Can nomograms improve our ability to select candidates for active surveillance for prostate cancer?

BACKGROUND

Our goal was to compare the ability of active surveillance (AS) criteria and preoperative nomograms to identify patients with pathologically low-risk prostate cancer.

METHODS

The study cohort consisted of 402 radical prostatectomy patients with Gleason 6 prostate cancer on at least 10-core biopsy. In this group, we analyzed the ability of Kattan and Truong nomograms to select patients with Gleason 3+3 or 3+4 organ-confined prostate cancer, and compared it with that of AS criteria of John Hopkins (JH) and University of California at San Francisco (UCSF) medical centers, and Prostate Cancer Research International: Active Surveillance (PRIAS) study. The performance of each tool was evaluated with respect to discrimination and predictive accuracy.

RESULTS

About 30% of patients were upgraded and 8% were upstaged in the final pathology. The nomograms demonstrated slightly higher discrimination in detecting organ-confined Gleason 3+3 and 3+4 disease. The predictive accuracy of the nomograms in selecting patients with low-grade organ-confined prostate cancer was superior to that of JH and UCSF criteria but not to PRIAS criteria. Furthermore, the nomograms were unable to select larger subgroups of patients with the same prevalence of Gleason 3+3 organ-confined prostate cancer as in men who met the PRIAS criteria. No difference was seen between the studied nomograms and AS criteria in their ability to identify patients with Gleason 3+4 organ-confined prostate cancer.

CONCLUSIONS

PRIAS criteria demonstrate optimal balance between sensitivity and specificity and are not inferior to the available pathological nomograms in selecting patients with low-grade organ-confined prostate cancer.