Development and validation of a nomogram for predicting positive surgical margins after robot-assisted radical prostatectomy
10.3760/cma.j.cn112330-20250217-00057
- VernacularTitle:预测机器人辅助根治性前列腺切除术后切缘阳性列线图的建立和验证
- Author:
Zhoujie YE
1
;
Jinpeng SHAO
1
;
Ziyan AN
1
;
Haoyu ZOU
1
;
Zongyu FU
1
;
Kun ZHAO
1
;
Zheng WANG
1
;
Weijun FU
1
Author Information
1. 解放军总医院第三医学中心泌尿外科医学部，北京 100039
- Publication Type:Journal Article
- Keywords:
Robot-assisted radical prostatectomy;
Prostate cancer;
Positive surgical margin;
Nomogram
- From:
Chinese Journal of Urology
2025;46(6):439-446
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To investigate the risk factors for positive surgical margins（PSM）after robot-assisted radical prostatectomy（RARP），and to develop and validate a predictive nomogram.Methods:We retrospectively analyzed the clinicopathological data of 874 prostate cancer patients who underwent RARP performed by a single surgeon at the First Medical Center of Chinese PLA General Hospital between January 2012 and December 2018. Patients were divided into positive surgical margin（n=327）and negative surgical margin（n=547）groups based on postoperative margin status.The PSM group had significantly higher preoperative median tPSA［31.200（19.050，54.400）ng/ml vs. 15.050（9.840，27.590）ng/ml， P<0.01］，higher proportion of patients with PSAD>1 ng/ml 2［49.5%（162/327）vs. 21.2%（116/547）， P<0.01］，biopsy Gleason score ≥8［33.3%（109/327）vs. 21.2%（116/547）， P<0.01］，ISUP grade 4-5［33.3%（109/327）vs. 21.2%（116/547）， P<0.01］，clinical T stage ≥cT 3［11.3%（37/327）vs. 4.2%（23/547）， P<0.01］，and high-risk classification［82.3%（269/327）vs. 55.9%（306/547）， P<0.01］compared to the negative surgical margin group. Conversely，the PSM group had a lower prevalence of hypertension［29.7%（97/327）vs. 40.2%（220/547）， P=0.002］.Patients were randomly split into a training cohort（n=656，75%）and an internal validation cohort（n=218，25%）. An external validation cohort included 71 patients who underwent RARP by different surgeons between January 2014 and December 2016. No significant differences in baseline characteristics were observed between cohorts（ P>0.05）.Univariate and multivariate logistic regression analyses identified independent predictors of PSM，which were incorporated into a nomogram. Predictive performance was assessed using receiver operating characteristic（ROC）curves，decision curve analysis（DCA），and calibration curve. Internal and external validations were performed. Results:The PSM group had longer postoperative hospitalization［6（5，8）vs. 6（5，7）days， P=0.028］，higher rates of pathologic Gleason score ≥8［41.5%（115/277）vs. 24.9%（111/446）， P<0.01］，ISUP grade 4-5［41.5%（115/277）vs. 24.9%（111/446）， P<0.01］，pT 3 stage［52.3%（171/327）vs. 17.4%（95/547）， P<0.01］，pN 1 stage［12.8%（42/327）vs. 3.8%（21/547）， P<0.01］，extracapsular extension［52.3%（171/327）vs. 17.4%（95/547）， P<0.01］，and seminal vesicle invasion［34.6%（113/327）vs. 9.1%（50/547）， P<0.01］.Multivariate analysis identified elevated tPSA（ OR=1.014，95% CI 1.004—1.024，P=0.006）and PSAD ≥0.15 ng/（ml/g）（ OR=11.638，95% CI 1.450—93.396，P=0.021）as independent risk factors for PSM. The area under the ROC curve（AUC）of the nomogram constructed based on the above variables was 0.770（95% CI 0.735—0.805）. The AUC values for the internal and external validation sets were 0.698（95% CI 0.630—0.767）and 0.643（95% CI 0.513—0.774），respectively. The calibration curve demonstrated good agreement between the predicted and observed outcomes，and the DCA indicated that the predictive model has potential clinical utility in decision-making. Conclusion:tPSA and PSAD were identified as independent risk factors for PSM. The nomogram constructed based on these two independent predictive variables effectively predicted PSM after RARP.
