Introduction

The optimal primary treatment for men with clinically high-risk prostate cancer (PCa) is controversial as both radical prostatectomy (RP) and radiotherapy (RT) are associated with potential advantages and disadvantages. Our objective was to compare the overall mortality-free survival of high-risk PCa patients treated with primary RP vs. primary RT with neoadjuvant/adjuvant androgen deprivation therapy [ADT], within the National Cancer Data Base (NCDB).

Methods

Within the NCDB, a total of 87,875 high-risk PCa patients fulfilled our pre-specified inclusion criteria (53,197 in RP group and 34,678 in RT+ADT group). We employed an instrumental variable analysis (IVA) approach using the yearly rate of RP as the instrument, to mitigate the impact of both observed and unobserved confounders. Multiple sensitivity analyses were performed, including stratification for age, comorbidity, ADT utilization and high dose (>75.6 Gy) RT. In addition, the overall mortality-free survival of RP was compared to that of RT reported in three recently published randomized controlled trails (RCTs), after selecting only RP patients who fitted inclusion/exclusion criteria of these RCTs

Results

On IVA adjusting for socio-demographic, facility- and tumor-specific covariates, RP was associated with lower overall mortality compared to RT+ADT (hazard ratio (HR) 0.52; 95% CI, 0.47-0.57; p<0.001) in the overall analysis, in patients with age ?65 years with CCI 0 (HR 0.48; p<0.001), in patients >65 years with CCI 0 (0.53; p<0.001), those receiving RT with neoadjuvant (HR 0.52; p<0.001) or adjuvant ADT (HR 0.47; p<0.001), or treated with high dose (?75.6 Gy) RT (HR 0.54; p<0.001). While the survival outcomes for patients treated with RT (+/-ADT) in the RCTs were not statistically different from similarly treated and appropriately selected patients within the NCDB, RP was associated with greater overall mortality-free survival than any of the arms represented in the RCTs.

Conclusions

Our results suggest that in patients with clinically high-risk PCa, primary RP is associated with greater overall mortality-free survival than primary RT+ADT in patients with clinically high-risk PCa, regardless of baseline characteristics. These findings, in lieu of a randomized trial, can guide the clinicians to carefully choose the primary modality of treatment for patients with high-risk PCa.

Funding

None