key: cord-0428418-70kltev7
authors: Lugg, A.; Schindle, M.; Sivak, A.; Tankisi, H.; Jones, K. E.
title: Nerve excitability as a biomarker for amyotrophic lateral sclerosis: a systematic review and meta-analysis.
date: 2022-02-13
journal: nan
DOI: 10.1101/2022.02.11.22270866
sha: b904447bfc735b65046c8cff42efa22c5524406f
doc_id: 428418
cord_uid: 70kltev7

Objective: To determine which nerve excitability outcome measures are potential biomarkers for ALS via systematic review and meta-analysis. Methods: Potential studies were identified from the following databases: MEDLINE, PubMed Central, CINAHL Plus, EMBASE, HealthSTAR, Scopus, and Web of Science up to March 2020. Only studies performed in human participants and assessing median motor axons were included. Forest Plot analyses using a random-effects model to determine pooled effect (Z-score), heterogeneity (I2) and Cohen's d were used to identify potential biomarkers. Results: From 2866 studies, 26 (patients=942, controls=719) were used in the systematic review and 23 in the meta-analysis. Seven nerve excitability indices met the three criteria for: significant Z-score, heterogeneity I2<40% and Cohen's d >0.2 (in descending rank order): TEd 90-100 ms, strength-duration time constant (SDTC), superexcitability, TEd 40-60 ms, resting I/V slope, 50% depolarizing, and subexcitability. A sensitivity analysis restricted to patients with *early* ALS indicated that four indices are potential early biomarkers of ALS (Z ranging from 2.99 to 2.16, in descending rank order): TEd 10-20 ms, TEd 90-100 ms, superexcitability, and SDTC. Conclusion: Seven excitability indices differentiate ALS patients from healthy controls, four of which may serve as early biomarkers for ALS. The candidate biomarkers may be used to monitor disease progression, predict survivability, and measure treatment response in clinical trials. High quality diagnostic test accuracy studies are warranted to firmly establish the utility of these indices in individuals suspected of an ALS diagnosis.

Amyotrophic lateral sclerosis (ALS) is a fatal neurodegenerative disorder characterized by progressive degeneration of upper motor neurons (UMNs) and lower motor neurons (LMNs) with an incidence of 1-2 per 100,000 and median survival of 3-5 years from symptom onset. [1] [2] [3] [4] [5] [6] The clinical features of ALS are highly variable depending on the extent of UMN and LMN involvement, location of disease onset, spread, and presence of extra-motor symptoms. 5, 7 The combined heterogeneity of ALS and the absence of established biomarkers renders a difficult diagnosis and limits individual patient prognosis, disease surveillance and therapeutic development.

Currently, no definitive diagnostic test for ALS exists; diagnosis is made if an individual meets the revised El Escorial, Awaji or the recent Gold Coast criteria: a set of clinical signs augmented by electromyography. 8, 9 Disease heterogeneity and lack of biomarkers contributes to misdiagnosis of ALS and diagnostic delay. [10] [11] [12] Diagnostic delay is a barrier to early intervention with disease-modifying drugs that may be most effective in early ALS. [13] [14] [15] [16] Therefore, there is a need to improve accuracy and speed of ALS diagnosis. Identifying prognostic biomarkers to monitor disease progression and treatment response is also important. The rate of disease progression and survival of ALS patients is highly variable, limiting the ability to predict functional decline and survival. 7, 17 Heterogeneity has implications for efficient development of novel therapeutics in ALS, with different ALS phenotypes having variable therapeutic responses. Identifying biomarkers sensitive to disease activity may permit stratified enrollment for improved efficiency in clinical trials.

