Promoter hypermethylation in plasma‐derived cell‐free DNA as a prognostic marker for pancreatic adenocarcinoma staging

• Published in: International journal of cancer 2017-12, Vol.141 (12), p.2489-2497
• Main Authors: Henriksen, Stine Dam, Madsen, Poul Henning, Larsen, Anders Christian, Johansen, Martin Berg, Pedersen, Inge Søkilde, Krarup, Henrik, Thorlacius‐Ussing, Ole
• Format: Article
• Language: English
• Subjects: Adenocarcinoma; Aged; biomarker; Blood; Cancer; Cell-Free System; Classification; Deoxyribonucleic acid; DNA; DNA - genetics; DNA Methylation; epigenetics; Female; Gene Regulatory Networks; Humans; hypermethylation; Logistic Models; Male; Medical research; Middle Aged; MLH1 protein; Neoplasm Staging; Pancreatic cancer; Pancreatic Neoplasms - genetics; Pancreatic Neoplasms - pathology; Prediction models; Prognosis; Promoter Regions, Genetic; Prospective Studies; Regression analysis; Sensitivity; Wnt protein
• ISSN: 0020-7136; 1097-0215
• DOI: 10.1002/ijc.31024

Correct staging of pancreatic cancer is paramount, as treatment is stage specific. However, minimally invasive tools to facilitate staging are lacking. DNA promoter hypermethylation is a hallmark of cancer. The aim of this study is to evaluate promoter hypermethylation in cell‐free DNA as a prognostic marker for stage classification of pancreatic adenocarcinoma. Consecutive patients with pancreatic adenocarcinoma were prospectively included. Plasma samples were obtained before diagnostic work‐up and treatment. Patients were staged according to the TNM classification. Methylation‐specific PCR of 28 genes was performed. Prognostic prediction models for staging of pancreatic adenocarcinoma were developed by multivariable logistic regression analysis using stepwise backwards elimination. Ninety‐five patients with pancreatic adenocarcinoma were included. The mean number of hypermethylated genes was identical for stage I, II and III disease (7.09 (95% CI; 5.51–8.66), 7.00 (95% CI; 5.93–8.07) and 6.77 (95% CI; 5.08–8.46)), respectively, and highly significantly different from stage IV disease (10.24 (95% CI; 8.88–11.60)). The prediction model (SEPT9v2, SST, ALX4, CDKN2B, HIC1, MLH1, NEUROG1, and BNC1) enabled the differentiation of stage IV from stage I‐III disease (AUC of 0.87 (cut point 0.55; sensitivity 74%, specificity 87%)). Model (MLH1, SEPT9v2, BNC1, ALX4, CDKN2B, NEUROG1, WNT5A, and TFPI2) enabled the differentiation of stage I‐II from stage III‐IV disease (AUC of 0.82 (cut point 0.66; sensitivity 73%, specificity 80%)). Cell‐free DNA promoter hypermethylation has the potential to be blood‐based prognostic markers for pancreatic adenocarcinoma, as panels of hypermethylated genes enables the differentiation according to cancer stage. However, further validation is required. What's new? The treatment of pancreatic cancer depends on accurate staging, and while tools to assist with tumor evaluation are available, the development of novel minimally invasive approaches could greatly facilitate staging. Here, analyses of plasma‐derived cell‐free DNA show that promoter hypermethylation accumulates during neoplastic progression and changes with advancing cancer stage, with the mean number of hypermethylated genes in stage IV differing significantly from stages I, II, and III. Prognostic prediction models successfully differentiated patients according to cancer stage, suggesting that cell‐free DNA hypermethylation could serve as a blood‐based prognostic marker for