Cortical Excitability as a Prognostic and Phenotypic Stratification Biomarker in Amyotrophic Lateral Sclerosis

Abstract

Objective: Despite its clinical heterogeneity, amyotrophic lateral sclerosis is unified by early and prominent alterations in cortical excitability, increasingly recognized as contributors to disease progression. This study assessed whether the ratio between motor evoked potential (MEP) amplitude, reflecting upper motor neuron integrity, and compound muscle action potential (CMAP) amplitude, indexing lower motor neuron function, could provide an accessible marker of corticospinal excitability to stratify patients by phenotype, stage, and survival.

Methods: In this multicenter retrospective study, 743 amyotrophic lateral sclerosis patients from 16 tertiary centers in Italy were analyzed. The MEP:CMAP ratio, recorded from upper limb muscles, was categorized as hyperexcitable, normal, or hypoexcitable. Phenotypes included progressive muscular atrophy (or lower motor neuron), flail arm/leg, classic, bulbar, patient with predominant upper motor neuron signs (or pyramidal), and primary lateral sclerosis. Disease stage was assessed using King's staging. Survival was analyzed using Kaplan-Meier curves and Cox regression models.

Results: The MEP:CMAP ratio differed significantly across phenotypes (p < 0.0001), with hyperexcitability predominating in lower motor neuron, flail, classic, and bulbar forms, and hypoexcitability in pyramidal and primary lateral sclerosis. Hypoexcitability increased in advanced King's stages (p < 0.0001). Hyperexcitable patients had shorter survival (p = 0.004), including when tested within 1 year of onset (p = 0.006). Cox regression identified the MEP:CMAP ratio as an independent survival predictor (HR 1.84, 95% CI 1.12-3.03, p = 0.016).

Interpretation: This real-world study supports the clinical value of the MEP:CMAP ratio as a scalable biomarker of cortical excitability in amyotrophic lateral sclerosis, with prognostic relevance across phenotypes and disease stages. ANN NEUROL 2025;98:801-813.

FIGURE 1

Flowchart of patient selection from a large cohort who underwent motor evoked potential examination for diagnostic purposes. ALSFRS‐R = Amyotrophic Lateral Sclerosis Functional Rating Scale Revised; CMAP = compound muscle action potential; LL = lower limb; MEP = motor evoked potential; PUMNS = Penn upper motor neuron score; UL = upper limb.

FIGURE 2

Progression of central motor conduction time (CMCT) alterations in relation to affected body regions and upper limb involvement. (A) Distribution of central motor conduction time is shown across the number of affected body regions (1–3) and (B) according to upper limb regional involvement at King's staging, defined by any reduction in Amyotrophic Lateral Sclerosis Functional Rating Scale‐Revised handwriting (item 4) or cutting food and handling utensils (item 5A) scores. The data indicate that pathological findings become increasingly prevalent both with disease stages (ie, from King's stage 1–3) and with upper limb involvement, in terms of either prolonged CMCT or corticospinal inexcitability (pathological findings, χ² < 0.0001 for both conditions). N = normal CMCT; P = prolonged CMCT; NEC = non‐evocable cortical response (with preserved peripheral response); NEP = non‐evocable peripheral response. [Color figure can be viewed at www.annalsofneurology.org]

FIGURE 3

Cortical excitability spectrum across region spreading and phenotypes in amyotrophic lateral sclerosis (ALS). (A) the motor evoked potential/compound muscle action potential (MEP:CMAP) ratio values are shown across ALS phenotypes, ranging from lower motor neuron phenotypes (eg, lower motor neuron [LMN] and flail) to upper motor neuron phenotypes (eg, pyramidal and primary lateral sclerosis [PLS]). The data reveal a significant gradient, with higher MEP:CMAP ratios observed in LMN‐dominant phenotypes, and progressively lower ratios in upper motor neuron (UMN)‐dominant phenotypes (Kruskal–Wallis test, p < 0.001). This pattern underscores the increasing burden of corticospinal tract degeneration and the reduction in cortical excitability associated with phenotypes predominantly characterized by UMN dysfunction. (B) MEP:CMAP ratio values are presented across anatomical regions classified by King's staging (1–3). The data demonstrate a significant decline in MEP:CMAP ratios as the number of anatomical regions involved increases (Kruskal–Wallis test, p < 0.001). This decline reflects the cumulative axonal degeneration of corticospinal tracts and a corresponding reduction in cortical excitability, which becomes more pronounced as ALS advances. (C) The percentage distribution of cortical excitability states, based on normative data (hyperexcitable, normal, and hypoexcitable), is illustrated across ALS phenotypes. The results show that hyperexcitability is significantly more prevalent in LMN‐dominant phenotypes, such as LMN and flail, whereas hypoexcitability is more frequently observed in UMN‐dominant phenotypes, including pyramidal and PLS (χ² < 0.001). (D) the percentage distribution of cortical excitability states is shown across the number of body regions involved (1–3). The data indicate that hyperexcitability is most common in early stages (King's stage 1) but diminishes as the disease progresses. Conversely, hypoexcitability becomes increasingly frequent in later stages (King's Stage 3, χ² < 0.001). In (A,C), the symbols ^, #, § and ° denote statistically significant differences between the PLS and pyramidal groups compared with the LMN, flail, bulbar, and classic groups. In (B,D), *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001. [Color figure can be viewed at www.annalsofneurology.org]

FIGURE 4

Kaplan–Meier survival analysis of amyotrophic lateral sclerosis (ALS) patients stratified by cortical excitability. (A) Kaplan–Meier survival curves for the entire ALS cohort, stratified by cortical excitability status. Patients classified as hyperexcitable (blue curve), defined by a motor evoked potential/compound muscle action potential (MEP:CMAP) amplitude ratio exceeding the 75th percentile of normative values, showed significantly reduced survival compared with hypoexcitable patients (red curve), whose ratios were below the 25th percentile (p = 0.004). (B) A subgroup analysis of patients with disease duration ≤24 months at the time of the transcranial magnetic stimulation assessment (n = 335). Hyperexcitability remained significantly associated with decreased survival (p = 0.012). (C) The analysis was further restricted to patients with disease duration ≤12 months (n = 201), reaffirming the association between hyperexcitability and poorer survival outcomes (p = 0.006). Collectively, these findings underscore the prognostic relevance of cortical excitability, particularly in the early stages of ALS. [Color figure can be viewed at www.annalsofneurology.org]