Presentation Authors: Alexander Haese*, Hamburg, Germany, Amy Tin, Daniel Sjoberg, Sigrid Carlsson, New York, NY, Dirk Pehrke, Berlin, Germany, Thomas Steuber, Hartwig Huland, Markus Graefen, Hamburg, Germany, Peter Scardino, New York, NY, Thorsten Schlomm, Berlin, Germany, Andrew Vickers, Hans Lilja, New York, NY, Guido Sauter, Hamburg, Germany
Introduction: A statistical model based on a panel of four kallikrein (4K) markers in blood-commercially available as 4Kscore-can accurately predict high-grade (â‰¥Grade Group [GrdGrp] 2) prostate cancer (PC) on biopsy. However, sampling error and variation in pathology reporting may miss aggressive disease. We evaluated the ability of the 4K panel to predict adverse pathology at radical prostatectomy (RP)-the gold standard for accurate histological diagnosis, and biochemical recurrence (BCR) after RP.
Methods: The 4K panel was measured in cryopreserved blood from 2,330 patients obtained before RP at a tertiary referral center between 2002-2010. Adverse pathology at RP was defined as: primary Gleason grade 4, any grade 5, seminal vesicle invasion, extracapsular extension or lymph node invasion. BCR was defined as prostate specific antigen (PSA) â‰¥ 0.20 Âµg/L. Multivariable (logistic and Cox) regression was used to study the association between the 4K panel, adverse pathology, and BCR, respectively. Discrimination was assessed by the area under the curve (AUC) and Harrel's concordance (C) index, comparing the improvement in discrimination by adding the pre-specified 4K panel-based model to a preoperative clinical base model (age, PSA, clinical stage and grade). Clinical utility was assessed using decision curve analysis. We prespecified that we would separately assess men with GrdGrp1-where the clinical decision concerns obtaining a confirmatory biopsy-and GrdGrp2-where the decision concerns active surveillance vs. treatment.
Results: The median age was 64 years and the median follow-up was 7.8 years. On multivariable logistic regression, the 4K panel was significantly associated with adverse pathology (OR 1.49; 95% CI 1.32, 1.67; p < 0.0001). The 4K panel added 0.045 to the discrimination of the clinical model (AUC of 0.718 vs. 0.673) within GrdGrp1 and 0.015 (AUC of 0.659 vs. 0.644) within GrdGrp2. Decision curve analysis confirmed the added value of the 4K panel across appropriate ranges for decision-making thresholds: 5-20% for GrdGrp1 and 20-60% for GrdGrp2. A higher 4K panel risk score was associated with higher risk of BCR on multivariable analysis (HR 1.16, 95% CI 1.06, 1.26; p=0.001). Adding the 4K panel risk score to the clinical base model improved the prediction of BCR within GrdGrp1 (Harrell's C from 0.642 to 0.667) but not among men with GrdGrp2.
Conclusions: The 4K panel is strongly associated with adverse pathology and BCR after RP. Use of the 4K panel in clinical practice could guide treatment decisions for men with GrdGrp1 (confirmatory biopsy) and GrdGrp2 (active surveillance vs. treatment).
Source of Funding: This work was supported in part by the National Institutes of Health/National Cancer Institute (NIH/NCI) with a Cancer Center Support Grant to Memorial Sloan Kettering Cancer Center [P30 CA008748], a SPORE grant in Prostate Cancer to Dr. H. Scher [P50-CA9