Myoclonus After Cardiac Arrest: Need for Standardization—A Systematic Review and Research Proposal on Terminology

Abstract

OBJECTIVES:

Although myoclonus less than or equal to 72 hours after cardiac arrest (CA) is often viewed as a single entity, there is considerable heterogeneity in its clinical and electrophysiology characteristics, and its strength of association with outcome. We reviewed definitions, electroencephalogram, and outcome of myoclonus post-CA to assess the need for consensus and the potential role of electroencephalogram for further research.

DATA SOURCES:

PubMed, Embase, and Cochrane databases.

STUDY SELECTION:

English-language adult (≥ 18 yr) studies from 1966 to May 31, 2024, reporting myoclonus, myoclonic status/status myoclonus (MyS/SM), myoclonic status epilepticus (MSE), and/or early Lance-Adams Syndrome (eLAS) less than or equal to 72 hours post-CA. All study designs were independently screened by two authors.

DATA EXTRACTION:

Data on patients presenting myoclonus, MyS/SM, MSE, and eLAS less than or equal to 72 hours post-CA, along with their definitions, electroencephalogram, and outcomes were extracted. The Newcastle-Ottawa Scale and Cochrane-Risk-of-Bias Assessment tool were used to evaluate study quality (PROSPERO n.CRD42023438107).

DATA SYNTHESIS:

Of 585 identified articles, 119 met the inclusion criteria, revealing substantial heterogeneity in definitions, electroencephalogram, and outcomes. Among 3881 patients, myoclonus was reported in 2659, MyS/SM in 883, MSE in 569, and eLAS in 40. Among patients with a defined outcome, a Cerebral Performance Category (CPC) scale of 1–2 was reported in 9.8% of patients with myoclonus, 5.8% with MyS/SM, 5.7% with MSE, and 82.0% with eLAS. Electroencephalogram was recorded in 2714 patients (69.9%). CPC of 1–2 was observed in 1.6% of patients with suppression/suppression burst (SB)/unreactive (U) electroencephalogram, 11.3% with non-SB/U electroencephalogram and status epilepticus (SE), and 22.3% with non-SB/U electroencephalogram without SE.

CONCLUSIONS:

Heterogeneity in definitions resulted in weak associations with outcomes. We propose to investigate myoclonus by including related electroencephalogram patterns: myoclonus associated with suppression/SB background electroencephalogram, myoclonus with nonsuppression/SB background but SE-electroencephalogram, and myoclonus with nonsuppression/SB background without SE-electroencephalogram. This pragmatic research approach should be validated in future studies.

KEY POINTS.

Question: Is there uniformity in defining myoclonus following cardiac arrest? How does the nomenclature impact on outcome assessment?

Findings: A systematic review of 119 studies revealed heterogeneity in myoclonus definitions, associated electroencephalogram patterns, and outcomes. Categorizing patients including electroencephalogram patterns demonstrated better association with outcomes than current myoclonus definitions.

Meaning: The substantial heterogeneity in myoclonus definitions emphasizes the need for international consensus. A research proposal on terminology facilitates the integration of electroencephalogram background and status epilepticus criteria.

Myoclonus is characterized by sudden, brief, and involuntary movements caused by muscle contractions/inhibitions and is often observed in comatose patients after cardiac arrest (CA) (1). While generally associated with poor outcome (2–4), a small subset of patients achieves good prognosis, indicated by a favorable Cerebral Performance Category (CPC) (Supplementary Table S1, http://links.lww.com/CCM/H618) (5–7).

Current literature presents a wide range of clinical definitions and associated electroencephalogram features of post-CA myoclonus, which poses a challenge for treatment decisions and accurate prognostication (6, 8, 9). Historically, myoclonus persisting for more than 30 minutes has been clinically defined as myoclonic status epilepticus (MSE) (3). The 2015 guidelines from the European Resuscitation Council/European Society of Intensive Care Medicine (ERC/ESICM) defined myoclonic status/status myoclonus (MyS/SM) similarly (10), highlighting its strong association with poor outcome (10). Consequently, Status Myoclonus less than or equal to 72 hours has been included among those poor outcome factors allowing a withdrawal of life-sustaining treatment (WLST) in the 2021 ERC/ESICM guidelines (10). Concurrently, the International League Against Epilepsy (ILAE) revised the time criterion for convulsive status epilepticus (CSE), reducing it from 30 to 5 minutes (11), and for non-CSE to 10 minutes.

