Encephalopathy and coma


3
Encephalopathy and coma


3.1 Nonspecific patterns of encephalopathy


There are many nonspecific changes in the EEG that occur during diffuse encephalopathies. Early changes include slowing of the alpha rhythm and excess slowing during wakefulness, first theta and then delta. This is followed by loss of the alpha rhythm, more prominent (mainly delta) slowing, loss of normal faster activity and loss or attenuation of normal sleep transients. Abnormal arousal patterns, such as the cyclic alternating pattern of encephalopathy (CAPE), and generalized rhythmic delta activity (GRDA) may also be seen. GRDA is most commonly seen in diffuse encephalopathy and is nonspecific. Examples of GRDA can be found in Chapter 5: Rhythmic and periodic patterns (RPPs).


As encephalopathy worsens, changes include loss of normal variability and state changes, loss of reactivity to external stimuli including pain, possibly burst suppression, diffuse attenuation and ultimately electrocerebral inactivity (a ‘flat’ tracing). It is important to recognize that most or all of these patterns can be produced in a normal brain via the use of high dose sedatives such as barbiturates, propofol and benzodiazepines.


3.2 Grading of encephalopathy


Grading the degree of encephalopathy is clinically valuable; and allows for the progress (improvement/deterioration) of a patient to be objectively monitored over time (especially if the clinical examination is unreliable). Although the term ‘encephalopathy’ is a clinical diagnosis (and therefore does not feature in the ICU EEG terminology), as a patient becomes progressively encephalopathic and enters coma there are a number of sequential (mostly generalized) EEG changes that suggest deeper states of diffuse cerebral dysfunction. Documenting the severity of encephalopathy also provides ancillary information that can assist in patient prognostication.


Two of the most validated encephalopathy grading scales were proposed by Synek, VM in 1988 and Young et al. in 1997. Both scales, which have been replicated in Tables 3.1 and 3.2, mostly described EEG changes that would fall under ‘background EEG’ in the latest terminology. The inter-rater reliability was good to very good for both scales, and the prognostic implications to the various ‘grades and subgrades’ of encephalopathy were determined. It should be noted that several specific clinical states, such as ‘alpha, theta or spindle coma’ do not feature in the ICU EEG terminology. Similar to ‘encephalopathy’, the terminology provides the descriptive tools to classify the EEG changes and avoids clinical interpretation. For example, the EEG finding of an ‘unreactive alpha record without state changes’ would infer the clinical state of ‘alpha coma’, and practically (including at times in this atlas) the EEG and clinical terms are used interchangeably.


Since the publication of the above EEG grading systems there have been several updates in the literature:



  1. The clinical importance of certain features (such as reactivity) has been further studied, and many have become important determinants of prognosis.

    TABLE 3.1 Synek EEG Coma Scale (1988)


































    Grade of encephalopathy Description Subgrades or subdivisions
    I Predominantly regular alpha activity with some scattered activity in the theta frequency range (reactive)
    II Predominantly theta range activity with some alpha and delta waves

    1. Normal voltage, reactive
    2. Low voltage, unreactive
    III Predominantly delta activity (regular or irregular) with little activity in other frequencies. Further divided into ‘subgrades’ consisting of:

    1. High voltage rhythmic delta activity that is maximal frontally (reactive) [Would now be described as GRDA].
    2. Spindle coma
    3. Low voltage irregular delta (non-reactive)
    4. Medium voltage irregular delta (non-reactive)
    IV

    1. Burst-suppression


    1. With epileptiform activity (epileptiform discharges consisting of polyspikes or clusters of sharp waves) [Now referred to as highly epileptiform bursts]
    2. No epileptiform activity



    1. Alpha coma


    1. With some reactivity
    2. Without reactivity



    1. Theta coma
    2. Low output EEG, voltages less than 20 μV but not suppressed
    V (Suppression) Absence of cerebral activity (all activity <2 μV).

    Adapted from Synek VM. EEG abnormality grades and subdivisions of prognostic importance in traumatic and anoxic coma in adults. Clin Electroencephalogr. 1988;19(3):160–166. The grading system was originally validated in patients with TBI and post anoxic brain injury, but later tested across a range of disease processes.


  2. Select patterns (e.g., highly epileptiform bursts) have been further characterized, but in order to limit heterogeneity they have been mostly studied in very specific patient cohorts (e.g., in the weaning of highly sedating anti-seizure medications used for treatment of status epilepticus, or in post anoxic brain injury).
  3. The definitions of certain terms have become standardized and some of these differ from historical definitions (such as the definition of burst suppression mentioned above).
  4. New patterns/terms have been identified (such as identical bursts) that were not included in these original classification systems.
  5. Old descriptions (such as triphasic waves) have been further assessed and their specific association with encephalopathy (or any other clinical scenario) placed in question.