Several candidate biomarkers have been identified through genetic testing, biological fluid analysis, neuroimaging and neurophysiology studies. 16, 18, 19 Genetic tests provide definitive diagnosis and pathogenetic mechanisms in approximately 10% of ALS cases but have limited utility for monitoring treatment response. 7, 20 Investigations for UMN pathology include neurofilaments, magnetic resonance imaging (MRI) and transcranial magnetic stimulation (TMS). Elevated neurofilaments in cerebrospinal fluid and plasma have diagnostic, prognostic and monitoring potential in ALS but this is not unique to ALS pathophysiology. 21, 22 MRI studies have corroborated the pathological elements of neurodegeneration that occur in ALS: degeneration of the motor cortex and corticospinal tract, as well as extramotor (frontotemporal) regions, with the extent of degeneration in these regions associated with shorter survival. [23] [24] [25] [26] [27] [28] TMS studies have identified cortical dysfunction as an important early and specific biomarker in ALS. 16, [29] [30] [31] [32] Recent developments for diagnostic tests of LMN pathology include fasciculations and ultrasound. 33, 34 The recent publication of consensus guidelines for nerve excitability has affirmed its use for quantifying LMN pathology. 35 Nerve excitability techniques assess the biophysical properties of the axonal membrane and ion channel function. 7, 36, 37 Nerve excitability testing is completed using a harmonized protocol, called TROND, and specialized software (QTrac, UCL Institute of Neurology, London, UK, available from Digitimer Ltd at www.Digitimer.com). 38 A test yields approximately 30 different excitability indices distributed across five subtests: stimulus-response (SR) curve, strength-duration properties, recovery cycle (RC), threshold electrotonus (TE), and current-voltage (I/V) relationship. 35, 39 ALS was the first pathophysiology for which the nerve excitability test was employed and studies have consistently identified pathological changes in sodium (Na + ) and potassium (K + ) channels. [40] [41] [42] [43] [44] [45] [46] However, a recent study suggests that approximately 99% of the differences between axons affected by ALS and healthy controls (HCs) could be attributed to a non-selective reduction in the expression of all ion channels, not exclusively Na + and K + channels. 47 Given that there are approximately 30 measures generated from a single excitability test, it is important to first determine which excitability indices distinguish axons of ALS patients from HCs and therefore have LMN biomarker potential.

Currently, only narrative reviews and qualitative syntheses comparing nerve excitability in ALS and controls are available. Therefore, we conducted a systematic review and meta-analysis of the existing studies to determine: which excitability indices distinguish ALS patients from HCs, the reliability . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) 

Patient engagement interviews were conducted with two individuals living with ALS prior to conducting this review. The purpose was to ground the research in the community and determine whether this review would be meaningful for individuals living with ALS. These individuals were consulted at each stage throughout the review to maintain patient engagement.

Studies comparing median motor (abductor pollicis brevis (APB) muscle) nerve excitability in ALS patients and HCs were considered for inclusion. Reporting of all excitability measures was not required for inclusion in the meta-analysis. Narrative reviews, theoretical or modelling studies, and animal studies were excluded. Studies not published in English were excluded. Peer-reviewed articles and published dissertations were included; abstracts were excluded.

Participants . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted February 13, 2022. ; https://doi.org/10.1101/2022.02.11.22270866 doi: medRxiv preprint Participants were individuals diagnosed with ALS according to the El Escorial 50 , revised El Escorial 8 , or the Awaji Criteria 51 and HCs. ALS patients must meet the classification for definite, probable, probable laboratory-supported, or possible ALS. 8, 50, 51 Both familial ALS (fALS) and sporadic ALS (sALS) were included. All ALS phenotypes were included (i.e., bulbar-onset versus limb-onset). Screening of the HCs was evaluated as part of the formal quality assessment.

The index test is the nerve excitability test, administered according to the standardized TROND protocol. 35 Studies using a TROND protocol developed prior to the 1999 protocol were evaluated for inclusion on a case-by-case basis. 37

The reference standard was the El Escorial 50 , revised El Escorial 8 , or the Awaji Criteria 51 .

A search was executed by an expert searcher/health librarian (AS) on the following databases:

OVID MEDLINE, PubMed Central, EBSCO CINAHL, OVID EMBASE, OVID HealthSTAR, Scopus, Web of Science -All Databases, and EBSCO SPORTDiscus. See Appendix A for the strategy used to search MEDLINE. The searches were completed using controlled vocabulary (e.g., MeSH) and key words representing variations of the concepts "amyotrophic lateral sclerosis" and "nerve/axonal excitability."

All searches were performed on March 12, 2020. Detailed search strategies for each database are available in the online supplemental tables 1-7.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted February 13, 2022. ; https://doi.org/10.1101/2022.02.11.22270866 doi: medRxiv preprint

Studies were screened using Covidence systematic review software (Veritas Health Innovation, Melbourne, Australia. Available at www.covidence.org). Search results (titles) were exported from the databases directly to Covidence (AS). Duplicate studies were removed when importing into Covidence.

