Analysis of the surgical approach in prostate cancer staging: results from the surveillance, epidemiology and end results program

Abstract

AbstractSurgery is not used as a criterion for staging prostate cancer, although there is evidence that the number of analyzed and affected lymph nodes have prognosis value. The aim of this study was to determine whether there are significant differences in staging criteria in patients who underwent prostatectomy compared to those who did not, and whether the number of affected and analyzed lymph nodes (LN) plays a prognostic role. In this retrospective study, a test cohort consisting of 404,210 newly diagnosed men with prostate cancer, between 2004 and 2010, was obtained from the 17 registries (Nov 2021 submission); a validation consisting of 147,719 newly diagnosed men with prostate cancer between 2004 and 2019 was obtained from the 8 registries (Nov 2021 submission). Prostate cancer-specific survival was analyzed by Kaplan–Meier curves, survival tables and Cox regression; overall survival was analyzed only to compare Harrell's C-index between different staging criteria. In initial analyses, it was observed that the prognostic value of lymph node metastasis changes according to the type of staging (clinical or pathological), which is linked to the surgical approach (prostatectomy). Compared with T4/N0/M0 patients, which are also classified as stage IVA, N1/M0 patients had a shorter [adjusted HR: 1.767 (1429–2184), p < 0.0005] and a longer [adjusted HR: 0.832 (0.740–0.935), p = 0.002] specific survival when submitted to prostatectomy or not, respectively. Analyzing separately the patients who were submitted to prostatectomy and those who were not, it was possible to obtain new LN metastasis classifications (N1: 1 + LN; N2: 2 + LNs; N3: > 2 + LNs). This new (pathological) classification of N allowed the reclassification of patients based on T and Gleason grade groups, mainly those with T3 and T4 disease. In the validation group, this new staging criterion was proven to be superior [specific survival C-index: 0.908 (0.906–0.911); overall survival C-index: 0.788 (0.786–0.791)] compared to that currently used by the AJCC [8th edition; specific survival C-index: 0.892 (0.889–0.895); overall survival C-index: 0.744 (0.741–0.747)]. In addition, an adequate number of dissected lymph nodes results in a 39% reduction in death risk [adjusted HR: 0.610 (0.498–0.747), p < 0.0005]. As main conclusion, the surgery has a major impact on prostate cancer staging, mainly modifying the effect of N on survival, and enabling the stratification of pathological N according to the number of affected LN. Such a factor, when considered as staging criteria, improves the prognosis classification.