DEVELOPMENTAL MEDICINE & CHILD NEUROLOGY
ORIGINAL ARTICLE
Visual field function in school-aged children with spastic unilateral cerebral palsy related to different patterns of brain damage LENA JACOBSON 1 | AGNETA RYDBERG 1 | ANN-CHRISTIN ELIASSON 2 | ANNIKA KITS 3 | OLOF FLODMARK 3 1 Clinical Neuroscience, Ophthalmology and Vision, Karolinska Institutet, Stockholm, Sweden. 2 Neuropaediatric Research Unit, Department of Women and Child Health, Karolinska Institutet, Stockholm, Sweden. 3 Clinical Neuroscience, Karolinska Institutet, Stockholm, Sweden. Correspondence to Dr Lena Jacobson at Unit of Paediatric Ophthalmology, Department of Neuropaediatrics, Astrid Lindgren Children's Hospital, Karolinska University Hospital, Solna S-171 76, Stockholm, Sweden. E-mail:
[email protected] This article is commented on by Guzzetta, p. 699 of this issue.
PUBLICATION DATA
AIM To relate visual field function to brain morphology in children with unilateral cerebral palsy
Accepted for publication 12th January 2010. Published online 30th April 2010.
METHOD Visual field function was assessed using the confrontation technique and Goldmann
LIST OF ABBREVIATION
WMDI White matter damage of immaturity
(CP). perimetry in 29 children (15 males, 14 females; age range 7–17y, median age 11y) with unilateral CP classified at Gross Motor Function Classification System (GMFCS) level I and Manual Ability Classification System levels I to III. The type and extent of brain lesions were determined using cerebral imaging. RESULTS Eighteen children had subnormal visual field function. The visual fields were severely restricted in six. The underlying brain lesions were malformation (n=7), white matter damage of immaturity (WMDI; n=13), and cortical–subcortical lesions (n=9). Visual field function could be correlated with the pattern of brain damage in children with cortical–subcortical lesions or extensive lesions caused by malformation or WMDI. Total homonymous hemianopia was common in the cortical–subcortical group but rare in children with malformation or WMDI. Five children had normal visual field function despite having malformation or WMDI involving parts of the brain usually encompassing the visual system. INTERPRETATION Visual field function may be preserved by plasticity of the immature brain in children with malformation and WMDI. Severely restricted visual fields were more often associated with lesions occurring later in the developing brain. All children with severely restricted visual fields were identified by the confrontation technique. Goldmann perimetry was a suitable method to identify relative visual field defects.
The cerebral lesions in children with cerebral palsy (CP) often involve parts of the brain that contain the posterior visual system (geniculate body, optic radiation, occipital cortex). Thus, visual field function may be affected in children with CP. Visual field function can be mapped using various techniques and the results expressed as the outer limits of the visual field, i.e. the furthest points at which an individual can see a strong and large stimulus, or, as the sensitivity with which an individual is able to see a small and weak object in the central field. Procedures in which a white ball is moved from the periphery towards the fixation line, with or without an arc perimeter, are referred to here as confrontation techniques, and can be used to identify the outer limits of the visual field. Goldmann perimetry is a manual perimetry technique in which changes in the size and intensity of a stimulus are used to detect relative visual field defects. Quinn et al.1 used Goldmann perimetry with V ⁄ 4 (25mm2) and II ⁄ 2 (1mm2) stimuli to assess visual fields in a e184 DOI: 10.1111/j.1469-8749.2010.03650.x
group of children born preterm who had been treated with retinal ablation because of retinopathy of prematurity. They found the method to be suitable for children aged from 6 years. It is well known that visual field deficits restrict daily life activities in adults with acquired hemianopic field defects.2 For example, overlapping total visual field defects, such as homonymous hemianopia, affect orientation and mobility. In addition, individuals with relative homonymous defects can see large and strong light stimuli, but not small and weak stimuli. This makes reading difficult and may lead to disqualification from driving. The long-term consequences of congenital visual field defects have not been well described. Assessment of visual function in children with CP may be hampered by their inability to understand or cooperate with tests. Studies of visual field function in children with CP have often been performed using confrontation techniques. However, visual field defects may be underestimated as confrontation techniques may fail to identify both scotoma within the ª The Authors. Journal compilation ª Mac Keith Press 2010
field and relative defects. Porro et al.3 examined 24 children with spastic hemiplegia aged from 1 to 10 years (mean age 5y 5mo) using Goldmann perimetry and a confrontation technique to assess the peripheral visual field along four diagonal meridians. Only four of the 24 children could cooperate in Goldmann perimetry. Overall, they found visual field defects in as many as 18 children. Guzzetta et al.4 used an arc perimeter and a white ball in 47 children (aged 8mo–4y 4mo; mean age 2y 1mo) with unilateral CP. Using this method, which is suitable for infants and toddlers, they found abnormal visual field function in 16 children. The same authors have shown that abnormal visual field function in infancy may normalize later in life.5 They assessed visual field function using a confrontation technique in 16 young schoolchildren with neonatal cerebral infarction and found abnormal visual field function in only two children. Hence, most previously published findings on visual field function in children with CP are based on methods that do not allow the identification of relative visual field defects, and the results of these previous studies are diverse. Brain imaging in children with unilateral CP reveals various pathologies, such as malformations, white matter damage of immaturity (WMDI), and cortical–subcortical damage (focal infarct, multicystic encephalomalacia, and basal ganglia and thalamic lesions).6 This heterogeneity of aetiologies, which reflect insults at different developmental stages, may explain differences in the prerequisites for reorganization and compensation, reflected in the visual field outcome. The purpose of this study was to assess visual field function using a confrontation technique and Goldmann perimetry, and to relate visual field function to the pattern of brain damage, determined by cerebral imaging, in a group of school-age children with unilateral CP.
