NEW YORK – Investigators at Oregon Health and Sciences University have developed a novel and simple assay they say could improve screening and detection of pancreatic cancer among individuals with elevated risk, such as those with diabetes or other disorders of the pancreas.
The team published a study last week in Science Translational Medicine, which described their discovery of a novel pancreatic cancer biomarker, a member of the MMP protease family, and a retrospective validation that demonstrated sensitivity over 70 percent with nearly 100 percent specificity.
Dubbed PAC-MANN, short for "protease activity-based assay using a magnetic nanosensor," the approach is unlike other proteomic tests that aim to quantify the levels of target molecules. Instead, it offers an indirect readout of the activity of the protease in question, via a simple fluorescence-based platform.
Jared Fischer, the study's corresponding author and a scientist at the OHSU Knight Cancer Institute’s Cancer Early Detection Advanced Research Center, said that pancreatic ductal adenocarcinomas secrete large amounts of certain proteases during their growth.
"Looking at protein levels … a lot of times the total amount of proteins isn't very different between cancer and normal … What makes what we were doing different is looking at the activity."
Proteases are initially inactive. "You don't want them to start doing their job until they get to the right location," added Jose Montoya Mira, the study's first author, also part of OHSU's CEDAR center.
"What we found was that the level of the inactive proteases was actually very similar between healthy and cancer. But in cancer there is activation of these proteases," which are part of the mechanisms that degrade the extracellular matrix and allow cancer cell invasion and dissemination into normal tissue, he added.
PAC-MANN is designed to capture these active proteases in the blood and isolate them from a background of inactive proteases to measure activity.
In the team's retrospective case-control validation, which included blood samples from 110 pancreatic cancer patients, 170 healthy controls and 76 patients with non-cancer diseases such as pancreatitis, PAC-MANN identified pancreatic cancer samples with specificity of 98 percents and sensitivity of 73 percent.
Notably, it distinguished 100 percent of patients with noncancer neoplasia from patients with pancreatic cancer, suggesting it could be especially reliable in the established clinical niches where suspicions of pancreatic cancer drive the need for reliable discrimination.
Broken down by stage, PAC-MANN alone detected 62 percent of stage I, 56 percent of stage II, 92 percent of stage stage III, and 85 percent of stage IV cancers.
Combining PAC-MANN with the clinically used biomarker CA 19-9, sensitivity improved across stages with 85 percent sensitivity for stage I, 81 percent for stage II, and 92 percent for stages III and IV.
Although CA 19-9 is approved by the US Food and Drug Administration, its low positive predictive value limits its use, and oncologists mostly test for the protein in the context of measuring disease burden during treatment not as an early-detection tool.
Unlike DNA-based approaches for early detection being explored by numerous academic and commercial groups that require costly and complex next-generation sequencing, PAC-MANN is based on relatively simple technology.
"We have a small nanoparticle, an iron oxide nanoparticle. On the particle, we have the mimic sequence of the target of that protease," said Montoya Mira. On the other side of that mimic peptide is a fluorescent molecule.
"We take that construct, and we take some serum, and we incubate it for 45 minutes. If the protease is there and it's active, it will chop that region off releasing the fluorescent molecule," Montoya Mira said.
A magnet is then used to remove all the magnetic particles. The remaining fluorescent molecules, having been severed from their magnetic vehicle, offer a measure of the presence of active proteases.
Fischer added that the foundational biology of proteases suggests that they could be a rich vein for cancer detection targets across tumor types, as they play a role throughout cancer progression from early growth through metastasis.
"If you're looking at levels of a single protein, that's just a single number, and what happens many times is that that number can be too low for you to detect … but if you look at activity … it's an exponential reaction, so your actual signal becomes much higher," Fischer said.
In their initial discovery work, he and his colleagues screened a series of protease-cleavable peptide probes before homing in on the one with the highest sensitivity. Different molecules might be optimal for other types of cancer.
For PAC-MANN, the next steps will be to further validate the assay in independent sample sets. Clinically, the team is focusing on tuning the test to utilize with high-risk individuals.
"Pancreatic cancer is so rare that if you start looking at the general population, your positive predictive value is just extremely low," Fischer said. Because the test uses such a small volume of blood it would also allow frequent screens that currently aren't practical for patients or affordable for the healthcare system.
"A lot of the companies are going toward the general population because that has more money involved, but the nice thing in being a hospital and a university is that we can focus on populations where it actually is the best use case because we're not trying to make a profit," he added.
Fischer highlighted the team's data on the test's discrimination of pancreatic cancer from non-cancer neoplasia, which showed nearly 100 percent accuracy. In this population, CA 19-9 is "basically a coin flip," he said, whereas PAC-MANN appears to be far from it.
"Even if our tests are positive, you'd still do the endoscopic ultrasound or other tests as a follow-up," Fischer added. But that is a very different proposition from using those types of tests as a regular screen. "I wouldn't want to have to get an endoscopic ultrasound constantly," he said, whereas, regular blood testing is an economic, patient-friendly, and practical proposition.
Moreover, the test could also be used serially to monitor those who test positive but don't show other signs of disease, with lingering positivity potentially leading to more extreme measures.
The OHSU team is not the only group to explore proteases for cancer detection. A team at MIT spun out a company about a decade ago, called Glympse Bio, to advance its technology, which involves introducing similar nanoparticles constructs to PAC-MANN's into a patient's body, and then detecting the cleaved substrate in urine as a measure of the level of protease activity. Glympse merged with Singapore-based biotech firm Sunbird Bio in 2023.
Another firm, Hawkeye Bio has also developed an approach to measuring proteases without having to introduce anything into the body, showing some strong data in lung cancer last year.
For their part, Fischer, Montoya Mira, and colleagues are already working on a follow-up validation in an independent cohort in collaboration with University College London. This will include pre-diagnostic samples from patients later diagnosed with pancreatic cancer, which will allow the group to explore how long before clinical diagnosis their biomarker might be able to detect emerging disease.
The team is also looking for other groups to run their own trials of the assay in different populations.