An experimental genetic screen that detects DNA methylation changes appears to discriminate pancreatobiliary cancers from benign disease with a high degree of accuracy, reported in the November 2008 issue of Clinical Gastroenterology and Hepatology.
Although the screen is yet not ready for clinical use, it could eventually help identify patients with microscopic preinvasive disease — particularly pancreatic intraepithelial neoplasia, said John Hopkins Medical Institute, Baltimore, and his coauthors. The method identifies aberrantly methylated DNA in brush samples obtained during endoscopic retrograde cholangiopancreatography (ERCP).
“Detecting pancreatic intraepithelial neoplasia is particularly important for patients with an inherited redisposition to pancreatic and other periampullary cancers . . . . The ability to reliably detect and quantify [the disorder] using molecular assays in high-risk individuals would help identify those needing more surveillance to detect advance pancreatic neoplasia, and could also open up the option of enrolling affected individuals in chemoprevention trials.”
DNA methylation is one of several genetic alternations that can arise during pancreatobiliary tumorigenesis. A number of methylated genes have been identified in such cancers, including TFP12, NPTX2, and cyclin D2, as well as FOXE1. Assessed the predictive value of methylation changes in these and other sequences, compared with cytology, in brush samples obtained endoscopically from 130 patients with biliary tract strictures in the biliary (118) or pancreatic (8) ducts, or both (4).
Pancreatic ductal adenocarcinoma was present in 41 of the samples. Other diagnoses were biliary tract adenocarcinoma (10), other periampullary malignancies (13), primary sclerosing cholangitis — associated strictures (27), and non-PSC-associated strictures (39).
The samples were examined with both conventional methylation-specific polymerase chain reaction (MSP) and quantitative real-time methylation-specific polymerase chanin reaction (QMSP).
The MSP analysis tested five gene sequences, and found significantly more methylated genes in the samples of patients with pancreatic cancer than in samples from patients with noncancerous conditions (64% of all genes in the group with cancer vs. 37% of all genes in the group without cancer). But none of the genes achieved 80% specificity, which is usually seen in cytology studies.
Even combining markers — for example, requiring two positive markers for a diagnosis — did not raise sensitivity and specificity above that associated with cytology.
The QMSP analysis was more useful, the authors said. This analysis used three genes: TFP12, NPTX2, and cyclin D2. Methylated genes were found in 735 of patients with pancreative adenocarcinoma, 80% of those with biliary tract cancer, and 23% of those with other pancreatobiliary neoplasms, but only 14% of those with benign disease — a significant difference.
QMSP was 19.5% more sensitive than was cytology for pancreatic adenocarcinoma and 30% more sensitive for periampullary cancer. In fact, combining cytology and QMSP did not improve the sensitivity above that QMSP alone.
The test might also help to distinguish patients whose strictures are caused by primary sclerosing cholangitis from those with strictures related to other conditions. “We found evidence that patients with [PSC-related] strictures have a higher prevalence of low-level methylation (41%) than those with other benign strictures (15%).” While cytology remains the preferred method for cancer diagnosis, “an accurate marker panel, such as QMSP, that has been extensively validated, could aid cytology in establishing a diagnosis of cancer.”