Five-year survival disparities in two population-based cohorts of adults with early-onset colorectal cancer.

Abstract

11155 Background: Early-onset colorectal cancer (EO-CRC) is now the leading cause of cancer-related death in those under the age of 50 despite advances in treatment. However, the impact of specific sociodemographic and clinical factors on long-term survival remains poorly characterized. We evaluated 5-year survival patterns and risk factors across two diverse, population-based cohorts to optimize risk reduction strategies and survivorship care delivery. Methods: We constructed two cohorts of adults < 50 years of age with newly diagnosed EO-CRC from 2015 to 2018 (stages I-III) and treated with curative-intent surgery. One cohort was derived from three SEER registries (Georgia, Los Angeles County, Kentucky; n = 2995) and one cohort was derived from the Veteran’s Health Administration (VHA) (n = 295). Pathologic-confirmed recurrence was identified in the follow-up period 6 months to five years after curative-intent surgery via human review of electronic pathology reports. Five-year survival was estimated from the date of curative-intent surgery to five years post-surgery. Adjusted risk of CRC-specific mortality was estimated in both populations using multivariable proportional hazards regression, adjusting for key demographic and clinical characteristics. Results: The proportion of patients surviving 5 years was 86% in the SEER cohort and 92% in the VHA cohort. Overall, 14% experienced recurrence in SEER and 18% in VHA, and recurrence was strongly associated with reduced survival in both cohorts (both p < 0.01). Five-year survival varied by clinical stage and was higher in the VHA across all stages (SEER: 93% stage I, 87% stage II, and 76% stage III vs. VHA: 97% stage I, 93% stage II, and 88% stage III). Mortality risk varied by race/ethnicity in the SEER cohort only: Hispanic patients (adjusted HR 1.6, 95% CI: 1.1–2.2) and Black patients (adjusted HR 1.9, 95% CI: 1.5–2.5) had a greater risk of 5-year mortality vs. white patients. Age at diagnosis, sex, and living in a rural area were not significantly associated with mortality in either cohort. Rectal tumors (vs. colon) were associated with an increased risk of mortality in VHA (adjusted HR: 2.5, 95% CI: 1.0-6.4), but not in SEER. Lymphovascular invasion was associated with an increased risk of mortality in both cohorts. Conclusions: There are critical sociodemographic and clinical disparities in survival among adults with stage I-III EO-CRC. These distinct patterns observed between the two cohorts underscore the influence of varying population-level risk profiles. Tailoring survivorship care to identify and support these high-risk subgroups is essential to mitigate disparities and improve long-term outcomes in this growing patient population.