January 2008: Volume 34, Number 1
The Quest for Faster Sepsis Diagnosis
Can New Markers Improve Survival Rates?
By Deborah Levenson
Sepsis can quickly turn severe—and fatal—when accompanied by organ dysfunction or failure. A range of clinical conditions caused by the body’s systemic response to an infection, sepsis strikes an estimated 750,000 people in the U.S. annually and kills 30–35% of them. While sepsis sometimes arises from an infection in an otherwise healthy person, all hospital patients are at risk. But identifying patients with sepsis is difficult because its clinical signs can mimic other conditions. And now, hospitals are seeing more cases of sepsis in part due to increasing numbers of immunosuppressed cancer and transplant patients. Early diagnosis is the key to saving more lives, but results from culture alone can take up to 48 hours. Consequently, some labs have started to offer faster tests for sepsis markers to speed diagnosis.
Currently, just three tests have been specifically cleared by the FDA for sepsis. The first, the Endotoxin Activity Assay (EAA), was cleared in 2003, and is marketed by Spectral Diagnostics (Toronto, Canada). The manual BRAHMS PCT LIA assay, developed by Brahms Diagnostics (Annapolis, Md.), was cleared in January 2005. The newest, the VIDAS BRAHMS PCT Assay, was cleared in October. Marketed by bioMérieux (Durham, N.C.), it uses technology developed and licensed by Brahms and runs in just 20 minutes on bioMérieux’s automated VIDAS platform. These and other tests are intended for use in conjunction with culture, but often are much more expensive than culture alone. Lab directors considering these and other sepsis marker tests must consider not only speed and cost, but also the test’s ability to aid diagnosis, determine whether patients will benefit from antimicrobial therapy, and monitor a patient’s response to interventions.
But the central question surrounding sepsis markers is their clinical utility. “Do we have markers that aid in good medical practice—diagnosis, prognosis, and monitoring? There is no individual marker that addresses any of these questions with 100% certainty,” explained Patrick St. Louis, PhD, Director, MDS Pharma Services Central Laboratory in Mississauga, Ontario. St. Louis recently gave an online presentation about sepsis on AACC’s Web site (See Box).
For More Information
- Patrick St. Louis’ online expert access presentation on sepsis is available on the AACC Web site.
- The Surviving Sepsis campaign offers background information about the syndrome, links to treatment guidelines, and information about educational opportunities at their Web site.
- The drug company Eli Lilly offers information about sepsis screening, diagnosis, and treatment at the following Web site.
Sepsis is not a specific disease, but rather a continuum of events triggered by the body’s inflammatory immune responses to a bacterial, viral, fungal, or parasitic infection. For many years, understanding of sepsis was limited in part by disagreement over its definition and lack of common terminology for the syndrome. In the early 1990s, a better definition emerged and clinicians started to approach sepsis as a continuum of clinical events.
In 1992, a consensus statement issued by the American College of Chest Physicians and the Society of Critical Care Medicine recognized progressive stages of sepsis, beginning with the initial systemic inflammatory response syndrome (SIRS) to infection. According to the consensus statement, severe sepsis occurs when SIRS compromises one or more vital organs and can lead to septic shock, which is marked by low blood pressure that does not respond to standard treatment, problems in vital organs, and oxygen deprivation. About half of patients who suffer septic shock die. In 2001, researchers at Henry Ford Hospital in Detroit, Michigan showed that early, goal-directed therapy can improve sepsis mortality rates (NEJM 2001, 345: 1368–1377).
PCT vs. CRP
CRP is a popular inflammation marker because it is easy and cheap at about $3 per result, according to St. Louis, while PCT is much more expensive, $25–$40. PCT and CRP are of roughly equal utility, according to St. Louis’s presentation. He pointed out that both markers are normally absent or present at very low levels in serum, have variable utility for diagnosis and prediction of severity and outcome, and have somewhat variable cutoffs.