Any additional duplicates found were removed by reviewers during screening. Two reviewers (AL and MS) independently screened titles and abstracts of all identified studies, with disagreements resolved by a third reviewer (KJ). The included citations moved to full-text review. Full-text articles were uploaded to Covidence; two reviewers (AL and KJ) independently assessed the articles, with conflicts resolved by a third reviewer (HT) to determine the final selection of articles.

A data extraction form was developed within Covidence. Two reviewers (AL and KJ) independently extracted the following variables from the included articles: author information, publication year, country of primary correspondence, study design, reference standard, eligibility criteria, number of participants, mean (or median) age of participants, sex ratio of participants, mean (or median) disease duration, medications, and stratification of patients. Data was also extracted for all reported nerve excitability indices. One reviewer (AL) checked the data entered in the form for errors or discrepancies and resolved such conflicts.

Methodological quality was assessed using a form created within Covidence. Two independent reviewers (HT and KJ) assessed the methodological quality of included studies using the Quality Assessment of Diagnostic Accuracy Studies (QUADAS-2) tool. 52 QUADAS-2 involves a structured assessment using signaling questions in four domains: (1) patient selection, (2) index test, (3) reference . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted February 13, 2022. ; https://doi.org/10.1101/2022.02.11.22270866 doi: medRxiv preprint standard, and (4) flow (of patients through the study) and timing (of the index test and reference standard). Each domain is assessed in terms of the risk of bias, and the first three domains also include an assessment of applicability concerns. Review-specific modifications to the QUADAS-2 tool were made; a domain for evaluating case-control studies based on the Newcastle-Ottawa Scale (NOS) was added (modified tool available in online supplemental Table 8 ). 53

Data was exported from Covidence to Review Manager, Version 5.4 (RevMan 5.4) software (available at https://training.cochrane.org/online-learning/core-software-cochrane-reviews/revman).

We then performed a meta-analysis to obtain a quantitative evaluation of nerve excitability in ALS patients and HCs for all sufficiently reported (n≥4) excitability indices. The results for each measure were represented with forest plots. A sensitivity analysis of excitability measures reported in ALS patients with preserved compound muscle action potential (CMAP) was performed to investigate potential early indicators. All statistical analyses were performed using the RevMan 5.4 software. We applied a random-effects model to reduce statistical heterogeneity in the extracted data and to obtain the mean of a distribution of true effects. 54 The mean difference and 95% confidence intervals were computed for each excitability measure. Statistical significance of the effect was determined using the Z-test for overall effect. The results were regarded as statistically significant if p<0.05. Cohen's d was calculated for each excitability measure to determine standardized differences in the means of ALS and HCs.

Indices with a Cohen's d below 0.2 have negligible effect size and were excluded as potential biomarkers. Heterogeneity was examined by applying the I 2 statistic. 55 I 2 values of 25%, 50%, and 75% correspond to low, moderate, and high heterogeneity between studies. 56 Indices with heterogeneity below 40% were considered acceptable. 57 Pooled means and standard deviations for each of the nerve excitability indices were calculated for ALS patients and HCs. Mean ± SD were calculated for sample size . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted February 13, 2022. ; https://doi.org/10.1101/2022.02.11.22270866 doi: medRxiv preprint and sex distribution (%male) across studies. Pooled means and standard deviations were calculated for participant age, disease duration, and revised ALS Functional Rating Scale (ALSFRS-R).

A total of 4867 articles were retrieved from the search: OVID MEDLINE (1065), PubMed Central (10). Results were exported from the databases to Covidence; 2001 duplicates were removed. We screened the title and abstract of 2866 articles. Following screening, the full-text eligibility of 43 articles was assessed. After full-text review, we selected 26 articles fitting the criteria for the systematic review, 23 of which were included in the meta-analysis. The procedure for study selection and reasons for exclusion are summarized in Figure 1 .

