Prognostic significance of microscopic positive margins for gastric cancer patients with potentially curative resection.

BACKGROUND: Most previous studies concerning the impact of positive margins on patient outcomes were based on patients with D1 lymphadenectomy. The prognostic significance of positive margins for patients with D2/D3 lymphadenectomy has not been investigated.
METHODS: Clinicopathologic and prognostic data were compared between 110 patients with positive margins and 2,159 patients with negative margins who underwent R0 surgery (D2/D3 lymphadenectomy). Multivariate analysis was performed to identify independent factors correlated with positive margins. Cox's proportional hazard model was applied to investigate whether a positive margin was an independent factor predicting poor outcome. Overall survival rates and the incidence of recurrence were compared between patients with positive and negative margins, stratifying by pT, pN, and TNM stage.
RESULTS: Tumor size, pT stage, and pN stage were independent factors associated with positive margins. Although the overall survival rate of patients with positive margins was significantly poorer than that of patients with negative margins, margin status was not an independent prognostic factor on multivariate analysis. Significant differences in survival between patients with negative and positive margins could be observed for those in pT1-2, pN0-1, and I-II stage, but not for those in pT3-4, pN2-3, and III-VI stage. Although locoregional recurrence was more frequently observed for patients with positive margins, the incidence of any recurrence was significantly higher only for patients in pT1-2, pN0-1, and I-II stage, but not for those in pT3-4, pN2-3, and III-VI stage.
CONCLUSION: A positive margin results in a significantly worse outcome for pT1-2, pN0-1, and I-II stage gastric cancer patients who undergo D2/D3 lymphadenectomy, but not for those in pT3-4, pN2-3, and III-IV stage.