Survival by lymph node ratio (IG) in colorectal cancer patients in Oncologos del Occidente from Caldas: Colombia 1997–2012.

Abstract

641 Background: Colorectal cancer is the third cause of cancer death in the western world and in Colombia the fifth leading cause. The Caldas State being one of the five with the highest mortality rates. There are many prognostic factors reported in the literature but the clinical stage is the most important mainly determined by the number of involved nodes but exist other factors in their evaluation. It has been proposed Lymph Node Ratio (IG) ( Nro. (+) nodes / Nro. nodes removed x 100) for the analysis of nodal stage but there is controversy about its usefulness in various contexts. Methods: We made a descriptive study of patients with colorectal cancer attending at ONCOLOGOS DEL OCCIDENTE from Caldas - Colombia, who applied the Lymph node ratio (IG) and related to demographic, clinical and others prognostic factors in order to see its relation with overall survival at 5, 10 and 15 years. Results: 333 patients with 66% colon cancer, under 40 years 12% , women 50%, 68% without ACE data prior treatment, 41% on right colon, 49% medium rectum, 44% localized stage, 2% pathological complete response and 3% no classification data. Obstruction (29%), ulceration (25%) and 30% without prognostic factors, 33% had 12 or more lymph nodes removed, 50% N0 (TNM), 21% N2; the resected nodes, positive nodes, relapse, prior ACE and TNM as significant variables, 27% relapsed mainly liver, 50% had nodal index group 1 (0%); group 2 (1-20%) 15%, 14% group 3 (21-40%) and 21% group 4 (> 40%) . Mean follow-up: 53.7 months (SD: 43.8, Range:0 -218 months) and overall survival 64% at 5 years and 54% at 10 and 15 years. Conclusions: With an inadequate number of lymph nodes removed (33%) and multiple prognostic factors, Lymph node ratio is proposed like an independent prognostic factor in patients with colorectal cancer associated with overall survival at 5, 10 and 15 years, on which a therapeutic decision can be based.