Validation of a nomogram predicting the probability of lymph node invasion based on the extent of pelvic lymphadenectomy in patients with clinically localized prostate cancer

OBJECTIVE To develop a multivariate nomogram to predict the rate of lymph node invasion (LNI) in patients with clinically localized prostate cancer according to the extent of extended pelvic lymphadenectomy (PLND), which is associated with significantly higher rate of LNI. PATIENTS AND METHODS The study comprised 781 consecutive patients (median age 66.6 years, range 45-85) treated with PLND and radical retropubic prostatectomy (RRP) for clinically localized prostate cancer. Their median ( range) prostate-specific antigen (PSA) level was 7 (1.03-49.91) ng/mL, and their clinical stages were T1c in 433 (55.4%), T2 in 328 (42%) and T3 in 20 (2.6%). Biopsy Gleason sums were <= 6 in 514 (65.8%), 7 in 204 (26.1%) and 8 - 10 in 63 (8.1%). Multivariate logistic regression models were used to test the association between predictors including PSA level, biopsy Gleason sum, clinical stage, number of nodes removed and the rate of LNI. Finally, regression coefficients were used to develop a nomogram, which was internally validated with 200 bootstrap re-samples. RESULTS The median (range) number of lymph nodes removed was 14 (2 - 40); LNI was detected in 71 patients (9.1%). The univariate predictive accuracy for total PSA level, clinical stage, biopsy Gleason sum and number of total nodes removed and examined was 64.2%, 59.8%, 74% and 62.9%, respectively. Except for PSA (P = 0.2), all variables were statistically significant multivariate predictors of LNI at RRP (P <= 0.001). A nomogram based on clinical stage, PSA level, biopsy Gleason sum and the number of total lymph nodes removed was 78.6% accurate, and 1.8% more accurate than a nomogram without the number of removed lymph nodes. CONCLUSIONS The extent of PLND is directly related to the probability of LNI. The risk of LNI increases linearly, and is proportional to the number of nodes removed and examined. The effect of the increased probability of LNI is weighted more heavily in men with more advanced clinical stage and grade.