New tumor-node-metastasis staging strategy for node-positive (stage III) rectal cancer: An analysis

Purpose The tumor-node-metastasis system for staging rectal cancer is based on invasion, number of involved nodes, and metastasis. Nodes are classified as N1 or N2 according to the number involved with metastases. Nodal positivity defines stage III regardless of depth of invasion or number of positive nodes. Our purpose was to analyze overall survival when node-positive patients were stratified into three new subsets. Methods We analyzed data entered into the National Cancer Data Base for 5,987 stage III patients with rectal cancer between 1991 and 1993. Survival was calculated using three new subgroups (IIIA: T1/2, N1; IIIB: T3/4, N1; IIIC: any T, N2). Survival following surgery and adjuvant therapy was assessed. The observed survival rates were calculated and compared using the log-rank method. The Cox regression model assessed subgroup differences. Results Five-year observed survival rates for stage III subcategories were 55.1% in IIIA; 35.3% in IIIB; and 24.5% in IIIC. Stratifying for treatment outcome, stage IIIA patients having surgery alone (n = 278) had poorer observed 5-year survival (39%) than patients treated with surgery and adjuvant chemotherapy or radiation therapy (chemo/XRT; n = 765; 60%). Similar outcomes occurred in IIIB (surgery-alone [n = 726; 21.7%] and chemo/XRT [n = 2,130, 40.9%] groups) and in IIIC (surgery-alone [n = 467; 12.2%] and chemo/XRT [n = 1,621, 28.9%] groups). Differences were significant (P < .0001) in all stages. Conclusion The traditional stage III designation of rectal cancer fails to account for invasion (T1-4) and number of involved nodes (NI, N2). The stratification of stage III patients into three subsets should be used in future analyses of rectal cancer, The effect of postoperative adjuvant therapy was beneficial in all subsets. (C) 2004 by American Society of Clinical Oncology.