ARTICLE IN BRIEF

Researchers found that serum tau was a more accurate prognostication tool for neurologic recovery after cardiac arrest than the biochemical marker neuron-specific enolase.

Measuring levels of tau in the blood may help predict whether a person who is left unconscious from cardiac arrest will have a poor neurological outcome, information that could help doctors and families decide whether to continue with intensive-care treatment or withdraw life support.

A new study from Sweden found that serum tau, an indicator of axonal injury, was a more accurate prognostication tool than the biochemical marker neuron-specific enolase (NSE), which is already in clinical use but is known to have some limitations.

The study, published in the October 5 online edition of Annals of Neurology, found that increased serum tau was associated with poor neurological outcomes at six months after cardiac arrest and was also a predictor of short survival.

Currently, the findings have no clinical applicability since the bioassay used in the study to measure serum tau is considered a research tool and is only available at specialized laboratories.

“Cardiac arrest accounts for a high proportion of mechanically ventilated patients admitted to intensive care units (ICU),” the study authors pointed out. “The length of ICU stay is increasing, but rates of mortality and neurological morbidity due to anoxic brain injury remain very high. There is a need for early and accurate prognostication methods, both to avoid prolonged treatment of patients where continued life-supporting measures are futile, and to ensure that patients with potential for recovery receive optimal management.”

“Prognostication of neurological outcomes after cardiac arrest is difficult, and new tools are needed to increase the prognostic accuracy,” said study coauthor Niklas Mattsson, MD, PhD, associate professor in the department of clinical sciences at Lund University in Sweden.

“Tau is released from injured and dying neurons. We therefore believe that the serum tau concentrations reflect the overall extent of the neuronal injury,” Dr. Mattsson said in an email to Neurology Today. “These results are important since they show that serum tau (and potentially other specific and sensitive brain-derived biomarkers) may provide unprecedented accuracy for prognostication after cardiac arrest.”

The study was supported by grants from the Swedish Research Council, the Swedish Heart and Lung Foundation, and other medical and science organizations and associations in Sweden. The authors reported no conflicts of interest.

MULTIPLE ASSESSMENTS USED

Clinicians typically use neurological examination, electroencephalography, short-latency somatosensory evoked potentials (SSEP), neuroimaging, automated pupillometry, and biochemical markers, among other tools, for determining the neurologic prognosis after cardiac arrest.

The paper noted that while NSE is the most widely studied biomarker for cardiac arrest and its use is recommended in guidelines for assessing patients, “NSE is susceptible to false positive values due to haemolysis.”

STUDY METHODS, FINDINGS

The assessment of tau as a biomarker tau was done as part of the Target Temperature Management-trial (TTMtrial), a large, prospective study on cardiac arrest. The study was designed to compare targeted temperature management at 33 degrees C versus 36 degrees C in patients who were unconscious after out-of-hospital cardiac arrest. Neurological prognostication and criteria for withdrawal of life-sustaining therapy were assessed as part of the study, which enrolled 819 patients at 29 European medical centers between November 2010 and January 2013. Treating physicians and other staff were blinded to the tau and NSE measurements collected during the study, which helped limit the self-fulfilling prophecy effect.

Tau was measured using a research grade kit from Quanterix, a Massachusetts company. Neurological outcome at six months was assessed face-to-face by the cerebral performance categories (CPC) scale, which ranges from 1 (no or minor cerebral disability) to 5 (brain death), and the modified Rankin scale (mRS), which ranges from 0 (no symptoms) to 6 (death). The main endpoint was “poor neurological outcome,” defined as CPC 3-5, compared with “good neurological outcome,” defined as CPC 1-2. All-cause mortality was assessed after a minimum follow-up of six months. A total of 689 patients with at least one tau measurement were included in the analysis.

“Poor outcome was associated with higher tau” at six months after cardiac arrest, the researchers reported. The median tau levels for those who had a poor outcome was 38.5 ng/L versus 1.5 ng/L for those with a good outcome at 72 hours (p<0.0001).

The researchers said the association between poor outcome and high tau levels was already present at 24 hours with median tau levels at 12.0 ng/L for poor outcome versus 2.4 ng/L for good outcomes (p<0.0001) and was even greater at 48 hours with median tau levels of 49.5 ng/L for poor outcomes and 1.9 ng/L for good outcomes (p<0.0001). There were no real differences in tau trends between the group maintained at a temperature of 36 degrees C versus those kept at 32 degrees C. There was also an association between high tau levels and poor scores on the mRS.

