切换至 "中华医学电子期刊资源库"

中华结直肠疾病电子杂志 ›› 2024, Vol. 13 ›› Issue (05) : 389 -393. doi: 10.3877/cma.j.issn.2095-3224.2024.05.006

论著

随机森林模型预测T1 期结直肠癌淋巴结转移的初步研究
张伟伟1, 陈启1, 翁和语1, 黄亮2,()   
  1. 1.211100 南京医科大学附属逸夫医院肛肠外科
    2.510655 广州,中山大学附属第六医院结直肠肛门外科
  • 收稿日期:2024-07-09 出版日期:2024-10-25
  • 通信作者: 黄亮
  • 基金资助:
    中山六院临床医学研究计划项目[No. 1010CG(2023)-07]

Preliminary study on prediction of lymph node metastasis of T1 colorectal cancer by random forest model

Weiwei Zhang1, Qi Chen1, Heyu Weng1, Liang Huang2,()   

  1. 1.Department of Colorectal Surgery, Sir Run Run Hospital, Nanjing Medical University, Nanjing 211100, China
    2.Department of Colorectal Surgery, the Sixth Affiliated Hospital of Sun Yat Sen University, Guangzhou 510655, China
  • Received:2024-07-09 Published:2024-10-25
  • Corresponding author: Liang Huang
引用本文:

张伟伟, 陈启, 翁和语, 黄亮. 随机森林模型预测T1 期结直肠癌淋巴结转移的初步研究[J]. 中华结直肠疾病电子杂志, 2024, 13(05): 389-393.

Weiwei Zhang, Qi Chen, Heyu Weng, Liang Huang. Preliminary study on prediction of lymph node metastasis of T1 colorectal cancer by random forest model[J]. Chinese Journal of Colorectal Diseases(Electronic Edition), 2024, 13(05): 389-393.

目的

探索T1 期结直肠癌淋巴结转移的危险因素,建立预测T1 期结直肠癌患者淋巴结转移的随机森林模型。

方法

收集2010 年8 月至2020 年6 月于中山大学附属第六医院进行根治性手术治疗的280 例T1 期结直肠癌患者的临床资料,运用R 软件建立预测T1 期结直肠癌患者淋巴结转移的随机森林模型。

结果

T1 期结直肠癌淋巴结转移发生率为11.07%(31/280)。随机森林模型显示,T1 期结直肠癌发生淋巴结转移的危险因素重要性排序为:年龄、脉管神经浸润、肿瘤位置、分化程度、性别、肿瘤大小和Ki67。随机森林模型预测T1 期结直肠癌淋巴结转移曲线下面积(AUC)为0.811,最佳阈值是0.072,特异度和敏感度分别为0.703 和0.778。

结论

随机森林模型有助于临床医师快捷地分析T1 期结直肠癌淋巴结转移风险,并制定合适的治疗方案。

Objective

In order to explore the risk factors of lymph node metastasis of stage T1 colorectal cancer and establish a random forest model to predict lymph node metastasis of stage T1 colorectal cancer patients.

Methods

We collected the clinical data of 280 patients with stage T1 colorectal cancer who underwent radical surgery in the Sixth Affiliated Hospital of Sun Yat-sen University from August 2010 to June 2020,and established a random forest model to predict lymph node metastasis of stage T1 colorectal cancer patients by R software.

Results

The incidence of lymph node metastasis in T1 stage colorectal cancer was 11.07% (31/280). The importance ranking of risk factors following the random forest model for lymph node metastasis in stage T1 colorectal cancer was as follows: age,vascular or neural invasion,tumor location,differentiation degree,gender,tumor size,and Ki67 level. The area under the curve (AUC) of our random forest model for lymph node metastasis in stage T1 colorectal cancer was 0.811,with an optimal threshold of 0.072,specificity of 0.703 and sensitivity of 0.778.

Conclusion

The random forest model is helpful for clinicians to quickly analyze the risk of lymph node metastasis in stage T1 colorectal cancer and make appropriate treatment plans.

