Presentation Authors: Sigrid Carlsson*, New York, NY, Marianne Månsson, Gothenburg, Sweden, Sue Moss, London, United Kingdom, Maciej Kwiatkowski, Franz Recker, Aarau, Switzerland, Teuvo Tammela, Tampere, Finland, Chris Bangma, Monique Roobol, Rotterdam, Netherlands, Anssi Auvinen, Tampere, Finland, Jonas Hugosson, Gothenburg, Sweden
Introduction: Differential treatment between trial arms has been suggested to bias prostate cancer mortality in the European Randomized Study of Screening for Prostate Cancer (ERSPC). The objective of this study was to assess the contribution of treatment differences on the observed prostate cancer mortality reduction between the screening arm and control arm.
Methods: We analyzed a total of 14,136 men with prostate cancer (7,310 men in the screening arm: 6,826 in the control arm) in the core age group (55-69 years) at 16 years of follow-up. We compared the observed versus estimated number of prostate cancer deaths by treatment allocation in the screening arm and control arm, respectively. Treatment allocation was modeled using multinomial logistic regression adjusting for: center, age, year, prostate-specific antigen level, Gleason score, and TNM stage. For each treatment, logistic regression models were fitted for risk of prostate cancer death, separately for the screening arm and control arm, and using the same covariates as for the treatment allocation model. Treatment probabilities were multiplied with estimated prostate cancer death risks for each treatment based on one arm, summed and compared to the observed number of deaths.
Results: The difference between the estimated and observed treatment distributions in the screening arm and control arm was marginal and ranged between -3.3% and 3.3%. The difference in the estimated and observed number of prostate cancer deaths was small: 0.01% (95% CI -0.28%, 0.24%) when comparing the estimated screening arm model to the observed number of prostate cancer deaths in the control arm, and 0.05% (95% CI -0.11%, 0.18%) when comparing the estimated control arm model to the observed number of prostate cancer deaths in the screening arm. A limitation of this study is that only data on primary treatment was available.
Conclusions: This study supports the notion that the effectiveness of PSA-screening in reducing prostate cancer mortality in the ERSPC trial is due to early detection, allowing for effective management, and is not attributable to differential treatment between trial arms.
Source of Funding: This work was supported in part by funds from the Sidney Kimmel Center for Prostate and Urologic Cancers, a Specialized Program of Research Excellence grant (P50-CA92629) from the National Cancer Institute to Dr. Howard Scher, a National Institutes of Hea