Characteristics and Outcomes of Idiopathic and Non-idiopathic Ocular Motor Apraxia in Children

Abstract

Purpose:

To systematically compare idiopathic and non-idiopathic ocular motor apraxia (OMA) in children.

Methods:

A retrospective chart review was conducted of all children (< 18 years) diagnosed as having OMA from 2010 to 2020. Demographics, clinical characteristics, and oculomotor outcomes were compared for children with idiopathic and non-idiopathic OMA.

Results:

Thirty-seven children were included, 17 (46%) with idiopathic OMA and 20 (54%) with non-idiopathic OMA. Among patients with non-idiopathic OMA, Joubert syndrome was the most frequent underlying diagnosis (30%). Strabismus (45% vs 12%, P = .04), nystagmus (30% vs 0%, P = .02), and vertical saccade involvement (25% vs 0%, P = .049) were significantly more common in non-idiopathic than idiopathic OMA, respectively. Neuroimaging abnormalities (90% vs 18%, P < .0001) and developmental delays (100% vs 59%, P = .002) were also more frequent in non-idiopathic than idiopathic OMA, respectively. Endocrine disorders (most commonly growth hormone deficiency) were diagnosed in 12% and 20% of children with idiopathic and non-idiopathic OMA, respectively (P = .67). On survival curve analysis, improvement in OMA occurred faster and more frequently in children with idiopathic than non-idiopathic OMA (median time to improvement 56 vs 139 months, respectively, P = .034).

Conclusions:

Non-idiopathic OMA is associated with a higher rate of vertical saccade involvement, nystagmus, and developmental delays. These findings should prompt neuroimaging in children with OMA. Additionally, endocrine disorders may be more frequent in children with OMA than the general pediatric population.

INTRODUCTION

Ocular motor apraxia (OMA) is a rare disorder characterized by an inability to initiate horizontal and/or vertical saccades with preserved smooth pursuit.^1,2 OMA was first described by Cogan³ and the eponym ‘Cogan’s ocular motor apraxia’ is sometimes used in cases of congenital OMA. Children with OMA typically present with attenuated visual responses early in infancy and may be initially misdiagnosed as having visual impairment. When head control develops, typically between 4 and 6 months old, children begin to use head thrusts to compensate for the failure of saccade initiation during target fixation. Over time, they may develop an exaggerated blinking response, which enables them to initiate saccades without head thrusting.^2,4 OMA in children is frequently idiopathic, but it may also occur patients with a genetic, metabolic, or neurodegenerative disorder, structural brain abnormality, or diffuse cortical injury, as in the case of perinatal hypoxia, meningoencephalitis, or post-cardiac surgery.^4-11

Although children with idiopathic OMA and non-idiopathic OMA have been previously compared with regard to neurodevelopment,⁴ other outcomes, such as the rate of improvement of OMA over time, have not been evaluated. The objective of this study was to describe characteristics and outcomes of idiopathic and non-idiopathic OMA in children evaluated in a pediatric neuro-ophthalmology clinic over a 10-year period.

PATIENTS AND METHODS

This study was approved by the local institutional board and adhered to the tenets of the Declaration of Helsinki and the U.S. Health Insurance Portability and Accountability Act of 1996. The medical records of children examined at Children’s Hospital Los Angeles between January 1, 2010 and April 1, 2020 who were diagnosed as having OMA were retrospectively reviewed. The electronic medical record was searched based on age younger than 18 years and International Classification of Disease (ICD)-9 and ICD-10 codes used at our institution for OMA (ICD-9 379.59, ICD-10 H51.8). OMA was diagnosed in children with impaired horizontal and/or vertical saccade initiation, intact vestibulo-ocular reflex, and “locking up” during opticokinetic nystagmus (OKN) testing. Locking up is characterized by the failure to generate quick phases of nystagmus, leading to tonic gaze ipsilateral to the direction of movement of the OKN stimulus. Head thrusting was not required for diagnosis due to poor head control in some patients with neurodegenerative conditions. Patients were excluded from our study if they did not meet criteria for OMA diagnosis or did not have a neuro-ophthalmic examination at our institution.