While myoclonus and MyS/SM are clinical entities, several attempts have been made to delineate their corresponding electroencephalogram features. The American Clinical Neurophysiology Society (ACNS) (12) implemented the Salzburg Criteria (13) for nonconvulsive status epilepticus (NCSE) including: 1) time-locked clinical correlate (i.e., jerks here), 2) epileptiform discharges greater than 2.5 Hz, 3) evolution in time/location/morphology, and 4) electroencephalogram and clinical improvement to IV anti-seizure medication (12–14) lasting at least 10 continuous minutes or for a total duration of at least 20% of any 60-minute recording (12).

In patients after CA, electroencephalogram has been classified into the categories highly malignant, malignant, and benign, irrespective of myoclonus (15): unequivocal electrographic seizures post-CA (with or without myoclonus) were considered among the malignant patterns only in the absence of suppression/suppression burst (S/SB) background (12).

Additionally, Lance-Adams syndrome (LAS) refers to generalized action myoclonus, persistent, usually with preserved of consciousness, typically developing days to weeks after CA (9). The myoclonus can be chronic and then (after 72 hr) usually associated with continuous background with narrow, midline-centered spike waves (9, 16–19), however, may appear early (≤ 72 hr post-CA), that is, early LAS (eLAS), even in comatose patients (9, 20, 21). For the purpose of this review, eLAS refers to those patients with myoclonus who are comatose in the first 72 hours and later develop LAS. These patients, due to their comatose state, cannot be clinically distinguished from patients with myoclonus less than or equal to 72 hours who later have poor outcome.

Given the above-described heterogeneity, we performed a systematic review to 1) assess the currently available approaches in terminology of post-hypoxic myoclonus regarding homogeneity and consistency, 2) determine whether these available definitions are related to outcome, 3) determine whether myoclonus stratified for associated electroencephalogram patterns had a higher discriminative ability regarding published outcomes compared with a purely clinical assessment, and 4) develop a research proposal on terminology derived from data analysis.

MATERIALS AND METHODS

This systematic review has been conducted according to the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) guidelines. The final study protocol was registered in PROSPERO (n.CRD42023438107).

Search strategy, selection criteria, data extraction, risk of bias assessment, and data analysis are reported in Supplementary Method S1 (http://links.lww.com/CCM/H618).

RESULTS

Study Selection and Quality Assessment

The PRISMA flowchart of studies investigated in this systematic review is shown in Figure 1 (22). One hundred nineteen papers reporting on myoclonus, MyS/SM, MSE, and eLAS met the inclusion criteria (Supplementary Table S2, http://links.lww.com/CCM/H618). Quality assessment score results for each study revealed an overall high to medium risk of bias (90/115 had a Newcastle-Ottawa Scale score < 7 and 4/4 had an overall high risk of bias on the Cochrane Risk-of-Bias assessment tool) and are provided in Supplementary Tables S3 and S4 (http://links.lww.com/CCM/H618).

Figure 1.

Preferred Reporting Items for Systematic reviews and Meta-Analyses flowchart. CA = cardiac arrest.

Myoclonus, MyS/SM, and MSE: Prevalence, Definitions, and Association With Outcome

Seventy-eight studies reported myoclonus (6, 8, 9, 16–21, 23–90) in comatose patients evaluated for prognostication, and 19 provided various semiological definitions (6, 8, 9, 19, 21, 28, 30, 31, 37, 44, 50–53, 56, 58, 64, 76, 83). Five studies reported the presence or absence of epileptiform discharges associated with myoclonus (6, 8, 9, 28, 52, 90). Most studies did not include information on the semiology of myoclonus or electroencephalogram features. Defined outcome was available in 2008 of 2659 patients (75.5%) with myoclonus, 9.81% achieved a CPC of 1–2.

