eP416 - Proximal vs Distal Colon Cancer Location: A Subset Analysis of the Minnesota Colon Cancer Control Study

Purpose/Background: Over the past decade, primary tumor location has been proposed as a prognostic factor in colon cancer patients, with proximal lesions being associated with an increased risk of mortality in modern cohorts. Several explanations have been attributed to this observation, including different genetic profiles, epigenetics, and microbiome environment, but the underlying reason remains unclear. The aim of the current study was to validate these findings in a historical cohort with long-term follow-up.Methods/Interventions: The Minnesota Colon Cancer Control Study randomized 46,551 patients 50-80 years to usual care, annual or biennial screening with fecal occult–blood testing (FOBT). Screening was performed from 1976-1982 and 1986-1992. Positive FOBT was followed by colonoscopy. The analysis was restricted to patients with a colon adenocarcinoma diagnosis before 1996 who were not diagnosed at death, did not have a Duke stage D tumor, and had known tumor location, which was grouped in proximal vs distal; proximal tumors were defined as those originating up to the splenic flexure. Stabilized inverse probability weighting (IPW) weights were calculated via logistic regression with main effects for age, gender, aspirin use, smoking history, Body Mass Index (BMI), treatment arm, and Duke stage (A, B, C). The subdistribution hazard ratio (SHR) of proximal tumors for disease-specific mortality with the competing risk of death from another cause was estimated using the Fine-Gray estimator with stabilized IPW weights. The hazard ratio (HR) of proximal tumors on overall survival was estimated through a weighted Cox proportional hazards regression model.Results/Outcomes: A total of 989 participants were diagnosed with colon adenocarcinoma without metastases; 56% had proximal tumors. Mean age at randomization was 65, while mean age at cancer diagnosis was 75 years old for proximal and 72 years old for distal cancers. The distribution of gender, smokers, aspirin use, and elevated BMI was similar between the two groups. However, patients with proximal adenocarcinoma presented with higher Duke stage. In adjusted models, there was no difference between proximal and distal tumors in disease-specific mortality (SHR 0.92, CI 0.71-1.2, p =0.54), although having a proximal cancer was associated with a small, reduced risk of death from any cause (HR 0.86, CI: 0.76-0.97, p < 0.01, Figure 1). Weighted 5-year overall survival was indeed 69% vs 67% for proximal vs distal tumors.Conclusion/

Discussion: In this subset of colon cancer patients from an epidemiologic randomized study in the pre-modern chemotherapy era, no differences were observed in disease-specific mortality based on tumor location, while there was a marginal increased risk of overall mortality for distal neoplasms. Although lacking granular data, these findings raise questions about the generalizability of the association between side of origin and prognosis identified in contemporary treatment-based trials.Figure 1. Weighted Overall Survival by Tumor Location.