Eye Movement Abnormalities in Amyotrophic Lateral Sclerosis in a Tunisian Cohort

ABSTRACT

Few studies have reported abnormal ocular movements in cases of amyotrophic lateral sclerosis (ALS) and their link with other disease features. Our study aimed to describe and analyse eye movement abnormalities in ALS patients. Specifically, we set out to investigate the correlation between non-motor signs and oculomotor impairment in order to understand the pathogenesis of the disease. All ALS patients seen from 2018 to 2020 in the department of Neurology of Razi hospital underwent the recording of saccadic eye movements. Results were compared with healthy controls. Sixty-two patients were included. Altered saccadic eye movements (72.6%) correlated with tongue atrophy and bladder dysfunction. The most common finding was altered smooth pursuit (56.5%), which showed correlation with bladder dysfunction and altered frontal assessment battery (FAB) scores. Prolonged latencies of horizontal saccades (34%) correlated with sensory and extrapyramidal signs. Our study is the first to examine the characteristics of eye movements in a large African cohort of ALS patients and to show correlations with extra-motor clinical signs. Our findings showed extra-motor cortex dysfunction in ALS with greater frequency of eye movement abnormalities in comparison with previous studies. Altered horizontal pursuit, the core abnormality, confirmed the extension of the neurodegenerative process to the frontal and prefrontal cortices. Prolonged horizontal saccade latencies reflect mainly the involvement of the parietal eye field. Anti-saccadic abnormalities were the least common finding and showed, paradoxically, no link with executive dysfunction.

Introduction

Amyotrophic lateral sclerosis (ALS) is a neurodegenerative disorder of upper motor neurons of the corticospinal tract and lower motor neurons in brainstem nuclei and the anterior horn of the spinal cord.¹ Classically, it is characterised by rapid worsening in motor function. Although oculomotor function, sphincter control and sensory tracts have been traditionally considered as spared in ALS, post-mortem studies have demonstrated that the pathology spreads to include astrocytes, microglia, macrophages, and potentially oligodendrocytes.² Extra-motor cortex involvement can explain the clinico-pathological overlap with certain neurodegenerative dementias, including fronto-temporal dementia (FTD) and progressive supranuclear palsy.^3–5 Prior studies have shown abnormal findings in eye movement recordings in ALS patients, such as altered fixation, saccades, and smooth pursuit eye movements. Ophthalmoplegia and nystagmus have also been reported.⁶ In our current study, we aim to describe oculomotor abnormalities in ALS patients and to investigate oculomotor impairment and its implications in understanding the affected oculomotor networks in ALS.

Materials and methods

A case-control observational study was conducted in the Department of Neurology of Razi hospital in Tunis, Tunisia, starting from January 2018 to October 2021. Patients who fulfilled the El Escorial revised criteria for ALS were considered eligible for the study.⁷ Patients with general co-morbidities and those under medications that may affect oculomotor function (diabetes mellitus, psychotropic drugs, etc.) were excluded.^8–10 Patients who lost their autonomy due to physical disability were also excluded. We analysed retrospectively clinical and neurophysiological data from patient files and a local database of patients with ALS. The presence of ALS was confirmed by history-taking, clinical examination reports and electromyographic studies. Only ALS patients who underwent the eye movement recordings were finally included. Subjects were categorised according to spinal or bulbar onset. The mean disease duration from the onset of symptoms to the time of the present study was recorded in months. The presence of non-motor symptoms was also reported (sensory symptoms, oculomotor dysfunction, bladder dysfunction, cognitive impairment, and dysautonomic symptoms). The patients were screened for cognitive impairment using the mini-mental state examination (MMSE).¹¹ Frontal dysfunction was evaluated using the frontal assessment battery (FAB) and the Edinburgh cognitive and behavioural ALS screen (ECAS).^11,12 The revised amyotrophic lateral sclerosis functional rating scale (ALSFRS-R) was assessed for all patients at the first visit at baseline.⁷ Respiratory deficits and requirements for non-invasive ventilation were reported. The data used in this study were made anonymous. Eye movement recordings and cognitive assessment were part of the routine management of our patients with ALS.

