Predicting the risk of lymph node invasion during radical prostatectomy using the European association of urology guideline nomogram: A validation study

• Published in: European journal of surgical oncology 2012-07, Vol.38 (7), p.624-629
• Main Authors: Schmitges, J, Karakiewicz, P.I, Sun, M, Abdollah, F, Budäus, L, Isbarn, H, Bianchi, M, Trinh, Q.-D, Schlomm, T, Chun, F.K, Graefen, M, Steuber, T
• Format: Article
• Language: English
• Subjects: Adult; Aged; Biomarkers, Tumor - blood; Calibration; Confounding Factors (Epidemiology); Europe; Germany; Hematology, Oncology and Palliative Medicine; Humans; Lymph Node Excision; Lymph node excision/statistics & numerical data; Lymph Nodes - pathology; Lymphatic Metastasis; Lymphatic metastasis/diagnosis; Male; Middle Aged; Neoplasm Grading; Neoplasm Staging; Nomograms; Pelvis; Practice Guidelines as Topic; Predictive Value of Tests; Prognosis; Prostate-Specific Antigen - blood; Prostatectomy; Prostatic neoplasm/pathology; Prostatic Neoplasms - immunology; Prostatic Neoplasms - pathology; Prostatic Neoplasms - surgery; Reproducibility of Results; Risk Assessment; Risk Factors; Societies, Medical; Surgery
• ISSN: 0748-7983; 1532-2157
• DOI: 10.1016/j.ejso.2012.04.005

Abstract Background The 2011 European Association of Urology (EAU) guidelines for prostate cancer recommend a pelvic lymph node dissection (PLND) at radical prostatectomy (RP) in all individuals with a nomogram predicted lymph node invasion (LNI) risk of >7%. Methods To test the performing characteristics for several thresholds (1–14%) and to examine the overall accuracy and calibration plot of the EAU nomogram at our institution. The study population consisted of 3081 patients treated with RP and PLND limited to the obturator fossa and the external iliac vein between 2008 and 2010 at a single European institution from Germany. More extensive PLNDs were performed at the surgeon’s discretion. Results Overall, 260 patients (9.2%) had LNI. The 7% threshold would have avoided 30% of PLNDs, at the cost of missing 8% of patients with LNI. The use of 8% and 9% threshold would have allowed the avoidance of respectively 39% and 48% of PLNDs, at the cost of missing respectively 12% and 14% of patients with LNI. The accuracy of the LNI nomogram was 78%, and the unadjusted departure from ideal calibration was 5.3%. Conclusions We confirmed adequate accuracy and calibration of the LNI nomogram. The 7% cut-off may be overly conservative. Better trade-offs between avoided PLNDs and missed LNI cases may be achieved with a limit of 8 or even 9%.