NEW YORK (GenomeWeb) – Moffitt Cancer Center researchers have shared early data suggesting a microRNA signature they discovered using Nanostring's nCounter platform has potential as a tool to detect and distinguish early pancreatic precancers noninvasively from blood samples.
In a study published this week in Cancer Prevention Research, the team, led by Moffitt research scientist Jennifer Permuth-Wey, described the 30-miRNA signature and shared data showing that it could distinguish individuals with intraductal papillary mucinous neoplasms (IPMNs) from healthy controls.
The group also identified a smaller set of five miRNAs that could discriminate between patients with high- and low-risk IPMNs.
Pancreatic cancer is typically diagnosed at a late stage, due to the absence of accurate methods for early detection or tools to distinguish cancers from other pancreatic conditions with similar symptoms.
Detecting and accurately evaluating early precancerous lesions, of which there are three main types, offers one potential avenue for reducing the disease's high mortality.
Besides IPMNs, pancreatic cancer precursor lesions include pancreatic intraepithelial neoplasia (PanIN) and mucinous cystic neoplasms (MCN). According to the study's authors, MCNs and IPMNs account for over half of the approximately 150,000 asymptomatic pancreatic cysts detected incidentally in the general population each year via radiologic imaging.
Once detected, fine needle aspirations are often performed to assess these lesions, but the biomarker analyses performed on the resulting tissue samples do not reliably predict disease severity.
Permuth-Wey told GenomeWeb this week that she and her colleagues had previously attempted to identify tissue-based miRNA biomarkers that could be useful to distinguish IPMNs and determine their relative risk.
In their newer study, the researchers decided to look instead at circulating miRNAs as a potential non-invasive diagnostic strategy.
The team used the nCounter technology to measure the levels of 800 miRNAs in plasma samples from 44 patients taken before they underwent surgery to remove IPMNs. The group also measured miRNAs in the blood of 25 healthy controls.
From this analysis, the researchers identified a panel of 30 miRNAs that distinguished individuals with IPMNs with an area under the receiver operating curve of 0.74. The candidate miRNAs had 2-fold to more than 4-fold higher expression in cases compared with controls, the authors wrote.
Interestingly, the signature contained novel miRNAs as well as miRNAs previously implicated in pancreatic cancer. MiR-145-5p was the circulating miRNA that was most significantly associated with IPMN, with an AUC of 0.79.
Permuth-Wey and her colleagues also studied a handful of matched plasma and tissue specimens to look at the correlation of miRNAs in these two sample types.
In 12 IPMN cases, among 160 miRNA probes evaluated in both specimen types, only three had expression levels that were significantly positively correlated, supporting growing evidence from liquid biopsy research suggesting that genomic and transcriptomic signals can be distinct in the tissue and circulation of the same individual.
Though the initial signature was not able to distinguish between high- and low-risk IPMNs, the team was able, by going back to their primary data, to also identify a smaller panel of five miRNAs that did so with an AUC of 0.73.
"Certainly as a former clinician I wish the accuracy was better, but I think … this is good preliminary data to suggest we should go further and design some more rigorous studies that are larger in scale," Permuth-Wey said.
"We wanted this to be a proof of principle," she explained, "to basically show that we could detect miRNAs in the blood plasma of these patients with pancreatic cancer precursors using this new technology from Nanostring."
According to Permuth-Wey, the ultimate goal for the group is to create an assay to help guide more personalized clinical care for detected IMPNs.
"We now know about these precursor lesions … which are radiologically detectable. But its kind of a double-edged sword in my opinion," she said. "It's great that you detect these precursors, but once you do its hard to distinguish between those that are low-grade, or low-risk, and those that are high-grade and that should be resected."
"In a perfect world we would have … for the medical team … a test they could order that would accompany the radiology findings, and increase accuracy to distinguish these groups … to add to the confidence in the decision for either surgery or watching over time," she said.
According to Permuth-Wey, the next major step for her team will be organizing a larger prospective validation of its miRNA signature.
The group recently received funding through the state ofFlorida, she said, for a multi-center study involving Moffitt as well as theUniversityofFloridaand theUniversityofMiamito validate the group's current findings in a prospective cohort of IPMN cases.
In this study, the partner institutions plan to recruit up to 100 cases over one year, Permuth-Wey said. Hopefully, they will also be able to extend the study to five years, and up to 500 cases, she added.
The researchers will continue to use the Nanostring platform for this continuing validation work. "Its one of those technologies that is available in numerous labs, so that was one of the attractive things about the technology," Permuth-Wey said.
If the validation is successful, she and her team would then work to develop a clinical diagnostic assay.
The group has submitted a patent application for a miRNA assay for detection and management of pancreatic cancer precursors.
"We'd plan to get things incorporated into our CLIA lab here at our institution," she said.
Nanostring, which is developing its own diagnostics in other cancer areas including breast cancer and lymphoma, said that it has not discussed publicly any internal plans in the pancreatic cancer space.