Bluestar Genomics has presented data from a study assessing its non-invasive pancreatic cancer test at the American Pancreatic Association (APA ) annual meeting.

The trial included the examination of whole blood samples from 917 patients, with 117 being pancreatic cancer patients and the remainder being made up of non-cancer control patients with and without new-onset diabetes.

The analytical performance data indicated that the company’s test detected a pancreatic cancer signal in patients of several age groups, including those with new-onset diabetes.

The new test uses a blood sample to evaluate whether the person has an abnormal DNA signal linked to pancreatic cancer.

It uses the company’s epigenomics technology platform, which leverages machine learning along with a DNA-based 5-hydroxymethylcytosine (5hmC) biomarker to detect early stage cancer.

This predictive model had a median of 51.5% sensitivity and 98% specificity.

The early detection of pancreatic cancer helps patients to access timely treatment options.

Bluestar Genomics CEO and chief scientific officer Samuel Levy said: “Along with our test’s recent designation by the FDA as a breakthrough device, the findings from this study presented at APA underscore the opportunity we have to make a positive difference for patients by bringing this pancreatic cancer test to market.

“Based on these results, we are moving forward with completing the analytical validation in the coming months and making a CLIA laboratory-developed test available next year.”

According to the company, about 25% of the estimated 60,000 pancreatic cancer patients detected in the US every year are found to have new-onset diabetes before being detected with pancreatic cancer.

Bluestar Genomics claimed that its non-invasive pancreatic cancer test has the potential to annually screen approximately one million adult patients with new-onset diabetes.

Additionally, the company is planning a large clinical trial for further validation of the liquid biopsy test.