Quality control of lymph node dissection in the Dutch Gastric Cancer Trial

Abstract

Abstract Background Current guidelines indicate that D2 resection is the standard of care for patients with locally advanced gastric cancer. To assess the impact of quality assurance of lymph node removal, non-compliance and contamination in the D1 and D2 study arms of the Dutch Gastric Cancer Trial were investigated with respect to recurrence and survival. Methods The location and numbers of lymph nodes detected at pathological investigation in the Dutch Gastric Cancer Trial were compared according to the guidelines of the Japanese Research Society for the study of Gastric Cancer. Non-compliance was defined as inadequate removal of lymph node stations. Contamination was defined as lymph nodes removed outside the intended level of resection. The dissection groups D1 and D2 were divided into non-compliance, compliance and contamination categories. Long-term overall survival was calculated for minor (2 or fewer lymph nodes) and major (more than 2 lymph nodes) non-compliance and contamination in the D1 and D2 group, using Kaplan–Meier plots. Results Some 1078 patients were included, of whom 711 with potentially curative surgical resections were evaluated. Overall non-compliance was 80·5 per cent in the D1 and 81·6 per cent in the D2 group. Major non-compliance occurred in 15·3 per cent of the D1 and 26·0 per cent of the D2 group. Major contamination hardly occurred. Overall 15-year survival rates in the randomized groups were 21·2 per cent (D1) and 29·0 per cent (D2) (P = 0·351). After exclusion of patients with major non-compliance and/or major contamination, survival rates were 23·2 per cent (319 patients) and 32·6 per cent (245) respectively (P = 0·261). Where there was major non-compliance, survival rates in the D1 (58 patients) and D2 (86) groups were 10 and 17 per cent respectively (P = 0·302). Survival in the D2 compliant + contaminated group (139 patients) was significantly better than that in the D1 group without contamination (282): 35·7 versus 19·9 per cent (P = 0·041). In the D2 group, there was a significant difference in survival between contaminated (95 patients) and non-contaminated (236) groups: 39 versus 25·1 per cent (P = 0·041). Conclusion Non-compliance in the D2 dissection group may have obscured a significant difference in survival between the randomized groups. A D2 dissection with contamination was associated with the best survival, suggesting that extended D2 lymph node dissections improve survival.