The result of the above has meant the above grading systems have been poorly adopted into routine clinical practice and the information included in them possibly outdated. The prior grading systems were not intuitively linear, e.g., in both systems a predominantly theta record (Synek grade II, Young grade I) was classified as either (1) reactive or (2) unreactive. However, if a record is unreactive this has a worse prognosis than GRDA (Synek grade IIIa) when this pattern was later studied; or unreactive vs those with epileptiform discharges (Young grade IV). A grading system for encephalopathy should ideally only make comment on the degree of diffuse dysfunction, i.e., should not have implications outside of this discussion. The prior grading systems included terms such as epileptiform discharges, which clearly have implications on the probability of subsequent seizures. One of the main limiting factors of historical studies has been that over the years the definitions of many of the terms have been debated, refined and standardized. Take the example of ‘burst suppression’. As mentioned in the legend of Table 3.2, Young et al.’s definition of burst suppression would have included many patients’ EEGs that would now be classified as either ‘nearly continuous’ or ‘discontinuous’. The implication of this is that the ‘severity’ and clinical prognosis attributed to many of these patterns may no longer be valid.


TABLE 3.2 Young et al.; EEG Classification for Coma (1997)

























Category Subcategory


  1. Delta/theta >50% of record (not theta coma)

  1. Reactivity
  2. No reactivity


  1. Triphasic waves


  1. Burst-suppression

  1. With epileptiform activity
  2. Without epileptiform activity


  1. Alpha/theta/spindle coma (unreactive)


  1. Epileptiform activity (not in burst-suppression pattern)*

  1. Generalized
  2. Focal or multifocal


  1. Suppression

  1. <20 μV, but >10 μV
  2. ≤10 μV

Reproduced from Young GB, et al., An Electroencephalographic Classification for Coma. Canadian Journal of Neurological Sciences. 1997;24(04):320–325.


It should be noted that some of the suggestions differ significantly from the updated definitions of the same term. For example, Young, et al., suggested that burst-suppression pattern should have generalized flattening at standard sensitivity for ≥1 second at least every 20 seconds. Under current terminology 1 second of suppression every 20 seconds equates to a suppression percent of 5% and therefore that record would be classified as ‘nearly continuous’ rather than burst suppression, which would require a suppression percent between 50–99% of the record. If more than one category applied, then it was suggested that the most ‘serious’ category should be selected. Therefore the 1997 classification did not account for multiple patterns to be included in the same record.


a ‘Epileptiform activity’ was not specifically defined; however, the example of activity included under this category consisted of a highly epileptiform pattern of fluctuating spiky 1.5–2-Hz GPDs on a suppressed background (i.e., a pattern that would currently be considered on the ictal-interictal continuum).


There is currently no universally accepted scale of encephalopathy. Surveys have shown that most institutions either use no specific scale or utilize center-specific grading systems that are much more basic. These grading systems often use the terms ‘mild’, ‘mild to moderate’, ‘moderate’, ‘moderate to severe’ and ‘severe’ to classify severity of diffuse dysfunction. The major benefit of this is that they are much easier to understand and integrate into clinical practice. For example, if the EEG findings for a given patient changed from ‘severe’ to ‘mild’, this intuitively suggests improvement. An example of such a grading system is the Yale Adult Background EEG Grading Scale 2021 (Table 3.3). This was generated after a series of multicenter surveys of experts about which features are most important, how many categories to use, etc., and has not yet been validated. These scales specifically only comment on features of the ‘background’ EEG and allow focal findings, sporadic epileptiform discharges and RPPs to stand independent of this. Additional features of the EEG, such as loss of physiologic sleep transients, are also markers of encephalopathy; however, these were not consistently utilized across centers and therefore were not specifically incorporated into the grading scale.


The grading should begin from severe to normal, i.e., if there is no PDR, no reactivity (with adequate testing), and no state changes the record automatically indicates ‘severe dysfunction’ irrespective of any other features such as the predominant frequency.


3.3 Findings in specific medical conditions


All metabolic encephalopathies can cause diffuse slowing, generalized periodic discharges (GPDs) including GPDs with triphasic morphology (also known as triphasic waves), and most predispose to seizures as well. Conditions that commonly lead to these findings are hepatic or renal failure, hyponatremia, Hashimoto’s encephalopathy and COVID-19-related encephalopathy. Neuroleptic malignant syndrome, serotonin syndrome and some medication toxicities (i.e., cefepime, baclofen, lithium, ifosfamide, CAR-T cell therapy and others) can cause similar patterns. One particular pattern of encephalopathy, the extreme delta brush (EDB) pattern, was initially described in the setting of anti-NMDA receptor mediated encephalitis. Since then, it has been confirmed to have reasonable specificity for the condition (but not 100%) and its features have now been defined in the 2021 ACNS terminology (Figures 3.18, 3.19 and 3.20).


