European Randomized Study of Screening for Prostate Cancer, ERSPC trial-Rotterdam: minimum requirements for final evaluation and policy decision making

Prostate cancer (PC) has become the most common non-cutaneous malignancy in men in several countries. The proportion of patients ultimately dying from the disease is similar or even higher than in breast, cervical or colorectal cancer, since symptoms often appear at late stages. Prostate specific antigen (PSA) is a biomarker capable of detecting 80% of PCs early, and trials have now shown surgical treatment to be beneficial. The effects, favorable and unfavorable, and costs of (nation-wide) screening for PC however are unknown. The European Randomized Study of Screening for Prostate Cancer (ERSPC trial) set up in 1994 has the power to show a 25% PC mortality reduction or more due to PSA screening at 4-year intervals. A 25% PC mortality reduction may indicate up to 600 PC deaths prevented per year in the Dutch setting. In eight European countries more than 150,000 men aged 55-69 have been randomized, of which Rotterdam contributes approximately one quarter. In the Rotterdam section of ERSPC, 3 rounds of screening have already been performed with a 90% attendance rate. In all centers, a 78% biopsy rate for men with high PSA-values has been achieved. The purpose of this final proposal is to complete follow-up evaluation of all men randomized in the Rotterdam section, estimate the magnitude of mortality reduction and unfavorable effects in a nation-wide setting, and analyse the cost-effectiveness of PC screening. Cumulative deaths and causes of death for the calendar years up to 2008 will be ascertained in both arms of the trial through record linkage with the Central Bureau for Genealogy (CBG) and Statistics Netherlands (CBS). Three independent medical doctors, blinded towards arm of the trial, will review files from all deceased PC patients, verified through linkage to the regional cancer registry, to correctly ascertain the cause of death (PC or other). According to the internationally agreed monitoring plan, PC mortality in each arm at the end of 2006, 2008 and 2010 latest (respectively 7.5, 9.5 and 11.5 mean years of follow-up and representing cumulative data of 2004, 2006 and 2008) will be compared for the ERSPC-trial as a whole. In case of a statistically significant interim result, the trial will be published. The point estimate for PC mortality reduction will be adjusted for compliance in the study arm and contamination in the control arm. The latter will be analyzed by regular linkages of trial participants to the Regional Clinical Chemistry and Pathology laboratories, and by model estimations on the increase in incidence and stage distribution of PC in the control arm. The long follow-up period and information collected on more than 2,500 PC patients in Rotterdam will allow to assess different screening strategies and to move towards identifying aggressive and less-aggressive prostate cancers. The MISCAN PC model has already been developed, based on the available detailed screen data and clinical data, and has estimated lead time, overdiagnosis, sensitivity, stage shifts due to screening and cost. In these final years, the magnitude of unfavorable effects, per alternative screening scenario, and its impact on health-related quality of life (HRQoL) will be assessed (based on the finished separate HRQol-project). We will extrapolate the ERSPC-performances to the population level, predicting the number of screens, false positives, extra incidence, treatment changes, quality-of-life decrements, PC deaths prevented, life-years gained, cost-changes, and cost-effectiveness under different realistic screening scenarios in the Netherlands. These estimates will give the answer on recommending or discouraging PSA screening for asymptomatic men. At present it is estimated that 45 million PSA tests are annually being sold worldwide.