Comparison of Pathological Outcomes of Active Surveillance Candidates Who Underwent Radical Prostatectomy Using Contemporary Protocols at a High-volume Korean Center

Abstract

We compared contemporary active surveillance protocols based on pathological outcomes in patients who underwent radical prostatectomy. We identified the experimental cohort from prostate cancer patients who underwent radical prostatectomy between 2001 and 2011, and who met the inclusion criteria of five published active surveillance protocols, namely Johns Hopkins Medical Institution, University of California at San Francisco, Memorial Sloan-Kettering Cancer Center, University of Miami and Prostate Cancer Research International: Active Surveillance. To compare each protocol, we evaluated the pathological outcomes and calculated the sensitivity, specificity and accuracy for each protocol according to the proportion of organ-confined Gleason≤6 disease. Overall, 376 patients met the inclusion criteria of the active surveillance protocols with 61, 325, 222, 212 and 206 patients meeting the criteria of the Johns Hopkins Medical Institution, University of California at San Francisco, Memorial Sloan-Kettering Cancer Center, University of Miami and Prostate Cancer Research International: Active Surveillance protocols, respectively. The sensitivity and specificity values of the five protocols, respectively, were 0.199 and 0.882 in Johns Hopkins Medical Institution, 0.855 and 0.124 in University of California at San Francisco, 0.638 and 0.468 in Memorial Sloan-Kettering Cancer Center, 0.599 and 0.479 in University of Miami, and 0.609 and 0.527 in Prostate Cancer Research International: Active Surveillance. In terms of both the sensitivity and specificity, Prostate Cancer Research International: Active Surveillance was the most balanced protocol. In addition, Prostate Cancer Research International: Active Surveillance showed a more accurate performance for favourable pathological outcomes than the others. However, using the area under the curve to compare the discriminative ability of each protocol, there were no statistically significant differences. The contemporary active surveillance protocols showed similar pathological characteristics in patients who had undergone radical prostatectomy. However, we concluded that the Prostate Cancer Research International: Active Surveillance protocol would be most helpful to Korean populations in choosing candidates for active surveillance considering the balance between sensitivity and specificity and the accuracy of diagnosis.