Multicenter evaluation of an investigational prostate cancer methylation assay

Prostate specific antigen tests have low specificity, which frequently results in unnecessary biopsy and typically limits screening to patients with prostate specific antigen greater than 4.0 ng/ml. We evaluated an investigational prostate cancer methylation specific polymerase chain reaction assay that detects aberrant methylation in 3 markers (GSTP1, RARbeta2 and APC) that indicate the presence of prostate cancer.

The assay was evaluated in 337 post-digital rectal examination urine samples (178 cancer and 159 noncancer) collected prospectively at a total of 9 clinical sites. Samples were processed wholly or after division into equal portions. Subject prostate specific antigen was 2.0 to 10.0 ng/ml. All subjects underwent transrectal ultrasound guided needle biopsy with 6 or greater cores sampled. Detection of 1 or greater markers indicated positivity.

Methylation specific polymerase chain reaction assay performance was better in whole than in divided urine cohorts (p = 0.035). Assay AUC was 0.72 in the whole urine cohort and 0.67 in the combined population. These values were higher than those of prostate specific antigen alone using 4.0 ng/ml as the cutoff (p = 0.00 and 0.01, respectively). Moreover, the assay together with the Prostate Cancer Prevention Trial risk calculator or a standard nomogram significantly improved AUC in the whole urine cohort and the combined population vs predictive algorithms alone (p <0.05). Assay positive predictive value was 54% in whole urine cohort with prostate specific antigen 2.0 to 4.0 ng/ml and negative predictive value was 87% with prostate specific antigen 4.1 to 10.0 ng/ml. Assay positive predictive value was higher in subjects with all 3 methylation markers positive.

These data demonstrate that this investigational assay used in conjunction with current screening algorithms may potentially add value to the biopsy decision making process.