4% vs. 4.1%, OR: 1.362 (95% CI: 0.513, 3.615), p=0.5348; 5.5 vs. 4.3%, OR: 1.167 (CI: 0.392, 3.472), p=0.7815). No significant differences in PFS were detected between patient with YOCRC and those with LOCRC in any of the patient cohorts before or after propensity score matching (all patients: p=0.743, p=0.639; patients receiving matched therapy: p=0.497, p=0.909; patients receiving non-matched therapy: p=0.999, p=0.62). Conclusions: YOCRC patients were more likely than LOCRC patients to achieve a response on matched therapy though this did not translate to improved PFS. Nevertheless, matched therapies are associated with increased response rate for YOCRC patients. Patient demographics. Trait n (%) Age at diagnosis (median, range) 51 (18-83) White/Caucasian 425 (74.7%) Black/African American 67 (11.6%) Asian 40 (6.9%) Hispanic/Latino 79 (13.7%) Microsatellite Instability-High Tumor 5 (0.9%) KRAS mutation 352 (61.0%) BRAF V600E mutation 38 (6.6%) Prior Unique Lines of Therapy (median, range) 4 (0-11)