The chart review yielded data on demographics, birth history, family history, neuroimaging findings, medical history, and ophthalmologic diagnoses, findings, treatments, and outcomes. We specifically recorded characteristics of OMA such as involvement of horizontal or vertical saccades, asymmetry of OMA between the right and left gazes, presence or absence of head thrusting, and ability to use blinking to initiate saccades. Standard ocular motor examination, including alternate cover with prism testing, was used to evaluate strabismus and nystagmus. Strabismus was diagnosed in patients with a manifest deviation in any gaze position. In assessing outcomes, OMA was considered improved if accurate volitional saccades of any amplitude could be generated in the affected direction with or without blinking. Children were categorized as having idiopathic or non-idiopathic OMA based on associated systemic abnormalities and work-up, including neuroimaging and genetic testing if indicated. All children with developmental delays or any other neurologic symptoms or signs underwent neuroimaging. Genetic testing was performed in patients with abnormal neuroimaging or systemic comorbidities suggestive of an underlying genetic disorder. Idiopathic OMA was diagnosed when onset was within the first few months of life and no genetic, metabolic, or neurologic disorder was diagnosed. In patients with developmental delay, unremarkable neuroimaging was required for OMA to be considered idiopathic. OMA was deemed non-idiopathic if it was acquired or associated with a genetic, metabolic, or neurologic disorder, including a structural brain abnormality.

Data were collected and statistical analyses were conducted in Microsoft Excel version 16.45 (Microsoft Corporation) and GraphPad Prism (version 8.4.3; GraphPad, Inc) software. Characteristics of children with idiopathic OMA and non-idiopathic OMA were compared using the chi-square test for categorical variables or Fisher exact test if any category had a value of 5 or less. The Mann-Whitney test was used to compare non-normal (skewed) continuous variables. Kaplan-Meier survival curve analysis was used to assess the proportion of patients with improvement in OMA over time, to account for variable follow-up duration. P values less than .05 were considered statistically significant.

RESULTS

Search of the electronic medical record using ICD codes yielded 37 patients diagnosed as having OMA who met the inclusion criteria for our study. Of those, 17 patients (46%) were idiopathic and 20 patients (54%) were non-idiopathic. Table 1 shows the underlying diagnoses in 20 children with non-idiopathic OMA. Three patients (15%) had an acquired cause of OMA. The remaining patients had congenital OMA secondary to a genetic disorder (8 patients, 40%) or structural brain abnormality (9 patients, 45%), with the most common underlying condition being Joubert syndrome in 6 patients (30%).

Table 1.

Systemic diagnoses in 20 children with non-idiopathic ocular motor apraxia (OMA).

Diagnosis	Number of patients
Acquired OMA
Cardiac surgery with post-operative stroke	1 (5%)
Hypoxic-ischemic encephalopathy	1 (5%)
Meningoencephalitis	1 (5%)
Congenital OMA
Genetic disorders
Chromosomal deletion or duplication	2 (10%)
Ataxia telangiectasia	1 (5%)
Congenital disorder of glycosylation	1 (5%)
Galactosemia	1 (5%)
Gaucher’s disease	1 (5%)
Mitochondrial disorder	1 (5%)
Undiagnosed neurodegenerative disorder	1 (5%)
Structural brain abnormality
Joubert syndrome	6 (30%)
Othera	3 (15%)

^aOther structural brain abnormalities included holoprosencephaly, midbrain malformation, polymicrogyria, and pachymicrogyria)

Demographics and clinical characteristics of children with idiopathic and non-idiopathic OMA are compared in Table 2. Age at presentation and sex did not significantly differ between groups. Children with non-idiopathic OMA were more likely to have involvement of vertical saccades than children with idiopathic OMA (25% vs 0%, respectively). The difference was borderline significant (P = .049). There was no difference in the rates of asymmetric horizontal OMA or presence of head thrusts at presentation.

Table 2.

Demographics and clinical characteristics of 37 children with idiopathic and non-idiopathic ocular motor apraxia (OMA).

	Idiopathic OMA n=17	Non-idiopathic OMA n=20	p-value
Demographics
Age at presentation, median (range)	19 months (5 months to 13 years)	23 months (4 months to 11 years)	0.96
Sex (M/F)	9 (53%) / 8 (47%)	12 (60%) / 8 (40%)	0.46
Ophthalmic characteristics
Head thrusts	15 (88%)	16 (80%)	0.67
Asymmetric OMA	3 (18%)	1 (5%)	0.32
Vertical OMA	0 (0%)	5 (25%)	0.049*
Strabismus	2 (12%)	9 (45%)	0.04*
Strabismus subtypes			0.35
Intermittent exotropia	1 (6%)	4 (20%)
Sensory exotropia	0	1 (5%)
Intermittent esotropia	0	2 (10%)
Constant esotropia	0	1 (5%)
Superior oblique palsy	1 (6%)	0
Skew deviation	0	1 (5%)
Nystagmus	0	6 (30%)	0.02*
Other ophthalmic diagnosesa	3 (18%)	6 (30%)	0.46
Systemic characteristics
Abnormal neuroimaging	3 (18%)	18 (90%)	<0.0001*
Developmental delay	10 (59%)	20 (100%)	0.0019*
Endocrinopathy	2 (12%)	4 (20%)	0.67

^aOther ophthalmic diagnoses included optic atrophy, chorioretinal coloboma, ptosis, and microphthalmia.

