ABSTRACT
<p><b>BACKGROUND</b>Prophylactic para-aortic nodal dissection (PAND) has no proven benefits for potentially curable advanced gastric cancer. However, the value of therapeutic PAND for involved para-aortic nodes (PANs) in patients with locally advanced gastric cancers has not been determined yet.</p><p><b>METHODS</b>Between 1998 and 2010, 157 gastric cancer patients with 1-3 involved PANs underwent extended D2 (D2+) lymphadenectomy plus PAND (PAND group, n = 69) or extended D2 lymphadenectomy alone (non-PAND group, n = 88). The clinicopathologic features and prognostic data were compared between the two groups. A propensity score-adjusted analysis was used for a balanced comparison.</p><p><b>RESULTS</b>The rate of PAN metastasis was 40.6% (28/69) in the PAND group. The 5-year survival rate was significantly higher in the PAND group than in the non-PAND group (43.7% vs. 31.8%, P = 0.044). Compared to the non-PAND group, the death hazard ratios in the PAND group were 0.45 (95% CI 0.274-0.739; P = 0.002) and 0.536 (95% CI 0.328-0.861; P = 0.0097) by multivariate analysis without and with propensity score adjustment respectively. Recurrence rate at 5 years was 39.1% in the PAND group and 43.2% in the non-PAND group (P = 0.628).</p><p><b>CONCLUSION</b>Extended D2 lymphadenectomy plus PAND is associated with superior outcomes for advanced gastric cancer patients with 1-3 involved PANs.</p>