Abstract
Mr. Jones is a 65-year-old man with a past medical history of coronary artery disease who is admitted after an unwitnessed pulseless electrical activity arrest at home. He achieved return of spontaneous circulation (ROSC) in the emergency department but now remains unconscious. Upon transfer to the intensive care unit, he is intubated, sedated, and undergoing targeted temperature management (TTM) targeting 36˚C. Over the next few days, sedation is discontinued as he completes TTM and controlled rewarming. Twenty-four hours after restoration of normothermia, he remains unconscious. His clinical examination is unchanged from prior: preserved pupillary and corneal reflexes bilaterally and a reflexive flexor response to painful stimuli. Continuous electroencephalography (EEG) reveals no seizures, preserved continuity, and unclear background reactivity to stimulation (Figure 10.1). Brain magnetic resonance imaging (MRI), obtained on day 4, is notable for restricted diffusion in the bilateral primary sensory cortices (Figure 10.2). Mr. Jones’s family requests a meeting to discuss his chances of achieving a “meaningful recovery.”
Mr. Jones is a 65-year-old man with a past medical history of coronary artery disease who is admitted after an unwitnessed pulseless electrical activity arrest at home. He achieved return of spontaneous circulation (ROSC) in the emergency department but now remains unconscious. Upon transfer to the intensive care unit, he is intubated, sedated, and undergoing targeted temperature management (TTM) targeting 36˚C. Over the next few days, sedation is discontinued as he completes TTM and controlled rewarming. Twenty-four hours after restoration of normothermia, he remains unconscious. His clinical examination is unchanged from prior: preserved pupillary and corneal reflexes bilaterally and a reflexive flexor response to painful stimuli. Continuous electroencephalography (EEG) reveals no seizures, preserved continuity, and unclear background reactivity to stimulation (Figure 10.1). Brain magnetic resonance imaging (MRI), obtained on day 4, is notable for restricted diffusion in the bilateral primary sensory cortices (Figure 10.2). Mr. Jones’s family requests a meeting to discuss his chances of achieving a “meaningful recovery.”
Figure 10.2 Brain MRI: hypoxic–ischemic changes in the bilateral primary sensory cortices seen on diffusion-weighted imaging (A–C; white arrows), apparent diffusion coefficient sequences (D–F; black arrows), and fluid-attenuated inversion recovery sequences (G–I; black arrowheads). Note the T2 hyperintensities in I (black circle) without associated diffusion restriction, representing findings likely unrelated to the hypoxic–ischemic injury.
Despite numerous advances in emergency and post-arrest care, the overall rates of survival after cardiac arrest remain low.1 A small proportion of patients recover immediately and suffer little to no neurologic dysfunction, whereas others remain unconscious for days to weeks after the restoration of forward flow. The vast majority of patients have an unclear prognosis: they lack many of the highly predictive poor outcome markers, yet they remain in an unconscious state with a likely protracted course of recovery. Thus, families and providers face a critical decision: to continue aggressive care in pursuit of a “meaningful” neurologic recovery – with the inherent risk of achieving only a profoundly disabled or vegetative state at best – or to withdraw life support. In the face of such uncertainty, neurologists and critical care specialists play a vital role in providing guidance to families. Communicating an accurate prognosis is paramount, as decisions of continued care versus withdrawal of life-sustaining therapy (WLST) may hinge on provider–family discourse.
The goals of care discussions after cardiac arrest should contain several key elements: the estimated severity of primary and secondary brain insults, an assessment of individual cerebral resilience and the potential for recovery, and a frank conversation delineating the levels of disability and quality of life that are acceptable to individuals. In this Chapter, we discuss the usefulness of various tools used in neuroprognostication after cardiac arrest, as well as other factors that may inform shared decision-making.
