[Construction and evaluation of a nomogram for predicting the prognosis of patients with colorectal cancer with peritoneal carcinomatosis treated with cytoreductive surgery plus hyperthermic intraperitoneal chemotherapy].

Objectives: To construct a nomogram incorporating important prognostic factors for predicting the overall survival of patients with colorectal cancer with peritoneal metastases treated with cytoreductive surgery (CRS) plus hyperthermic intraperitoneal chemotherapy (HIPEC), the aim being to accurately predict such patients' survival rates. Methods: This was a retrospective observational study. Relevant clinical and follow-up data of patients with colorectal cancer with peritoneal metastases treated by CRS + HIPEC in the Department of Peritoneal Cancer Surgery, Beijing Shijitan Hospital, Capital Medical University from 2007 January to 2020 December were collected and subjected to Cox proportional regression analysis. All included patients had been diagnosed with peritoneal metastases from colorectal cancer and had no detectable distant metastases to other sites. Patients who had undergone emergency surgery because of obstruction or bleeding, or had other malignant diseases, or could not tolerate treatment because of severe comorbidities of the heart, lungs, liver or kidneys, or had been lost to follow-up, were excluded. Factors studied included: (1) basic clinicopathological characteristics; (2) details of CRS+HIPEC procedures; (3) overall survival rates; and (4) independent factors that influenced overall survival; the aim being to identify independent prognostic factors and use them to construct and validate a nomogram. The evaluation criteria used in this study were as follows. (1) Karnofsky Performance Scale (KPS) scores were used to quantitatively assess the quality of life of the study patients. The lower the score, the worse the patient's condition. (2) A peritoneal cancer index (PCI) was calculated by dividing the abdominal cavity into 13 regions, the highest score for each region being three points. The lower the score, the greater is the value of treatment. (3) Completeness of cytoreduction score (CC), where CC-0 and CC-1 denote complete eradication of tumor cells and CC-2 and CC-3 incomplete reduction of tumor cells. (4) To validate and evaluate the nomogram model, the internal validation cohort was bootstrapped 1000 times from the original data. The accuracy of prediction of the nomogram was evaluated with the consistency coefficient (C-index), and a C-index of 0.70-0.90 suggest that prediction by the model was accurate. Calibration curves were constructed to assess the conformity of predictions: the closer the predicted risk to the standard curve, the better the conformity. Results: The study cohort comprised 240 patients with peritoneal metastases from colorectal cancer who had undergone CRS+HIPEC. There were 104 women and 136 men of median age 52 years (10-79 years) and with a median preoperative KPS score of 90 points. There were 116 patients (48.3%) with PCI≤20 and 124 (51.7%) with PCI>20. Preoperative tumor markers were abnormal in 175 patients (72.9%) and normal in 38 (15.8%). HIPEC lasted 30 minutes in seven patients (2.9%), 60 minutes in 190 (79.2%), 90 minutes in 37 (15.4%), and 120 minutes in six (2.5%). There were 142 patients (59.2%) with CC scores 0-1 and 98 (40.8%) with CC scores 2-3. The incidence of Grade III to V adverse events was 21.7% (52/240). The median follow-up time is 15.3 (0.4-128.7) months. The median overall survival was 18.7 months, and the 1-, 3- and 5-year overall survival rates were 65.8%, 37.2% and 25.7%, respectively. Multivariate analysis showed that KPS score, preoperative tumor markers, CC score, and duration of HIPEC were independent prognostic factors. In the nomogram constructed with the above four variables, the predicted and actual values in the calibration curves for 1, 2 and 3-year survival rates were in good agreement, the C-index being 0.70 (95% CI: 0.65-0.75). Conclusions: Our nomogram, which was constructed with KPS score, preoperative tumor markers, CC score, and duration of HIPEC, accurately predicts the survival probability of patients with peritoneal metastases from colorectal cancer treated with cytoreductive surgery plus hyperthermic intraperitoneal chemotherapy.目的： 评估影响肿瘤细胞减灭术（CRS）加腹腔热灌注化疗（HIPEC）治疗结直肠癌腹膜转移患者总体生存的重要预后因素，并生成有效的列线图模型，以预测该类患者的总生存率。 方法： 采用回顾性观察性研究的方法，回顾性收集2007年1月至2020年12月期间，在首都医科大学附属北京世纪坛医院接受CRS+HIPEC综合治疗的结直肠腹膜转移患者的临床资料与预后信息。纳入的病例均为确诊结直肠癌腹膜转移且腹膜转移为单一转移部位，未合并其他部位远处转移者，排除因梗阻或出血行急诊手术，合并其他恶性肿瘤，合并严重的心、肺、肝、肾疾病而不能耐受治疗，以及失访者。本研究观察指标包括：（1）基本临床病理特征；（2）CRS+HIPEC治疗情况；（3）总生存期和总体生存率；（4）影响总体生存的独立预后因素；（5）基于独立预后因素构建列线图预测模型，并评估模型的性能。本研究采用的评价标准包括：（1）卡氏评分（KPS）：评估肿瘤患者生活质量，评分越低则患者状态越差；（2）腹膜癌指数（PCI）：即将整个腹腔分为13个区域，每个区域最高得分为3分，分数越低，治疗价值越大；（3）肿瘤细胞减灭程度评分（CC）：CC-0和CC-1指完全肿瘤细胞减灭，CC-2和CC-3为不完全肿瘤细胞减灭；（4）列线图预测模型的验证和评估：从原始数据中随机抽样（Bootstrap）1 000次组成内部验证数据集，使用该数据集进行列线图内部验证，当C-index值为0.70~0.90时，表示模型预测准确度较高；用校准曲线判断预测符合度，预测风险越接近于标准曲线，则模型的符合度越好。 结果： 共纳入240例接受CRS+HIPEC综合治疗的结直肠癌腹膜转移患者，女性104例，男性136例；中位年龄52（10~79）岁；术前中位KPS评分90分；PCI≤20分者116例（48.3%），>20分者124例（51.7%）；术前肿瘤标志物异常175例（72.9%），正常38例（15.8%）。HIPEC时长30 min、60 min、90 min和120 min的患者分别有7例（2.9%）、190例（79.2%）、37例（15.4%）和6例（2.5%）；CC评分0~1分142例（59.2%），2~3分98例（40.8%）。术后Ⅲ~Ⅴ级不良事件发生率21.7%（52/240）。全组患者中位随访15.3（0.4~128.7）个月，中位总生存期为18.7个月，1、3和5年总体生存率分别为65.8%、37.2%和25.7%。多因素分析结果显示，KPS评分、术前肿瘤标志物、CC评分和HIPEC时长均为影响CRS+HIPEC综合治疗结直肠癌腹膜转移预后的独立因素（均P<0.05）。采用上述4个变量构建的列线图模型，发现其在预测1、2、3年生存率方面表现良好。在1、2、3年生存率校准曲线中预测值和实际值一致性良好，C-index值为0.70（95%CI：0.65~0.75）。 结论： 以KPS评分、术前肿瘤标志物、CC评分和HIPEC时长构建的列线图，可有效预测结直肠癌腹膜转移患者接受CRS+HIPEC后的生存率。.