Residency · Residency · Neurology
Neuroprognostication After Cardiac Arrest
Introduction
Hypoxic-ischemic brain injury following cardiac arrest is one of the most common reasons for neurological consultation in the ICU. Accurate prognostication is critical because it guides decisions about withdrawal of life-sustaining therapy, which accounts for the majority of deaths in comatose cardiac arrest survivors. A multimodal approach using clinical examination, electrophysiology, neuroimaging, and biomarkers minimizes the risk of self-fulfilling prophecy.
Pathophysiology of Hypoxic-Ischemic Brain Injury
Global cerebral ischemia during cardiac arrest leads to energy failure, excitotoxicity, calcium influx, and oxidative stress. Primary injury occurs during the no-flow and low-flow periods. Secondary injury involves reperfusion injury, inflammation, and delayed neuronal death occurring hours to days after return of spontaneous circulation (ROSC). Selectively vulnerable structures include hippocampal CA1 neurons, cerebellar Purkinje cells, cortical layers 3, 5, and 6, and the basal ganglia. White matter injury (delayed leukoencephalopathy) can develop days to weeks after the initial insult.
Timing of Prognostication
Targeted temperature management (TTM) must be completed and adequate time for rewarming and drug clearance must pass before prognostication. Current guidelines recommend waiting at least 72 hours after return to normothermia before making prognostic assessments. Residual sedation and neuromuscular blockade must be excluded, with drug levels measured when possible. Serial examinations over days are more reliable than a single assessment.
Clinical Examination
Motor Response
Absent or extensor motor response (GCS motor score 1-2) at 72 hours or more post-ROSC is an unfavorable sign but has a false positive rate of approximately 10-20% when used alone. Status myoclonus (continuous, generalized, stimulus-sensitive myoclonus within 72 hours) is strongly associated with poor outcome, though rare exceptions exist (Lance-Adams syndrome). Isolated myoclonic jerks are less specific and should not be used in isolation.
Pupillary Reflexes
Bilaterally absent pupillary light reflexes at 72 hours or more have a false positive rate of less than 1% for poor outcome prediction. Automated infrared pupillometry provides quantitative, reproducible measurements, with a neurological pupil index less than 2 being an unfavorable sign.
Corneal Reflexes
Bilaterally absent corneal reflexes at 72 hours or more have a false positive rate of approximately 2-5%. They should be tested with direct contact (cotton wisp or saline drops), not air puff alone.
Electrophysiology
Electroencephalography (EEG)
Highly malignant patterns at 72 hours or more include burst suppression without reactivity, suppressed background without reactivity, and status epilepticus. Malignant patterns include discontinuous background and absence of reactivity. Benign patterns include continuous and reactive background and sleep architecture. EEG reactivity to external stimulation is a key prognostic feature, with absence of reactivity at 72 hours being unfavorable. Continuous EEG monitoring detects seizures in 10-30% of comatose cardiac arrest survivors.
Somatosensory Evoked Potentials (SSEPs)
Bilateral absence of cortical N20 responses with median nerve stimulation at 24-72 hours post-ROSC has a false positive rate for poor outcome of less than 1%, making it one of the most robust individual predictors. Intact peripheral and spinal cord conduction (N9, N13 responses) must be confirmed. SSEPs cannot be reliably interpreted during therapeutic hypothermia in some protocols.
Neuroimaging
CT Head
Loss of gray-white matter differentiation suggests diffuse cerebral edema. The quantitative gray-to-white matter ratio (GWR) at the basal ganglia level is prognostically useful, with GWR less than 1.18 associated with poor outcome. CT is widely available and rapid.
MRI Brain
Diffusion-weighted imaging (DWI) is the most sensitive modality for detecting hypoxic-ischemic injury. Extensive cortical and deep gray matter restricted diffusion at 2-7 days is highly predictive of poor outcome. Whole-brain mean ADC values less than 650-700 x 10^-6 mm^2/s are associated with poor outcome. MRI may detect injury missed by CT, particularly in the cortex, hippocampi, and brainstem.
