A simple scoring system for predicting inguinal lymph node involvement and survival in patients with penile squamous
Yepeng Guo1, Zijian Cai2, Antonio Augusto Ornellas3
1Department of Urology, Shenzhen People's Hospital, The Second Clinical Medical College, Jinan University, Shenzhen, China.
Purpose:
Currently, no standardized predictive model is widely adopted in clinical practice for assessing outcome and lymph node metastasis (LNM). This study aimed at predicting inguinal lymph node involvement and mortality in patients suffering penile squamous cell carcinoma (PSCC) by developing a simple and clinically applicable scoring system.
Methods:
Clinical and histopathological data from 11 centers, collected between 2000 and 2021, were reclassified (the Eighth Edition of the AJCC Cancer Staging Manual). Data from 8 centers (517 patients) were randomized into training and internal validation cohorts (7:3), and data from the remaining 3 centers contributed 201 patients for external validation. A dual-outcome prediction model was established with univariate and multivariate logistic regression for LNM prediction, and Cox regression for outcome assessment. Kaplan-Meier curves were drawn to estimate the disease-specific survival (DSS), alongside the log-rank test conducted on the difference in survival. Model performance was evaluated through the AUC, calibration plots, Hosmer-Lemeshow goodness-of-fit test, along with decision curve analysis (DCA).
Results:
The study enrolled 718 patients with penile cancer aged at 55 years (IQR: 46-65). Perineural invasion (PNI), lymphovascular invasion (LVI), tumor grade and T stage were independently associated with both LNM and survival (all P < 0.05). For LNM prediction, the scoring system achieved AUCs of 0.80 (95% CI: 0.75-0.84), 0.79 (95% CI: 0.72-0.86) and 0.76 (95% CI: 0.68-0.83) in the training, internal validation, and external validation cohorts, respectively. For 3-year DSS prediction, the time-dependent AUCs were 0.86 (95% CI: 0.81-0.91), 0.84 (95% CI: 0.75-0.92), and 0.87 (95% CI: 0.79-0.94) in respective cohorts. The scoring system was adopted to categorize patients into different risk groups for both LNM and 3-year DSS (Log-rank P < 0.05). Calibration plots revealed a strong consistency between predicted and actual probabilities for both LNM and 3-year DSS across all cohorts. DCA confirmed the favorable predictive accuracy of the scoring system for both inguinal lymph node involvement and mortality in PSCC.
Conclusion:
This simple scoring system demonstrates robust predictive accuracy for inguinal LNM and DSS in PSCC, which practically benefits the risk stratification and personalized management decisions in penile cancer care.
