Ratio of metastatic to resected lymph nodes for prediction of survival in patients with inadequately staged gastric cancer

BACKGROUND: Staging is inadequate in up to 70 per cent of patients with gastric cancer in Western countries owing to the small number of lymph nodes dissected during surgery. The aim was to determine whether using the ratio of metastatic to resected lymph nodes (LNR) might improve accuracy. METHODS: Data were analysed from patients with gastric cancer who had gastrectomy in several centres between 1986 and 1998, with dissection of 15 or fewer lymph nodes. LNRs and other prognostic factors were evaluated. RESULTS: From a total of 738 patients, the median number of resected nodes was 8 (range 1-15) and median LNR was 42.8 per cent. The number of metastatic nodes significantly affected survival only in univariable analysis. In a Cox proportional hazards model, patient age, depth of tumour infiltration, tumour location, and LNR were identified as independent prognostic factors. Compared with node-negative patients, the hazard ratio for an LNR of 0.1-40.0 per cent was 1.85 (P < 0.001), increasing to 2.93 (P < 0.001) when the LNR exceeded 40.0 per cent. CONCLUSION: The LNR cannot be used as a substitute for staging with adequate lymphadenectomy. It may help to stratify patients in terms of prognosis when the number of resected lymph nodes is limited.