Serum Biomarkers
Neuron-Specific Enolase (NSE)
NSE is the most extensively studied biomarker for cardiac arrest prognostication. Elevated levels at 48-72 hours post-ROSC correlate with poor outcome. A threshold of greater than 33 mcg/L is commonly cited, but assay-specific thresholds should be used, and trending values over time is more informative than single measurements. Hemolysis causes false elevation because NSE is present in red blood cells.
Other Biomarkers
Serum neurofilament light chain (NfL) is emerging as a highly accurate prognostic marker, with elevated levels at 24-72 hours strongly predicting poor outcome. S100B protein is less specific than NSE and serves as an astroglial marker. Glial fibrillary acidic protein (GFAP) reflects astrocytic injury.
The Multimodal Approach
| Modality | Timing | Finding Predicting Poor Outcome | False Positive Rate |
|---|---|---|---|
| Pupillary reflexes | ≥72 hours post-ROSC | Bilaterally absent | <1% |
| Corneal reflexes | ≥72 hours post-ROSC | Bilaterally absent | 2-5% |
| Motor response | ≥72 hours post-ROSC | GCS-M 1-2 (absent/extensor) | 10-20% (alone) |
| SSEPs (N20) | 24-72 hours | Bilaterally absent cortical N20 | <1% |
| EEG | ≥72 hours | Burst suppression without reactivity; status epilepticus | Low (pattern-dependent) |
| MRI (DWI) | 2-7 days | Extensive cortical/deep gray restricted diffusion | Low |
| CT (GWR) | 24-48 hours | GWR <1.18 at basal ganglia | Low |
| NSE | 48-72 hours | >33 mcg/L (assay-specific) | Variable |
| NfL | 24-72 hours | Markedly elevated | Very low (emerging data) |
| Status myoclonus | <72 hours | Continuous, generalized, stimulus-sensitive | Rare exceptions (Lance-Adams) |
No single test should be used in isolation to predict poor outcome. The 2023 ERC/ESICM guidelines recommend beginning with clinical examination at 72 hours or more post-ROSC (after rewarming and drug clearance). If GCS motor score is 1-2, at least two concordant prognostic tests should be added. Concordant unfavorable findings from two or more modalities (clinical, SSEP, EEG, imaging, biomarkers) provide high confidence of poor outcome. If any test yields indeterminate or discordant results, observation should be extended and assessments repeated.
Clinical Pearls
Prognostication should never be based on a single modality; a multimodal approach must always be used to minimize the risk of self-fulfilling prophecy. At least 72 hours after normothermia must pass, with confirmed clearance of sedatives, before any prognostic assessment. Bilateral absence of N20 on SSEPs is one of the most reliable individual predictors but must be interpreted in conjunction with other findings. Status myoclonus is strongly associated with poor outcome, but rare survivors with good outcomes exist (Lance-Adams syndrome), so it should not be the sole basis for withdrawal of care. Serum NfL is emerging as a powerful prognostic biomarker that may outperform NSE in accuracy.
References
- Sandroni C, Nolan JP, Andersen LW, et al. ERC-ESICM guidelines on temperature control after cardiac arrest in adults. Intensive Care Med. 2022;48(3):261-269.
- Wijdicks EFM, Hijdra A, Young GB, et al. Practice parameter: prediction of outcome in comatose survivors after cardiopulmonary resuscitation. Neurology. 2006;67(2):203-210.
- Moseby-Knappe M, Mattsson-Carlgren N, Engerstrom A, et al. Serum neurofilament light as a biomarker of outcome after cardiac arrest. Neurology. 2023;100(23):e2371-e2383.
- Cronberg T, Greer DM, Lilja G, et al. Brain injury after cardiac arrest: from prognostication of comatose patients to rehabilitation. Lancet Neurol. 2020;19(7):611-622.