TABLE 3.3 Yale Adult Encephalopathy Scale (2021)





image

a LVF = low voltage beta (or faster) activity diffusely: In awake patients with meaningful interaction, LVF can be part of a normal background, mild dysfunction or mild-moderate dysfunction based on other features. In patients without meaningful interaction, LVF should be classified as moderate to severe if EEG is reactive, and severe if unreactive and without state changes.


b NR: Not required. These features are not part of the definition of that level of dysfunction, but all records will be reactive and have state changes if a long enough recording is obtained.


c Complete suppression: No discernible cerebral rhythms whether or not standards for determination of electrocerebral silence are met. NA: Not Applicable


3.4 Medication effects


Many medications exert their effect on the brain and hence can cause changes to the EEG. The number of medications that can alter the EEG is exhaustive. In broad terms these medications can be split into (1) those that cause an intended effect on the brain or (2) those that have an unintended adverse effect on the brain.


Medications with an intended effect


The most commonly encountered medications with effects on the brain are the highly sedating anti-seizure medications (ASMs), which are often administered intravenously while the cEEG is being recorded in order to control seizures or status epilepticus. Sequentially increasing the dose of any sedative or anesthetic agent will eventually result in diffuse dysfunction of increasing severity (as discussed above). Agents reported to result in marginally more specific EEG changes are:



  1. benzodiazepines and barbiturates, which lead to an excess of fast activity, primarily beta range
  2. dexmedetomidine (selective alpha-2-adrenoreceptor agonist), which is predominantly given for non-opioid sedation in the setting of agitation, induces a state that resembles N2 sleep with prominent spindle-like activity
  3. propofol, which at lower levels causes prominent beta activity, followed by moderate to high amplitude 2–3 Hz delta, and then emergence of 14-Hz spindle-like activity if dosing is maintained and
  4. ketamine, which results in increases in delta activity but with some laboratory evidence of retention of gamma range activity, while attenuating activity in other frequencies.

An important caveat is that most of this information comes from documented use of anesthetic agents in healthy brains. In the critical care setting these medications are being administered to patients with severely altered systemic and cerebral physiology with highly abnormal EEG patterns to start. In this context, administration often leads to an alteration of the EEG pattern, but not necessarily the emergence of the typical EEG response. The second point is that the above descriptions are not absolutely specific. At high enough levels, almost all the agents above can cause burst suppression followed by electrocerebral inactivity.


Medications with an unintended adverse effect


There are also many medications that can cause an unintended adverse effect on the brain and EEG. These medications are administered for an action that is not controlling seizures, but can result in diffuse slowing, epileptiform discharges, GPDs and seizures (similar to the changes described in section 3.3 above that occur in the metabolic encephalopathies). The medications most established to cause such changes are the centrally acting agents of baclofen, lithium and clozapine, as well as the fourth-generation cephalosporin antibiotics, such as cefepime. Medications with a modest association with encephalopathic changes include metronidazole, isoniazid, theophylline, cyclosporine and tacrolimus.


Figure list



EEGs throughout this atlas have been shown with the following standard recording filters unless otherwise specified: LFF 1 Hz, HFF 70 Hz, notch filter off.


Suggested reading



  1. Akeju O, Song AH, Hamilos AE, et al. Electroencephalogram signatures of ketamine anesthesia-induced unconsciousness. Clin Neurophysiol. 2016; 127(6):2414–2422.
  2. Antony AR, Haneef Z. Systematic review of EEG findings in 617 patients diagnosed with COVID-19. Seizure. 2020; 83:234–241.
  3. Bahamon-Dussan JE, Celesia GG, Grigg-Damberger MM. Prognostic significance of EEG triphasic waves in patients with altered state of consciousness. J Clin Neurophysiol 1989; 6:313–319.
  4. Bickford RG, Butt HR. Hepatic coma: the electroencephalographic pattern. J Clin Invest 1955; 34:790–799.
  5. Blume WT. Drug effects on EEG. J Clin Neurophysiol 2006; 23:306–311.
  6. Capparelli FJ, Diaz MF, Hlavnika A, Wainsztein NA, Leiguarda R, Del Castillo ME. Cefepime- and cefixime-induced encephalopathy in a patient with normal renal function. Neurology 2005; 65:1840
  7. Chatrian GE, White LW, Daly D. Electroencephalographic patterns resembling those of sleep in certain comatose states after injuries to the head. Electroencephalogr Clin Neurophysiol 1963; 15:272–280.
  8. Dhakar MB, Sheikh ZB, Desai M, Desai RA, Sternberg EJ, Popescu C, Baron-Lee J, Rampal N, Hirsch LJ, Gilmore EJ, Maciel CB. Developing a Standardized Approach to Grading the Level of Brain Dysfunction on EEG. J Clin Neurophysiol.

    Only gold members can continue reading. Log In or Register to continue

    Related posts:

    Vision The Deep Tendon or Muscle Stretch Reflexes Default ThumbnailTransient Global Amnesia Dysmorphic Vertebral Body Default ThumbnailArteries of Spinal Cord of the Forebrain

    Stay updated, free articles. Join our Telegram channel
Tags:
Mar 23, 2024 | Posted by in NEUROLOGY | Comments Off on Encephalopathy and coma

Full access? Get Clinical Tree
Get Clinical Tree app for offline access