Predictors of local recurrence and its impact on survival after glansectomy for penile cancer: time to challenge the dogma?

Objectives: To identify predictive pathological factors for local recurrence (LR) and to study the impact of LR on survival in patients treated with glansectomy for penile squamous cell carcinoma (pSCC). Patients and Methods: We retrospectively studied patients treated with glansectomy at international, high-volume reference centres. We analysed histopathological predictors of LR, stratified patients into risk groups based on the number of risk factors present, and studied the impact of LR on survival outcomes using Kaplan–Meier survival analysis and stepwise Cox proportional hazards regression models. Subsequently, we performed sensitivity analyses excluding margin-positive cases, pT3 disease, and cN+ disease, or all of these factors. Results: Across nine institutions, 897 patients were included, of whom 94 experienced LR. On multivariable analysis, presence of high-grade disease and pT3 stage were independent predictors of LR. LR-free survival rates significantly differed according to the number of risk factors present, with a hazard ratio (HR) of 1.90 (95% confidence interval [CI] 1.17–3.07; P = 0.01) for the intermediate-risk group (one risk factor) and 6.11 (95% CI 3.47–10.77; P < 0.001) for the high-risk group (two risk factors), using the low-risk group (no risk factors) as reference. Patients who experienced LR had significantly worse overall survival (OS; HR 2.89, 95% CI 2.02–4.14; P < 0.001) and cancer-specific survival (CSS; HR 5.64, 95% CI 3.45–9.22; P < 0.001). LR (HR 3.82, 95% CI 2.14–6.8; P < 0.001), lymphovascular invasion and cN status were significant predictors of decreased CSS. LR remained a strong predictor of both OS and CSS in all sensitivity analyses. Conclusions: Pathological T3 stage and presence of high-grade disease were independent histopathological predictors of LR after glansectomy for primary pSCC, which allowed risk stratification into three groups with significantly different risk of developing LR. Additionally, LR is related to poor OS and CSS, indicating that LR is a manifestation of underlying aggressive disease and clearly challenging the dogma of using organ-sparing surgery whenever possible since survival is unaffected by higher LR rates.