High levels of 5-hydroxymethylcytosine (5hmC) is an adverse predictor of biochemical recurrence after prostatectomy in ERG-negative prostate cancer

• Published in: Clinical epigenetics Vol. 7; no. 1; p. 111
• Main Authors: Strand, Siri H, Hoyer, Soren, Lynnerup, Anne-Sofie, Haldrup, Christa, Storebjerg, Tine Maj, Borre, Michael, Orntoft, Torben F, Sorensen, Karina D
• Format: Journal Article
• Language: English
• Published: Germany BioMed Central Ltd 15.10.2015; BioMed Central
• Subjects: Analysis; Epigenetic inheritance; Methylation; Prognosis; Prostate cancer; Surgery; 5hmC; 5-hydroxymethylation; Biomarker; DNA methylation; Epigenetics; ERG; Prostate cancer; Prognostic
• ISSN: 1868-7075; 1868-7083; 1868-7083; 1868-7075
• DOI: 10.1186/s13148-015-0146-5

Prostate cancer (PC) can be stratified into distinct molecular subtypes based on TMPRSS2-ERG gene fusion status, but its potential prognostic value remains controversial. Likewise, routine clinicopathological features cannot clearly distinguish aggressive from indolent tumors at the time of diagnosis; thus, new prognostic biomarkers are urgently needed. The DNA methylation variant 5-hydroxymethylcytosine (5hmC, an oxidized derivative of 5-methylcytosine) has recently emerged as a new diagnostic and/or prognostic biomarker candidate for several human malignancies. However, this remains to be systematically investigated for PC. In this study, we determined 5hmC levels in 311 PC (stratified by ERG status) and 228 adjacent non-malignant (NM) prostate tissue specimens by immunohistochemical analysis of a tissue microarray, representing a large radical prostatectomy (RP) cohort with long clinical follow-up. We investigated possible correlations between 5hmC and routine clinicopathological variables and assessed the prognostic potential of 5hmC by Kaplan-Meier and uni- and multivariate Cox regression analyses in ERG+ (n = 178) vs. ERG- (n = 133) PCs using biochemical recurrence (BCR) as endpoint. We observed a borderline significant (p = 0.06) reduction in 5hmC levels in PC compared to NM tissue samples, which was explained by a highly significant (p < 0.001) loss of 5hmC in ERG- PCs. ERG status was not predictive of BCR in this cohort (p = 0.73), and no significant association was found between BCR and 5hmC levels in ERG+ PCs (p = 0.98). In contrast, high 5hmC immunoreactivity was a significant adverse predictor of BCR after RP in ERG- PCs, independent of Gleason score, pathological tumor stage, surgical margin status, and pre-operative prostate-specific antigen (PSA) level (hazard ratio (HR) (95 % confidence interval (CI)): 1.62 (1.15-2.28), p = 0.006). This is the first study to demonstrate a prognostic potential for 5hmC in PC. Our findings highlight the importance of ERG stratification in PC biomarker studies and suggest that epigenetic mechanisms involving 5hmC are important for the development and/or progression of ERG- PC.