Impact of early-onset colorectal cancer on utilization of chemotherapy and outcomes in patients with stage II disease.

Abstract

85 Background: Rising rates of early-onset colorectal cancer (EOCRC) pose a dilemma for clinicians when deciding how to treat early-stage patients to maximize outcomes. While standard-of-care for Stage II colon cancer is largely surgical resection, evidence suggests that treatment selection may differ by patient age. We sought to determine whether rates of adjuvant chemotherapy administration differ between early and later-onset patients with Stage II CRC. Methods: Data were derived from the Flatiron National Database spanning 1/1/2003 to 8/1/2021. Patients 18 years or older with Stage II CRC were grouped into those aged 18-49 (EOCRC) and those aged 50 or older (LOCRC). Demographic characteristics, ECOG score, stage and site of tumor, and chemotherapy were included for all patients. Primary outcomes of interest included rates of adjuvant chemotherapy administration by age and ethnicity. Univariate and multivariable logistic regression models were used to evaluate relationships between chemotherapy administration, age groups, and ethnicity while adjusting for significant covariates. Results: Of 2133 patients with Stage II CRC, 1606 patients with complete data were included. A secondary analysis of 1065 patients with colon cancer was performed to address potential confounding factors related to neoadjuvant and/or adjuvant chemotherapy given in patients with stage II rectal cancer. Mean age of EOCRC patients was 45.0 years (range: 41.0-48.0) vs. 68.0 years (60.0-75.0) for LOCRC. Adjusting for ethnicity, gender, site, and ECOG score, multivariate analysis showed EOCRC patients received chemotherapy significantly more often than LOCRC patients for stage II CRC (adjusted odds ratio 1.85, 95% CI 1.32-2.60, p < 0.001). Similar findings were observed in the colon cancer only cohort (adjusted OR 2.02, 95% CI 1.31-3.09, p < 0.001). By ethnicity, non-Hispanic patients received chemotherapy at significantly lower rates than Hispanic patients in both cohorts (adjusted odds ratio 0.58, 95% CI 0.39-0.88, p = 0.009 and adjusted odds ratio 0.55, 95% CI 0.34-0.91, p = 0.018). In a subgroup analysis of Stage IIA patients, multivariate logistic regression adjusting for gender, ECOG, site, and ethnicity showed that patients with EOCRC were more likely to receive chemotherapy than patients with LOCRC (adjusted odds ratio 1.91, 95% CI 1.21-2.99, p = 0.005). Updated survival data will be presented. Conclusions: Adjuvant chemotherapy is given preferentially in Stage II EOCRC, even in Stage IIA disease, despite deviation from established guidelines. This may expose patients at low risk for recurrence to unnecessary toxicities and reveals potential provider bias in favor of younger patients in aggressively treating CRC, despite unclear evidence for any outcome benefit.