Utility of promoter hypermethylation in malignant risk stratification of intraductal papillary mucinous neoplasms

Ankit Chhoda, Anup Sharma, Bethsebie Sailo, Haoyu Tang, Nensi Ruzgar, Wan Ying Tan, Lee Ying, Rishabh Khatri, Anand Narayanan, Shrikant Mane, Bony De Kumar, Laura D. Wood, Christine Iacobuzio-Donahue, Christopher L. Wolfgang, John W. Kunstman, Ronald R. Salem, James J. Farrell, Nita Ahuja

Research output: Contribution to journalArticlepeer-review


Background: Intraductal papillary mucinous neoplasms (IPMNs), a type of cystic pancreatic cancer (PC) precursors, are increasingly identified on cross-sectional imaging and present a significant diagnostic challenge. While surgical resection of IPMN-related advanced neoplasia, i.e., IPMN-related high-grade dysplasia or PC, is an essential early PC detection strategy, resection is not recommended for IPMN-low-grade dysplasia (LGD) due to minimal risk of carcinogenesis, and significant procedural risks. Based on their promising results in prior validation studies targeting early detection of classical PC, DNA hypermethylation-based markers may serve as a biomarker for malignant risk stratification of IPMNs. This study investigates our DNA methylation-based PC biomarker panel (ADAMTS1, BNC1, and CACNA1G genes) in differentiating IPMN-advanced neoplasia from IPMN-LGDs. Methods: Our previously described genome-wide pharmaco-epigenetic method identified multiple genes as potential targets for PC detection. The combination was further optimized and validated for early detection of classical PC in previous case–control studies. These promising genes were evaluated among micro-dissected IPMN tissue (IPMN-LGD: 35, IPMN-advanced neoplasia: 35) through Methylation-Specific PCR. The discriminant capacity of individual and combination of genes were delineated through Receiver Operating Characteristics curve analysis. Results: As compared to IPMN-LGDs, IPMN-advanced neoplasia had higher hypermethylation frequency of candidate genes: ADAMTS1 (60% vs. 14%), BNC1 (66% vs. 3%), and CACGNA1G (25% vs. 0%). We observed Area Under Curve (AUC) values of 0.73 for ADAMTS1, 0.81 for BNC1, and 0.63 for CACNA1G genes. The combination of the BNC1/ CACNA1G genes resulted in an AUC of 0.84, sensitivity of 71%, and specificity of 97%. Combining the methylation status of the BNC1/CACNA1G genes, blood-based CA19-9, and IPMN lesion size enhanced the AUC to 0.92. Conclusion: DNA-methylation based biomarkers have shown a high diagnostic specificity and moderate sensitivity for differentiating IPMN-advanced neoplasia from LGDs. Addition of specific methylation targets can improve the accuracy of the methylation biomarker panel and enable the development of noninvasive IPMN stratification biomarkers.

Original languageEnglish (US)
Article number28
JournalClinical Epigenetics
Issue number1
StatePublished - Dec 2023


  • Methylation-specific biomarker
  • Pancreatic cancer
  • Pancreatic cyst

ASJC Scopus subject areas

  • Genetics(clinical)
  • Genetics
  • Molecular Biology
  • Developmental Biology


Dive into the research topics of 'Utility of promoter hypermethylation in malignant risk stratification of intraductal papillary mucinous neoplasms'. Together they form a unique fingerprint.

Cite this