Predictors of disease-free survival in rectal cancer patients undergoing curative proctectomy

Objective: To identify the factors that affect the disease-free survival (DFS) of rectal cancer patients. Method: Patients from an IRB approved rectal cancer database were reviewed (1990-2000). All patients underwent either abdominoperineal resection or low anterior resection using total mesorectal excision with curative intent. Univariate and multivariate analyses were performed to analyse the factors that influenced DFS. Results: A total of 304 patients were reviewed (mean age 64, 52% male). Seventy-seven per cent of patients received neoadjuvant therapy (28.6% short-course radiation therapy (RT), 35.5% long-course RT, 12.5% chemo-RT). The radial margin was involved with tumour in 5.2% of patients (final pathology). The overall survival rate was 85.2% with a mean follow-up time of 33 ± 26 months. The mean time to death was 34.8 ±26.8 months. Local recurrence (± distant recurrence) occurred in 4%. Anastomotic leaks occurred in 3.6% of patients. Overall pathologic stage, pathologic T stage, nodal status, the use of adjuvant chemotherapy, tumour fixation, involvement of the radial margin, the presence of mucin, and lymphatic and perineural invasion (PNI) were predictors of DFS by univariate analysis. Of note, anastomotic leaks and obstructing cancers did not influence DFS. Using multivariate analysis with backward elimination, overall pathologic stage, radial margin status, adjuvant chemotherapy, and PNI predicted the DFS. Conclusion: Major predictors of DFS in rectal cancer are the overall pathologic stage, adjuvant chemotherapy, radial margin status and PNI. Radial margin status may be a marker of tumour aggressiveness and should be considered in deciding on adjuvant chemotherapy.