Association of Clinically Evident Eye Movement Abnormalities With Motor and Cognitive Features in Patients With Motor Neuron Disorders

Abstract

Background and objectives: Although oculomotor abnormalities (OMAs) are not usually considered prominent features of amyotrophic lateral sclerosis (ALS), they may represent potential clinical markers of neurodegeneration, especially when investigated together with cognitive and behavioral alterations. The aim of our study was to identify patterns of clinically evident OMAs in patients with ALS and to correlate such findings with cognitive-behavioral data.

Methods: Three consecutive inpatient cohorts of Italian patients with ALS and controls were retrospectively evaluated to assess the frequency of OMAs and cognitive-behavioral alterations. The ALS population was divided into a discovery cohort and a replication cohort. Controls included a cohort of cognitively impaired individuals and patients with Alzheimer disease (AD). Participants underwent bedside eye movement evaluation to determine the presence and pattern of OMAs. Cognitive assessment was performed with a standard neuropsychological battery (discovery ALS cohort and AD cohort) and the Italian Edinburgh Cognitive and Behavioural ALS Screen (ECAS) (replication ALS cohort).

Results: We recruited 864 individuals with ALS (635 discovery, 229 replication), 798 who were cognitively unimpaired and 171 with AD. OMAs were detected in 10.5% of our ALS cohort vs 1.6% of cognitively unimpaired controls (p = 1.2 × 10^-14) and 11.4% of patients with AD (p = NS). The most frequent deficits were smooth pursuit and saccadic abnormalities. OMA frequency was higher in patients with bulbar onset, prominent upper motor neuron signs, and advanced disease stages. Cognitive dysfunction was significantly more frequent in patients with OMAs in both ALS cohorts (p = 1.1 × 10^-25). Furthermore, OMAs significantly correlated with the severity of cognitive impairment and with pathologic scores at the ECAS ALS-specific domains. Last, OMAs could be observed in 35.0% of cognitively impaired patients with ALS vs 11.4% of patients with AD (p = 6.4 × 10^-7), suggesting a possible involvement of frontal oculomotor areas in ALS.

Conclusion: Patients with ALS showed a range of clinically evident OMAs, and these alterations were significantly correlated with cognitive, but not behavioral, changes. OMAs may be a marker of neurodegeneration, and bedside assessment represents a rapid, highly specific tool for detecting cognitive impairment in ALS.

Figure 1. OMAs Distribution According to Presence of Cognitive or Behavioral Alterations in the Replication Cohort (A) and ROC Curve for OMAs According to the Revised Strong Criteria (B)

Black bars indicate patients with ocular movement abnormalities (OMAs); gray bars indicate patients with normal ocular movements. Bonferroni-adjusted p values for pairwise comparisons between different Edinburgh Cognitive and Behavioural ALS Screen categories: amyotrophic lateral sclerosis (ALS) vs ALS with cognitive impairment (ALSci), ALS vs ALS with frontotemporal dementia (ALS-FTD), ALS with behavioral impairment (ALSbi) vs ALS-FTD (p < 0.0001); ALS vs ALS with cognitive and behavioral impairment (ALScbi) (p = 0.049); ALSbi vs ALSci (p = 0.032); ALScbi vs ALS-FTD (p = 0.013); ALSci vs ALS-FTD (p = 0.004); and ALS vs ALSbi, ALSbi vs ALScbi, and ALScbi vs ALSci (p = NS). Area under the receiver operating characteristic curve is 0.92 (95% confidence interval 0.89–0.96).

Figure 2. Association of Oculomotor Abnormalities With Cognitive-Behavioral Alterations and Site of Onset in Patients With ALS From the Replication Cohort

Cerebral areas most prominently involved in the different cognitive functions investigated by neuropsychological evaluation were language (green, A), verbal fluency and executive functions (orange, A), memory (yellow, C), and visuospatial functions (blue, A), as well as the primary motor cortex (light blue strip, A and B). (A) Lateral aspect of brain. (B) Medial aspect. (C) Coronal section at the level of the hippocampi. Verbal fluency and executive functions are considered together from a topographic standpoint because they are both localized in the dorsolateral prefrontal cortex (DLPFC). In addition, the orbitomedial prefrontal cortex is highlighted in purple as the most relevant cortical area for behavior. Within each cortical area and corresponding cognitive domain, the percentage frequencies of oculomotor abnormalities among patients with and without impairments, respectively, in the relative domains are reported (black). Within the orange area, the percentages written above pertain to verbal fluency, while those written below pertain to executive functions. For behavioral alterations, the behavior screen scores of patients with vs without oculomotor abnormalities are reported. Comparisons resulting in statistically significant differences are in bold, and the p value is shown. In the primary motor cortex, the percentage frequencies of oculomotor abnormalities in patients with disease onset in different body segments are reported in dark blue, according to the somatotopic representation of motor control within the motor strip (bulbar region and upper limbs: lower and upper parts of the motor strip in the lateral aspect of the brain, respectively, A; lower limbs: part of the motor strip lying in the upper part of the medial aspect of the hemisphere, B). In addition to the DLPFC (delimited by the continuous red line), the frontal eye field (FEF; located within the DLPFC itself) and the supplemental eye field (SEF) are depicted (areas within the dashed red line and within the pointed red line, respectively). Image created with BioRender.