Twenty-four studies reported MyS/SM (7, 29, 31, 50, 52, 91–109) in comatose patients post-CA and 20 of them provided definitions which differed in clinical description and duration (7, 31, 50, 52, 91–97, 99–103, 105–108) (Supplementary Table S2, http://links.lww.com/CCM/H618). Out of 883 patients with MyS/SM, defined outcome was reported in 487 (55.2%), 5.8% reached a CPC of 1–2.

Thirty-one studies reported MSE (7, 15, 42, 50, 51, 58, 71, 86, 87, 95, 110–130) in comatose patients, of which 17 studies (7, 15, 50, 58, 86, 95, 110, 111, 113, 117–119, 121, 122, 124, 128, 130) provided different definitions (7, 95, 110, 111, 113, 119, 121, 122) (Supplementary Table S2, http://links.lww.com/CCM/H618). Despite the more frequent use of electroencephalogram to establish the definition of MSE compared with MyS/SM, associated electroencephalogram features were variable among studies, and sometimes purely clinical definitions were reported. Among 437 patients (76.8%) with MSE and reported outcome, 5.7% accomplished a CPC of 1–2.

CPC distribution for each study with reported outcomes based on the definitions of myoclonus, MyS/SM, and MSE is presented in Figure 2, while pooled data are shown in Figure 3.

Figure 2.

Cerebral Performance Category (CPC) 1–2 (good outcome) distribution for each study with reported outcome based on definition of myoclonus, myoclonic status/status myoclonus, and myoclonic status epilepticus. Circle size is proportioned to cohort size (e.g., myoclonus: Lybeck et al [52]: 321 patients; Reynolds et al [86]: three patients).

Figure 3.

Cerebral Performance Category distribution based on definitions and electroencephalogram (EEG) categories. Percentage of good outcome with Cerebral Performance Category (CPC) 1–2 (good outcome) scores with myoclonus, myoclonic status/status myoclonus, and myoclonic status epilepticus (A) and across the EEG-based clustering (B). SB/U = suppressed, suppression-burst or unreactive background, SE = status epilepticus.

Early Lance Adams Syndrome

Twenty-one studies (9, 16–21, 23, 45, 51, 82, 84, 88, 89, 91, 92, 102, 110, 118, 122, 128) reported 40 patients with eLAS, that is, developing LAS after acute myoclonus, MyS/SM, or MSE (Supplementary Table S5, http://links.lww.com/CCM/H618). CPC 1–2 was reported in 27 of 33 patients (81.8%) with reported outcome.

Electroencephalogram Categories: Prevalence and Association With Outcome

Electroencephalogram was recorded in 2715 of 3881 patients (69.9%).

Seven hundred eighty-nine patients had an suppression burst/unreactive (SB/U) electroencephalogram, 474 had a non-SB/U electroencephalogram fulfilling criteria of SE (non-SB/U + SE), and 480 had a non-SB/U electroencephalogram without SE (non-SB/U), with a good outcome (CPC 1–2) reported in 1.6% (CPC reported in 577 patients), 11.2% (CPC reported in 391), and 22.2% (CPC reported in 404), respectively. We included only patients reported as CPC 1–2.

CPC distribution for each study with reported outcomes based on electroencephalogram features is presented in Figure 4, while pooled data across studies are shown in Figure 3.

Figure 4.

Cerebral Performance Category (CPC) 1–2 (good outcome) distribution for each study with reported outcome based on different electroencephalogram (EEG) pattern. Circle size is proportioned to cohort size (e.g., myoclonic status epilepticus [MSE] EEG: Seder et al [71], 102 patients; Beekman et al [111], one patient). SB/U EEG = suppressed, suppression-burst or unreactive background electroencephalogram, SE = status epilepticus.

Clustering Metrics

Qualitative analysis of the clustering metrics of the pooled data revealed greater segregation (0.07 vs. 0.02) and cohesion (0.12 vs. 0.07) of the categorization including electroencephalogram compared with current terminology.

DISCUSSION

This systematic review highlights a huge heterogeneity in the definitions of myoclonus, MyS/SM, MSE, and eLAS, which explains the substantial variability and discrepancy among reported association with outcome. At the same time, homogeneous electroencephalogram patterns (presence or absence of NCSE electroencephalogram) and background in myoclonic patients are associated with more homogenous outcomes.

