J Natl Cancer Inst 2017 01 25;109(6). Epub 2017 Jan 25.
Affiliations of authors: Department of Urology (XG, LYL, JP, MKC, ZC, JGQ), Department of Pathology (DH, CKS, ZLS), and Department of Radiology (JW), The Third Affiliated Hospital, Sun Yat-sen University, Guangzhou, China; Department of Urology, St. Elisabeth Hospital, University of Leipzig, Leipzig, Germany (JR); Department of Urology, The First Affiliated Hospital, Nanchang University, Nanchang, China (WPL); Department of Urology, Changhai Hospital, The Second Military Medical University, Shanghai, China (SCR); Department of Urology, Cancer Center, Sun Yat-sen University, Guangzhou, China (FJZ); Department of Urology, Guangzhou First Municipal People's Hospital, Guangzhou Medical University, Guangzhou, China (KJX); Department of Urology, The Second Affiliated Hospital, Guangzhou Medical University, Guangzhou, China (XZ); Department of Urology, Renmin Hospital of Wuhan University, Wuhan, China (HJQ); Department of Urology, Cancer Hospital, Guangxi Medical University, Nanning, China (XZB); Department of Urology, Guangdong General Hospital, Guangzhou, China (JML); Department of Urology, Shenzhen People's Hospital, Shenzhen, China (JGY); Department of Medical Statistics and Epidemiology, School of Public Health, Sun Yat-sen University, Guangzhou, China (LL).
Background: For patients with prostate cancer (PCa), the presence of pelvic lymph node metastasis (LNM) is a strong predictor of poor outcome. However, the approaches with promising sensitivity and specificity to detect LNM are still lacking. We investigated the value of collapsin response mediator protein 4 (CRMP4) promoter methylation in biopsies as a predictor for LNM.
Methods: CRMP4 promoter methylation at two previously identified CpG sites was determined in 80 case-matched biopsy samples (the training set) using bisulfite pyrosequencing. The predictive cutoff value was independently validated using cohort I of 339 PCa patients (Southern China) and cohort II of 328 case patients (Germany, across China). Mann-Whitney U test, the receiver operating characteristic curve, McNemar's test, and logistic regression were used to assess data. All statistical tests were two-sided.
Results: In the training set, CRMP4 promoter methylation (≥15.0% methylated) was statistically significantly associated with LNM (P < 001). Successful validations were achieved in both cohorts I and II (sensitivity = 92.3%, 95% confidence interval [CI] = 79.3 to 97.9, and sensitivity = 92.2%, 95% CI = 81.1 to 97.8, respectively; specificity = 92.7%, 95% CI = 80.2 to 99.1, and specificity = 91.3%, 95% CI = 87.4 to 94.4, respectively). The sensitivity of CRMP4 promoter methylation is superior to conventional MRI (cohort I: 92.3% vs 26.2%, P < 001; cohort II: 92.2% vs 33.3%, P < 001). CRMP4 promoter methylation is an independent predictor of LNM (cohort I: hazard ratio [HR] = 8.35, 95% CI = 5.64 to 12.35, P < 001; cohort II: HR = 12.46, 95% CI = 5.82 to 26.70, P < 001) in a multivariable analysis model.
Conclusion: CRMP4 promoter methylation in diagnostic biopsies could be a robust biomarker for LNM in PCa.