Long-term results of radical surgery for rectal cancer: multivariate analysis of prognostic factors influencing survival and local recurrence

Abstract Aims To evaluate consistent radical surgery performed over a 13-year period for rectal cancer in terms of local tumour control and long-term survival. Methods Radical surgical procedure principally using total mesorectal excision (TME) for middle and lower rectal tumours, high ligation of the inferior mesenteric artery and sphincter-saving resections (SSR) whenever possible, has been performed prospectively since January 1984. Results Tumour resection was possible in 98.8% (636/644), potentially curative resections (UICC/AJCC R0 resection) in 85.7% (552/644) and sphincter preservation in 71.7% (462/644). Five- and 10-year observed survival rates, surgical mortality not excluded, for all patients were 49.2% and 37.4%. Tumour-adjusted 5- and 10-year survival rates were 60.5% and 55.3%. For curatively operated patients (UICC/AJCC R0) 5- and 10-year observed survival rates were 56.3% and 42.6% and tumour-adjusted survival rates were 68.6% and 62.7%. The 5- and 10-year local recurrence rates for R0 resected patients were 12.0% and 12.6%. Post-operative hospital mortality was 3.1%. Conclusions Multivariate analysis using Cox's model identified increasing pT category and pN category, old age and low tumour location as detrimental factors having independent influence on survival. For local tumour failure only pT and pN category as well as adjuvant radiation therapy were identified in the Cox model as having an independent detrimental influence.