Twenty-six articles consisting of a total of 942 ALS patients and 719 HCs were included in the systematic review. Studies were published in the last 25 years, from 1996 to 2020. The included studies were conducted in five countries: Australia (n=14), Japan (n=9), Germany (n=1), Korea (n=1), and Portugal (n=1). All studies used either the El Escorial (n=5), revised El Escorial (n=11), or Awaji (n=5) criteria as the reference standard. Three studies used both the revised El Escorial and Awaji criteria. One study used a combination of genetic testing (C9orf72) and Awaji criteria. One study did not report the specific reference standard used. Study characteristics are summarized in online supplemental Table 9 .

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted February 13, 2022. ; https://doi.org/10.1101/2022.02.11.22270866 doi: medRxiv preprint

The range of sample sizes included in the studies was broad, as reported in Table 1 . The pooled mean age of the ALS patients was older than the HCs with a modestly larger proportion of males in the ALS group (Table 1 ). Ten studies reported average disease duration (17.7 months, range, 12.5 to 25.9) while seven studies reported median disease duration (ranging from 9.50 to 16.1 months). The pooled mean ALSFRS-R was 40.5 (8 studies, range, 39.0 to 42.1). A summary of the participant characteristics is provided in Table 1 (more detail in online supplemental Table 9 ). 17 .7(14.9) 12.5-25.9 10 n/a n/a n/a Mean ALSFRS-R 40.5(5.00) 39.0-42.1 8 n/a n/a n/a

Sixteen out of 30 nerve excitability indices were reported by four or more studies and were analyzed. Ten indices showed significant pooled effect (Z ranging from 9.88 to 2.81, in descending rank order): TEd 90-100 ms, strength-duration time constant (SDTC), superexcitability, CMAP, TEd 40-60 ms, TEd 10-20 ms, resting I/V slope, 50% depolarizing, subexcitability, and rheobase. CMAP (I 2 = 88%) and

TEd 10-20ms (I 2 = 51%) were above the cut-off of 40% heterogeneity. Rheobase was below the small effect size cut-off based on Cohen's d. Therefore, seven excitability indices are being proposed as . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted February 13, 2022. candidate biomarkers. Table 2 gives a summary of the number of studies reporting a measure, mean difference between ALS and HCs, heterogeneity, overall effect and p-value, and Cohen's d. Additional forest plots for other excitability measure are available in online supplemental figures 1-12. Pooled means(SD) for nerve excitability indices are reported in online supplemental Table 10 .

Not all indices generated by an nerve excitability test were fully reported; a bias was to report indices that demonstrated significant differences between ALS patients and HCs. Two out of twelve (17%) measures with significant pooled effect were reported in fewer than four studies, and nine out of fifteen (60%) measures that did not have significant pooled effect were reported in fewer than four studies. Measures that were not different between ALS and HCs were often reported without raw data and could not be included in the analysis. 29, 30, 41, 46, 58, 60, 61, 65, 67, 68, 70 The sensitivity analysis showed no difference in hyperpolarizing I/V slope.

QUADAS-2 results for individual articles included in the systematic review (n=26) are displayed in online supplemental Table 11 ; a summary of the author's judgements across all included articles is . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted February 13, 2022. ; https://doi.org/10.1101/2022.02.11.22270866 doi: medRxiv preprint shown in online supplemental Figure 13 . Most studies (n=21) were judged to be high risk of bias in the patient selection domain due to the use of non-consecutive sampling and case-control design. The remaining studies were assessed as unclear (n=3) or low (n=2) risk of bias. Case-control is an expected study design early in the career of a medical test; therefore, an additional domain was added to the QUADAS-2 tool to assess case-control design based on the NOS criteria. The risk of bias assessment was as follows: low (n=9), unclear (n=14), or high (n=3). The unclear and high risk of bias assessments were due to inadequate reporting of healthy control selection, screening and exclusion criteria. The majority of studies (n=24) scored low for index test risk of bias; two studies had unclear risk of bias. Most studies (n=23) were judged as low risk of bias in the reference standard domain. The remaining studies had unclear (n=1) or high (n=2) risk of bias. Most studies (n=16) had low risk of bias in the flow and timing domain, while eight and two studies had unclear and high risk of bias, respectively. Seventeen studies scored low for applicability concerns related to patient selection. Eight studies were judged as unclear primarily due to long disease durations. A single study scored high for applicability concerns due to the use of axonal excitability testing on single axons; this study was excluded from the meta-analysis. Most studies (n=21) were judged as low for applicability concerns in the index test domain. The remaining studies had unclear (n=3) or high (n=2) applicability concerns. The majority of studies (n=24) had low applicability concerns for the reference standard.