图1 T1 期结直肠癌发生淋巴结转移危险因素的重要性排序。注:Mean Decrease GiNi 表示基尼指数;Predictor 表示预测指标;AGE 表示年龄;V.N_Infi 表示血管神经浸润;Tum_loc 表示肿瘤位置;Dif_len 表示肿瘤分化程度;SEX表示性别;Tum_len 表示肿瘤大小
表1 训练组和测试组的基线资料比较[例(%)]
图2 随机森林预测模型的ROC 曲线图。注:Specificity表示特异度;Sensitivity 表示敏感度
图3 随机森林预测模型的校准曲线。注:Predicted Probability 表示预测概率;Observed Probability 表示实际概率
图4 随机森林预测模型的DCA。注:Cost:Benefit Ratio 表示成本收益比;High Risk Threshold 表示高风险阈值;Net Benefit 表示净收益值
图5 随机森林预测模型的CIC。注:Cost:Benefit Ratio 表示成本收益比;High Risk Threshold 表示高风险阈值;Number high risk(out of 1 000)表示高风险人数(‰)
[1]
Bray F,Laversanne M,Sung H,et al. Global cancer statistics 2022:GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries[J]. CA Cancer J Clin,2024,74(3): 229-263.
[2]
国家卫生健康委员会医政司,中华医学会肿瘤学分会. 国家卫健委中国结直肠癌诊疗规范(2023 版)[J]. 中国实用外科杂志,2023,43(6): 602-630.Hospital Authority of National Health Commission of the People's Republic of China; Chinese Society of Oncology,Chinese Medical Association. Chinese protocol of diagnosis and treatment of colorectal cancer of the National Health Commission (2023 edition)[J]. Chinese Journal of Practical Surgery,2023,43(6): 602-630.
[3]
Ichimasa K,Kudo SE,Miyachi H,et al. Risk stratification of T1 colorectal cancer metastasis to lymph nodes: current status and perspective[J]. Gut Liver,2021,15(6): 818-826.
[4]
Saitoh Y,Inaba Y,Sasaki T,et al. Management of colorectal T1 carcinoma treated by endoscopic resection[J]. Dig Endosc,2016,28(3): 324-329.
[5]
Bosch SL,Teerenstra S,de Wilt JH,et al. Predicting lymph node metastasis in pT1 colorectal cancer: a systematic review of risk factors providing rationale for therapy decisions[J]. Endoscopy,2013,45(10):827-834.
[6]
Shin-Ei K,Katsuro I,Benjamin V,et al. Artificial intelligence system to determine risk of T1 colorectal cancer metastasis to lymph node[J].Gastroenterology,2021,160(4): 1075-1084.e2.
[7]
Miyazaki K,Wada Y,Okuno K,et al. An exosome-based liquid biopsy signature for pre-operative identification of lymph node metastasis in patients with pathological high-risk T1 colorectal cancer[J]. Mol Cancer,2023,22(1): 2.
[8]
Vermeer N,Backes Y,Snijders HS,et al. National cohort study on postoperative risks after surgery for submucosal invasive colorectal cancer[J]. BJS Open,2019,3(2): 210-217.
[9]
黄亮,蔡永华,张兴伟,等. 中低位直肠癌新辅助放化疗后病理完全缓解预测因素分析[J]. 中国实用外科杂志,2019,39(7): 694-697.Huang L,Cai YH,Zhang XW,et al. Clinical verification of sensitivity to neoadjuvant chemoradiotherapy in cases of locally advanced rectal cancer[J]. Chinese Journal of Practical Surgery,2019,39(7): 694-697.
[10]
Wolters U,Stützer H,Keller HW,et al. Colorectal cancer--a multivariate analysis of prognostic factors[J]. Eur J Surg Oncol,1996,22(6): 592-597.
[11]
Miyachi H,Kudo SE,Ichimasa K,et al. Management of T1 colorectal cancers after endoscopic treatment based on the risk stratification of lymph node metastasis[J]. J Gastroenterol Hepatol,2016,31(6): 1126-1132.
[12]
