Sugimoto Kiichi, Sakamoto Kazuhiro, Tomiki Yuichi, Goto Michitoshi, Kotake Kenjiro, Sugihara Kenichi
Department of Coloproctological Surgery, Juntendo University Faculty of Medicine, Tokyo, Japan,
Ann Surg Oncol. 2015 Feb;22(2):528-34. doi: 10.1245/s10434-014-4015-9. Epub 2014 Aug 27.
We retrospectively examined the optimal lymph node ratio (LNR) cutoff value and attempted to construct a new classification using the LNR in stage III colon cancer.
The clinical and pathological data of 4,172 patients with histologically proven lymph node metastasis who underwent curative surgery for primary colon cancer at multiple institutions between 1995 and 2004 were derived from the multi-institutional database of the Japanese Society for Cancer of the Colon and Rectum (JSCCR). We determined independent prognostic factors and constructed a new classification using these factors. Finally, we compared the discriminatory ability between the new classification and the TNM seventh edition (TNM 7th) classification.
The optimal LNR cutoff value was 0.18. Multivariate analysis revealed that year of surgery, age, gender, histological type, TNM 7th T category, lymphatic invasion, venous invasion, TNM 7th N category, and LNR were found to be significant independent prognostic factors. We attempted to construct a new classification based on the combination of TNM 7th T category and LNR. As a result, the cancer-specific survivals were well stratified (P < .0001). According to the Akaike's information criteria value, the new classification was judged to be superior to the TNM 7th classification with respect to both a better fit and lower complexity.
The optimal LNR cutoff value that was found using the Japanese multi-institutional database and the new classification using LNR are considered to be extremely significant. Therefore, these findings strongly support the application of LNR in the stage classification in stage III colon cancer.
我们回顾性研究了Ⅲ期结肠癌的最佳淋巴结比率（LNR）临界值，并尝试使用LNR构建新的分类方法。
1995年至2004年间，在多家机构接受原发性结肠癌根治性手术且经组织学证实有淋巴结转移的4172例患者的临床和病理数据，来自日本结直肠癌学会（JSCCR）的多机构数据库。我们确定了独立的预后因素，并使用这些因素构建了新的分类方法。最后，我们比较了新分类方法与第七版TNM（TNM 7th）分类方法的鉴别能力。
最佳LNR临界值为0.18。多变量分析显示，手术年份、年龄、性别、组织学类型、TNM 7th T分期、淋巴管侵犯、静脉侵犯、TNM 7th N分期和LNR是显著的独立预后因素。我们尝试基于TNM 7th T分期和LNR的组合构建新的分类方法。结果，癌症特异性生存率得到了很好的分层（P <.0001）。根据赤池信息准则值，新分类方法在拟合优度和复杂度方面均优于TNM 7th分类方法。
使用日本多机构数据库得出的最佳LNR临界值以及使用LNR的新分类方法被认为具有极其重要的意义。因此，这些发现有力地支持了LNR在Ⅲ期结肠癌分期分类中的应用。
