BERKELEY, CA (UroToday.com) - The Prostate Cancer Prevention Trial (PCPT) randomized over 18,000 men to placebo or finasteride to determine if there was a decrease in prostate cancer (CaP) risk.
The placebo arm of PCPT has provided data to develop a CaP risk calculator (PCRC), to model the overall risk of CaP, and Gleason score 7 or higher CaP - at biopsy. Variables included in the multivariate logistical regression analysis to develop the model included PSA, abnormal DRE, family history of CaP, black race and age. The biopsy scheme in PCPT was a sextant 6- core pattern. Others have validated the risk calculator in the extended biopsy era. Dr. Joseph Presti and his group at Stanford University School of Medicine assessed the PCPT risk calculator in their population of patients undergoing a 12-core scheme using the same statistical methods as the PCRC to model the risk of a positive biopsy and risk of high-grade disease.
The Stanford University cohort included 636 men who underwent initial prostate biopsy between 1999 and 2004. Compared with the cohort used to develop the PCRC, the Stanford patients were younger (20.3% younger than 55 years old), while the PCRC had half of its cohort older than 70 years. Stanford had more Hispanic and Asian men and a comparable number of African-American men and men with a positive family history of CaP. Men in the Stanford database had a median PSA of 5.7ng/ml, compared with 1.5ng/ml in the PCRC group. Overall, 46.6% of Stanford men were found to have CaP compared with 21.9% in the PCRC. CaP Gleason score 7 or higher was diagnosed in 34.9% of Stanford men and 4.7% of the PCRC group. In the Stanford regression analysis for high-grade CaP, PSA, abnormal DRE, and age were significant, but black race was not. Receiver operator characteristic curves resulted in an AUC of 0.664 for overall risk of CaP and 0.509 for high-risk CaP. The corresponding values in PCPT were 0.702 and 0.698, respectively.
The investigators conclude that the PCRC must be used with caution in contemporary populations. However, they also point out that the differences may be due to variations in the referral nature of the populations, biopsy grading and biopsy schemes.
Ngo TC, Turnbull BB, Lavori PW, Presti JC Jr.