^*p<0.05

Strabismus was more prevalent in the non-idiopathic OMA group than the idiopathic OMA group (45% vs 12%, respectively) (P = .04). There was no significant difference between groups in subtypes of strabismus (P = .35). Nystagmus was also more common in the non-idiopathic OMA group than the idiopathic OMA group (30% vs 0%, respectively) (P = .02). The types of nystagmus were seesaw or torsional in 2 patients with Joubert syndrome, sensory in 1 patient with optic atrophy, gaze-evoked in 1 patient, downbeat in 1 patient, and infantile nystagmus syndrome in 1 patient. The two groups had similar rates of other ophthalmic diagnoses (18% of patients with idiopathic OMA and 30% of patients with non-idiopathic OMA, P = .46), including optic atrophy, chorioretinal coloboma, ptosis, and microphthalmia.

With regard to systemic findings, neuroimaging abnormalities were more common in children with non-idiopathic OMA (P < .0001). Three patients (18%) with idiopathic OMA had findings on neuroimaging that were likely incidental (few nonspecific white matter changes in 2 patients and an arachnoid cyst in 1 patient). However, 8 patients (47%) with idiopathic OMA did not undergo neuroimaging, due to normal development and absence of neurologic symptoms and signs. In the non-idiopathic OMA group, 18 patients (90%) had neuroimaging abnormalities related to the underlying neurologic or genetic condition associated with OMA (eg, Joubert syndrome). The 2 patients with non-idiopathic OMA who had normal neuroimaging were diagnosed as having metabolic disorders (Gaucher’s disease and galactosemia).

Developmental delays were also more common in children with non-idiopathic OMA than idiopathic OMA (100% vs 59%, respectively) (P = .0019). In the idiopathic OMA group, the 10 patients who were diagnosed as having developmental delay all had motor involvement, and 3 patients (30%) also had language delay.

Endocrine abnormalities were diagnosed in 2 patients (12%) with idiopathic OMA and 4 patients (20%) with non-idiopathic OMA (P = .67). In the idiopathic OMA group, endocrine disorders included growth hormone deficiency in 1 patient and secondary hyperparathyroidism in 1 patient with chronic kidney disease. In the non-idiopathic OMA group, endocrine disorders included growth hormone deficiency and hypothyroidism in 1 patient, hypothyroidism alone in 1 patient, hyperinsulinemic hypoglycemia in 1 patient, and hypophosphatemic rickets in 1 patient.

As seen in Figure 1, outcomes between the idiopathic and non-idiopathic OMA groups were significantly different on survival curve analysis. Patients with idiopathic OMA had earlier and higher rates of improvement in OMA than patients with non-idiopathic OMA (P = .034). The median time to improvement was 56 months in the idiopathic group and 139 months in the non-idiopathic group. As described previously, OMA was considered improved if saccades of any amplitude could be generated in the affected direction with or without blinking. We analyzed improvement rather than resolution due to low rates of complete resolution in both idiopathic (24%) and non-idiopathic (10%) OMA groups, presumably because follow-up ceased after improvement was demonstrated. We also excluded 3 patients from this comparison whose OMA was already improved when they presented to the clinic, because we could not accurately determine when improvement occurred in these patients.

Figure 1.

Survival curve analysis of the proportion of children with idiopathic and non-idiopathic ocular motor apraxia (OMA) who did not experience improvement in OMA over time. Children with idiopathic OMA had significantly faster improvement in OMA (P = .034).

DISCUSSION

Our study confirms prior reports of high rates of underlying conditions in children with OMA, with 54% of children falling into the non-idiopathic group. Previous studies reported non-idiopathic etiology in 59% to 73% of children with OMA.^1,4,12 Joubert syndrome was the most common underlying etiology in the current study, which was also similar to prior studies.¹ Our findings are also consistent with previous literature indicating a high rate of developmental delays and neuroimaging abnormalities in children with OMA. In the current study, 59% of children with idiopathic OMA had developmental delays (mostly motor), which was similar to previous reports ranging from 60% to 80%.^1,4,12 All children with non-idiopathic OMA were developmentally delayed. As expected, neuroimaging abnormalities were seen in almost all patients with non-idiopathic OMA (90%), but a few patients with idiopathic OMA (18%), due to our definition of non-idiopathic OMA including patients with structural brain abnormalities and genetic, metabolic, and neurodegenerative conditions. Overall, 21 patients (57%) in the current study had abnormal neuroimaging, which is remarkably similar to prior reports ranging from 52% to 63%.^1,4,10,12