This systematic review and meta-analysis demonstrated that there are seven candidate biomarkers of LMN pathology in ALS and that four of these are present at earlier stages of the disease.

The candidate biomarkers come from different subtests of the TROND nerve excitability protocol [37, 40] and include: TEd 90-100ms, SDTC, superexcitability, TEd 40-60ms, resting I/V slope, 50% depolarizing current, and subexcitability.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted February 13, 2022. ; https://doi.org/10.1101/2022.02.11.22270866 doi: medRxiv preprint

A decrement in CMAP is well established in ALS patients with considerable axonal degeneration. 47, 73, 74 Therefore, it is unsurprising that the meta-analysis demonstrated that CMAP is significantly lower in ALS patients. However, a reduction in CMAP requires substantial axonal loss before becoming clinically apparent due to collateral sprouting and reinnervation. 74 While some studies have proposed a reduction in CMAP as a biomarker for ALS, the results were highly variable across studies and this change is not specific to ALS. A decrement in CMAP is observed in other motor neuron disorders, such as PMA and Hirayama disease, as well as carpal tunnel syndrome. 39, 75 Given that reduced CMAP is not apparent until later in the disease stage and is not specific for ALS, its utility as a biomarker is limited.

Studies have consistently identified a prolonged SDTC in ALS, a finding attributed to upregulation of persistent Na + conductance. 29, 41, [43] [44] [45] [46] [59] [60] [61] [62] 65, 67, 71, 72 Prolonged SDTC has been shown to be a strong, independent predictor of rapid functional decline and shorter survival in ALS patients. 76, 77 Further, SDTC may function as an important prognostic biomarker in ALS. In contrast, the utility of rheobase as a biomarker is limited. Despite having significant pooled effect, the mean difference in rheobase was small and demonstrated 99% overlap between ALS and controls. Therefore, rheobase is not a strong biomarker for ALS.

In ALS patients, the threshold for action potential generation is lower during the superexcitable period compared to HCs. Similar to SDTC, longitudinal changes in superexcitability are associated with shorter survival in ALS patients. 59, 76 The change of the subexcitable phase in ALS may arise from a . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted February 13, 2022. ; https://doi.org/10.1101/2022.02.11.22270866 doi: medRxiv preprint combination of passive electrotonic and active components underlying the depolarizing afterpotential. 78, 79 Threshold Electrotonus (TE)

Studies have consistently identified greater threshold changes in depolarizing TE of ALS patients compared to HCs. In contrast, threshold changes in hyperpolarizing TE are similar for ALS patients and HCs. Three measures of TEd were significantly different between ALS and HCs: TEd 10-20ms, TEd 40-60ms, and TEd 90-100ms. However, TEd 10-20ms had marked heterogeneity across studies, and may therefore be limited as a biomarker for ALS. These changes in TEd have been shown to become more pronounced upon follow-up, and reflect a reduction in K + channel conductance, particularly slow K + conductance. 59, 80 Further, threshold changes in TEd may serve as important diagnostic and prognostic indicators, specifically at 90-100ms, which had the largest pooled effect amongst all indices analyzed 59, 76 .

Resting I/V slope is significantly reduced in ALS patients compared to HCs, suggesting a reduction in conductance, and therefore, ion channels in the axonal membrane. This finding is consistent with the previous interpretation of reduced ion channel conductance in ALS. 47 Additionally, there is a greater threshold reduction in response to 50% depolarizing current in ALS, consistent with the observation in TEd, suggesting reduced K + conductance to oppose the depolarizing conditioning pulse.

