CLINICAL PERFORMANCE OF SERUM ISOFORM [-2]PROPSA (P2PSA) AND ITS DERIVATIVES, NAMELY %P2PSA AND PHI (PROSTATE HEALTH INDEX) IN MEN WITH FAMILY HISTORY OF PROSTATE CANCER. RESULTS FROM A MULTICENTRIC EUROPEAN STUDY (PROMETHEUS PROJECT)

Lazzeri M.1, Lughezzani G.1, Lista G.1, Larcher A.1, Cestari A.1, Buffi N.1, Haese A.2, Renter A.2, De La Taille A.3,Bosset P.O.3, Le Corvoisier P.3, McNicholas T.4, Fowler L.4, Roux J.W.4, Palou J.5, Breda A.5, De La Torre P.6, BiniV.7, Graefen M.2, Guazzoni G.11Università Vita e Salute San Raffaele, Dept. of Urology, Milan, Italy, 2Martini-Klinik, Prostatakarzinomzentrum AmUKE Universitätsklinikum Hamburg Eppendorf, Dept. of Urology, Hamburg, Germany, 3CHU Mondor, Faculté deMédecine De Créteil, Dept. of Urology, Creteil, France, 4South Bedfordshire & Hertfordshire Urological Cancer Centre,Dept. of Urology, Stevenage, United Kingdom, 5Fundació Puigvert, Dept. of Urology, Barcelona, Spain, 6FundacióPuigvert, Dept. of Radiology, Barcelona, Spain, 7University of Perugia, Dept. of Medicine, Perugia, ItalyINTRODUCTION & OBJECTIVES: Recently the EMA and FDA approved serum isoform [-2]proPSA (p2PSA) and itsderivate PHI (prostate health index) for discriminating between men with or without prostate cancer (PCa). To datethere have been no published studies characterizing p2PSA and PHI in men with a PCa affected first degree relative(father, brother, son). We test the hypothesis that p2PSA and its derivates are more accurate than tPSA, fPSA andf/tPSA in detecting cancer in men with a family history of PCa.MATERIAL & METHODS: The analysis consisted of a nested case-control study from the PRO-psa MulticentricEuropean Study (PROMEtheusS) project (detailed description of study design, setting, centres and patients areavailable at http://www.controlled-trials.com ref. ISRCTN04707454). The primary outcomes were the measures ofsensitivity, specificity and accuracy of serum p2PSA, %p2PSA {([-2]proPSA/10)/fPSA)} and Beckman Coulter PHI(prostate health index) {[[-2]proPSA/fPSA] x squart square root PSA}, which were compared to the accuracy ofestablished PCa serum tests (tPSA, fPSA and %fPSA) in men who had undergone prostatic biopsy for suspected PCaand who have had a PCa affected first degree relative (father, brother, son). The potential reduction of unnecessarybiopsies and the characteristics of missing PCa were reported as secondary outcomes. Multivariable logisticregression models were complemented by predictive accuracy analysis and decision curve analysis.RESULTS: Over 1036 patients were enrolled in the PROMEtheusS cohort, 157 (15.5%) had a PCa affected firstdegree relative. PCa cancer was found in 71 subjects (42.2%). p2PSA, %p2PSA and PHI values were significantlyhigher (p<0.0001), and %fPSA values significantly lower (p<0.0001) in patients with PCa. At univariate accuracyanalysis, %p2PSA (AUC: 0.733) and PHI (AUC: 0.733) were the most accurate predictors and significantlyoutperformed tPSA (AUC: 0.549), fPSA (AUC: 0.489) and %fPSA (AUC: 0.600) in the prediction of PCa at biopsy(p≤0.001). For %p2PSA a cut-off of 1.66 provided the optimal trade-off between sensitivity and specificity(respectively 70.4 and 70.1%; 95%C.I: 58.4-80.7 and 59.4-79.5 respectively). A cut-off of 40 for PHI showed the best848 Clinical performance of serum isoform [-2]proPSA (p2PSA) and itsderivatives, namely %p2PSA and PHI (Prostate Health Index) in men withfamily history of prostate cancer. Results from a multicentric Europeanstudy (PROMEtheuS project)Eur Urol Suppl 2013;12;e848