We found that vertical involvement (25% non-idiopathic and 0% idiopathic), strabismus (45% non-idiopathic and 12% idiopathic), and nystagmus (30% non-idiopathic and 0% idiopathic) were significantly more common in children with non-idiopathic than idiopathic OMA. Although previous studies have reported rates of vertical involvement, strabismus, and nystagmus in children with OMA, they have not specifically compared idiopathic and non-idiopathic cases. Harris et al¹ reported vertical involvement in 11% of children with OMA, all with underlying conditions (herpetic encephalitis, agenesis of the corpus callosum, and Gaucher’s disease). Strabismus was diagnosed in 31% of children and nystagmus was diagnosed in 28% of children, which is within the same range as the current study. Similar to our findings, Harris et al¹ also reported no consistent subtype of strabismus or nystagmus in children with OMA, but nearly all patients with nystagmus (19 of 21, or 90%) had an underlying condition. Therefore, the presence of nystagmus in a child with OMA should prompt work-up for an underlying condition.

Besides strabismus and nystagmus, 30% of children with non-idiopathic OMA and 18% of children with idiopathic OMA had a variety of other ophthalmic diagnoses, including ptosis, chorioretinal coloboma, optic atrophy, and microphthalmia. Some of these diagnoses, such as optic atrophy, may be related to the underlying neurologic condition in patients with non-idiopathic OMA. Our case series is too small to determine whether these associations are random or systematic.

To our knowledge, this study is the first to evaluate rates of improvement in pediatric OMA and compare this outcome in patients with idiopathic and non-idiopathic OMA. Prior small case studies have observed that congenital OMA improves or resolves over time, but the time course is not well-delineated.⁷ Using survival analysis, we found that improvement in OMA (defined as ability to generate saccades in the affected direction with or without blinking) occurred at a median of 56 months after onset in patients with idiopathic OMA and 139 months in patients with non-idiopathic OMA. Most patients were discharged from our clinic when improvement in OMA was noted, so we could not accurately compare rates of complete resolution between groups.

Another new finding from the current study was the high rate of endocrine disorders in children with both idiopathic (12%) and non-idiopathic (20%) OMA. Epidemiologic studies suggest that the prevalence of endocrine disorders in the general pediatric population is much lower, ranging from 9.6 to 29 per 100,000¹³. The reason for this increase in endocrine disorders among children with OMA is unclear, but it does challenge traditional assumptions regarding the pathophysiology of OMA. Based on the structural brain abnormalities observed in many patients with OMA, brainstem and/or cerebellar dysfunction has been suggested as the cause of the saccade initiation failure.^1,2,5 However, our findings suggest that OMA may involve more widespread dysfunction of the CNS including the hypothalamic-pituitary axis. Due to a small number of patients, our results are not robust enough to recommend endocrine work-up in all children with OMA. However, we do suggest screening for endocrinopathies by history, with particular attention to rate of growth because growth hormone deficiency was the most common endocrinopathy in our cohort.

Our study has limitations, including its retrospective design and small sample size. Therefore, our data are subject to bias inherent in small sample sizes. Additionally, we did not conduct neuroimaging, genetic testing, and endocrine work-up in all patients. Neuroimaging was not performed in 47% of patients with idiopathic OMA who had normal development and no neurologic abnormalities, and genetic testing and endocrine work-up were also only conducted if clinically indicated. Additionally, as noted previously, we only analyzed the rate of improvement in OMA and not resolution, due to low rates of complete resolution that were likely due to our practice of discharging patients after documented improvement in OMA. Finally, there was variable patient follow-up, although we accounted for this using Kaplan-Meier survival curve analysis when evaluating our outcome (improvement in OMA over time).

Our study confirms prior reports of high prevalence of developmental delays and neuroimaging abnormalities in children with OMA, provides evidence for a possible association between OMA and endocrine abnormalities, and highlights the increased rates of systemic and ophthalmologic comorbidities and worse oculomotor outcomes in children with non-idiopathic OMA. We suggest that children presenting with OMA undergo neuroimaging if they are found to have developmental delay, neurologic symptoms, vertical saccade involvement, or nystagmus, with additional genetic, metabolic, and/or endocrine work-up based on the presence of systemic findings such as dysmorphic features and growth delay.