But PCT has a slight diagnostic edge over CRP, in St. Louis’s opinion. As part of a research team that conducted a meta-analysis of studies examining PCT’s diagnostic value, he found PCT was better than CRP for differentiating bacterial from noninfective causes of inflammation, with PCT levels showing a sensitivity of 88% (95% CI, 80%–93%) versus 75% (95% CI, 62%–84%) for CRP. PCT’s specificity was 81% (95% CI, 67%–90%) versus 67% (95% CI, 56%–77%) for CRP. The sensitivity for differentiating bacterial from viral infections was also higher for PCT markers, 92% (95% CI, 86%–95%) versus 86% (95% CI, 65%–95%), while the molecules’ specificities were comparable at 73% (95% CI, 42%–91%) and 70% (95% CI, 19%–96%), respectively (Clinical Infectious Diseases 2004; 39:206–217).
Slides that accompanied St. Louis’ presentation point to a meta-analysis that examined 15 studies using both markers and concluded that the ROC curves for procalcitonin were better. In the 15 studies using both markers, the Q* value was 0.78 for PCT, versus 0.71 for CRP. Based on these findings, researchers concluded that PCT should be included in diagnostic guidelines for sepsis and in clinical practice in intensive care units (Critical Care Medicine 2006; 34:1996–2003).
In a more recent study of 82 patients with intraoperatively proven secondary peritonitis, European researchers from five sites concluded that “procalcitonin monitoring is a fast and reliable approach to assessing septic multiorgan dysfunction syndrome and overall prognosis in secondary peritonitis. This single-test marker improves stratification of patients who will develop clinically relevant complications” (Archives of Surgery 2007; 142:134–142).
PCT has yet to become a popular diagnostic tool for sepsis in the U.S., as it is in Europe, where it’s used routinely in clinical decision making and has been on the market for about 10 years, according to St. Louis. “I think it will take off with more North American studies and marketing. The reality is that if the company is selling in North America, it needs clinical data from North America.”
EAA, the other FDA-cleared assay, measures endotoxin, the highly toxic liposaccharide component of the wall of Gram-negative bacteria that can be a potent initiator of immune cascades leading to severe sepsis. Endotoxins can seep into the bloodstream from either infectious bacteria or a compromised gut, and the lipid portion of the molecule is identical among all Gram-negative species.
Spectral’s EAA was cleared by the FDA in 2003 and is indicated for risk-stratifying patients for impending sepsis. The test relies on the reaction of the lipid portion of endotoxin with a highly specific antibody in a test format that uses patients’ own neutrophils in a chemiluminescence platform, according to Debra Foster, Director of Spectral Diagnostics’ Sepsis Program. She recommends running the test on a SmartLine TL instrument (Berthold Detection Systems USA, Oak Ridge, Tenn.) “The test’s strength is that it identifies a trigger of inflammation, as opposed to a response factor such as CRP or PCT,” she noted. At a cost of $20 to $30 per result, EAA uses whole blood, can run on a STAT or urgent basis, and requires only a fraction (1.0 mL) of the blood that culture does. Foster pointed out that endotoxemia indicates more than Gram negative bacteremias. “Patients with pneumonia, peritonitis, and other severe infections may be endotoxemic and therefore at high risk for sepsis,” she explained.
However, no studies have directly compared EAA to other markers of sepsis, Foster admitted. But endotoxemia is common in the ICU, she noted, and research has identified it as an important risk factor for sepsis. Foster pointed to research that links high EAA levels with sepsis and mortality. In one study, 57.2% of 857 American, Canadian, and European patients admitted to ICUs had either intermediate or high endotoxin activity. For patients with low, intermediate, and high EA levels, rates of severe sepsis were 4.9%, 9.2%, and 13.2%, respectively, while ICU mortality in these groups was 10.9%, 13.2%, and 16.8%, respectively (Journal of Infectious Diseases 2004; 190: 527–534).
