Princess Margaret Cancer Centre, University of Toronto, University Health Network
Introduction: The use of adjuvant androgen deprivation therapy (ADT) in selected cohorts of patients with prostate cancer (PC) treated with radiation therapy (RT) is supported by multiple studies showing a benefit in both cancer specific survival (CSS) and in overall survival (OS). Recently, concerns regarding the physiological ,cardiovascular and metabolic effects of ADT, particularly among older men have become topical. Unfortunately, the side effect associated with this treatment limits its use in daily practice. We set out to determine whether adjuvant ADT is beneficial for elderly subpopulations of men with high risk treated with primary XRT.
Methods: We conducted a pooled analysis among 603 patients older than 75 years of age from prospective Radiation Therapy Oncology Group (RTOG) trials that randomized patients to receive adjuvant ADT + RT vs. RT alone. Inverse probability weighting (IPW) analysis was utilized to optimize group comparison. Kaplan-Meier curves and competing risk analyses were used to express survival outcomes.
Results: Our initial cohort consisted of 1045 men over age 75 years. After exclusion of patients with T4,N1 ,M1 and PSA> 30 ng/ml, 603 were available for final analysis (62% received ADT). Median age was 77 (IQR 75-78), with relatively good performance status (median Karnofsky score 100). A total of 37% of the patients developed biochemical recurrence (BCR) with a median time from RT of 5.4 years (IQR 3.3-9.15). 9.4% developed distant metastasis (DM) with a median time of 8.2 years (4.8-8.7). 1.8% had cancer specific mortality (CSM) (median time 7.9 years, IQR 4.5-8.5) and overall mortality occurred in 8.2% (median time 8.5 years, IQR 5-8.4). In the Kaplan-Meier curves ADT patients had less BCR (p= 0.0018) and less CSS (p= 0.0078), but there was no difference in OS (p= 0.79). In the competing risk analysis ADT still conferred better CSS (p= 0.0141) but there was no difference in the competing risk curve (p= 0.155). We then used IPW to remove the bias due to group differences and repeated the analysis, again finding a benefit of ADT on CSS (HR 0.556, 95% CI 0.513-0.741, p< 0.001) but not on OS (HR= 0.989, 95% CI 0.873-1.121. p= 0.863), and it actually increases other cause mortality (HR 1.25, 95% CI 1.08-1.44, p= 0.002) .
Conclusions: As expected, ADT results in a decrease in BCR and even CSS, but not OS as was noted within the entire cohort. Competing risks in part due to ADT-related adverse effects among men >75 years old question the use of ADT for this population. Source of