December 12, 2011

Landmark DNA study finds easier way to diagnose pancreatic cysts

Scientists at Johns Hopkins have surveyed the DNA in four common types of pancreatic cysts and have determined that each type bears a distinct pattern of gene mutations. Pancreatic cysts are present in about 2 percent of U.S. adults and can, in some cases, require surgical removal and microscopic analysis to determine their type and likelihood of turning cancerous.

“The findings should let us identify most pancreatic cysts accurately without major surgery, sparing patients unnecessary and hazardous procedures—and enabling those with dangerous cysts to have them removed long before they turn into pancreatic tumors,” said Bert Vogelstein, the Clayton Professor of Oncology at the Johns Hopkins Kimmel Cancer Center and co-director of Johns Hopkins’ Ludwig Center for cancer biology research.

Reporting Dec. 5 in the online early edition of the Proceedings of the National Academy of Sciences, the Johns Hopkins scientists also discovered clues to the cysts’ possible origins: a new tumor suppressor gene and disruptions to the ubiquitin protein-disposal system.

Kenneth W. Kinzler, professor of oncology at the Johns Hopkins Kimmel Cancer Center and co-director of the Ludwig Center, said, “Our results should help physicians manage the increasingly large number of patients with pancreatic cysts, and should help biologists understand how such cysts develop and turn cancerous.”

Pancreatic cysts are relatively common after middle age; most are not worrisome and, until recently, people who had such cysts seldom knew that they were there. However, the increasing use of MRI, CT and ultrasound scans of the abdomen means that these cysts are often found incidentally, posing a dilemma for doctors and patients.

“We would like to know which cysts are dangerous, but the clinical diagnosis of pancreatic cysts can be wrong unless the cyst is directly examined after being removed surgically,” said Ralph Hruban, professor of pathology and oncology and director of the Sol Goldman Pancreatic Cancer Research Center at the Johns Hopkins University School of Medicine, who designed the study in conjunction with Vogelstein. “Moreover,” he said, “such surgery can have complications, sometimes disturbing digestive and other pancreas-dependent functions, so we really need to find a way to diagnose these cysts without operating.”

One way would be to harvest fluid and cells from cysts during an ultrasound endoscopy of the pancreas—a relatively inexpensive and safe outpatient procedure—and check the DNA in the sample for mutations that are characteristic of the cyst type.

Vogelstein and Kinzler have helped pioneer the study of gene mutations that underlie cancer and, with Hruban, recently began applying this molecular pathology approach to pancreatic cysts. In a study reported in Science Translational Medicine in July, their research teams surveyed 169 cancer-related genes in the most common type of pancreatic cyst, intraductal papillary mucinous neoplasms, or IPMNs, and found that two genes were frequently mutated in these cysts.

In the new study, the team went further by surveying all the protein-coding regions of DNA in eight sample IPMNs as well as eight samples of the three other common pancreatic cyst types: serous cystadenomas, or SCAs; mucinous cystic neoplasms, MCNs; and solid pseudo-papillary neoplasms, SPNs.

Each turned out to have its own pattern of DNA abnormalities that was readily distinguishable from the others. The SCAs typically showed signs of a missing fragment of chromosome 3, disrupting the activity of the known tumor-suppressor gene VHL. The MCNs commonly had mutations in the growth-driving oncogene KRAS, as well as mutations in or losses of the gene RFN43, newly identified here as a tumor suppressor.

Surprisingly, Vogelstein says, SPNs bore hardly any mutations, except in the gene CTNNB1, which hints at an unusual mechanism of cancer development, since most tumors appear to arise only after multiple mutations. The IPMNs, as in the previous study, commonly had mutations to the oncogenes KRAS and GNAS, but this wider survey revealed that they typically (six out of eight samples) bore RFN43 mutations, too.

“Looking at the mutation patterns in these five genes and perhaps a few others should allow physicians to do a much better job of diagnosing and managing these cysts,” Hruban said.

The study has implications for basic cancer research as well, Kinzler notes. It identifies RFN43 as a tumor suppressor for the first time, and scientists may pursue whether RFN43 plays a role in other cancers. The study also points to the disruption of the ubiquitin system in cells as a key factor in the development of cysts. Three of the genes whose loss or mutation appears to underlie the development of pancreatic cysts—VHL, CTNNB1 and RFN43—are known to interact with the ubiquitin protein-disposal system, which helps remove abnormal proteins and helps keep normal proteins at healthy levels. VHL and RFN43 code for proteins that act as ubiquitin ligases, which means that they mark unwanted proteins for disposal by tagging them with ubiquitin molecules.

Disruptions of ligase activity could mean that the proteins they normally tag with ubiquitin are allowed to rise to unsafe levels. Similarly, the CTNNB1 mutations found in SPN cysts are known to make the gene’s product, the protein beta-catenin, resistant to normal disposal by ubiquitin-tagging. Disruptions to the ubiquitin system that leave the levels of key proteins unregulated have been implicated before in cancers, but they now seem particularly important in pancreatic cysts. “It’s an intriguing finding that will certainly stimulate further research,” Vogelstein said.

Major support for the study was provided by the Lustgarten Foundation, a private nonprofit dedicated to funding pancreatic cancer research. Other funding was provided by the Virginia and D.K. Ludwig Fund for Cancer Research, the Sol Goldman Pancreatic Cancer Research Center, the Joseph L. Rabinowitz Fund, the Michael Rolfe Foundation, the Stringer Foundation, the family of Ted and Julie Smith, the Indiana Genomics Initiative of Indiana University, which is supported in part by Lilly Endowment Inc., The J.C. Monastra Foundation, Swim Across America and grants from the National Science Foundation, the National Institutes of Health and the National Cancer Institute.

In addition to Vogelstein, Kinzler and Hruban, the authors of the report are Jian Wu, Yuchen Jiao, Marco Dal Molin, Anirban Maitra, Roeland F. deWilde, Laura D. Wood, James R. Eshleman, Michael Goggins, Christopher L. Wolfgang, Marcia L. Canto, Richard D. Schulick, Barish H. Edil, Michael A. Choti, Alison P. Klein, Hannah Carter, Rachel Karchin, Luis A. Diaz Jr. and Nickolas Papadopoulos, all of Johns Hopkins; Volkan Adsay, of Emory University; David S. Klimstra and Peter J. Allen, both of Memorial Sloan-Kettering Cancer Center; G. Johan A. Offerhaus, of University Medical Center of Utrecht in the Netherlands; Levy Kopelovich, of the National Cancer Institute; C. Max Schmidt, of Indiana University; and Yoshiki Naito, of Kurume University School of Medicine in Japan.