The sensitivity analysis on patients with early-stage disease (based on CMAP) identified four of the seven indices: TEd 10-20ms, TEd 90-100ms, superexcitability, and SDTC. This finding is consistent with findings of previous studies comparing early ALS (preserved CMAP) and late ALS (decrement in . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted February 13, 2022. ; https://doi.org/10.1101/2022.02.11.22270866 doi: medRxiv preprint CMAP). 43, 45 These early indicators may aid diagnosis of ALS as the abnormalities are observed prior to a reduction in CMAP, which requires substantial axonal loss before becoming clinically apparent. 47, 74, 81 These measures may also be useful for monitoring disease progression and response to treatment, especially if early intervention is achieved. As the disease advances, it appears that ALS patients may experience progressive abnormalities in axonal excitability, demonstrated by an increase in the number of excitability measures that differentiate ALS patients from HCs. This inference is consistent with findings that axonal dysfunction is more prominent in patients with a greater reduction of CMAP. 45 However, longitudinal excitability studies including follow-up from the time of early disease stage are necessary to test this hypothesis.

Disturbances in membrane excitability and axonal ion channel function have been identified since the earliest application of axonal excitability techniques in ALS. 40 Specifically, excitability abnormalities in ALS have been attributed to increased Na + and decreased K + conductances. 45, 46, 59, 80 Upregulation of persistent Na + currents produces a drive towards depolarization. This depolarizing drive is further augmented by a reduction in the hyperpolarizing slow and fast K + currents, resulting in membrane hyperexcitability. 46 Axonal excitability studies have consistently identified prolonged SDTC in ALS, suggesting an upregulation of nodal persistent Na + currents. 43, 45, 46, 59, 63 This increase in SDTC has been associated with axonal degeneration and provides evidence for the pathological role of persistent Na + channels in ALS. [40] [41] [42] [43] 46, 59, 67 However, reduced K + conductance has also been suggested to lengthen SDTC. 43 Several studies have found evidence for a reduction of K + channels, as demonstrated by the abnormalities in superexcitability, subexcitability, depolarizing TE, and threshold reduction to 50% depolarizing current. 43, 45, 46, 63 Abnormalities in Na + and K + channel function may reflect several of the axonal excitability changes observed in ALS patients. However, a recent study suggests that altering the function of persistent Na + channels and slow and fast K + channels alone cannot explain the observed . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted February 13, 2022. ; https://doi.org/10.1101/2022.02.11.22270866 doi: medRxiv preprint changes in ALS. 47 Rather, the changes in ALS may be best attributed to a non-selective reduction in all axonal ion channels caused by a lack of supply of axonal proteins. This study suggests that the axonal dysfunction occurring in ALS may be caused by a protein homeostasis issue, resulting in abnormal membrane excitability. This finding is supported by the reduction in resting I/V slope found in this metaanalysis, suggesting reduced axonal membrane conductance, specifically K + channels open at rest.

Although the exact mechanisms by which axonal excitability changes are involved in neurodegeneration in ALS remain to be elucidated, the findings of this meta-analysis substantiate the circulating thought in narrative reviews that ALS patients experience abnormalities in ion channel conductance and resulting axonal excitability. 35, 82, 83 Further, this meta-analysis provides evidence beyond a narrative review for the use of specific nerve excitability measures as LMN biomarkers for ALS.

There are limitations to this systematic review and meta-analysis. First, the literature is saturated with case-control studies comparing axonal dysfunction in diagnosed ALS patients and HCs; the inclusion of disease controls is lacking. Therefore, while some excitability measures clearly differentiate ALS patients from HCs, due to the absence of disease controls we are unable to determine whether these measures are specific to ALS, or are representative of conditions affecting axonal health.

Future diagnostic test accuracy studies investigating axonal excitability changes in ALS should include comparison to disease controls and calculate measures of sensitivity and specificity. Second, not all measures generated during an axonal excitability test were reported equally across studies; a bias was to report indices that demonstrated significant differences. This limitation may be addressed in the future by encouraging sharing of raw data and conduction of mega-analyses. Additionally, some patients were reported more than once in studies by the same authors. We were unable to separate these repeated patients from the results. However, the number of repeated patients was minimal, and we do . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted February 13, 2022. ; https://doi.org/10.1101/2022.02.11.22270866 doi: medRxiv preprint not suspect this skewed the results. Lastly, the data included in the meta-analysis was cross-sectional;

we are unable to determine longitudinal changes in excitability. Future studies investigating axonal excitability longitudinally are warranted to fully elucidate the prognostic potential of the indices.

The potential for biomarkers to transform ALS diagnosis, prognostic stratification, and future therapeutic development has made the hunt for biomarkers a priority in ALS research. The seven candidate biomarkers derived from axonal excitability tests identified in this study may be useful 

None declared.