10.1 Severity of Hypoxic–Ischemic Injury
Neuroprognostication begins with the collection of objective data to quantify the overall burden of hypoxic–ischemic brain injury. The etiology of the arrest, initial nonperfusing rhythm, performance of bystander cardiopulmonary resuscitation and prehospital care, and total downtime before ROSC all correlate with the severity of the initial injury from cardiac arrest.2, 3 After ROSC is achieved, other early markers of injury that reflect the postcardiac arrest syndrome include the serum lactate level4 and its clearance rate,5 hemodynamic instability and degree of vasopressor requirement,6 markers of extracerebral organ damage (which commonly include myocardial dysfunction, shock liver, and kidney injury),7 and initial EEG findings.8
Additionally, hypoxia and acidemia, hyperthermia, seizures, and rearrest are not infrequent. These factors set the stage for ongoing and secondary injury, especially because impaired cerebral autoregulation after ROSC leads to a mismatch between cerebral energetic demand and supply.9, 10 These additional insults can exacerbate the total damage suffered, thus challenging further the individual’s physiologic resilience – currently a theoretical concept, because there are no quantifiable measures validated in practice.
10.2 Cerebral Resilience
Resilience refers to positive adaptive mechanisms after exposure to trauma or hardship. Although rooted in psychology, this concept also has relevance to neurology. In addition to the severity of injury after cardiac arrest, the patient’s background characteristics also inform recovery after hypoxic–ischemic brain injury; these characteristics include age, premorbid functional state, comorbidities, prior brain injuries, and past experiences with rehabilitation. These factors interact to produce an individualized trajectory of recovery, even before the initiation of post-arrest care. Neuroimaging, obtained at any point, may help to quantify the degree of cerebral injury accrued thus far over the life course; microvascular changes, old insults, and the degree of atrophy can be visualized as surrogate markers of cerebral plasticity. Although difficult to quantify clinically, greater baseline pathology almost certainly attenuates recovery potential.
Similarly, early EEG findings and changes over time reflect both resilience and the severity of hypoxic–ischemic injury. Pathologic patterns after cardiac arrest – in particular, burst suppression in the absence of anesthetics – arise from ischemia-mediated disruptions in neurotransmission and subsequent synaptic failure.11 Whether or not synaptic failure is reversible may depend on the degree of baseline neuronal pathology as well as the degree of acute injury. The resolution of synaptic failure and return to a continuous pattern within the first 12 hours suggests a greater recovery potential, whereas a lack of improvement within 24 hours likely represents a poorer capacity for recovery.12
10.3 TTM
Beginning in the early 2000s, therapeutic hypothermia after cardiac arrest quickly became standard practice, based on the results of two independent clinical trials that demonstrated improved survival and favorable outcomes in patients randomized to hypothermia, compared to those assigned to no temperature management.13, 14 Notably, neuroprognostication was not standardized in either study, and the vast majority of patients with poor outcome died during hospitalization, mainly in the setting of WLST.
After the implementation of therapeutic hypothermia, the TTM trial of 2013 randomized 939 patients with out-of-hospital arrests to a TTM goal of either 36˚C or 33˚C and, after a mean follow-up duration of 8.5 months, found no significant difference in all-cause mortality or favorable outcome between these groups.15 The TTM trial raised the standards of cardiac arrest research by implementing a more rigorous and transparent methodology of neuroprognostication: all sites followed a standardized, multimodal approach.
In 2019, the HYPERION trial – the first large trial comparing hypothermia (32.5˚C–33.5˚C) with controlled normothermia (36.5˚C–37.5˚C) in nonperfusing rhythms – demonstrated a shift toward favorable neurologic survival with lower temperatures.16 Following the example set by the TTM trial, a multimodal neuroprognostic algorithm was employed, and data on decision-making processes were collected.
Although the exact mechanism is not fully understood, the neuroprotective effect of TTM is likely driven by decreased cerebral metabolic activity, as well as decreased free radical generation and reperfusion injury; however, it remains unclear whether simply avoiding hyperthermia is the key component. Regardless, the modern era of TTM has redefined the post-cardiac arrest landscape. Because TTM may slow neuronal recovery, the natural course of global recovery is likely prolonged, and delayed clearance of confounding medications, including sedatives, may further postpone improvement. Thus, our understanding of the time course of neurologic recovery has been reset.
