A Critical Analysis of the Tumor Volume Threshold for Clinically Insignificant Prostate Cancer Using a Data Set of a Randomized Screening Trial

Top Cited Papers
Open Access
Abstract
Purpose: The identification of clinically insignificant prostate cancer could help avoid overtreatment. Current criteria for insignificant prostate cancer use a tumor volume threshold of less than 0.5 ml for the index tumor. In this study we reassess this tumor volume threshold for clinically insignificant prostate cancer using an independent data set. Materials and Methods: The rate of insignificant prostate cancer was calculated by modeling lifetime risk estimates of prostate cancer diagnosis in screened and nonscreened participants in a randomized prostate cancer screening trial. Using lifetime risk estimates 50.8% of screen detected prostate cancer was calculated to be clinically insignificant and the 49.2% largest tumor volume of 325 prostatectomy specimens was used to determine the threshold tumor volume for insignificant prostate cancer. Because stage and grade represent the strongest determinants of cancer aggressiveness, we also calculated the tumor volume threshold for insignificant cancer after the selection of patients with organ confined prostate cancer without Gleason pattern 4/5. The analyses were performed for total tumor volume and for index tumor volume. Results: The minimum threshold tumor volume of the index tumor and total tumor was 0.55 and 0.70 ml, respectively. After accounting for tumor stage and grade we obtained a threshold volume for the index tumor and total tumor of 1.3 and 2.5 ml, respectively. Conclusions: We confirmed the original value of the index tumor volume threshold of 0.5 ml for insignificant prostate cancer, and we demonstrated that clinically insignificant prostate cancer may include index Gleason score 6, pT2 tumors with volumes up to at least 1.3 ml. These results suggest a reconsideration of current methods and nomograms used for pretreatment risk assessment.

This publication has 19 references indexed in Scilit: