551 D'AMICO RISK STRATIFICATION OUTPERFORMS PUBLISHED ACTIVE SURVEILLANCE SELECTION CRITERIA FROM THE USA AS A WAY TO IDENTIFY PATIENTS WITH INDOLENT PROSTATE CANCER IN A RELATIVELY UNSCREENED UK POPULATION

Abstract

You have accessJournal of UrologyProstate Cancer: Localized (II)1 Apr 2013551 D'AMICO RISK STRATIFICATION OUTPERFORMS PUBLISHED ACTIVE SURVEILLANCE SELECTION CRITERIA FROM THE USA AS A WAY TO IDENTIFY PATIENTS WITH INDOLENT PROSTATE CANCER IN A RELATIVELY UNSCREENED UK POPULATION Greg Shaw, Ben Thomas, Dawson Sarah, Sarah Vowler, Vincent Gnanapragasam, Nimish Shah, and David Neal Greg ShawGreg Shaw Cambridge, United Kingdom More articles by this author , Ben ThomasBen Thomas Cambridge, United Kingdom More articles by this author , Dawson SarahDawson Sarah Cambridge, United Kingdom More articles by this author , Sarah VowlerSarah Vowler Cambridge, United Kingdom More articles by this author , Vincent GnanapragasamVincent Gnanapragasam Cambridge, United Kingdom More articles by this author , Nimish ShahNimish Shah Cambridge, United Kingdom More articles by this author , and David NealDavid Neal Cambridge, United Kingdom More articles by this author View All Author Informationhttps://doi.org/10.1016/j.juro.2013.02.1947AboutPDF ToolsAdd to favoritesDownload CitationsTrack CitationsPermissionsReprints ShareFacebookTwitterLinked InEmail INTRODUCTION AND OBJECTIVES Eligibility criteria for active surveillance (AS) are based on analysis of data from screened populations. We aim to evaluate the suitability of 6 different methods to identify indolent disease in a relatively unscreened UK cohort. METHODS 848 patients (445 patients with Gleason 3+3 prostate cancer on biopsy) underwent robotic radical prostatectomy between July 2007 and October 2011. F/U data was available for 841 of these. 5 published criteria designed to identify insignificant disease and therefore candidates for AS were identified and the ability to predict insignificant disease and low grade organ confined disease compared with D'Amico risk stratification (Gleason score 6, no grade 4/5, PSA<=10, cT1/T2a) using univariable Cox regression. AIC (Akaike information criterion) and BIC (Bayesian information criterion) were used to compare the six non-nested models. RESULTS AIC and BIC are two methods for comparing relative goodness of fit. Given a set of models, the one with the lowest value for AIC and/or BIC is preferred (table 1). Area under the curve (AUC) of the receiver operator characteristic curve is shown in table 2. CONCLUSIONS D'Amico outperforms the other methods for predicting insignificant (modified definition) or Gleason 6 organ confined disease according to AIC, BIC and AUC, however, all 6 methods have a poor PPV. Our dataset included only 2 classically defined insignificant tumours of less than 0.5cm3 rendering further analysis impossible and this reflects differences between our UK cohort and screened USA and European cohorts in which these criteria were developed and validated. These patients were all diagnosed by TRUS biopsy. Low AUR values underline the need for a better biomarker of indolent prostate cancer, and justify further investigation with confirmatory transrectal biopsy, transperineal biopsy or MRI prior to embarking on an active surveillance program. Approach AIC BIC Tosoian 2011 817.04 821.78 Adamy 2011 812.88 817.61 Van den Bergh 2009 816.37 821.11 Whitson 2011 810.51 815.25 Soloway 2010 817.26 821.99 D'Amico 1998 805.58 810.32 Approach Patients with Gleason 3+3 on biopsy meeting criteria for AS n(%) Insignificant cancer, updated definition Organ-confined low grade cancer Sensitivity Specificity Positive predictive value Negative predictive value Area under curve Sensitivity Specificity Positive predictive value Negative predictive value Area under curve Tosoian 2011 46 (10.3) 0.17 0.94 0.28 0.89 0.55 0.14 0.95 0.52 0.75 0.55 Adamy 2011 157 (35.3) 0.45 0.76 0.21 0.91 0.60 0.43 0.79 0.43 0.79 0.61 Van den Bergh 2009 67 (15.1) 0.22 0.91 0.26 0.89 0.56 0.20 0.93 0.51 0.76 0.56 Whitson 2011 184 (41.4) 0.51 0.69 0.19 0.91 0.60 0.48 0.72 0.39 0.79 0.60 Soloway 2010 114 (25.6) 0.35 0.84 0.23 0.90 0.59 0.30 0.85 0.43 0.77 0.58 D'Amico 1998 286 (64.3) 0.56 0.69 0.21 0.92 0.63 0.58 0.75 0.47 0.83 0.67 © 2013 by American Urological Association Education and Research, Inc.FiguresReferencesRelatedDetails Volume 189Issue 4SApril 2013Page: e226-e227 Advertisement Copyright & Permissions© 2013 by American Urological Association Education and Research, Inc.MetricsAuthor Information Greg Shaw Cambridge, United Kingdom More articles by this author Ben Thomas Cambridge, United Kingdom More articles by this author Dawson Sarah Cambridge, United Kingdom More articles by this author Sarah Vowler Cambridge, United Kingdom More articles by this author Vincent Gnanapragasam Cambridge, United Kingdom More articles by this author Nimish Shah Cambridge, United Kingdom More articles by this author David Neal Cambridge, United Kingdom More articles by this author Expand All Advertisement Advertisement PDF downloadLoading ...