In this study, the SEARCH database authors focus on this cohort of men and look at the role of primary treatment on subsequent outcomes. Obviously, this is a retrospective study with its inherent biases, particularly selection bias. The SEARCH database utilized data from men treated at Veterans Affairs (VA) hospitals in the United States – hospitals designed for the care of patients with prior military service. In this study, the authors looked at all men diagnosed with mCRPC after the year 2000 and stratified them based on primary prostate treatment: none, prostatectomy ± radiation or radiation alone. ADT utilization with radiotherapy was not clearly identified. The primary outcomes were overall and cancer-specific survival – and time 0 is defined as the diagnosis of metastatic disease.
The identified 729 men diagnosed with mCRPC during that time, of which 284 (39%) underwent no local treatment, 176 (24%) underwent RP ± XRT and 269 (37%) underwent XRT alone.
On multivariable analysis (other variables in the study included: age, race, PSA, PSADT, Grade group, Time from ADT to CRPC, time from CRPC to metastases, type of metastases, and treatment center), men with prior RP ± XRT had improved OS (HR, 0.71; p=0.005) and CSS (HR, 0.55; p<0.001) compared to men with no prior local therapy. These effects were noted on both univariate and multivariable analyses.
Interestingly, this decreased risk of OS (HR, 0.89; p=0.219) and CSS (HR, 0.87; p=0.170) was not seen in men with prior XRT alone. It was noted on univariate analyses but was not borne out on MV analyses.
After further adjusting for comorbidity with Charlson Comorbidity Index, patients with prior local therapy with RP ± XRT still had reduced OS (HR, 0.70; p=0.003) while this was not seen in patients who received prior XRT alone (HR, 0.88, p=0.185). No comment was made on CCS.
In this study, which is the first to assess the benefit of primary therapy on outcomes specifically in mCRPC patients, it would appear that prior RP but not XRT alone was associated with overall and cancer-specific survival benefit, regardless of comorbidities. Some explanations for this include biologic rationale (removal of the primary may reduce metastases seeding), selection bias, and lead-time bias (patients treated with local therapy may have had better follow-up and therefore diagnosed earlier and treated better).
However, despite the authors acknowledging selection bias, it is again important to reiterate that patients who underwent RP were likely better selected for surgery (for factors perhaps not captured by the dataset), this conclusion should be taken with some caution!
Some additional points brought up in the discussion:
1. Patient selection for radiation or surgery – the reasons for choosing one or the other may not be adequately captured and may contribute to differences in outcomes.
Questions that still need to be answered:
1. Why were patients with RP grouped together regardless of XRT adjuvant/salvage therapy? Can these be separated in the database?
2. Can the use of adjuvant ADT be assessed?
Presented by: Devin Patel, MD Los Angeles, CA
Co-Authors: Shalini Jha, Lauren Howard, Durham, NC, Christopher Amling, Portland, OR, William Aronson, Los Angeles, CA, Matthew Cooperberg, San Francisco , CA, Christopher Kane, San Diego , CA, Martha Terris, Augusta, GA, Brian Chapin, Houston, TX, Stephen Freedland, Los Angeles, CA
Written by: Thenappan Chandrasekar, MD, Clinical Fellow, University of Toronto, | twitter: @tchandra_uromd at the 2018 AUA Annual Meeting - May 18 - 21, 2018 – San Francisco, CA USA