Liquid biopsy shows promise for early detection of pancreatic cancer
A liquid biopsy has shown promise for the early detection and staging of pancreatic ductal adenocarcinoma.
The test — developed by researchers at University of Pennsylvania’s Perelman School of Medicine, Abramson Cancer Center and School of Engineering and Applied Science — could play a valuable role in guiding treatment of the most common form of pancreatic cancer. If detected early, curative-intent surgery is possible.
“One important decision point in the care of a patient with pancreatic cancer is determining whether the tumor is localized to the pancreas, such that the patient is a candidate for surgical resection,” Erica L. Carpenter, MBA, PhD, director of the liquid biopsy laboratory at Abramson Cancer Center and senior author of the multidisciplinary study, told Healio. “We worked with Charles M. Vollmer, MD, and Mark H. O’Hara, MD, who are co-authors on this study, to identify patients in our clinic for whom blood samples and imaging data could be used for test development.”
Carpenter spoke with Healio about how the test was developed, its efficacy and its potential to improve care for patients with pancreatic cancer.
Question: This test seems to have an unprecedented ability to detect pancreatic cancer before it metastasizes. How did you develop the test?
Answer: The development of the test originated in conversations with our clinical team. As we were developing the test for early detection of pancreatic cancer, they pointed out that a noninvasive test such as ours might have clinical utility even after pancreatic cancer is diagnosed — if it can provide information beyond current standard-of-care diagnostics, such as imaging. Using a workflow similar to that used for the early diagnosis test, we used machine learning to identify the elements of a multi-analyte signature with the highest accuracy to distinguish patients with occult metastases — ie, those whose metastatic disease was below the level of detection of standard-of-care imaging.
Q: How did you establish this panel of biomarkers to detect pancreatic cancer? What role do these biomarkers play in pancreatic cancer?
A: We have combined three different types of biomarkers, each with a different role. Extracellular vesicles recently were discovered to be shed from tumor and other tissue in the bodies of patients with cancer. They circulate in blood and thus are easily obtainable with a simple blood draw. Importantly, they have been shown to provide a means for cells to communicate with each other and may play a functional role in the development and progression of small tumors. Extracellular vesicles are known to contain RNA, DNA and proteins that are representative of the tumor cells from which they were shed, meaning they provide a way of “sampling” the tumor without having to perform an invasive biopsy.
Second, we analyzed cell-free DNA (cfDNA) circulating in a patient’s blood by extracting the DNA and measuring its concentration. This simple measure of the amount of total cfDNA in blood often correlates with a patient with cancer’s disease burden, as well as prognosis. Next, we performed a droplet digital PCR (ddPCR) analysis on the cfDNA to identify mutations in the gene KRAS that are found in most pancreatic tumors. We then used the ddPCR results to measure the proportion of the total cfDNA that likely came from the tumor, also known as KRAS allele fraction. This also has been correlated with disease burden and worse outcome.
Finally, we added a protein analyte known as CA19-9, which typically is measured in clinical blood work for patients with pancreatic cancer. However, one of the limitations of the CA19-9 test is that not all patients with pancreatic cancer will have detectable amounts of the protein. We reasoned that a multi-analyte approach, combining extracellular vesicles, DNA, and CA19-9, would provide a more comprehensive and accurate picture of the patient’s disease than any one analyte alone.
Q: For whom would this test be appropriate?
A: As with all biomarkers, our test would have to be validated in a larger, independent study before it could be used in a clinical setting. However, assuming that validation is eventually achieved — and this is something we are actively working on — we envision the test to be used first for patients at risk, including those with inherited mutations in genes such as BRCA, those with other familial risk, and those with conditions such as diabetes or pancreatitis. To build the foundation for performing this sort of testing, we are working closely with co-authors Bryson William Katona, MD, PhD, and Ben Z. Stanger, MD, PhD, who lead efforts here at Penn Medicine in pancreatic cancer risk management and at the Pancreatic Cancer Research Center, respectively.
Q: What is the next step for research into liquid biopsy for pancreatic cancer?
A: An important next step for us, and for the field, is to start to integrate liquid biopsy into the clinical management of patients. Again, this assumes that the essential validation steps have been performed and the accuracy and reproducibility of the tests established. This will mean integrating liquid biopsy with, for example, endoscopic ultrasound and other standard-of-care tests used to monitor patients at risk. It will also mean proper and timely referral of patients at risk from primary care physicians to risk management programs.
Q. How might this test improve the prognosis for pancreatic cancer?
A: Patients with earliest-stage pancreatic cancer who are candidates for curative-intent surgery have the highest chance of achieving longer-term OS and PFS. Therefore, having a test that will shift detection to earlier stages of disease may lead to improved overall prognosis. It also will allow for investigation and clinical trials of additional therapeutic interventions for early-stage disease.
For more information:
Erica L. Carpenter, MBA, PhD, can be reached at Abramson Cancer Center, Perelman School of Medicine at University of Pennsylvania, SPE 8th Floor, Room 8-323, 3400 Civic Center Blvd., Philadelphia, PA 19104; email: firstname.lastname@example.org.