Not commissioned; externally peer reviewed.

All data used in this meta-analysis are available either in the main article or in supplementary materials available on Figshare https://figshare.com/s/76a237bcc221604f8a2f.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted February 13, 2022. ; https://doi.org/10.1101/2022.02.11.22270866 doi: medRxiv preprint . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted February 13, 2022. ; https://doi.org/10.1101/2022.02.11.22270866 doi: medRxiv preprint Figure 1 . Prisma flow chart of study selection and reasons for exclusion.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted February 13, 2022. ; https://doi.org/10.1101/2022.02.11.22270866 doi: medRxiv preprint Figure 2 . Forest plot of SDTC in ALS patients and HCs. Overall effect size was significantly different between ALS patients and HCs. Diamond represents overall effect size.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted February 13, 2022. ; https://doi.org/10.1101/2022.02.11.22270866 doi: medRxiv preprint Figure 3 . Forest plot of superexcitability in ALS patients and HCs. Overall effect size was significantly different between ALS patients and HCs. Diamond represents overall effect size.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted February 13, 2022. ; https://doi.org/10.1101/2022.02.11.22270866 doi: medRxiv preprint . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted February 13, 2022. ; https://doi.org/10.1101/2022.02.11.22270866 doi: medRxiv preprint Figure 5 . Forest plot of TEd 90-100ms in ALS patients and HCs. Overall effect size was significantly different between ALS patients and HCs. Diamond represents overall effect size.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted February 13, 2022. ; https://doi.org/10.1101/2022.02.11.22270866 doi: medRxiv preprint

Global epidemiology of amyotrophic lateral sclerosis: a systematic review of the published literature

Epidemiology of amyotrophic lateral sclerosis: a review of literature

Prognosis in amyotrophic lateral sclerosis: a population-based study

Amyotrophic lateral sclerosis. The Lancet

Amyotrophic lateral sclerosis

Amyotrophic lateral sclerosis

The phenotypic variability of amyotrophic lateral sclerosis

El Escorial revisited: revised criteria for the diagnosis of amyotrophic lateral sclerosis

Electrodiagnostic criteria for diagnosis of ALS

Predictors of delay in the diagnosis and clinical trial entry of amyotrophic lateral sclerosis patients: a population-based study

Factors affecting the diagnostic delay in amyotrophic lateral sclerosis

A proposal for new diagnostic criteria for ALS

An analysis of extended survival in patients with amyotrophic lateral sclerosis treated with riluzole

Riluzole disease stage and survival in ALS

An outcome study of riluzole in amyotrophic lateral sclerosis

Biomarkers in amyotrophic lateral sclerosis

Prognosis for patients with amyotrophic lateral sclerosis: development and validation of a personalised prediction model

ALS biomarkers for therapy development: state of the field and future directions

A systematic review of suggested molecular strata, biomarkers and their tissue sources in ALS

Genetic Convergence Brings Clarity to the Enigmatic Red Line in ALS

Exosomal neurofilament light: A prognostic biomarker for remote symptoms after mild traumatic brain injury?

Plasma Neurofilament Light Chain as a Predictive Biomarker for Poststroke Cognitive Impairment: A Prospective Cohort Study

Combined structural and neurochemical evaluation of the corticospinal tract in amyotrophic lateral sclerosis

Cerebral degeneration predicts survival in amyotrophic lateral sclerosis

A voxel-based morphometry study of patterns of brain atrophy in ALS and ALS/FTLD

Hyperintensity of the corticospinal tract on FLAIR: a simple and sensitive objective upper motor neuron degeneration marker in clinically verified amyotrophic lateral sclerosis

Linear associations between clinically assessed upper motor neuron disease and diffusion tensor imaging metrics in amyotrophic lateral sclerosis

Evaluating the cerebral correlates of survival in amyotrophic lateral sclerosis

Cortical dysfunction underlies the development of the split-hand in amyotrophic lateral sclerosis

Pathophysiology and diagnosis of ALS: insights from advances in neurophysiological techniques

Neurophysiological biomarkers in amyotrophic lateral sclerosis

The rise and fall of fasciculations in amyotrophic lateral sclerosis

A model incorporating ultrasound to predict the probability of fast disease progression in amyotrophic lateral sclerosis