Tailoring Therapy with EAA
Recently published research suggests that EAA may one day be used to tailor therapy for patients with severe sepsis. A pilot study of 345 ICU patients by researchers at Spectral Diagnostics, University of Toronto, and University of Ottawa found that variable EAA activity occurred independently of infection status and was a marker of increased severity of illness (Shock 2007, 28: 524–529).
Meanwhile, the Japanese healthcare system has embraced a sepsis therapy that relies on EAA results, Foster noted. Once sepsis is confirmed using the assay, patients undergo a dialysis process that strips endotoxin from the blood. A Japanese company, Eisai (Ridgefield Park, N.J.) is also integrating EAA into drug development .
Integrating Sepsis Biomarkers and Therapy
An assay specifically cleared by the FDA to diagnose sepsis may someday be used to personalize therapy. In an ongoing, phase III clinical trial of eritoran (E5564 ), a drug for severe sepsis, Eisai Co. (Ridgefield Park, N.J.) is now using Spectral Diagnostics’ EAA at 250 sites in the U.S. and several other countries. E5564 is a second-generation synthetic lipodisaccharide that blocks the ability of Gram-negative bacteria to stimulate human cytokine production in whole blood.
Currently only one drug, Xigris, is FDA-approved specifically for sepsis therapy. Marketed by Eli Lilly (Indianapolis, Ind.), it’s the focus of another study, RESPOND, which in November 2006 began tailoring dose and duration of Xigris therapy based on serial levels of protein C, which is involved in inflammation, systemic sepsis, and the concomitant intravascular coagulopathy. According to the company, the study involves 500 patients at 50 sites in 11 countries. The study is part of an effort with Biosite, now being acquired by Inverness, to develop a point-of-care test that could help adjust Xigris treatment based on patients’ protein C levels.
However, some controversy surrounds Xigris. Approved in 2001 for patients with severe sepsis and high risk of death based on data that showed that Xigris significantly reduced mortality, a subsequent trial ended early because it demonstrated little beneficial treatment effect and increased incidence of serious bleeding. (NEJM 2005; 353:1332–1341). The European Union’s Committee for Medicinal Products for Human Use, an FDA equivalent, has asked for more data for Xigris’s license renewal process. Beginning in the first quarter of 2008, Eli Lilly will launch a 2.5- year trial to fulfill that request.
However, EAA has its detractors, too. “Endotoxin is difficult to measure, so it hasn’t caught on in clinical labs. Traditionally Gram-negative bacteria has been seen as a primary culprit behind sepsis and these organisms produce endotoxin,” explained James Versalovic, MD, PhD, Director of the Division of Molecular Pathology and the Microbiology Laboratories at Texas Children’s Hospital and Associate Professor of Pathology at Baylor College of Medicine in Houston. He is using arrays to design a test that identifies multiple etiologic agents of sepsis. “Plus there are organisms, like Gram-positive bacteria and fungi, that can cause sepsis but may not produce endotoxins at all. So EAA may be used to assess gut translocation and accounts for only part of the story,” Versalovic asserted.
Working Toward a Multi-Marker Assay
At least one diagnostic manufacturer is combining markers for sepsis diagnosis. Biosite (San Diego, Calif.), now being acquired by Inverness Medical Innovations (Waltham, Mass.), is developing a multi-marker panel. A team at Henry Ford Hospital in Detroit, led by Emanuel P. Rivers, MD, Director of Research in the Department of Emergency Medicine, helped develop the panel, which includes neutrophil gelatinase-associated lipocalin (NGAL), CRP, and macrophage inflammatory protein-3 (MIP-3). Rivers, who discussed the panel during a Biosite-sponsored event at the AACC’s 2007 Annual Meeting last July, said the panel is intended as a point-of-care test to assess risk of sepsis progression within 72 hours of emergency department presentation. The multi-marker assay is the result of an extensive Biosite screening program that evaluated about 150 possible markers for both their ability to detect sepsis and risk stratify patients. The company has launched a prospective multicenter study involving 1,000 patients at 10 U.S. centers to validate the clinical utility of the data and compile it for submission to the FDA, according to Rivers. He noted, however, that panel is now undergoing refinement and the markers could change.
