Urine DNA methylation markers identify prostate cancer patients
MedWire News: Prostate cancer patients can be accurately discriminated from biopsy-negative patients by analysis of DNA methylation biomarkers in urine, say scientists who suggest the use of these biomarkers in predictive nomograms.
Although the low specificity of prostate-specific antigen (PSA) for detecting prostate cancer has been partly addressed by PSA derivative analysis, the lack of accuracy means the search for prostate cancer biomarkers continues.
Exploring the utility of DNA methylation biomarkers, Shannon Payne, from Epigenomics in Seattle, Washington, USA, and colleagues assessed plasma and urine samples from 142 patients referred for prostate biopsy and 50 asymptomatic males aged <30 years.
Based on a previous genome-wide scan, the team analyzed levels of four candidate DNA methylation markers: glutathione S-transferase pi (GSTP1), Ras association (RalGDS/AF-6) domain family 2 (RASSF2), histone cluster 1, H4k (HIST1H4K), and transcription factor AP-2 epsilon (activating enhancer binding protein 2 epsilon) (TFAP2E).
Prostate cancer was detected in 91 biopsy-referred patients, leaving 51 biopsy negative. GSTP1 methylation was found in 81% of prostate cancer urine samples and 39% of prostate cancer plasma samples, compared with 59% and 31%, respectively, of samples from biopsy negative patients, and 6% and 20%, respectively, of samples from healthy controls.
Methylation of the other genes studied was found in 92%–100% of prostate cancer urine samples and 18%–31% of prostate cancer plasma samples. Methylation was also detected in the two control groups at lower frequencies, except for TFAP2E in urine samples.
The area under the receiver operating characteristics curve (AUC) for predicting prostate cancer using DNA methylation in urine samples versus healthy controls ranged from 0.86 for TFAP2E to 0.91 for HIAT1H4K. Compared with biopsy negative patients, the AUC ranged from 0.64 for HIST1H4K to 0.69 for GSTP1. DNA methylation in plasma samples had lower AUCs and substantially lower sensitivity than that from urine samples.
Although the biomarkers in urine DNA correlated significantly with each other, they demonstrated minimal correlation with PSA levels, with none reaching significance. There were also no associations with organ-confined disease.
The team notes in The Prostate that paneling of the biomarkers either with each other or with PSA-based biomarkers, did not significantly improve performance over assessing the DNA methylation biomarkers alone.
They conclude: “These biomarkers provided information independent of PSA and may serve as a useful tool in discrimination of prostate cancer from benign prostatic hyperplasia in prostate biopsy-negative males.”
MedWire (www.medwire-news.md) is an independent clinical news service provided by Current Medicine Group, a part of Springer Science+Business Media. © Current Medicine Group Ltd; 2009
By Liam Davenport