Definitions

Myoclonus definitions were different among studies (6, 33, 37, 52, 86) and several studies did not distinguish among myoclonus, MyS/SM, and MSE and used terms interchangeably (25, 31, 39, 47, 52, 58, 71, 75, 113). The difference between MyS/SM and MSE based on electroencephalogram features appeared consistent in some studies (7, 95). In contrast, others adopted heterogeneous definitions (reporting either malignant [115] or highly malignant electroencephalogram pattern [115] or were ambiguous regarding their criteria [7, 71, 86, 94, 95, 99, 110, 111, 113, 116, 117, 122, 126, 128, 130]).

However, almost two thirds (64.0%) of the screened studies described only the presence or absence of myoclonus. In those, the information on whether an electroencephalogram was performed was missing (30, 46, 59, 78, 80) or specific correlations between myoclonus and electroencephalogram were not reported (24, 38, 40, 55, 57, 61, 63, 65, 67, 73, 79, 100, 104, 131, 132). Data were unclear as to whether myoclonic jerks were time-locked to periodic discharges on a continuous (36, 37) or suppressed background (49) or time-locked to bursts (74). The post hoc analysis of a multicenter Targeted Temperature Management trial described this specific dilemma: seizures were defined without an electroencephalogram correlate and MyS/SM was defined regardless electroencephalograms including electrographic seizures (52).

Finally, patients typically present LAS few days to weeks after CA, although it can present less than or equal to 72 hours (eLAS) (9, 20, 21). The continuity of electroencephalogram background is of central importance in LAS. Patients with LAS have already suffered from eLAS (9, 16–21, 45, 51, 72, 82, 84, 88, 89), MyS/SM (91, 92, 102), and MSE (91, 92, 102) in the acute setting or may have developed myoclonus only in the chronic phase. The electroencephalogram background has only rarely been reported in eLAS: it was normal or nearly normal, with or without superimposed epileptiform discharges ((16–19). These findings illustrate the wide spectrum of the clinical and electrophysiology presentations: whereas LAS includes the defining feature of a continuous background after 72 hours, it remains to be elucidated whether this continuous background needs to be essentially present already less than or equal to 72 hours (eLAS). Therefore, we encourage to investigate the electroencephalogram patterns identified less than or equal to 72 hours in patients who later develop LAS.

The main shortcoming of the current terminology is that SM and its synonymous term MyS are purely clinical entities and are therefore wide umbrella terms including several other categories without the possibility for further association studies on outcome. MSE should include validated electroencephalogram criteria for SE to do reliable research on it. Ideally, there should be no overlap of definitions in the highly dynamic condition of post hypoxic encephalopathy.

The Outcome

A few studies have focused on the electroencephalogram patterns associated with myoclonus and their relation to outcome. Cohort studies have highlighted the overall strong association of MyS/SM for negative outcomes. Still, they are at least partly influenced by a self-fulfilling prophecy due to its inclusion in the decisions for WLST (10). However, case reports suffer from publication bias indicating the exceptional positive outcome of MyS/SM, for example, potentially evolving to LAS (91, 92, 97).

Myoclonus with rhythmic and periodic patterns (RPPs) was associated with poor outcomes in discontinuous/or suppressed electroencephalogram, whereas few survivors were seen in the presence of a continuous background (4/47 patients) (62). However, it was unclear whether the myoclonus was time-locked with RPPs and if so, should be considered a SE according to ACNS terminology. A SB background was frequently described in the context of myoclonus, which may be associated mostly with 1) bad outcome (8, 17, 23, 35, 41, 43, 77), 2) rarely good outcome (33), and 3) interpreted as a 4) ictal (23) or 5) interictal (70) phenomenon.

MyS/SM was often defined only clinically as myoclonus lasting greater than 30 minutes (91, 101, 102, 105–108, 133) and was mostly associated with poor outcomes (93, 98, 105–108). Nevertheless, cases of good outcomes have also been reported, for example, patients with MyS/SM evolving into LAS (91, 102), or where MyS/SM was defined despite an epileptiform electroencephalogram potentially compatible with MSE (92, 97, 103).

