RESUMO
BACKGROUND: Prognosis assessment of node-positive colorectal cancer patients by Astler-Coller (AC) and TNM classifications is suboptimal. Recently, several versions of lymph node ratio (LNR; ratio metastatic/examined nodes) have been proposed but are still mostly unused. METHODS: The prognostic value of several criteria, including LNR (two classes-LNR1 and LNR2-identified by a 15% cut-off) was studied in 761 consecutive patients, from 2000 through 2010. The relationships between total examined nodes, N, T and LNR were also analysed. LNR1 and LNR2 patients' survival was analysed within AC and TNM subgroups, and then coupled with them. RESULTS: Age, tumour location and LNR are independent factors predicting survival. The relationships between LNR, N stage and T stage with examined nodes suggest confusing factors. LNR allows for identification of subgroups with different survival within AC and TNM classifications (p < 0.0001). Patients with LNR class discordant from AC stage (LNR1-C2 and LNR2-C1) have a similar 5-year survival (54 and 57%, respectively). LNR2 and TNM stage IIIC define a poor 5-year prognosis (33%). CONCLUSIONS: LNR is a powerful prognosis predictor, easily integrated with TNM and AC classifications to improve prognosis assessment and facilitate clinical use. Possible confusing factors should be considered in future studies.
