Lymphadenectomy in gastric cancer: influence on prognosis of lymph node count.

To better understand the role of the number of lymph nodes retrieved on long-term outcome of gastric cancer treatment, 154 patients who had undergone curative resection, with dissection of >15 nodes were retrospectively studied. Dissection of perigastric and extraperigastric lymph nodes, defined as 'extended' (>26 nodes dissected) in 39 cases and 'limited' (< or = 26 nodes dissected) in 115 cases, was performed. A total of 3479 lymph nodes (mean 22.6 per specimen), were dissected and of these 721 showed metastases. A mean of 8.1 lymph node metastases, per metastatic case, was found. Regression analysis showed no independent factor associated with the extent of lymphadenectomy. Depth of wall invasion (p=0.000) and histological growth pattern (p=0.044) were independently associated with the number of lymph nodes involved (pN0, pN1 1-7, pN2 >7). The cumulative 5-year survival rate was 47% in patients without lymph node metastases; 29% in those with 1-7 nodes involved and 17% in those with >8 nodes involved (p=0.002). Receiver operating characteristic (ROC) curve analysis, in 65 nodenegative cancer cases, demonstrated an area under the curve for vital status (alive or dead) of 0.602 (95% CI: 0.473 - 0.721). All node-negative cases with a number equivalent to or exceeding the cutoff point of 23 nodes were alive. ROC analysis showed 11 to be the cutoff number of metastasized lymph nodes in correlation with vital status. Almost all those patients in whom the number of positive nodes was equivalent to, or exceeded the cutoff point had died (area under the ROC curve 0.633; 95% CI: 0.524 - 0.733). ROC analysis showed that the cutoff lymph node ratio, in relation to vital status, was 0.33. The majority of patients at or above this cutoff point had died (area under ROC curve 0.682; 95% CI: 0.574 - 0.776). Multivariate survival analysis showed that lymph node ratio was the only independent prognostic factor (p=0.001). The present findings suggest that, in lymphadenectomy with at least 15 nodes, the number and status of regional nodes dissected, irrespective of the location, provide reliable prognostic information on curatively resected gastric carcinomas.