Bianco F,De Franciscis S,Belli A,et al. T1 colon cancer in the era of screening: risk factors and treatment[J]. Tech Coloproctol,2017,21(2):139-147.
[13]
Fang WL,Chang SC,Lin JK,et al. Metastatic potential in T1 and T2 colorectal cancer[J]. Hepatogastroenterology,2005,52(66): 1688-1691.
[14]
Backes Y,Elias SG,Groen JN,et al. Histologic factors associated with need for surgery in patients with pedunculated T1 colorectal carcinomas[J]. Gastroenterology,2018,154(6): 1647-1659.
[15]
Ozawa T,Kandimalla R,Gao F,et al. A microRNA signature associated with metastasis of T1 colorectal cancers to lymph nodes[J].Gastroenterology,2018,154(4): 844-848.e7.
[16]
Kang J,Choi YJ,Kim IK,et al. LASSO-based machine learning algorithm for prediction of lymph node metastasis in T1 colorectal cancer[J]. Cancer Res Treat,2021,53(3): 773-783.
[17]
Ichimasa K,Kudo SE,Mori Y,et al. Artificial intelligence may help in predicting the need for additional surgery after endoscopic resection of T1 colorectal cancer[J]. Endoscopy,2018,50(3): 230-240.
[18]
陈小鹏,王会东. 基于随机森林模型的乳腺癌患者术后并发下肢深静脉血栓预测模型的初步研究[J]. 血管与腔内血管外科杂志,2023,9(8): 987-991,1024.Chen XP,Wang HD. Preliminary study on the prediction model of postoperative deep vein thrombosis of lower extremity in breast cancer patients based on random forest model[J]. Journal of Vascular and Endovascular Surgery,2023,9(8): 987-991,1024.
[1] 明昊, 肖迎聪, 巨艳, 宋宏萍. 乳腺癌风险预测模型的研究现状[J]. 中华乳腺病杂志(电子版), 2024, 18(05): 287-291.
[2] 吴令杰, 陈瑞烈, 肖湘明, 郭耿龙, 林钟滨, 张海生, 周敏, 陈妙华. 2005至2021年广东省汕头市某综合医院人类免疫缺陷病毒感染/获得性免疫缺陷综合征患者流行病学特征及高效抗逆转录病毒治疗后生存状况影响因素[J]. 中华实验和临床感染病杂志(电子版), 2024, 18(04): 207-214.
[3] 罗青杉, 梅海涛, 郝家领, 蔡锦锋, 周润楷, 温玉刚. 连接蛋白43通过调控细胞周期抑制结直肠癌的增殖机制研究[J]. 中华普通外科学文献(电子版), 2024, 18(05): 344-349.
[4] 罗文斌, 韩玮. 胰腺癌患者首次化疗后中重度骨髓抑制的相关危险因素分析及预测模型构建[J]. 中华普通外科学文献(电子版), 2024, 18(05): 357-362.
[5] 李国煜, 丛赟, 祖丽胡马尔·麦麦提艾力, 何铁英. 急性胰腺炎并发门静脉系统血栓形成的危险因素及预测模型构建[J]. 中华普通外科学文献(电子版), 2024, 18(04): 266-270.
[6] 贺斌, 马晋峰. 胃癌脾门淋巴结转移危险因素[J]. 中华普外科手术学杂志(电子版), 2024, 18(06): 694-699.
[7] 黄一博, 李至彦, 林晨, 陶亮, 王萌, 管文贤. 胃癌根治术中淋巴结示踪剂的研究进展[J]. 中华普外科手术学杂志(电子版), 2024, 18(05): 586-588.
[8] 韩婧, 郝少龙, 康骅. 北京市单中心甲状腺癌患者临床特征的回顾分析[J]. 中华普外科手术学杂志(电子版), 2024, 18(05): 490-493.
[9] 黄俊龙, 刘柏隆, 罗瑞翔, 李晓阳, 李文双, 柳政, 陈嘉良, 周祥福. 联合盆底彩超数据和临床资料探讨压力性尿失禁的危险因素[J]. 中华腔镜泌尿外科杂志(电子版), 2024, 18(04): 323-330.
[10] 周艳, 李盈, 周小兵, 程发辉, 何恒正. 不同类型补片联合Nissen 胃底折叠术修补食管裂孔疝的疗效及复发潜在危险因素[J]. 中华疝和腹壁外科杂志(电子版), 2024, 18(05): 528-533.
[11] 任佳琪, 刁德昌, 何自衍, 张雪阳, 唐新, 李文娟, 李洪明, 卢新泉, 易小江. 网膜融合线导向的脾曲游离技术在左半结肠癌根治术中的应用[J]. 中华结直肠疾病电子杂志, 2024, 13(05): 362-367.
[12] 张迪, 王春霞, 张学东, 李发馨, 庞淅文, 陈一锋, 张维胜, 王涛. 梗阻性左半结直肠癌自膨式金属支架置入后行腹腔镜手术与开腹手术的短期临床疗效比较[J]. 中华结直肠疾病电子杂志, 2024, 13(05): 375-380.
[13] 张蔚林, 王哲学, 白峻阁, 黄忠诚, 肖志刚. 利用TCGA数据库构建基于miRNA的结直肠癌列线图预后模型[J]. 中华结直肠疾病电子杂志, 2024, 13(05): 381-388.
[14] 王曦娅, 尹弘青, 丁伟, 徐滨, 于海源, 马东升, 邵军. 桥本背景下甲状腺乳头状癌多参数分析预测大容量淋巴结转移[J]. 中华临床医师杂志(电子版), 2024, 18(06): 548-554.
[15] 黄镪, 孙金梅, 韩燕飞, 张拥波. 脑源性与非脑源性疾病所致应激性溃疡相关胃肠道出血的影响因素及临床预后差异:一项回顾性队列研究[J]. 中华脑血管病杂志(电子版), 2024, 18(04): 309-316.
阅读次数
全文


摘要