Shogo Hayashi1, Mitsuro Kanda2, Seiji Ito3, Yoshinari Mochizuki4, Hitoshi Teramoto5, Kiyoshi Ishigure6, Toshifumi Murai7, Takahiro Asada8, Akiharu Ishiyama9, Hidenobu Matsushita1, Chie Tanaka10, Daisuke Kobayashi10, Michitaka Fujiwara10, Kenta Murotani11, Yasuhiro Kodera10. 1. Department of Surgery, Tosei General Hospital, Seto, Japan. 2. Department of Gastroenterological Surgery (Surgery II), Nagoya University Graduate School of Medicine, 65 Tsurumai-cho, Showa-ku, Nagoya, 466-8550, Japan. m-kanda@med.nagoya-u.ac.jp. 3. Department of Gastroenterological Surgery, Aichi Cancer Center, Nagoya, Japan. 4. Department of Surgery, Komaki Municipal Hospital, Komaki, Japan. 5. Department of Surgery, Yokkaichi Municipal Hospital, Yokkaichi, Japan. 6. Department of Surgery, Konan Kosei Hospital, Konan, Japan. 7. Department of Surgery, Ichinomiya Municipal Hospital, Ichinomiya, Japan. 8. Department of Surgery, Gifu Prefectural Tajimi Hospital, Tajimi, Japan. 9. Department of Surgery, Okazaki City Hospital, Okazaki, Japan. 10. Department of Gastroenterological Surgery (Surgery II), Nagoya University Graduate School of Medicine, 65 Tsurumai-cho, Showa-ku, Nagoya, 466-8550, Japan. 11. Biostatistics Center, Graduate School of Medicine, Kurume University, Kurume, Japan.
Abstract
BACKGROUND: The prognostic significance of the number of retrieved lymph nodes (RLNs) in gastric cancer remains controversial. Therefore, we designed a multicenter collaborative database to investigate the correlation between the number of RLNs and prognosis of patients with advanced gastric cancer after curative resection. METHODS: We retrospectively analyzed 1103 patients who underwent gastrectomy for stage II/III gastric cancer between 2010 and 2014. Lymph nodes, which were retrieved by surgeons from surgically resected specimens, were validated by pathologists. A target population and the optimal cutoff were determined using receiver operating characteristic (ROC) curve analysis. After propensity score matching of eight variables, including splenectomy and adjuvant chemotherapy, the prognostic significance of RLNs was evaluated. RESULTS: According to ROC curve analysis, the optimum cutoff score for predicting postoperative survival was 40. After matching, the backgrounds of patients in the RLN < 40 and RLN ≥ 40 groups (n = 87 each) became well-balanced. The RLN < 40 group experienced significantly shorter relapse-free and overall survival. The prevalence of peritoneal recurrence was significantly increased in the RLN < 40 group. RLN < 40 was an independent prognostic factor in multivariable analysis, although pathological N status was not. A forest plot revealed that the RLN < 40 group was at greater risk of recurrence in most subgroups. CONCLUSIONS: RLN < 40 was associated with an adverse prognosis of patients with stage III gastric cancer who underwent total gastrectomy.
BACKGROUND: The prognostic significance of the number of retrieved lymph nodes (RLNs) in gastric cancer remains controversial. Therefore, we designed a multicenter collaborative database to investigate the correlation between the number of RLNs and prognosis of patients with advanced gastric cancer after curative resection. METHODS: We retrospectively analyzed 1103 patients who underwent gastrectomy for stage II/III gastric cancer between 2010 and 2014. Lymph nodes, which were retrieved by surgeons from surgically resected specimens, were validated by pathologists. A target population and the optimal cutoff were determined using receiver operating characteristic (ROC) curve analysis. After propensity score matching of eight variables, including splenectomy and adjuvant chemotherapy, the prognostic significance of RLNs was evaluated. RESULTS: According to ROC curve analysis, the optimum cutoff score for predicting postoperative survival was 40. After matching, the backgrounds of patients in the RLN < 40 and RLN ≥ 40 groups (n = 87 each) became well-balanced. The RLN < 40 group experienced significantly shorter relapse-free and overall survival. The prevalence of peritoneal recurrence was significantly increased in the RLN < 40 group. RLN < 40 was an independent prognostic factor in multivariable analysis, although pathological N status was not. A forest plot revealed that the RLN < 40 group was at greater risk of recurrence in most subgroups. CONCLUSIONS: RLN < 40 was associated with an adverse prognosis of patients with stage III gastric cancer who underwent total gastrectomy.
Entities:
Keywords:
Gastric cancer; Number of retrieved LNs; Prognosis; Total gastrectomy