In this systematic review, patients with any myoclonic presentation had a good prognosis in 6.2%. Our findings demonstrate that in the absence of standardized definitions, the terms myoclonus, MyS/SM, and MSE were often used interchangeably, consequently leading to similar outcomes: CPC 1–2 in 9.8% of patients with myoclonus, 5.8% of patients with MyS/SM, and 5.7% of patients with MSE. Whereas outcomes based on variable definitions of myoclonus reduce generalizability and data homogeneity, the approach including electroencephalogram proved to be more clinically relevant, as also qualitatively shown by cohesion, and separation metrics. CPC 1–2 rates were 1.6%, 11.2%, and 22.2% in SB/U, non-SB/U + SE, and non-SB/U, respectively. Hence, the outcome starts poor with SB/U and becomes increasingly better with patients who have a continuous background but still suffer from the burden of SE (non-SB + SE) and finally reaches best subgroup results in those who have continuous reactive background and are free of SE.

Research Proposal on Terminology

The clustering analysis of two parameters determines whether these are loosely distributed or whether there are points at which samples aggregate. This clustering analysis showed that myoclonus in the previous terminology only poorly co-localized with outcome across studies. However, electroencephalogram was associated much closer to outcome. We here propose a practical research approach to assess the role of electroencephalogram in this context, which should be evaluated in post-CA patients presenting with myoclonus within 72 hours (Fig. 5). This procedure uses a stepwise method, where “1” indicates the presence and “0” indicates the absence of a predefined entity.

Figure 5.

Research proposal on terminology based on electroencephalogram (EEG) corelate of myoclonus (M1). Note of caution: this classification needs future validation by an international expert panel and integration into a multimodal concept of outcome estimation. *Definitions of status epilepticus (SE) referring to Salzburg criteria within the American Clinical Neurophysiology Society (ACNS) (12–14): 1) time-locked clinical correlate (i.e., jerks here), 2) epileptiform discharges greater than 2.5 Hz, 3) evolution in time, location, or morphology, and 4) EEG and clinical improvement to diagnostic IV anti-seizure medication. 0 = absent, 1 = present, B = background, CA = cardiac arrest, M1-B0 = myoclonus associated with suppression/suppression burst background electroencephalogram, M1-B1-SE0 = myoclonus with nonsuppression/suppression burst background without status epilepticus-electroencephalogram, M1-B1-SE1 = myoclonus with nonsuppression/suppression burst background but status epilepticus-electroencephalogram.

Scenario A

If video-electroencephalogram is not performed/reported/described, no firm association of myoclonus, MyS/SM, or MSE with outcome may be proposed (“myoclonus without electroencephalogram data”).

Scenario B

Myoclonus associated with S/SB background electroencephalogram (M1-B0, “myoclonus without continuous, nearly continuous, or discontinuous background, i.e., representing S/SB background”) should be investigated for outcome. Although previous studies pointed toward a serious prognosis (15), the consideration of confounders such as timing of occurrence from return of spontaneous circulation, sedation, and temperature effect is obligatory. From a research perspective, it would be worthwhile to evaluate which suppression percent represents a reasonable threshold for good or poor outcome (12). While the electroclinical SE characterized by myoclonus time-locked with the electroencephalogram pattern is referred to as SE (12), in more severe cases, the outcome is primarily determined by the extent of brain damage reflected by the burst suppression electroencephalogram background in the absence of the effect of sedating medications (134).

Scenario C

Along the wide spectrum of tissue damage, the less severe injuries may be not accompanied by burst-suppression. In these, subsequent determinants of outcome become relevant. We propose to define this scenario as myoclonus with continuous or nearly continuous background (M1-B1). If the myoclonus in M1-B1 is associated with an electroencephalogram pattern fulfilling the Salzburg/ACNS criteria of SE (12, 14), the electroclinical picture should be categorized as myoclonus with non-S/SB background but SE-electroencephalogram (M1-B1-SE1), and myoclonus with non-S/SB background without SE-electroencephalogram (M1-B1-SE0). The M1-B1-SE1 category represents the group where prompt initiation of antiseizure treatment to mitigate MSE might improve overall patient prognosis. However, only future studies will determine whether such therapy leads to better patient outcomes and whether specific subtypes of electroclinical features are associated with distinct outcomes, as suggested by some reports (37, 90, 135).

The characterization of the electroencephalogram pattern (presence/absence of RRP, frequency, background frequency band, reactivity, etc.) may also be provided.

