Sparing Unnecessary Surgery: A Model with Intraductal Carcinoma Safely Reduces Extended Lymph Node Dissection by 31% in Prostate Cancer Patients

Background Accurate preoperative prediction of lymph node invasion(LNI) is critical for optimizing the use of extended pelvic lymph node dissection(ePLND) in prostate cancer(PCa) management.While established nomograms,such as those from Briganti(2012,2017) and the MSKCC,are widely used for this purpose,their predictive accuracy within Chinese PCa populations remains insufficiently validated.Furthermore,these models often lack systematic integration of lymphovascular invasion(LVI),a key histopathological feature strongly associated with adverse prognostic outcomes,including increased risk of lymph node metastasis.To address these limitations,this study aimed to develop and validate a novel nomogram for estimating LNI risk,specifically taliored to a Chinese high-risk PCa cohort.This model incorporates LVI alongside other critical clinical and pathological parameters,with the goal of improving predictive accuracy and guiding more personalized surgical decision-marking. Methods We conducted a retrospective analysis of 325 Chinese patients with prostate cancer who underwent radical prostatectomy with extended pelvic lymph node dissection (ePLND). In this cohort, we performed external validation of three established nomograms: the Briganti 2017, Briganti 2012, and Memorial Sloan Kettering Cancer Center (MSKCC) models. Independent predictors of lymph node invasion (LNI) were identified using multivariable logistic regression analysis. The newly developed nomogram was evaluated for discriminative ability using the area under the receiver operating characteristic curve (AUC) and for calibration using calibration plots. Clinical utility was further assessed with decision curve analysis (DCA). Results Our nomogram demonstrated excellent discriminative ability with an AUC of 0.89 (95% CI: 0.85–0.93) (Figure 2), significantly outperforming the established Briganti 2017 (AUC: 0.76), Briganti 2012 (AUC: 0.81), and MSKCC (AUC: 0.79) models. Decision curve analysis revealed a substantial net clinical benefit across threshold probabilities of 5%–80% (Figure 3). Critically, at the clinically relevant 7% risk threshold, application of our model would have avoided 31.3% (102/325) of ePLND procedures while missing only 1.7% (2/114) of lymph node invasion cases (Table 3). This represents a marked improvement over existing nomograms, which achieved avoidance rates of ≤7.4% at comparable thresholds. Conclusions We developed and internally validated a novel nomogram incorporating lymphovascular invasion (LVI) for preoperative prediction of lymph node invasion in Chinese high-risk prostate cancer patients. With superior discriminative accuracy (AUC: 0.89) and favorable clinical utility, this model enables safe reduction of unnecessary ePLND procedures by 31.3% at a 7% risk threshold, minimizing surgical morbidity without compromising oncologic safety. This tool addresses a critical gap in personalized nodal staging strategies for high-risk prostate cancer.