American Association for Clinical Chemistry
Better health through laboratory medicine
March 8, 2007

In this Issue...

Distinguishing Sepsis from Systemic Inflammation:
Microarrays Show Diagnostic Promise
By Julie McDowell


Even though sepsis is a swift-moving infection that can cause organ failure and death within days, there is currently no rapid diagnostic test to confirm the infection. In addition, it’s difficult to distinguish sepsis from non-infectious systemic inflammation, which presents similar symptoms of fever and an elevated white blood cell count. Recently, however, researchers at Washington University School of Medicine (St. Louis, Mo.) used molecular profiling to analyze gene expression patterns in the blood of mice, and found that the technique had an accurate discrimination rate of 94% between sepsis and non-infectious inflammation. This issue of Strategies examines this recent study, and what this means for a potential bedside test that could quickly diagnose sepsis.

Of the more than 700,000 Americans affected by sepsis every year, about 30% die, according to the National Institutes of Health. Accurate and rapid diagnosis is important—as well as treatment with appropriate antibiotics and eradicating the source of the infection. This is especially challenging in a hospital’s intensive care unit, where not only are patients suffering from severe illness, but they also cannot communicate their symptoms because they are often intubated or sedated, explained J. Perren Cobb, MD, Director of Washington University’s Center for Critical Illness and Health Engineering in St. Louis, Mo.

“Because it’s difficult to diagnose in the intensive care unit, our strategy for several decades has been to treat patients with a very broad spectrum of antibiotics,” he said. “ For the patient who is dying of sepsis, that is appropriate, but for those who do not have an infectious source, that results in eradicating all of the good organisms in the body. They become like Petri dishes for all of the evil organisms that live in intensive care units. This broad use of antibiotics also results in the bacteria that live in the ICU becoming very resistant organisms.”

In his over 20 years practicing critical care medicine, Cobb has also struggled with the inability to obtain a rapid diagnosis for sepsis and non-infectious systemic inflammation. Currently, the diagnosis is made through lab analysis of cultured blood, sputum, or urine, which can take days. But Cobb was hopeful that using molecular tools might provide a diagnosis within hours. “In the ICU, having a test like troponin I has really helped us to discern, very rapidly, who is having a heart attack and who is not,” he said. “Imagine if we had a diagnostic like that, a single lab test that would allow us to differentiate who is septic and who is not, and even, if they are septic, what they are septic with. If the strategy for using a molecular diagnostic can be effective in this patient population for something equally difficult to diagnose, say an acute heart attack, how about for infections?”

Applying GeneChip Technology

In their research on sepsis and molecular diagnostics, Cobb and his colleagues used GeneChip arrays, manufactured by Affymetrix (Santa Clara, Calif.), as well as software from Partek Incorporated (St. Louis, Mo.) to analyze molecular profiles in four groups of mice: three groups who had sepsis of varying severity, and a fourth group who had non-infectious inflammation. At 24 hours after injury, researchers were able to use blood microarray profiles to distinguish the mice with sepsis with a rate of 94% accuracy. In addition, they were able to analyze genes that are common responders to inflammation and infection in mice. Some of these genes have previously been identified by other researchers studying inflammation and infection in humans, but had not previously been linked with the development of sepsis. These findings were published in the November 2006 issue of the Journal of the American College of Surgeons [203(5):585-598].

While this research was based on an animal model, Cobb believes that these findings are an important first step in developing a similar test for human patients. In his current research, he and his colleagues are using gene expression profiles from human circulating leukocytes to derive graphs of patients’ clinical trajectory as they recover from illness complicated by pulmonary sepsis, or pneumonia. The researchers presented their early results last month at the annual meeting of the Society of Critical Care Medicine in Orlando, Fla., and an abstract of the study was published in a recent issue of Critical Care Medicine [34 (suppl.):A47, 2007].

“Our preliminary data suggest that these graphs of RNA information from leukocytes, which we call 'riboleukograms', may provide a powerful new quantitative tool for clinicians faced with diverse and clinically complex patients,” said Cobb. In addition, the researchers are currently looking at narrowing genetic markers in patients that would further the development of a bedside test to screen for symptoms of septic infection, as well as whether gene chips can identify the specific pathogen present in the septic infection.

Revolutionary Potential

While this data is preliminary and still needs to be extensively validated in human subjects, a molecular test would revolutionize the treatment for sepsis, said Herbert P. Wiedemann, MD, Chairman of Cleveland Clinic’s Department of Pulmonary, Allergy and Critical Care Medicine.

“If these researchers could detect the identity of specific infecting organisms within a few hours using the GeneChip technology, that would be a home run,” he said. “We would know exactly what antibiotics to use. We often have to treat very broadly because we don’t know what is present, and the use of broad spectrum antibiotics is associated with significant problems. Because you are using so many agents, the chance of getting a specific toxicity is higher and the chance of developing resistant strains is higher, along with the chance of developing life threatening side effects—such as certain types of colitis.”

However, Wiedemann is concerned about the accuracy rate in Cobb’s research—which was found to be 94% successful in distinguishing sepsis from non-infectious inflammation. If the infection was present, yet was not treated with antibiotics, the patient could be in grave danger. “I think that test discrimination might have to be higher than 94% to really influence the decision about whether to give antibiotics or not,” said Wiedemann. Nevertheless, the use of molecular diagnostics to produce rapid diagnoses in these areas would have a profound effect not only on the clinical laboratory, but on medicine in general, he added.

“We could look at gene expression rather than the need to culture an organism, which can take a long time, "he explained." This would really revolutionize clinical laboratories. It’s an exciting time because this could be a whole new way of doing things.”

Three of Cobb’s co-authors of the research published in the Journal of the American College of Surgeons are employees of Partek Incorporated.

Please help us improve Strategies by rating how useful this article is to your job.