Our proposal does not provide specific time points due to insufficient evidence. We encourage investigating whether a lower limit of frequency (measured in Hz) of myoclonus time-locked to electroencephalogram discharges exists. However, the criteria for duration should align with the recommendations of the ACNS (12) and ILAE (11). Additionally, using muscle relaxants, sedatives, and antiseizure medications can blur the distinction between SE and non-SE presentations.

This research proposal of terminology should primarily be used to establish objective and rigorous electroclinical criteria for acute myoclonus following CA and should be investigated regarding a potential future role within the context of multimodal prognostication and integration with other prognostic tools. Since SM/MyS is one of the poor outcome factors in the multimodal prognostication assessment, the correct definition of myoclonic entities is of utmost importance. Future studies are needed to directly assess the association between these electroclinical categories, patient outcomes, and other prognostic factors.

Limitations

This study has several limitations. The available literature is extremely heterogeneous and often contradictory, preventing to conduct comprehensive statistical analyses and meta-analyses. Additionally, the studies exhibit significant biases, including those related to outcome reporting, due to the self-fulfilling prophecy and strong publication biases. We could not address the difference between cortical and subcortical myoclonus. However, only one study investigated this difference in myoclonus with equivocal results (86).

As a note of caution, our proposal is incomplete due to limited data to perform sound statistical analysis. The current proposed classification primarily focuses on electroencephalogram, overlooking the semiology and location of the myoclonus. We planned to conduct subgroup analyses based on myoclonus semiology; however, the limited available data hindered our ability to proceed with this analysis. Myoclonus post-CA may exhibit variable and dynamic semiological patterns, both within the same patient and over time, due to medications or spontaneous changes. Therefore, it is inherently less definable without electroencephalogram.

Our proposal is not derived from a prospective validation study and forms just the core or a starting point of an outcome-orientated classification, which should be extended and completed as new evidence emerges. The underlying coding system provides easy and well-structured opportunities for this, for example, M1 (G, 1 Hz) would denote generalized myoclonus occurring with 1 Hz, whereas M1 (S, 0.5 Hz) denotes segmental myoclonus at 0.5 Hz. Regarding multimodality, the research proposal can be further extended, for example, by evoked potentials (EPs). For instance, M1-B1-SE0-EP1 denotes EPs being present in a patient with myoclonus, continuous background and no status epilepticus (SE).

In our research proposal, a pattern is defined as M1B0 in a patient with suppressed background, whereas the relevance of the frequency (Hz) of discharges needs to be investigated. The question whether continuity of background or SE is of higher importance emerged. Especially the situation when jerks are time-locked to bursts requires clarification as to whether the suppression burst pattern, the SE, or the combination of both has the strongest impact on outcome. We acknowledge that the background is not always clear and may not be visible due to high spike frequency. In this case the pattern should be considered M1B1-SE1 to still allow for the possibility of treatment.

We aim to prevent patients from being falsely classified with bad outcome due to insufficient testing whereas we tolerate patients being falsely classified as having a potentially good outcome. For instance, negative testing of reactivity might result from underlying sedation. Furthermore, testing for responsiveness is associated with a high inter-rater variability even within one center (136).

We do not convey that myoclonus does not matter from a prognostic standpoint in relation to electroencephalogram. We have used myoclonus as the starting point of our systematic review and found that extensive research is warranted to explore the role of the electroencephalogram in the management of patients with post-hypoxic myoclonus.

CONCLUSIONS

Our systematic review reveals substantial heterogeneity in the definitions, associated electroencephalogram features and outcomes of less than or equal to 72 hours myoclonus post-CA.

While heterogeneity in definitions leads to inconsistent and discrepant associations with outcomes, homogeneous electroencephalogram correlate and background perform much better. We therefore propose to stratify myoclonus together with its specific electroencephalogram patterns, adopting the following categories: M1-B0, M1-B1-SE1, and M1-B1-SE0. This proposal does not aim to change current clinical practice or guidelines. However, it offers an evidence-based research approach facilitating the creation of study designs, in which further research is highly warranted.

Supplementary Material

ccm-53-e410-s001.docx ^{(111.9KB, docx)}