Survival after surgical or radiotherapeutic treatment for high-risk localized prostate cancer: A National Cancer Database analysis with comprehensive treatment group imbalances adjustments.

Abstract

55
Background: There are no randomized trials to guide a high-risk localized prostate cancer patient’s treatment decision between radiotherapeutic and surgical options. Comparative studies have been limited by their ability to match patients based on pre-treatment prognostic variables and to adjust for the cancer–related, medical and socioeconomic differences between patients who choose radiotherapeutic or surgical approaches. Methods: We analyzed the outcome of all patients in the National Cancer Database with high-risk clinically localized prostate cancer with complete prognostic data who were treated with either radical prostatectomy (RP), external beam radiotherapy (EBRT) combined with androgen deprivation (AD) or EBRT+brachytherapy (brachy)±AD. Inverse probability of treatment weighting (IPTW) was used to adjust for covariate imbalance between treatment groups. Time-dependent Cox proportional hazards (TDCPH) model was then used to evaluate differences between treatment groups. Because of imbalances in the frequency of pathologic lymph node sampling between the treatment groups, a predictive model of pathologic nodal (pLN) status was built using PSA, Gleason score and clinical T stage and predicted pLN status was used to repeat the IPTW and TDCPH model. Results: 42,765 patients were analyzed. There was no statistically significant difference in survival between RP and EBRT+brachy±AD (HR 1.17 (0.88, 1.55)). However, EBRT+AD was associated with higher mortality than RP (HR 1.53 (1.22, 1.92)). Adjustment for predicted pLN status did not change the results. Subset analysis of EBRT+AD patients who received ≥7920 cGy narrowed the difference but remained significant (HR 1.32 (1.02, 1.66)). Conclusions: After comprehensively adjusting for imbalances in prostate cancer prognostic factors, other medical conditions and socioeconomic factors, this analysis showed no difference in survival between patients treated with RP vs. EBRT+brachy±AD. EBRT+AD was associated with lower survival. Randomized trials comparing these three methods of treatment are urgently needed.