The Clinical Utility of the Genomic Prostate Score in Men with Very Low to Intermediate Risk Prostate Cancer

Purpose: We retrospectively investigated the Genomic Prostate Score® assay in clinical practice at an urban tertiary care academic center.

Materials and methods: We reviewed all Genomic Prostate Score results acquired during a 3-year period. Changes in patient NCCN® (National Comprehensive Cancer Network®) risk group, including very low, low, intermediate or high risk, and ultimate management decisions were recorded.

Results: Genomic Prostate Score risk stratification was performed in 134 men. According to the NCCN Guidelines®, 31 of the 134 men (23.1%) were at very low risk, 45 (33.6%) were at low risk and 58 (43.3%) were at intermediate risk. After adding the score the risk group changed in 32 of 134 patients (23.9%). The risk group did not change in the 31 men at very low risk. However, in the low risk group the risk changed in 19 of the 45 men (42.2%), including in 15 to very low and in 4 to intermediate risk. Also, in the intermediate risk group the risk changed in 13 of the 58 men (22.4%), including to low in 12 and to high risk in 1. Nine of the 15 men (60%) in whom risk changed from low to very low elected active surveillance. Nine of the 12 patients (75%) at intermediate risk in whom risk changed to low risk elected active surveillance, 2 (16.7%) elected definitive therapy and in 1 (8.3%) the choice was unknown. Of the 45 men at intermediate risk in whom risk was unchanged 28 (62.2%) elected definitive therapy, 12 (26.0%) elected active surveillance and in 5 (11.1%) the choice was unknown. Of the 4 men upgraded from low to intermediate risk after adding the genomic prostate score 2 elected definitive therapy and 2 chose active surveillance.

Conclusions: The Genomic Prostate Score has limited clinical usefulness in patients at very low risk since the NCCN risk group did not change. While it may be more useful for men at low and intermediate risk, for 32 (31%) of whose risk group was reclassified, clinical management decisions did not always appear to reflect these changes.