Study
PLCO
(Learn more about this study)
Project ID
2010-0042
Initial CDAS Request Approval
Aug 15, 2012
Title
Evaluation of a four kallekrein panel to detect prostate cancer
Summary
A positive prostate-specific antigen (PSA) test (GT 4 ng/ml) is known to have moderate specificity for prostate cancer. Most men with elevated PSA do not have prostate cancer, leading to an estimated 750,000 unnecessary prostate biopsies per year in the US alone. Total PSA as measured is a combination of several different isoforms - complexed PSA, free PSA and nicked PSA - and there is accumulating evidence that measurement of these specific PSA-forms along, with hK2 -which converts proPSA to catalytic PSA - can improve the specificity of the PSA test. Using data from the European Randomized Study of Prostate Cancer screening (ERSPC), we have shown that a panel of these four kallikreins is a highly accurate predictor of prostate biopsy outcome in previously unscreened men. We also demonstrated that basing biopsy decisions on cancer risk predicted by this panel would lead to improved clinical outcomes. We subsequently found that the statistical prediction model developed using the panel is able to predict, with a high degree of accuracy, biopsy outcome in numerous other cohorts, including men with prior screening, those with prior biopsy, and those subject to clinical work-up before biopsy. To date, the kallikrein panel has been investigated, with highly consistent results, in 7 separate studies involving 10,575 biopsies and 2860 cancers. However, to date, the panel has only been investigated in European cohorts. Given data showing that important differences may exist between US- and European-biopsy cohorts, we plan to evaluate the kallikrein panel using samples from the PLCO. Subjects in the screening arm of the PLCO Trial with biopsies following a positive PSA test (GT 4 ng /ml) will be selected and kallikrein levels measured on aliquots stored from the PSA positive associated blood samples. Predictions from the pre-specified prediction model will be compared with biopsy outcome. Clinical implications of using the kallikrein panel to determine biopsy referral in US populations will be explored using decision analysis. In addition, prior work by our group has indicated that MSP the protein product of MSMB and certain SNPs in the kallikrein and MSMB pathways can modify the association between a blood marker and cancer status. We will therefore explore whether adding MSP and selected SNPs to our model will improve predictive accuracy.
Aims
1. Determine the predictive accuracy for prostate biopsy outcome of a pre-specified model consisting of four kallikrein markers - total PSA, free PSA, intact PSA and hK2 - compared to total PSA alone in men with elevated PSA. a. For all cancers. b. For high-grade cancers (Gleason score 7 or more) 2. Determine the clinical implications of using the model to determine referral to biopsy using decision analytic methods. a. For all cancers. b. For high-grade cancers (Gleason score 7 or more) 3. Determine whether the accuracy of the model is affected by race. 4. Determine whether adding SNPs affecting kallikrein production can improve accuracy of the four-kallikrein model. 5. Determine whether a combination of MSMB and an MSP-related SNPs can improve accuracy of the four-kallikrein model.
Collaborators
Jiyoung Ahn (NYU School of Medicine)
Richard B. Hayes (NYU Cancer Institute)
Andrew Vickers (Memorial Sloan-Kettering Cancer Center)
Related Publications
