Abstract
INTRODUCTION: Widespread use of PSA screening has raised concerns of overdiagnosis of low risk and underdiagnosis of high grade cancer. This is primarily due to the low sensitivity and specificity of PSA. PCA3, a non-coding large chain ribonucleic acid, is significantly over-expressed in cancer tissue and quantitatively measured by a novel urinary assay. The objective of this NCI EDRN trial was to conduct a comprehensive, independent validation of the PROGENSA PCA3 Assay for the detection of prostate cancer. METHODS: A prospective, PROBE-compliant NCI validation trial was undertaken at 11 clinical sites to evaluate PCA3 positive predictive value (PPV, PCA3 score >60) in the initial biopsy setting and negative predictive value (NPV, PCA3 score <20) in the repeat biopsy setting. PCA3 was obtained prior to biopsy but following an attentive DRE. We hypothesized that PPV for an initial prostate biopsy to be at least 55% and that NPV for a repeat biopsy to be at least 75%. The accuracy of PCA3 in detecting any prostate cancer and secondarily, high grade cancer (Gleason > 7), was compared to PCPT risk calculator through ROC curve analysis. RESULTS: 880 eligible men (mean age 62 years) were enrolled; 305 had a prior negative prostate biopsy. 99% had an informative PCA3 test. For the detection of any cancer, PPV was 80% (95% CI: 0.72 - 0.86) in the initial biopsy group while NPV was 88% (95% CI: 0.81 - 0.93) in the repeat biopsy group. PCA3 performance was superior to PCPT risk estimation and improved upon the detection of any cancer (p<0.006) and high grade cancer (p<0.02) when combined with the PCPT risk model. CONCLUSIONS: PCA3 ability to detect prostate cancer was superior to PCPT risk calculator in both the initial (p < 0.0001) and repeat biopsy setting (p=0.001)Independent validation of PCA3 demonstrated a high PPV in the initial biopsy setting and a high NPV in the repeat biopsy setting. Given the significant improvements in risk estimation over PCPT, PCA3 is expected to greatly enhance clinical decision making.
Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr 4451. doi:1538-7445.AM2012-4451