“Tau improved prediction of poor outcome compared to using clinical information, and was more accurate than serum NSE,” the researchers concluded. “Tau cut-offs had low false positive rates for good outcome while retaining high sensitivity for poor outcome.”

Dr. Mattsson, who is a specialist in clinical chemistry and resident in neurology, said tau, if further proven to be accurate, would be used in conjunction with other evaluations.

“Assessment after cardiac arrest remains multifaceted, including clinical testing, neuroimaging and neurophysiology, and biochemistry,” he said. “The acceptance for errors in prognostication after cardiac arrest is very low, since the prognostication is used for making decisions about discontinuation of intensive care.”

He added that his group is continuing to investigate tau and does not currently use the test in clinical care.

EXPERT COMMENTARY

Ariane Lewis, MD, assistant professor of neurology and neurosurgery at NYU Langone Medical Center, said “it would be great if we had this biomarker as part of our arsenal of tools to evaluate patients after cardiac arrest...to help us give families the most accurate prognostic information possible.”

“Prognostication after cardiac arrest is challenging,” she said, cautioning that one single test, whether tau or something else, could never be the lone determinant of whether to continue or discontinue care for a comatose person.

“To assess the relationship between tau and outcome, it is necessary for clinicians to be blinded to the tau level,” Dr. Lewis said. “Otherwise, if the tau level is bad, they may be more inclined to withdraw care and that would be a self-fulfilling prophecy.”

David M. Greer, MD, FAAN, professor and chair of neurology at Boston University, said he thought the report from Sweden was “a fantastic study” based on a “very robust data set from a well-conducted clinical trial.”

“Where there is severe global brain injury you would love to have a chemical biomarker to tell you how much injury has occurred and you want that biomarker to be specific to the brain,” he said.

Dr. Greer said a shortcoming of NSE as a biomarker for brain injury is that it is also found in red blood cells and platelets, which are often broken down in cardiac arrest patients, leaking NSE and making measurements somewhat questionable. Tau is also being evaluated as a biomarker for some other neurological conditions, including Alzheimer's and traumatic brain injury.

Dr. Greer cautioned, however, that tau, or any other potential biomarkers, should not be seen as holding all the answers when it comes to making decisions about comatose patients in the aftermath of cardiac arrest.

“It truly is a matter of life and death in this situation and you want to make sure a test does not lead to a self-fulfilling prophecy,” he said. “The more data you have going in a consistent direction, the more confident you can be when talking to a patient's family.”

Eelco F. Wijdicks, MD, PhD, FAAN, professor of neurology and chair of the division of critical care neurology at Mayo Clinic in Rochester, MN, said there has been a fascination for a long time with finding a means to do “early prognostication for comatose survivors after cardiac arrest.”

He said the new study, while well done, has “zero clinical application” at the moment because “the test is not widely available and certainly not on a timely basis, which would be necessary to be helpful.”

He said it was noteworthy that the study found that serum tau was a better predictor of poor neurological outcome than NSE, a test he said he rarely uses in assessments and decision-making.

He said among the questions that need to be answered is whether tau is “truly an independent variable or a co-variable” in predicting neurological outcome after cardiac arrest.

“It could very well be that serum tau is elevated in the patients we already know are going to do poorly,” said Dr. Wijdicks, who is collaborating with the Swedish team on an upcoming study (TTM2).

He said his experience in the ICU indicates that in about 20 percent of comatose patients after cardiac arrest, it is possible to reliably predict whether a patient will have a poor recovery. With the other 80 percent of patients, at least half have life support withdrawn at the family's request after days pass and no improvement is seen or if there is a major cardiac problem. In the remaining patients, “the outcome is simply unclear and we do not know what will happen with certainty so soon after the event.”

Dr. Wijdicks said a comprehensive neurologic examination, which includes a serial unconfounded neurologic evaluation, preferably by a neurologist or a neurointensivist, remains the most important tool for prognostication.

“Very few tests can improve or question clinical diagnosis,” he said. “I don't think we will have some sort of magic serum test that will tell us.”

EXPERTS: ON SERUM TAU FOR PREDICTING NEUROLOGY OUTCOME AFTER CARDIAC ARREST