Septic shock is a severe systemic infection and major cause of death for the old and young alike. Unfortunately, researchers say testing new drug regimens to stop the infection is confounded because clinical trials include patients who are either too sick to be saved by experimental therapies or not sick enough to warrant the treatments.
In a study published in the April edition of Critical Care Medicine, researchers at Cincinnati Children's Hospital Medical Center and the University of Cincinnati report a new blood test that helps solve the dilemma by identifying low-risk and high-risk patients. Based on genome-wide studies of septic shock patients, the tool measures five protein biomarkers in the blood that are combined to accurately estimate the risk of mortality and determine which patients should receive experimental therapeutic intervention.
The study involved 882 adults in intensive care units in medical centers located in the United States, Finland and Canada. The new study follows one published earlier this year by the same research group that developed a similar test for children. The new study included three different groups of patients to develop, test and retest the tool. This helps to make sure it works in different patient mixes -- often a limiting factor in large clinical trials.
"Substantial resources are invested in trying to find new treatments for septic shock, but the vast majority of them fail when they get to clinical trials," explains Hector Wong, MD, a lead investigator and director of Critical Care Medicine at Cincinnati Children's. "There are many reasons for this, but a consistent one is that the baseline mortality risk varies widely in septic shock patients, which muddies the water."
Helping lead the study was Wong's collaborator, Christopher Lindsell, PhD, Professor of Emergency Medicine and Associate Dean for Clinical Research at the University of Cincinnati College of Medicine.
Septic shock can cause inflammation and damage to critical organs. Organ damage is a key driver of the infection's deadly nature. The authors say a significant part of the challenge in determining who is at greatest risk of death is the complex medical variation among patients with septic shock -- who differ in age, underlying medical conditions and general overall health.
Researchers selected the biomarkers used in the stratification tool because of the clear links they have to a person's infection and inflammation. As well as biomarkers, the researchers included measurements of patient age, lactate levels, and the number of chronic diseases patients had. They then used a mathematical model to combine the information into a single tool that separates high- and low-risk patients.
The authors report that in general the tool accurately determined that patients who tested positive had less than a 50 percent chance of surviving, but patients who tested negative had more than a 90 percent chance of surviving. More detailed analyses lets the authors identify patients with a more than a 98 percent chance of surviving, and those with more than 75 percent chance of dying.
When the researchers tested the tool repeatedly in different groups of patients, it continued to perform well. It also performed better than what is already available, the APACHE II/III (Acute Physiology and Chronic Health Evaluation) testing. The authors say this is because APACHE II/III does not use blood tests and relies more heavily on measurements with considerable variation, such as blood pressure.
One important finding is that three of the biomarkers helpful in stratifying risk for adults were the same as those used for stratifying children. The researchers say having the same three biomarkers verified in both groups mean that, in addition to identifying those patients unlikely to survive, the biomarkers consistently point to the same underlying problems caused by sepsis, providing potential targets for treatment. Not only can the tool help identify which patients should be included in clinical trials, it also points to specific treatments that should be tested.
Cite This Page: