Introduction

There are currently no validated DNA based biomarkers available for routine clinical use to predict prostate cancer recurrence after prostatectomy. The Genomic Evaluators of Metastatic Cancer of the Prostate (GEMCaP) assay was developed to predict recurrence using a tumor genotype derived from copy number for a set of genomic loci. We aim to validate the GEMCaP assay using a separate cohort from the development cohort.

Methods

We randomly selected 203 patients who had undergone radical prostatectomy at the Cleveland Clinic (CC) or the University of Rochester (UR) Cancer Centers from 2000-2005 and had tumor tissue available for research. After pathology review cancer tissues were macrodissected and DNA was extracted and subjected to high resolution array comparative genomic hybridization (aCGH). A high GEMCaP score was defined as >20% of the genomic loci exhibiting copy number gain or loss in a given tumor, as in previous studies. Cox regression was used to evaluate associations between the GEMCaP score and risk of biochemical recurrence.

Results

We report the results from 140 patients, 54 from the CC cohort and 86 from the UR cohort. Overall, 38% of patients recurred with a median time to recurrence of 45 months. Based on the CAPRA-S score, 39% were low-risk, 42% were intermediate-risk and 19% were high-risk. Thirty-one percent of the cohort had a high GEMCaP score (?20%). A high GEMCaP score was associated with higher risk of biochemical recurrence (HR 2.69, 95% CI 1.51-4.77) and remained associated with biochemical recurrence after adjusting for the CAPRA-S score (HR 1.91, 95% CI 1.05-3.48). The C-index for GEMCaP alone was 0.64, and improved when combined with CAPRA-S (C-index = 0.76).

Conclusions

In this validation study, a high GEMCaP score was associated with biochemical recurrence in two external cohorts. This remained true after adjusting for clinical and pathologic factors as accounted for by the CAPRA-S score. The GEMCaP biomarker could be an efficient and effective clinical risk assessment tool to identify prostate cancer patients for early adjuvant therapy.

Funding

NIH