Subjects with no family or personal history of neurological diseases and/or psychiatric illness were included as healthy controls (HC). Their neurological examination and cognitive assessment were normal. They did not have hearing impairment and visual abnormalities, nor did they receive any medications that may affect oculomotor function. They willingly participated in the study and were matched in age and gender with the ALS patients.

Both ALS patients and HCs underwent video-oculographic recording. Eye movements were recorded monocularly, using infrared video-oculography and Ulmer’s program for video-nystagmography. The recording was performed in a quiet, dark room. All participants underwent a standardised evaluation by a single skilled examiner. The patient was seated with the head fixed, at a distance of 75 cm from a television screen (Figure 1). The eye movements were recorded in the horizontal plane to assess smooth pursuits, voluntary saccadic latencies and anti-saccadic tasks. All records were visually inspected for quality assurance. Smooth pursuits were studied using a white spot that moved horizontally with a constant velocity (less than 10° per second). To study saccades, we used the zero-gap method. Central and peripheral targets consisted of a white square of 4°. The central target went out when the peripheral one appeared. The latter appeared randomly at 20° to the right or to the left every 30 seconds. The subjects were trained to look at a central point and then at an eccentric target, which they then had to look at as quickly as possible.

Figure 1.

A 70-year-old amyotrophic lateral sclerosis patient undergoing the eye movement recording in our department.

Since normative values used to interpret the oculomotor findings are heterogeneous and age and population dependent,^13,14 we used our previously validated normative values specific to our Tunisian population and to our laboratory and device in order to avoid any potential biases. These values have been established based on a validation protocol that included 100 healthy controls with normal cognitive function. The group of healthy controls included in the current study represents a supplementary group matched by age and sex with our ALS patients to increase the relevance of the considered cut-offs and to check the significance of the abnormal findings. We measured the latencies of the voluntary saccades for all subjects. The cut-off value for the normal latency of voluntary horizontal saccades was 300 ms. This finding was consistent with previous studies where normal horizontal saccade latency ranged from 250 to 300 ms.^13,15,16 Errors in anti-saccadic tasks were calculated and were considered abnormal when the error rate exceeded 10% of the recorded trials. This cut-off percentage was also adopted by a previous study, which also found that the error rate is highest among children and decreases to around 10% by the age of 15 years.¹⁵ Regarding smooth pursuits, they were evaluated visually based on the accordance of its morphology with a reference line. Morphology of smooth pursuits is one among the validated criteria for their assessment, in addition to gain and left–right asymmetry.¹⁷

Statistical analysis was performed using Statistical Product and Service Solutions (version 22.0, SPSS Inc., Chicago, IL, USA). Quantitative variables were expressed as means and standard deviations if they were normally distributed and as medians and interquartile ranges (IQR) if there were not. Regarding the analytic study, the Pearson’s correlation coefficient and chi-squared test were applied to continuous and categorical variables, respectively, with p < .05 considered statistically significant.

Results

Descriptive analysis

Population of study

During the recruitment period, 62 out of the 300 ALS patients followed up in our department met the inclusion criteria and consented to participate in the study. Reasons for incomplete follow-up were: death; financial difficulties for patients living in rural regions and thus difficulty reaching our department, which is located in a university hospital in the capital; loss of autonomy and total disability during the advanced stages of the disease making the eye movement recording hard to achieve.

