Heterogeneity in D׳Amico classification–based low-risk prostate cancer: Differences in upgrading and upstaging according to active surveillance eligibility

Abstract

BackgroundTo date, no study has examined clinical, pathological, and surgical characteristics of D׳Amico low-risk patients according to active surveillance (AS) eligibility. Material and methodsWe relied on patients with low-risk prostate cancer, who were classified based on the D׳Amico classification, treated with radical prostatectomy (RP) between 2008 and 2013 at the Martini-Clinic Prostate Cancer Center. We assessed differences in clinical, pathological, and surgical characteristics in D׳Amico low-risk patients according to AS eligibility (prostate-specific antigen [PSA]≤10ng/ml, Gleason score≤3+3, ≤2 positive cores,≤50% tumor content per core, and≤cT1-2a). Multivariable logistic regression analyses targeted 2 end points: (1) presence of either intermediate- or high-risk characteristics (Gleason score≥3+4 or≥pT3 or pN1) or (2) exclusive presence of high-risk characteristics (Gleason score≥4+4 or≥pT3 or pN1) at RP. ResultsOf 1,331 patients low-risk prostate cancer classified based on the D׳Amico classification, 825 (62%) men were eligible for AS. AS candidates were less frequently either upgraded (55% vs. 78%, P<0.001) or upstaged (8% vs. 15%, P<0.001). Similarly, at final pathology, AS candidates less frequently harbored either intermediate- or high-risk (56% vs. 78%, P<0.001), or exclusive high-risk characteristics (9% vs. 16%, P<0.001). Tumor involvement per core (>50%) (most powerful), number of positive cores, PSA values, and age were independent predictors for either intermediate- or high-risk characteristics at RP. Tumor involvement per core and PSA values were independent predictors for exclusive high-risk characteristics at RP. ConclusionsD׳Amico low-risk patients did not have a homogeneous histology at RP. Especially, non-AS candidates were at a higher risk of either upgrading or upstaging at final pathology. Tumor involvement greater than 50% per core was the most powerful indicator of adverse pathology. Therefore, DʼAmico low-risk criteria are not safe enough to identify AS candidates.