Measurement of axonal excitability: Consensus guidelines

Threshold tracking techniques in the study of human peripheral nerve

Excitability of human axons

Multiple measures of axonal excitability: a new approach in clinical testing

Nerve excitability: a clinical translation

Axonal ion channel dysfunction in amyotrophic lateral sclerosis

Axonal excitability properties in amyotrophic lateral sclerosis

Strength-duration properties of sensory and motor axons in amyotrophic lateral sclerosis

Altered axonal excitability properties in amyotrophic lateral sclerosis: impaired potassium channel function related to disease stage

Increased nodal persistent Na + currents in human neuropathy and motor neuron disease estimated by latent addition

Axonal dysfunction precedes motor neuronal death in amyotrophic lateral sclerosis

Axonal ion channel dysfunction in c9orf72 familial amyotrophic lateral sclerosis

In vivo evidence for reduced ion channel expression in motor axons of patients with amyotrophic lateral sclerosis

Preferred reporting items for a systematic review and meta-analysis of diagnostic test accuracy studies

Cochrane Handbook for Systematic Reviews of Diagnostic Test Accuracy. 10.0. The Cochrane Collaboration

Awaji diagnostic algorithm increases sensitivity of El Escorial criteria for ALS diagnosis

QUADAS-2: a revised tool for the quality assessment of diagnostic accuracy studies

The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomized studies in meta-analysis

A basic introduction to fixed-effect and random-effects models for meta-analysis

Quantifying heterogeneity in a meta-analysis

Measuring inconsistency in meta-analyses

Cochrane Handbook for Systematic Reviews of Interventions. 6.2. Cochrane

Axonal hyperpolarization in amyotrophic lateral sclerosis

Progressive axonal dysfunction and clinical impairment in amyotrophic lateral sclerosis

Riluzole exerts transient modulating effects on cortical and axonal hyperexcitability in ALS. Amyotrophic Lateral Sclerosis and Frontotemporal Degeneration

ALS pathophysiology: insights from the split-hand phenomenon

Ischemic resistance of cutaneous afferents and motor axons in patients with amyotrophic lateral sclerosis

Distal excitability changes in motor axons in amyotrophic lateral sclerosis

Distal motor axonal dysfunction in amyotrophic lateral sclerosis

Differences in nerve excitability properties between isolated bulbar palsy and bulbar-dominant amyotrophic lateral sclerosis

Split hand syndrome in amyotrophic lateral sclerosis: different excitability changes in the thenar and hypothenar motor axons

Upregulation of persistent sodium conductances in familial ALS

Novel threshold tracking techniques suggest that cortical hyperexcitability is an early feature of motor neuron disease

Fatigue and activity dependent changes in axonal excitability in amyotrophic lateral sclerosis

Riluzole exerts central and peripheral modulating effects in amyotrophic lateral sclerosis

Utility of recovery cycle with two conditioning pulses for detection of impaired axonal slow potassium current in ALS

Increased variability of axonal excitability in amyotrophic lateral sclerosis

Neurophysiological measures in amyotrophic lateral sclerosis: Markers of progression in clinical trials

Functional compensation in partially denervated muscles

Pattern difference of dissociated hand muscle atrophy in amyotrophic lateral sclerosis and variants

Motor axonal excitability properties are strong predictors for survival in amyotrophic lateral sclerosis

Increased motor axonal persistent sodium currents predict rapid functional declines in amyotrophic lateral sclerosis

Intracellular recording from vertebrate myelinated axons: mechanism of the depolarizing afterpotential

After-effects of near-threshold stimulation in single human motor axons

Acute Effects of Riluzole and Retigabine on Axonal Excitability in Patients With Amyotrophic Lateral Sclerosis: A Randomized, Double-Blind, Placebo-Controlled, Crossover Trial

Electrodiagnostic studies in amyotrophic lateral sclerosis and other motor neuron disorders

Axonal excitability in amyotrophic lateral sclerosis

The puzzling case of hyperexcitability in amyotrophic lateral sclerosis

We thank Dr. Meghan Sebastianski (Alberta SPOR Support Unit) for providing advice on the metaanalysis procedure. We would also like to thank the members of the ALS community for their input in shaping this review.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)