OBJECTIVETo evaluate the survival, complication and postoperative mortality after D(1) or D(2) lymph node dissection for gastric cancer.

METHODSAll the randomized clinical trials about nodal dissection for gastric cancer published within the last 20 years were collected. Quality assessment was done on each trial and relevant data were extracted from qualified trials. Meta-analysis was performed with the use of RevMan 4.2 (Cochrane) for statistic analysis.

RESULTSThree hundred and ninety-four trials were yielded at the initial search. Four trials, recruited 1316 cases of gastric cancer in total, were included after quality assessment. Results of Meta-analysis showed that standard D(2) dissection could effectively improve patients' long-term survival [RR 1.35, 95%CI(1.12-1.62), NNT=9] as compared with D(1) dissection. If splenectomy (or pancreatico-splenectomy) was involved, D(2) dissection only improved the clinical outcome of T(3)-staged cases [RR 1.80,95%CI(1.03-3.15), NNT=13]. D(2) dissection produced higher rates of postoperative complication [RR 1.72,95%CI(1.46-2.03), NNT=6] and mortality [RR 2.12,95%CI(1.39-3.25), NNT=21] than D(1) dissection.

CONCLUSIONSStandard D(2) dissection can increase the overall survival rate when compared with D(1) dissection. If splenectomy (or pancreatico-splenectomy) cases are involved,D(2) dissection can only improve the survival rate of T(3)-staged patients. D(2) dissection yields higher postoperative morbidity and mortality than D(1) dissection.