Investigational blood biomarker panel may improve detection of pancreatic cancer

A biomarker blood panel that incorporates four different proteins—ANPEP, PIGR, CA19-9, and THBS2—enhanced the detection of pancreatic ductal adenocarcinoma (PDAC) compared to measuring CA19-9 levels alone, according to a study led by researchers from the Perelman School of Medicine at the University of Pennsylvania and published in Clinical Cancer Research, a journal of the American Association for Cancer Research (AACR).

Pancreatic ductal adenocarcinoma (PDAC), which forms from the exocrine cells in the pancreas, is the most common form of pancreatic cancer, accounting for about 95 percent of cases. When detected at a localized, early stage, the five-year relative survival is about 44 percent, but once the disease has metastasized, the five-year survival drops to 3 percent. Most cases of pancreatic cancer are diagnosed at late stages.

“Pancreatic cancer usually doesn't present with symptoms until it's too late for surgery, when the cancer has already metastasized to other parts of the body,” said Kenneth S. Zaret, PhD, senior author of the study and a professor of Cell and Developmental Biology. “Our goal was to look for biomarkers in the blood that appear in early-stage PDAC patients, to catch the disease early,” he explained.

CA19-9 is a protein that is released into the blood by both cancer cells and normal pancreatic cells, and elevated levels of CA19-9 in the blood can be a sign of pancreatic cancer or other diseases. THBS2 (thrombospondin 2) is a protein that is expressed in PDAC tumor tissue, and when screened for in blood alongside CA19-9 using a bioassay, enables the test to better discriminate between PDAC and healthy patients, according to previous work from Zaret and colleagues. However, Zaret noted that these biomarkers are insufficient for routine early detection.

“CA19-9 is widely used to monitor diagnosed pancreatic cancer but isn’t recommended as a standalone screening test—benign conditions can elevate it in some people, while others may have low levels, even if they have pancreatic cancer. THBS2 is investigational and can complement CA19-9, but its prediagnostic performance has been mixed,” he said.

Biomarkers identified from patient cohorts at two institutions

To identify novel biomarkers of pancreatic cancer, Zaret and colleagues analyzed plasma samples from two cohorts: 537 samples from the Mayo Clinic (Mayo) and 135 from the Hospital of the University of Pennsylvania (Penn). The cohorts included patients with confirmed pancreatic cancer, healthy individuals, and patients with benign pancreatic disease, which enabled the researchers to evaluate the ability of candidate biomarkers to differentiate between pancreatic cancer and benign conditions that could mimic cancer.

By comparing protein levels across the plasma samples, the researchers identified two proteins, ANPEP (aminopeptidase N) and PIGR (polymeric immunoglobulin receptor), that showed elevated levels in samples from patients with early-stage PDAC, compared with samples from patients without cancer.

Zaret and colleagues then developed a panel that measured the blood levels of four biomarkers: the newly identified ANPEP and PIGR, along with the previously known biomarkers CA19-9 and THBS2.

Across the two independent cohorts, Mayo and Penn, the four‑biomarker blood panel showed area under curve (AUC) values of 0.97 and 0.96, respectively, when comparing stage 1-2 PDAC and healthy controls. The AUC is a measure of the test’s ability to distinguish between two groups, with 1.0 being a perfect score.

The panel was also able to distinguish cancer from benign pancreatic conditions with an AUC 0.87 for early‑stage PDAC and 0.91 for all stages in the Mayo cohort.

Four-biomarker panel outperforms single-biomarker testing

The four-biomarker panel correctly detected 91.9 percent of pancreatic cancers across all stages and 87.5 percent of early‑stage cases, compared to testing for the CA19-9 biomarker alone, which identified 82.7 percent of the PDAC cases overall and 76.2 percent of the early-stage cases. The improvement in all-stage cancer detection was statistically significant, whereas the gain in early‑stage detection did not reach statistical significance in spite of the 11.3 percent improvement in sensitivity.

If confirmed in larger, prospective studies, the four-biomarker blood panel could improve the ability to identify which individuals with a high risk for pancreatic cancer would benefit from follow-up imaging, explained Zaret, adding that this could allow clinicians to detect more pancreatic cancers at an earlier, more treatable stage.

“With the addition of ANPEP and PIGR, the panel helps to overcome known limitations associated with CA19-9 and THBS2 testing—such as patients who genetically underexpress CA19-9 or tumors that present as different molecular subtypes—and could therefore reduce the number of missed cancer cases while keeping false positives low,” he said.

One limitation of this study is that the cohorts did not include individuals at increased risk for pancreatic cancer, such as those with a family history, germline BRCA mutations, or new-onset diabetes, introducing a bias in test performance, noted Zaret.

Another limitation is the retrospective nature of the study, which means that larger prospective studies are needed to confirm the panel’s real-world screening performance, Zaret explained.

The study was supported by the National Institutes of Health, Penn Pancreatic Cancer Research Center, A Love for Life, the Lustgarten Foundation, the Centene Foundation, and the Mayo Clinic Comprehensive Cancer Center. Zaret declares no conflicts of interest.