10.4 Outcome Measures
The Cerebral Performance Category (CPC) scale is the most widely used outcome measure after cardiac arrest. The scale (Table 10.1) ranges from 1 to 5, wherein 1 represents a full recovery and 5 represents death. A score of 3 constitutes a state of severe neurologic disability requiring daily support despite regaining of consciousness17; although usually considered an unfavorable outcome, some studies do report CPC 3 as a good outcome because it reflects “awakening.” Despite being widely used, the CPC scale lacks a standardized tool with which to assign scores.
Table 10.1. CPC Scale
The cardiac arrest literature is highly heterogeneous, and the most commonly used time points for outcomes diverge from those of other acute brain injuries, which frequently use the modified Rankin scale at 90 days as the primary outcome (see Chapter 11, “Decompressive Craniectomy for Stroke Patients”). In contrast, post-cardiac arrest neuroprognostic studies tend to employ early outcome measures such as survival to discharge or discharge CPC score, whereas therapeutic trials favor assessing mortality and neurologic outcome at 6 months or less frequently at 12 months. Understanding the timing of outcomes in these studies and applying this information to clinical practice are key aspects of neuroprognostication.
10.5 Prognostic Tools
The neurologic examination is the cornerstone for assessing prognosis after cardiac arrest. Motor response, once considered a reliable predictor, has recently been shown to carry an unacceptably high false-positive rate, regardless of TTM treatment and timing of assessment.18, 19 In contrast, absent pupillary and corneal reflexes remain robust predictors of poor outcome at 72 hours after arrest or after rewarming in TTM-treated patients. The use of a pupillometer improves inter-rater reliability,20 and a meticulous technique with escalating noxious stimuli (from saline squirt to tactile pressure using a cotton swab applicator at the limbus) should be used to confirm absent corneal reflexes.21
An EEG – and whenever possible, continuous monitoring or serial short studies to capture evolution of background activity – remains another mainstay of care after cardiac arrest. Although postanoxic seizures and status epilepticus have historically been considered pathognomonic of a poor outcome, treating these conditions promptly may be beneficial,22 and up to 50% of patients with post-arrest myoclonic status epilepticus may achieve a functional recovery.23 However, these findings do influence the pace of recovery, and a delayed awakening is to be expected. Unreactive EEG background and persistent burst suppression are also considered malignant findings that may foreshadow a poor recovery. Nonetheless, there is significant variability in assessing reactivity, and the presence of confounders (including sedation and temperature) affects the reliability of these predictors in clinical practice.
Somatosensory evoked potentials (SSEPs) are considered one of the more reliable prognostic tools, wherein absence of N20 waves at 72 hours is considered a strong indicator of a poor outcome.24 However, these results may be influenced by temperature and sedation and thus should be considered in conjunction with other tests.25
Neither a computed tomography (CT) scan nor an MRI to assess the extent of hypoxic–ischemic injury has been validated in a randomized controlled trial; however, both modalities offer potential insight into recovery. Although not sensitive, early loss of grey–white differentiation on a head CT scan is highly specific for global hypoxic–ischemic injury, with quantitative measures of ischemic changes offering greater predictive value.26 A brain MRI, in contrast, offers greater sensitivity for ischemic changes, but at the expense of specificity.27
The serum neuron-specific enolase (NSE) level, a promising biomarker of neuronal injury, is not well-validated at standardized time points, and the threshold at which levels are elevated remains ill-defined. As with other tests, NSE has greater usefulness when combined with other findings,24, 28 and may offer more insight when trended serially,29 with persistently increasing levels suggestive of ongoing secondary brain injury. More recently, serum neurofilament light chain levels have garnered interest as another biomarker, and although the data are limited currently, the initial results have demonstrated higher sensitivity and specificity for poor outcome in the first 24–48 hours compared with NSE performance.30
Related posts:
Chapter 2 – How Much Does the Family Want to Be Involved in Decision-Making?
Chapter 5 – The Do-Not-Resuscitate Order
Chapter 6 – The Do-Not-Intubate Order
Chapter 14 – Potentially Inappropriate Treatment and Conscientious Objection
Chapter 20 – Brain Death Discussions
Chapter 13 – Severe Traumatic Spinal Cord Injury
Full access? Get Clinical Tree
