Predict Lymph Node Metastasis in Penile Cancer Using Clinicopathological Factors and Nomograms.

OBJECTIVE

To investigate the predictive factors of lymph node metastasis (LNM) and evaluate the usefulness of prediction nomograms.

METHODS

This study included 300 patients diagnosed with penile squamous cell carcinoma at West China Hospital (WCH) of Sichuan University (Chengdu, China) and 412 cases acquired from the Surveillance, Epidemiology, and End Results (SEER) program. Logistic regression analysis was performed on these cohorts to investigate the predictive factors of LNM. We evaluated a recently developed prediction nomogram for LNM, which was established based on the National Cancer Database (NCDB). Moreover, we developed a novel nomogram using cases from the WCH for the prediction of lymphatic metastasis.

RESULTS

Logistic analysis identified that younger age at diagnosis, invasion of the penis body, poorer pT stage, cN stage, nuclear grade and the presence of lymph vascular invasion (LVI) were significantly correlated with LNM in WCH cases; however, only race, poorer T stage and cN stage were significantly associated with LNM among the cases from the SEER. Multivariate analysis demonstrated that younger age, poorer T stage, cN stage and nuclear grade were independent predictors of LNM. Receiver operating characteristic curve analysis of WCH cases showed that the tumor T stage 8th edition has better area under the curve than 7th stage (0.672 vs 0.636, respectively). Moreover, well AUC was seen in external validation of NCDB nomogram in WCH cohorts and SEER series (0.833 vs 0.795). The new nomogram included the aforementioned independent predictors and the bootstrap-corrected concordance was 0.876.

CONCLUSION

Younger diagnose age, poorer pT stage, cN stage, nuclear grade and LVI were the most important predictors of LNM in patients with penile cancer. 8th T stage performed better than 7th version in predicting LNM. NCDB nomogram has some application values in both WCH and SEER cases, and our novel model further improved the predictive accuracy.