Description:
<jats:p> 10 </jats:p><jats:p> Background: As finasteride could reduce by 25% the number of prostate cancers (PCa) diagnosed in the U.S. annually, the public health benefit of prostate cancer prevention could be enormous. We performed a survival analysis to assess for any evidence of increased risk of death in men randomized to finasteride, a potential indicator of a ‘true’ increased risk of high grade (HG) disease in the PCPT. Methods: A Social Security Death Index search was conducted on all randomized men to ascertain date of death. Cox proportional hazards models adjusting for known risk factors for overall survival and survival from time of diagnosis of PCa overall, low grade (LG) and HG were used to estimate hazard ratios (HR) and construct 95% confidence intervals to determine if survival on finasteride and placebo were equivalent. Results: A total of 5,128 deaths have been reported; 2,584 men on finasteride and 2544 on placebo. 15-year survival rates for all randomized men in each arm is 78%. The HR for overall survival on finasteride compared to placebo is 1.04 (95% CI 0.96, 1.10, p=.19). 10-year survival from diagnosis for men with PCa was slightly higher for men randomized to finasteride (10-year survival 83% vs. 81%) but not statistically significant (HR= 0.87, 95% CI 0.73 - 1.04, p=.14). For the men with HG PCa, there was no evidence of worse survival on finasteride (HR= 1.01; 95% CI 0.73, 1.14, p=.97) while those diagnosed with LG disease finasteride had superior survival (HR= 0.73; 95% CI 0.57, 0.94, p=.01). Conclusions: For men in PCPT with PCa there was no difference in survival from diagnosis date, a slightly-superior 10-year survival with finasteride and a statistically-superior survival among men with LG tumors in the finasteride group. A potential explanation for this phenomenon could be a lead-time bias. Arguing against this bias is the identical survival of HG PCa in both groups. Another potential explanation is that the men with LG PCa on placebo include a greater number with undetected HG disease; HG tumors in men on finasteride were more likely detected due to the improved performance of prostate biopsy. With follow-up of 18 years, finasteride administration for 7-years does not appear to affect mortality but significantly reduces the risk of a PCa diagnosis. Clinical trial information: NCT00288106. </jats:p>