Socio-demographic and clinical characteristics of ALS patients

The study included 62 patients with 34 males and 28 females. Their mean age was 60.5 ± 11 years. More than half of our patients (59%) did not have any co-morbidities. Essential hypertension was the most common associated condition (22.6%). Dyslipidaemia was present in 3.2% of the cases. As for medication, none of our patients were treated with benzodiazepines or anti-psychotics that may have an impact on eye movement recording.^9,10

The mean age of the onset of ALS was 56.6 ± 11 years. The mean disease duration was 57 ± 60 months. The mean follow-up period was 46 ± 75 months. All patients fulfilled the El Escorial revised criteria for ALS. Consanguinity was present in 46.8% of cases. The mean diagnostic delay was 1.7 ± 3 years. The ALS-plus spectrum was identified among three of our patients: one case of FTD-ALS spectrum; a second case of ALS with corticobasal degeneration; and a third case of ALS associated with dystonia in a patient with a family history of confirmed DYT5 dystonia. Only one case of juvenile ALS was included, with an age at onset of 10 years. Spinal onset was found in 88.7% of cases and bulbar onset in 11.3%. Bulbar involvement, during the progression of the disease, was clinically evident in 45.2% of cases (tongue atrophy and fasciculations). Swallowing difficulties, dysphonia and respiratory difficulties were identified in 64.5% of cases. Non-motor signs were present in 46.8% of cases: sensory symptoms in 22.6%; bladder dysfunction in 17.7%; autonomic dysfunction in 9.7%; cognitive impairment in 9.7%; and extrapyramidal signs in 4.8%. However, oculomotor function was clinically preserved in all of our patients.

The median MMSE score was 26 points (IQR 6, range 18–30). The MMSE score was abnormal in 17 (27%) of our patients. Sixteen patients had mild dementia and only one had moderate dementia with an MMSE score of less than 19. In regards to executive function, the median FAB score was 15 points (IQR 7, range 4–18) and was abnormal in 50% of our patients. The ECAS evaluation showed a median score of 74 points (IQR 51.5, range 31–100).

The median ALSFRS-R score during the first consultation was 19 (IQR 17, range 10–46) and was correlated with the mean duration of the disease (p = .043). Spirometry was evaluated in 32 patients and showed a restrictive pattern in 46%, an obstructive pattern in 42%, and a normal pattern in 12% of patients. However, none of the patients needed non-invasive ventilation by the time the eye movements recording took place. Riluzole was used in 95.2% of the patients.

Eye movements findings

Abnormalities in eye movement recordings were found in 72.6% of ALS patients (Table 1). The study of horizontal voluntary saccades showed prolonged latencies in 34% of cases. The horizontal smooth pursuit was abnormal in 56.5% of our patients. Altered anti-saccades were observed in 24.2% of our patients. No oculomotor abnormality was clinically evident, however. Eighty-five HCs were included, matched for age and sex. The comparison between the different eye movement parameters (gain, saccades latencies, pursuit, anti-saccadic error rates) in both groups is presented in Table 2. Examples of normal eye movement recordings are illustrated in Figure 2.

Table 1.

Subjects’ demographics and clinical characteristics according to the eye movement recording results

	All ALS patients(n = 62)	Normal eye movement recording(n = 17)	Abnormal eye movement recording(n = 45)	p value
Gender (male: female)	34: 28	10: 7	24: 21	.69
Mean age ± SD (years)	60.5 ± 11	60 ± 10	61 ± 4	.318
Mean age of onset ± SD (years)	56.6 ± 11	54 ± 17	57.6 ± 9	.024
Mean duration of disease ± SD (months)	57 ± 60	54 ± 60	58 ± 54	.879
Disease onset (spinal: bulbar)	55: 7	16: 1	39: 6	.48
ALSFRS-R score	32 ± 9	33	31	.166
Riluzole intake	59 (95.2%)	16	43	.814

ALS = amyotrophic lateral sclerosis

ALSFRS-R = revised amyotrophic lateral sclerosis functional rating scale

SD = standard deviation

Table 2.

Comparison of the findings of eye movement recordings between amyotrophic lateral sclerosis patients and healthy controls

	ALS patients(n = 62)	Healthy controls(n = 85)	p value
Median age (years)	60.5 ± 11	45.27 ± 15	.111
Sex ratio (male: female)	1.21	0.83	.098
Percentage of abnormal eye movement recording	72.6%	12.8%	.000
Median saccadic latencies ± IQR (ms)	290 ± 37	279 ± 33	.102
Percentage of prolonged latencies	34%	12.9%	.002
Percentage of altered horizontal smooth pursuit	56.5%	21.2%	.000
Median gain ± IQR	0.54 ± 0.17	0.68 ± 0.17	.004
Median anti-saccadic error rate ± IQR	13 ± 6	3 ± 4	.002
Percentage of altered anti-saccades	24.2%	0	.000

ALS = amyotrophic lateral sclerosis

IQR = interquartile range

Figure 2.

