The impact of smoking on cancer recurrence and survival in patients with stage III colon cancer: Findings from intergroup trial CALGB 89803.

4039 Background: Although tobacco use is linked to the development of colon cancer, little is known about its impact in colon cancer survivors. METHODS: We prospectively collected data on prior and current cigarette smoking on 1,045 patients with stage III colon cancer enrolled in a phase III adjuvant chemotherapy trial (bolus 5-FU/leucovorin ± irinotecan). Patients (pts) reported tobacco use in self- report questionnaires during (Q1) & 6 months after completion of adjuvant therapy (Q2). Smoking status was defined as never, current, or past. Lifetime pack years (pyr) were defined as # of packs of cigarettes over pt's lifetime. Since there was no difference in efficacy between the two treatments, data for all patients were combined. Cox proportional hazards were computed for disease-free survival (DFS, our primary endpoint) and overall survival (OS). DFS and OS were measured from completion of Q1 to event of interest, excluding events within the 1(st) 90 days to avoid biases of change in behavior due to impending event. RESULTS: Data on smoking history were captured on 1,045 patients of 1,264 in the phase III trial (460 [44%] past smokers; 107 [10%] current smokers; 478 [46%] never smokers). In analyses adjusted for age, N and T stage, grade of differentiation, and presence of bowel obstruction at dx, past smokers had hazard ratio (HR) for DFS of 1.15 (95% CI, 0.92-1.43; p=0.2) and current smokers had HR for DFS of 1.20 (95% CI, 0.85-1.70; p=0.3), compared to never smokers. Consistent with prior analyses, we defined pyr categories as 0, 0-10, 10-20, 20 or more. In unadjusted analyses, HR for DFS were 1.16 (95% CI, 0.89-1.52), 1.24 (95% CI, 0.85-1.83) and 1.33 (95% CI, 1.03-1.73) for lifetime pyr 0-10, 10-20 and 20+, respectively, compared to never smoking (p trend 0.04). In adjusted analyses, HR for 20+ pyr was 1.25 (95% CI, 0.96-1.63) compared to never smoking (p trend 0.15). HRs for OS were similar as DFS for smoking status and lifetime pyr. CONCLUSIONS: Lifetime total of tobacco usage may influence outcomes in patients with stage III colon cancer. Further research with larger datasets is warranted. No significant financial relationships to disclose.

Duke Scholars

Author Donna Niedzwiecki Biostatistics & Bioinformatics, Division of Biostatistics