Assessing the extent of ischemic brain injury in cardiac arrest patients in a coma is a complex clinical challenge, particularly if they have been treated with therapeutic hypothermia. During hypothermia, patients are mechanically ventilated, sedated, and given muscle relaxants to avoid shivering; as a consequence of hypothermia, the metabolism of drugs is prolonged. A clinical neurological examination can therefore become difficult to interpret and may be less reliable.
Neuron-specific enolase (NSE), an enzyme associated with brain cell death, was recommended as a prognostic marker after cardiac arrest in the 2006 AAN practice parameters. But the recommendation was not extended to patients treated with therapeutic hypothermia and the usefulness of measuring NSE in those cases has been controversial.
Now, a paper in the July 20 online edition of Neurology shows that measurement of NSE correlates well with other markers of ischemic brain injury and may be used to help predict survival in patients treated with hypothermia.
“Our study shows for the first time how well NSE correlates to other measures of the burden of ischemic cell death in the brain after cardiac arrest,” study author Tobias Cronberg, MD, PhD, associate professor of neurology at Lund University in Sweden, told Neurology Today. “Similar to troponin after cardiac infarction, NSE actually seems to quantify the extent of brain damage.
“If one combines NSE and continuous EEG-monitoring, the two add up to a powerful prognostic instrument,” Dr. Cronberg said. “While I would not recommend withdrawal of care based on these two methods alone, they can predict which patients have good chances of waking up with or without only minor neurological handicap and they can identify patients who probably have a very poor prognosis that can be confirmed by other measurements such as somatosensory evoked potentials.”
The investigators based their recommendations on a detailed analysis of 34 hypothermia-treated cardiac arrest patients who had been in a coma for four to five days after cardiac arrest. They monitored them with continuous amplitude-integrated EEG; collected samples of NSE; performed brain MRI and somatosensory evoked potentials (SSEP). Finally, they did a post-mortem brain exam in six patients who died.
Six of the 17 patients with peak NSE-levels less than 33μg/L at 48 hours regained the capacity to obey verbal commands. But 17 patients with peak NSE-levels exceeding 33 μg /L did not fare well: all 17 patients failed to recover consciousness; 10 showed extensive brain injury on MRI, and 12 lacked cortical responses on SSEP. All six patients who underwent autopsy had extensive severe histologic damage.
“NSE is a useful method to quantify the burden of ischemic brain damage after cardiac arrest if one takes methodological reservations into account,” Dr. Cronberg said. “This means that the level of NSE should be viewed in conjunction with other methods such as clinical examination, EEG, SSEP and neuroimaging. Today we can get NSE measurements on a 24-7 service and thereby have early prognostic information which could aid in the clinical decision making and provision of information to relatives.”
Dr. Cronberg and colleagues also identified a small group of patients who remained in coma with a very poor response to painful stimuli but who had low NSE, normal MRI and SSEP, and no pronounced brain damage post-mortem. All of these patients were found, however, to have had an electrographic postanoxic status epilepticus, a condition with a very poor prognosis. “It is tempting to speculate that this is a condition that may be reversible with the right treatment but that clearly needs to be studied further,” he said.
Despite the encouraging results, experts in the field who reviewed the report for Neurology Today continue to believe prognosis for hypothermia-treated patients will remain a clinical conundrum until more tests with greater specificity and sensitivity are developed.
“I think the jury is still out on the usefulness of NSE in this population,” said Bryan Young, MD, professor of neurology and critical care medicine at the University of Western Ontario in Canada.
In an interview with Neurology Today, Dr. Young said he is aware of several studies still in press, including a large multicenter Dutch study, which find that NSE measurements vary considerably so that the threshold for sensitivity must be adjusted rendering its use questionable. “NSE is not useless, but I think we need to wait until further studies are published before strongly endorsing this,” he said.
Dr. Cronberg and colleagues noted that there has been variability in the literature regarding results on “cut-off levels for a poor neurological outcome.” They pointed out there are probably standardization issues that need to be solved before NSE can be generally recommended as a prognostic marker.
But they added “a normal or modestly elevated NSE should alert the attending physician to a potentially treatable cause of prolonged coma after cardiac arrest.”
Dr. Young urged clinicians to have at least two tests that are concurrent for a poor outcome before making a prognosis. “If we can use biomarkers, that's good but I think the motivation is still get more and more sensitive and specific tests,” he said.
He added that indicators of favorable outcome should be used as well and noted that EEG reactivity is an under-recognized test. “We should not only be focused on indicators of poor outcome,” he said.
Neurocritical care expert Andrea Rossetti, MD, senior tenured lecturer in the Department of Neurology at the University Hospital and Faculty of Biology and Medicine in Lausanne, Switzerland, agreed with that sentiment. He cited the relatively small sample size of the study and the fact that peak times for measuring NSE levels may vary among patients. The relatively short biological half-life of the NSE biomarker may be an issue of particular concern with patients having NSE peaks above a given threshold before 48 hours, but who may partially “recover” lower levels afterwards, he said.
“The study appears important from a scientific point of view, as it strongly suggests that NSE represents a valuable biological marker in this clinical setting,” Dr. Rossetti told Neurology Today in an e-mail. “However, from a clinical point of view, before application of NSE as an independent predictor after therapeutic hypothermia and considering several independent reports disclosing the risk of false-predictions in this setting, additional, larger studied are needed. In the context of a decision about life or death, the currently most reasonable option continues to be to apply multimodal assessments and avoid concentrating on a single predictor.”
Cronberg T, Rundgren M, Friberg H, et al. Neuron-specific enolase correlates with other prognostic markers after cardiac arrest. Neurology 2011; E-pub 2011 Jul 20.
Mayer SA. Editorial: Outcome prediction after cardiac arrest. New game, new rules. Neurology 2011: E-pub 2011 Jul 20.
Wijdicks EF, Hijdra A, Young GB, Bassetti CL, Wiebe S. Practice parameter: prediction of outcome in comatose survivors after cardiopulmonary resuscitation (an evidence-based review): report of the Quality Standards Subcommittee of the American Academy of Neurology. Neurology