Endpoint Health’s approach identified two patient groups with significantly different treatment impact on 28-day mortality
Endpoint Health, Inc., a precision-first therapeutics company dedicated to addressing urgent needs in immune-driven critical and chronic illnesses, today announced findings from retrospective analyses published in the peer-reviewed medical journal SHOCK. The research showed that, in certain septic shock patients who possess similar gene expression signatures, the administration of a commonly used treatment may result in significantly greater 28-day mortality than in patients with the gene expression signature who did not receive the therapy. Septic shock is the leading cause of noncardiac death in intensive care units (ICUs) in the United States.
These findings indicate it may be possible to advance care for patients suffering from septic shock by assessing gene expression and developing personalized treatment plans that are precisely attuned to those gene expression signatures.
A retrospective analysis pooled publicly available data from various gene expression datasets of patients with sepsis and septic shock to develop a gene expression-based classifier that was then retrospectively tested, using data from two randomized placebo-controlled studies in patients with vasodilatory shock, including a trial in septic shock patients (VANISH) and a study in patients with severe burns (Burn study).
Using Endpoint Health’s proprietary approach to unsupervised clustering, the researchers identified two patient subclassifications, which were based on the expression of only 15 genes that reflect the state of a patient’s immune system activity. Patients were classified as immune-innate prevalent (IN-P), if their innate immune system was predominantly active, or immune-adaptive prevalent (IA-P), if they predominantly exhibited adaptive immune system activity.
The primary outcome was the assessment of 28-day mortality.
In the retrospective analysis of the VANISH study, hydrocortisone therapy was associated with significantly increased 28-day mortality in the IA-P subgroup: 43.3% for hydrocortisone vs. 14.7% for placebo, and in the IN-P subgroup, this outcome was reversed: 32.1% for hydrocortisone vs. 40% for placebo; for the interaction between hydrocortisone and the subgroups.
The same mortality trends for the two subgroups were seen in the Burn study, but the differences were not statistically significant for the interaction between hydrocortisone and the subgroups.
In addition, in the Burn study, blood specimens were collected before and after hydrocortisone therapy, and analysis of these specimens showed that hydrocortisone significantly suppressed the gene expression signature associated with adaptive immune function.
“While corticosteroids are currently recommended as a therapeutic option in patients with septic shock, we know there have been conflicting clinical findings regarding their impact on mortality,” said Mitchell Levy, MD, Director of Critical Care Medicine at Lifespan, and Chief of the Division of Pulmonary, Critical Care and Sleep Medicine, and co-author of the Surviving Sepsis Campaign International Guidelines for the Management of Sepsis and Septic Shock. “This research is a meaningful step towards exploring the possibility that subgroups of septic shock patients may respond differently to hydrocortisone therapy depending on their biological characteristics.”
“We are excited to share this scientific analysis in septic shock in SHOCK, which comes on the heels of our recently published ARDS study in BMJ Open,” added Jason Springs, co-founder and Chief Executive Officer of Endpoint Health. “These latest findings further illustrate the potential of our proprietary AI and machine learning platform to inform personalized care for critically ill patients, based on electronic health data and gene expression scoring.”
