PURPOSE: We clarified that lead time bias in screen detected prostate cancer is important for evaluating the outcome of any individual screening system. MATERIALS AND METHODS: Between 1992 and 2001, 195 and 958 prostate cancer cases with clinical T1c/T2N0M0 and T3N0M0 disease were enrolled in the current study as screen detected and outpatient clinic detected prostate cancer, respectively. Log10 prostate specific antigen velocity was calculated using log10 prostate specific antigen at diagnosis and at the most recent screening before cancer detection. Lead time in screen detected cancer was then estimated as the year when log10 prostate specific antigen in screen detected cancer would increase to the levels of log10 prostate specific antigen in outpatient clinic detected prostate cancer. RESULTS: Median log10 prostate specific antigen was 0.87 and 1.08 ng/ml for clinical T1c/T2N0M0 disease, and 1.14 and 1.53 ng/ml for T3N0M0 disease in screen and outpatient clinic detected cancer, respectively. The 25th, 50th and 75th percentiles of log10 prostate specific antigen velocity before cancer detection in the screening population were 0.05, 0.08 and 0.14 for T1c/T2N0M0 disease, and 0.07, 0.13 and 0.21 for T3N0M0 disease, respectively. The 25th, 50th and 75th percentiles of expected lead time in screen detected cancer were 1.9, 3.3 and 5.2 years for T1c/T2N0M0 disease and 1.4, 2.2 and 4.1 years for T3N0M0 disease, respectively. CONCLUSIONS: The lead time of screen detected cancer in our screening system is not as long as previously thought. This new methodology for lead time estimation may be useful for evaluating treatment outcomes of screen detected prostate cancer in individual screening systems done in various regions worldwide.
PURPOSE: We clarified that lead time bias in screen detected prostate cancer is important for evaluating the outcome of any individual screening system. MATERIALS AND METHODS: Between 1992 and 2001, 195 and 958 prostate cancer cases with clinical T1c/T2N0M0 and T3N0M0 disease were enrolled in the current study as screen detected and outpatient clinic detected prostate cancer, respectively. Log10 prostate specific antigen velocity was calculated using log10 prostate specific antigen at diagnosis and at the most recent screening before cancer detection. Lead time in screen detected cancer was then estimated as the year when log10 prostate specific antigen in screen detected cancer would increase to the levels of log10 prostate specific antigen in outpatient clinic detected prostate cancer. RESULTS: Median log10 prostate specific antigen was 0.87 and 1.08 ng/ml for clinical T1c/T2N0M0 disease, and 1.14 and 1.53 ng/ml for T3N0M0 disease in screen and outpatient clinic detected cancer, respectively. The 25th, 50th and 75th percentiles of log10 prostate specific antigen velocity before cancer detection in the screening population were 0.05, 0.08 and 0.14 for T1c/T2N0M0 disease, and 0.07, 0.13 and 0.21 for T3N0M0 disease, respectively. The 25th, 50th and 75th percentiles of expected lead time in screen detected cancer were 1.9, 3.3 and 5.2 years for T1c/T2N0M0 disease and 1.4, 2.2 and 4.1 years for T3N0M0 disease, respectively. CONCLUSIONS: The lead time of screen detected cancer in our screening system is not as long as previously thought. This new methodology for lead time estimation may be useful for evaluating treatment outcomes of screen detected prostate cancer in individual screening systems done in various regions worldwide.
Authors: Scott E Eggener; Alex Mueller; Ryan K Berglund; Raj Ayyathurai; Cindy Soloway; Mark S Soloway; Robert Abouassaly; Eric A Klein; Steven J Jones; Chris Zappavigna; Larry Goldenberg; Peter T Scardino; James A Eastham; Bertrand Guillonneau Journal: J Urol Date: 2009-02-23 Impact factor: 7.450