Analysis of risk factors and construction and verificantion of prediction model for tumor recurrence in upper tract urothelial carcinoma patients after radical nephroureterectomy combined with intravenous chemotherapy
10.3760/cma.j.cn115396-20200807-00245
- VernacularTitle:上尿路上皮癌行根治术联合静脉化疗后肿瘤复发的风险因素分析及预测模型的构建与验证
- Author:
Jimeng RUAN
1
;
Miaomiao WANG
;
Xiangyu WANG
;
Weisi XING
;
Meiyuan CHEN
;
Xin TONG
;
Jing XIAO
Author Information
1. 首都医科大学附属北京友谊医院泌尿外科 100050
- From:
International Journal of Surgery
2020;47(9):584-589,f3
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To explore the risk factors of tumor recurrence after radical nephroureterectomy combined with Gemcitabine and Cisplatin(GC) systemic intravenous chemotherapy for upper tract urothelial carcinoma (UTUC), establish a recurrence risk prediction model, and conduct preliminary verification.Methods:One hundred and one cases of UTUC were analyzed from January 2013 to October 2019 in Beijing Friendship Hospital, Capital Medical University retrospectively. All patients underwent radical nephroureterectomy+ bladder cuff resection, and were treated with GC intravenous adjuvant chemotherapy, among which 19 underwent preoperative neoadjuvant chemotherapy. The study collected general information and clinical characteristics of the patients, and follow up the patient's recurrence. Tumor recurrence and relapse free survival (RFS) were the main observation indexes. The patients were divided into the recurrent group and the non-recurrent group according to their recurrence. Kaplan-Meier and Log-rank methods were used to estimate and compare the RFS rates of the two groups. Univariate difference analysis was used to identify the indicators that were significantly different between patients in the recurrence group and the non-recurrence group, and the COX proportional hazard model was further used to explore the correlation between each factor and the tumor recurrence. According to the weights of relevant risk factors, an individual prognostic index (PI) equation was established, a recurrence prediction model was constructed, and the receiver operating characteristic (ROC) curve was used for verification.Results:One hundred and one patients were followed up for 2-82 months, with median 22 months. 40 patients had recurrence, including 32 in the bladder and 8 in the contralateral upper urinary. One-year RFS was 82.10%, two-year RFS was 68.90% and 5-year RFS was 42.10%. COX proportional risk model results showed that tumor hydronephrosis (X1), lymphovascular invasion (X2) and tumor T stage (X3) were independent risk factors, while neoadjuvant chemotherapy (X4) and chemotherapy cycle (X5) were independent protective factors. Individual PI equation =0.964X1+ 0.688X2+ 0.508X3-1.566 X4-0.675X5. The ROC curve was drawn to show that the optimal pointcut value was 179.5 when the Youden index was 0.537, the sensitivity of the model was 0.750, the specificity was 0.787, and the area under the curve (AUC) was 0.838(95% CI: 0.758-0.918). Conclusions:Hydronephrosis, tumor T stage, lymphovascular invasion, neoadjuvant chemotherapy and chemotherapy cycle are independent factors affecting the recurrence of UTUC patients. The multi-factor risk prediction model is suitable for evaluating the possibility of tumor recurrence after radical surgery combined with GC chemotherapy in UTUC patients, which can provide scientific evidence for the prognosis assessment of patients.
