University startup gets distribution deal for sepsis AI diagnostic tool

Apr. 10—CHICAGO — A biotech company with local roots has landed a distribution deal for its AI tool for sepsis diagnosis, with the news coming just one week after the technology received authorization from the Food and Drug Administration.

Prenosis, which is based in Chicago, has announced that it will work with Roche Diagnostics to bring its Sepsis ImmunoScore tool to hospitals across the country. The technology is the first AI tool for sepsis diagnosis to receive marketing authorization from the FDA, the company said.

"They (Roche) have thousands of sales representatives across the United States and deep relationships with many of these hospitals, and so they can really help us get it out there and get to save as many patients as possible as quickly as possible," said Prenosis CEO Bobby Reddy Jr., a University of Illinois Urbana-Champaign alum who co-founded the company alongside Rashid Bashir, dean of the Grainger College of Engineering.

According to the Centers for Disease Control and Prevention, sepsis is a "life-threatening medical emergency" that occurs when the body has an "extreme response" to an infection.

"In a typical year, at least 1.7 million adults in America develop sepsis and at least 350,000 adults who develop sepsis die during their hospitalization or are moved into hospice care," the CDC reported.

Roche vetted several sepsis algorithms before choosing Sepsis ImmunoScore as its exclusive partner for a five-year distribution contract, Prenosis officials said. Sepsis ImmunoScore will be available on Roche's navify Algorithm Suite platform.

The diagnostic tool uses lab-based biomarkers, clinical data and AI to evaluate a patient's likelihood of having or developing sepsis within 24 hours of assessment.

Prenosis has spent 10 years working with hospital systems to build a proprietary biobank, dataset and Immunix precision medicine platform that led to the development of Sepsis ImmunoScore.

According to Reddy, the first hospital they worked with to build this collection was Carle Foundation Hospital.

"I started as a post-doc in Rashid's group and we had a technology out of the university that we thought was really cool and we thought that it was going to help lots of different patients across many different diseases," he said. "... We actually had our headquarters there. For the first couple of years, it was actually within the cancer center at Carle."

During their time shadowing medical professionals, Reddy and his colleagues noticed a major need within the field.

"One of the biggest problems that we saw in the hospitals was just that there wasn't a way for clinicians to get deep insight into patient biology, and there wasn't really an understanding of the different types of biology that patients could have," he said.

They decided to pivot to focus on addressing this issue, which led to the creation of an extensive biobank and dataset.

Prenosis now has a collection of over 100,000 blood samples from over 25,000 patients at 10 health systems across the country, including Carle.

The company said its proprietary dataset "enables the rapid development of a pipeline of AI diagnostics and precision therapeutics for acute care, including the Sepsis ImmunoScore."

Sepsis ImmunoScore uses up to 22 biological parameters to determine a patient's level of risk, company officials said.

Prenosis Vice President of Health System Management Robin Carver, who has experience in the field of acute care, said that the tool is meant to "compliment" clinicians' workflow and assessments.

"As clinicians, when patients come into the emergency department, for example, we begin to triage them," she said. "We take their blood pressure, their temperature, their vital signs. We take some initial bloodwork from them. But once we suspect the patient has sepsis, we have to make a decision: Do I treat this patient or do I not treat this patient if I suspect them of sepsis? But we don't have really great ways of accurately telling us that information."

If a hospital treats a patient for sepsis who does not in fact have the condition, it can cause more problems, she said. For instance, if they treat the individual with antibiotics, it can create antimicrobial resistance.

Carver said Prenosis has worked to provide a solution that can give an accurate diagnosis so that clinicians can determine if they need to treat a patient for sepsis, as well as their next steps for helping that individual.

According to Reddy, Sepsis ImmunoScore places patients in four different risk categories for sepsis: low, medium, high or very high. The tool also predicts outcomes such as whether a patient is expected to die in a hospital, how long their stay might be, whether they will go to the ICU and what interventions might be necessary.