Researchers at Duke have discovered a biomarker that can be used to diagnose and develop effective treatments for sepsis.

Sepsis is the runaway immune response to infection and is responsible for the deaths of more than 250,000 Americans annually, accounting for a considerable number of hospital deaths.The group of investigators found that the presence of a molecule called methylthioadenosine, or MTA, could serve as a strong predictor of which patients are at an elevated risk of dying from sepsis—a considerable advantage given the difficulty of treating it.

“This area has been a graveyard for the pharmaceutical industry, with more than 100 failed clinical trials of therapies that target the body’s abnormal response to infection,” said Dr. Dennis Ko, assistant professor of molecular genetics and microbiology, in a Duke Today release.

Ko explained that the MTA biomarker is still far from being implemented into everyday clinical practice. He added that further studies are warranted to better evaluate the significance of MTA levels in helping patients. Despite these cautions, however, Ko said he remains confident based on his team’s findings so far.

“As of now, our research demonstrates that calculating risk of death can be done with fairly high accuracy, meaning that we can determine which patients require more aggressive care, based on their risk of death,” Ko said.

The group identified the inflammatory response that causes sepsis, demonstrating that differences in levels of MTA can be used to differentiate between those who will survive the infection and those that will likely pass away as a result. Among patients studied, those with higher levels of the biomarker tended not to survive.

Ko also tested whether modifying levels of MTA in mice could affect sepsis prognosis, finding that mice given MTA before being infected with Salmonella lived longer than those without any intervention.

The researchers hope that the results can eventually be connected to a particular therapy or to clear clinical strategies.The idea is that the failure of drugs to be effective is not indicative of the therapy itself, Ko said, but instead narrows down the pool of potential patients to those who would benefit from treatment.

“We want to couple those drugs produced by biotech and pharmaceutical companies that have failed with the biomarker to identify a therapy that could work in given scenarios given a certain subset of the inflammatory response that is inflicting a patient,” Ko said.