NGAL, one of the panel’s potential markers, is also being used as an experimental marker of acute kidney injury, the most common cause of sepsis in the ICU. Right now, acute kidney injury is usually a delayed diagnosis based on serum creatinine levels, which can vary considerably among individuals and require many days to reach a steady state, according to Prasad Devarajan, MD, Professor of Pediatrics and Director of Nephrology at Cincinnati Children’s Hospital, who also spoke at Biosite’s event. His institution holds an exclusive license for NGAL to detect acute kidney injury and is developing a urine test for Abbott Diagnostics (Abbott Park, Ill.) and a plasma version for Biosite. Both assays detect the rise of NGAL within 2 to 6 hours of an event that leads to acute kidney injury.
No Clear Winner
With no one sepsis marker clearly recognized as the preferred marker, whatever emerges as a leading tool has the potential to alleviate the huge burden sepsis creates for the healthcare system. Spectral’s Foster likened the current landscape of sepsis diagnosis to the state of cardiac markers prior to troponin. “Right now, sepsis care is in the same zone that cardiac care was before troponin became the leading test in diagnosis of acute myocardial infarction. Patients are dying at high rates. There are many markers, but no one test is dominant and so far, uptake in the U.S. has been slow. These markers need to be used and challenged to determine the best fit into the differential diagnosis of sepsis,” she explained. Those in the sepsis field are hopeful that research will yield tools that better diagnose the condition, guide therapy, and save more lives.
Addendum: Some Observations on Procalcitonin and Sepsis Testing
Congratulations on a well-written and informative article on “The Quest for Faster Sepsis Diagnosis” (CLN Jan. 2008). At ARUP, we are setting up the Brahms’ procalcitonin test in our hospital stat lab to serve the needs of the Salt Lake City. There is considerable interest among some local physicians in the test. We are also in line to get one of the first Kryptors in the U.S.
I would like to share some additional information with CLN readers. In the late 1970's, I developed one of the first radioimmunoassays for calcitonin, and we offered the test through the Laboratory Procedures division of the Upjohn Company in our Kalamazoo, Michigan laboratory. At the time, I was contacted by Dr. Russell Chesney, an endocrinologist at the University of Wisconsin who said that ER doctors had contacted him because many patients with toxic shock syndrome they were treating had been found to have hypocalcemia. The ER docs didn't know why, so they were asking their endocrinologists. Chesney sent me some specimens, and they all had very significantly elevated calcitonin levels. He also involved two other prominent endocrinologists who had been approached by their ER docs with the same puzzle. My colleague, Francis P. DiBella, PhD, and I not only tested all of these specimens, but then characterized the calcitonin immunoreactivity in these patient specimens by gel filtration and discovered that the immunoreactivity was of a larger molecular size. In the publication of our findings, we speculated that the calcitonin in the serum of toxic shock patients was of a larger size, perhaps a polymeric form of calcitonin (J Lab Clin Med, 1983: 101(4), 576-585). To put this in context, one has to realize that the concepts of prohormones and pre-prohormones were just barely being developed 25 years ago.
As I anticipate that interest in procalcitonin as a marker of sepsis is likely to grow among clinicians, I would also like to refer CLN readers to an excellent 2004 review on the subject in the Journal of Clinical Endocrinology & Metabolism (89(4):1512-1525). Procalcitonin testing may represent another opportunity to improve healthcare through laboratory medicine.
Charles D. Hawker, PhD, MBA, FACB
Scientific Director, Automation and Special Projects
ARUP Laboratories, Salt Lake City, Utah