Examples of abnormal oculomotor functions in our patients with amyotrophic lateral sclerosis (ALS) compared with healthy controls. (a) Horizontal voluntary saccades in a healthy control and (b) in an ALS patient showing prolonged latency. (c) Horizontal pursuit in a healthy control and (d) in a 67-year-old-man with ALS showing impaired horizontal pursuit. (e) Healthy control with normal anti-saccadic tasks compared with (f) a woman with ALS who makes errors on anti-saccadic tasks.

Analytic study

Having an abnormal eye movement recording correlated with the age of ALS onset (p = .024). It also correlated with the presence of tongue atrophy (p = .004), reflecting bulbar involvement, and with the presence of associated bladder dysfunction (p = .029). However, it did not correlate with having bulbar onset ALS (p = .48) (Table 3).

Table 3.

Correlations (p values) of the alerted eye movement recording findings with the clinical data

	Altered eye movement recording	Altered smooth pursuit	Prolonged latencies	Altered anti-saccades
Gender	.698	.921	.424	.893
Mean age of onset	.024	.637	.780	.304
Form of onset	.48	.396	.245	.221
Mean disease duration	.948	.789	.384	.879
Clinical findings: Tongue atrophy ALSFRS-R score	.004 .696	.089 .563	.285. 026	.960 .818
Non-motor signs: Sensory symptoms Bladder dysfunction Autonomic dysfunction Cognitive dysfunction Extrapyramidal symptoms	.021 .051 .029 .733 .733 .275	.003 .244 .036 .229 .737 .408	.087 .031 .780 .380 .380 .013	.992 .664 .678 .582 .120 .705
Cognitive assessment: Median MMSE score Altered MMSE score Median FAB score Altered FAB score	.917 .617 .495 .089	.117 .039 .026 .011	.712 .724 .459 .903	.752 .452 .380 .289

ALSFRS-R = Revised amyotrophic lateral sclerosis functional rating scale

FAB = Frontal assessment battery

MMSE = Mini-mental state examination

Having altered horizontal pursuit correlated with the presence of extra-motor signs (p = .003) and specifically to bladder dysfunction (p = .036). It also correlated with having altered FAB (p = .011) and MMSE (p = .039) scores.

Having prolonged horizontal saccadic latencies correlated with the presence of extrapyramidal (p = .013) and sensory signs (p = .031) and with the ALSFRS-R score at baseline (p = .026). However, having altered anti-saccades showed no correlation with any of the clinical features, the presence of extra-motor signs or with the cognitive evaluation parameters.

Lastly, there were no significant correlation between the oculomotor findings and gender, age, disease duration and form of onset.

Discussion

Our study showed a greater rate of eye movement abnormalities in Tunisian ALS patients compared with HCs, supporting further the importance of oculomotor impairment in ALS. To our knowledge, this is the first African study to address this topic and the first to investigate the link between non-motor signs and oculomotor impairment in ALS patients. Our population of study showed a higher frequency (72.6%) of abnormal eye movement recordings compared with previous studies.^4,5,18–21 Gorges et al. concluded that 60% of the 68 ALS patients had eye movement abnormalities.²¹ Kang et al. investigated 37 Korean ALS patients and eye movement abnormalities were reported in 64% of the cases.¹⁹ The most recent Italian study, including 200 ALS patients, reported 70% having oculomotor abnormalities.²⁰ Our study showed that the presence of clinical bulbar signs in ALS patients, all forms of onset included, was frequently associated with altered eye movement recordings. This finding is in line with previous studies indicating that eye movement abnormalities are a reflection of more severe and advanced disease.^19,20

