Four experts led a standing-room only session Tuesday morning on the future of identifying and treating sepsis. The session, Sepsis 2020: What is Down the Pike, reviewed recent and ongoing research on biomarkers, big data, vitamin C, and fluids.
Biomarkers in Sepsis
Moderator Steven Q. Simpson, MD, FCCP, professor of medicine in the division of pulmonary, critical care, and sleep medicine at the University of Kansas, kicked off the session with a review of sepsis biomarkers—both novel and familiar.
While the word biomarker typically brings to mind new technologies, Dr. Simpson, CHEST President-elect, reminded the audience that many things commonly used in sepsis are, in fact, biomarkers. Everything from white blood counts and platelet counts, to lactate and procalcitonin, fit the definition, he said.
However, Dr. Simpson focused much of his brief talk on reviewing data related to biomarkers recently approved by the US Food and Drug Administration (FDA)—a gene expression assay that measures four genes associated with infection and monocyte distribution width.
“The monocyte distribution width works on this principle,” he said. “Monocytes are a first line of defense against infection. When exposed, they crank up to begin releasing cytokines. As they crank up, they swell.”
He showed data comparing the size of monocytes in septic patients and nonseptic patients. In the septic individual, the monocytes have a longer tail and larger size.
“The advantage of the monocyte distribution width is that once your hospital puts it in your software, it will come with every single CBC that is sent,” Dr. Simpson said. “So it will behave as a sepsis early warning sign.”
Big Data: The Future of Sepsis Recognition
Clinical information expert Matthew M. Churpek, MD, MPH, PhD, an associate professor at the University of Wisconsin, studies predication models to detect early clinical deterioration. In his work, he uses electronic health record data and machine learning to identify patients at risk for clinical deterioration and sepsis.
“If you take a step back and think about it, sepsis is a combination of patients who are critically ill and patients who are infected,” he said. “From a machine learning perspective or a modeling perspective, it’s really two different prediction problems. When we think about critical illness, we think about things like dropping blood pressure and high respiratory rate. With predicting infection, we think about white count and temperature.”
During his talk, Dr. Churpek briefly reviewed traditional methods for predicating sepsis, including multiple scoring methods and clinical judgement, and asked the question: Can we do better with machine learning?
He reviewed a study of 1,600 patients admitted to the University of Chicago that showed a random forest algorithm had a much higher accuracy for predicting infection compared to Systematic Inflammatory Response Syndrome and International Prognostic Score (IPS). His team has now reviewed more than 4,000 charts.
“The final question is ‘Do these algorithms actually improve care?’” Dr. Churpek said. “The data suggests real-time alerts do increase the prescription of antibiotics. That’s pretty consistent. However, the results on the impact on outcomes like mortality have been quite mixed. In those studies where you have the highest mortality in the control group—in other words, those caring for the sickest patients—you actually see quite a difference in mortality after the intervention. So I think there is a potential here for automating care using big data for early identification and potential treatment recommendations.”
Vitamin C, Thiamine and Hydrocortisone: Ready for Prime Time?
Jon Sevransky, MD, MHS, FCCP, a principal investigator on the VICTAS (Vitamin C, Thiamine, and Steroids in Sepsis) trial, provided an overview of what is known about this combination that has critical care clinicians optimistic.
Dr. Sevransky, professor of medicine at Emory University, said he believes one of the reasons people have been so captivated by the positive study of vitamin C in sepsis is because it was the first new intervention in decades to show a significant reduction in mortality.
He began his talk with a review of the biological rational for the use of vitamin C, thiamine, and steroids, and then provided an overview of the VICTAS trial, which aims to provide new evidence regarding their efficacy.
The participants are adult patients with confirmed or suspected infection and evidence of respiratory or cardiovascular organ dysfunction. They receive triple therapy every 6 hours. In the comparison group, patients receive placebo unless the clinical team thought the patient should receive steroids, Dr. Sevransky said.
The primary outcome measure was days alive off of vasopressors and ventilation. Data was set at 0 if the patient died before day 30. He said he’s hopeful results will be available within the next few months. In the meantime, he recognized clinicians may already be using Vitamin C in some situations.
“If you decided that vitamin C is right for a patient, you need to pay careful attention to how you measure glucose,” Dr. Sevransky said. “Because of its chemical makeup, vitamin C can cause higher glucose readings. If you’re doing a Hail Mary, please make sure that you measure glucose with something that uses whole blood, like a blood gas, because otherwise, you might get an inaccurate result.”
Early Fluids or Early Vasopressors
Finally, Todd Rice, MD, MSc, a member of the protocol committee for the CLOVERS (Crystalloid Liberal or Vasopressors Early Resuscitation in Sepsis) study, discussed the conflicting evidence on fluids and vasopressors.
“If you’ve been in the sepsis world long enough—even 5 years—you know this pendulum swings back and forth,” said Dr. Rice, an associate professor of medicine in allergy, pulmonary, and critical care medicine at Vanderbilt University. “It swings from too little fluid to too much fluid to too little fluid. I think many of us wonder where that sweet spot is. I’m not sure we know where that sweet spot is right now. Each, as you can imagine, has its advantages.”
The aim with the CLOVERS trial is to try to answer that question by avoiding issues that have confounded previous studies.
Dr. Rice reviewed the study design and methodology. CLOVERS is a multi-center, prospective, phase 3 randomized, nonblinded, interventional trial of fluid treatment strategies in the first 24 hours for patients with sepsis-induced hypotension. The study has two treatment groups.
“The protocol for the ‘restrictive’ group calls for vasopressors first to try to minimize or restrict the amount of fluids given,” Dr. Rice said. “It’s truly a fluids first, vasopressors as a rescue later or a vasopressors first, fluids as a rescue later kind of comparison. We’re about halfway through enrolling patients. The goal is to finish early in 2021 and have an answer to this question.”