Gastrectomy with D3 Lymph Node Dissection

Abstract

Japan Clinical Oncology Group (JCOG) carried out a phase III study on prophylactic nodal dissection of para-aortic lymph nodes (PAN) for potentially curable T3/T4 advanced gastric cancer [1]. This study, JCOG9501, is a multi-institutional prospective randomized control study carried out among 24 Japanese hospitals to evaluate superiority of D2 + PAN dissection (D) over D2 alone. Eligibility criteria are shown in Table 8.1. The primary end point was overall survival (OS), and the secondary were recurrent-free survival (RFS) and surgery-related complications and hospital death. The sample size was projected as 520 to detect 8% increase of 5-year survival rate for PAND with a one-sided alpha level of 0.05 and power of 80%. Between July 9501 and April 2001, 523 patients were randomly assigned to D2 alone (263) or D2 plus PAND (260) (Fig. 8.1). All except one underwent allocated nodal dissection and followed without any adjuvant treatment. In terms of patients’ characteristics and prognostic factors, there was no difference between the two groups. Hospital death in both groups was as low as 0.8%. Incidence of major surgical complication such as anastomotic leak, intra-abdominal abscess, or pancreatic juice leak was the same between two groups, but that of total morbidity including minor complication was significantly higher after D2 + PAND than D2 [2]. Both OS and RFS curves were nearly completely overlapped, and HR was 1.03 (95% IC; 0.78–1.37; p = 0.83) (Fig. 8.2) and 1.08 (95% CI; 0.83–1.42; p = 0.72) (Fig. 8.3), respectively. Sites of initial recurrent did not show any difference between the two groups. In subgroup analysis, pathological tumor stage and lymph node stage (negative/positive) showed statistically significant interaction, but both of them cannot be known when performing surgery, and reasonable explanation could not be given. Although the interaction was not significant, hazard ratio for tumors of the upper third of the stomach was 0.58, while that of other locations was 1.10. Five-year OS of 22 out of 260 (8.5%) patients who had PAN metastasis was 18.2%, which was almost same as our expectation. In summary, PAND should be avoided in patients with potentially curable T3/4 tumors without any clinical evidence of PAN metastasis.