Altered horizontal smooth pursuits were the most common eye movement abnormality among our patients. However, results do diverge when it comes to studying smooth pursuits in ALS patients. For example, Shaunak and collaborators found that smooth pursuits were normal in all of the 17 patients involved in their study.²² Gizzi and co-authors found them altered, but only in patients who had associated Parkinsonism.¹⁸ Recent studies conducted on the Korean population by Kang et al. found, similar to our study that altered horizontal pursuits were the main abnormality (64% of cases).¹⁹ Smooth pursuit eye movements are generated through the cerebro-ponto-cerebellar pathway. Cortical regions are considered as the generators of the pursuit movement with the pons being the final destination. The frontal lobe is implicated via the frontal eye fields (FEF), which encode and predict the pursuit trajectories.^23,24 These anatomical data explain the established correlations of altered smooth pursuits with the presence of non-motor signs in general and specifically with bladder dysfunction and executive impairment. Such a combination may not be so surprising since both executive and bladder dysfunction are consistent with frontal lobe pathology.

As for saccadic eye movements, prolonged latency was present in one-third of our ALS patients. Voluntary saccades are initiated by two main cortical areas, the FEF and the parietal eye field (Figure 3). Once visual targets are identified, an excitatory signal from both cortical areas is directly transmitted to the superior colliculus (SC), then to the brainstem oculomotor network giving the order to initiate the orientated well-measured saccadic eye movement. Yet, this movement is subject to control determined by the substantia nigra, which inhibits the SC with a tonic discharge of GABA-ergic neurons. Saccadic control is managed by the dorsolateral prefrontal cortex through projections to the caudate nucleus. Defective saccade initiation is due to a dysfunction of the FEF.^23,24 Such a finding shows that the neurodegenerative process during ALS spreads beyond the motor cortex to include the parietal eye field in the posterior parietal cortex and its connections with the FEF. The link between prolonged latencies and extra-motor sensory signs suggests that even the anterior part of the parietal field is involved in the neurodegenerative process. Another interesting finding is the positive correlation of altered voluntary saccades with the presence of extrapyramidal signs. This is possibly due to the role of the basal ganglia in initiating saccades and in maintaining fixation (Figure 3).

Figure 3.

Networks managing saccadic eye movements including cortical regions (visual cortex, parietal eye field, frontal eye field, and dorsolateral prefrontal cortex) and subcortical regions (caudate nucleus, substantia nigra, superior colliculus, and oculomotor nucleus).

InhibitoryExcitatory

Anti-saccade errors, which reflect the distractibility of ALS patients, were found in 24.2% of cases, which is lower than the 31.9% anti-saccade error rate found in the study by Shaunak et al.²² Such finding is also congruent with the distractibility factor (33%) reported by Evdokimidis et al. in a more recent study including 51 Greek patients diagnosed with ALS.²⁵ It is mainly due to failure of the dorsolateral prefrontal cortex, which is responsible for the inhibition of the unwanted reflexive saccades.²⁵ Unlike previous studies,²⁶ altered anti-saccades in or study showed no correlations with cognitive findings.

Our study was conducted in only one neurology department out of seven in the country. However, the neurology department of Razi Hospital is considered as a referral centre for neurodegenerative diseases in Tunisia, which makes this sample relatively representative of the Tunisian population. Our work is the first to study the characteristics of oculomotor eye movements in an African population. Frontal dysfunction among our patients was established based on both altered horizontal pursuits and abnormal FAB test evaluation and supported by the presence of other clinical frontal signs such as bladder dysfunction, which also represents a distinctive finding. Pathways for oculomotor control are intricate. Nonetheless, our study allowed us to confirm the involvement of the frontal, prefrontal and parietal cortex. Thus, the study of eye movement recordings is a valuable potential tool to evaluate the extension of the neurodegenerative process in ALS even in the sub-clinical stages of extra-motor involvement, and should be included as an assessment method for ALS patients.

Funding Statement

The author(s) reported there is no funding associated with the work featured in this article.

Disclosure statement

No potential conflict of interest was reported by the author(s).