METHOD Participants A hospital-based group of 34 children (20 males, 14 females; age range 7–17y, median age 11y) with unilateral spastic CP was recruited. All the participants were classified at level I on the Gross Motor Function Classification System (GMFCS),7 and most of them were enrolled in a study on motor function of the paretic hand and, as part of that study, had undergone magnetic resonance imaging (MRI). Their hand function was classified according to the Manual Ability Classification System8 at levels I to III. The gestational age of participants ranged from 25 to 42 weeks (median 37wks). Twenty-five children attended normal schools and nine had learning disability.* According to the parents, the majority of the children attending normal schools had difficulties with reading. All children were examined at the Astrid Lindgren Children’s Hospital, Karolinska University Hospital in Stockholm, by one paediatric ophthalmologist (LJ) and one orthoptist (AR). Conclusive Goldmann perimetry could be performed in 29 of 34 children, and these 29 children (15 males) constituted the study group. Five males, all with learning disability, could *North American usage: mental retardation.
What this paper adds • The majority of children with unilateral spastic CP have visual field defects. • Goldmann perimetry is a suitable method to assess visual field function at school age. • Plasticity of the immature brain may preserve visual field function.
not cooperate with Goldmann perimetry and were excluded. We estimate that Goldmann perimetry can be performed satisfactorily in children with a developmental age of 7 years. Twenty-seven participants had been examined using MRI and two using computed tomography (CT) of the brain. All children and parents gave written informed consent before enrolment in the study. The study was approved by the Regional Ethics Committee Stockholm North in accordance with ethical standards on human experimentation and with the Helsinki Declaration of 1975, as revised in 1983.
Visual field function Confrontation technique A white ball of 40mm diameter was moved from the periphery towards the midline in four diagonal meridians while the child fixated a small target at a distance of 1m. Fixation was controlled by one of two examiners. Goldmann perimetry Kinetic Goldmann perimetry was performed in a standardized way.9 The extent of the visual field was primarily determined using the II ⁄ 4 (1mm2) stimulus; if a visual field restriction was detected, a larger stimulus, V ⁄ 4 (25mm2), was used. For the central visual field, even weaker and smaller stimuli, I ⁄ 3, I ⁄ 2, or I ⁄ 1 were used. The examiner continuously monitored the fixation and cooperation of each individual. In all 29 children central visual field function was good enough to see the fixation target. The children were told to say ‘yes’ when they saw the stimulus. A few children had problems saying when they saw the stimulus. In these cases, the field was measured by means of eye movements, as described by Quinn et al.10 Some children needed a break between testing of the first and second eye. The test session rarely exceeded 20 minutes. In seven children, perimetry had to be repeated after 1 year or more, to obtain a conclusive result because of lack of cooperation, tiredness, and problems with maintaining fixation. All children were examined by one perimetrist (AR). The visual field outcome was grade as levels 0 to 4 (Table I; Figs S1 and S2, available online only). Cerebral imaging MRI (n=27) and CT (n=2) of the brain were performed at various ages (9mo–15y; median 10y) using different scanners and imaging protocols. Two children had undergone CT at the age of 9 months; one had bilateral schizencephaly and the other had middle cerebral artery infarction. Ten participants were scanned for clinical purposes, but 19 children underwent MRI within 24 months of assessment of the visual field function because they were participating in another study. Images were analysed retrospectively, and the cerebral pathology was described in terms of the type and extent of the lesion by two Visual Field Function in Children with Spastic Unilateral CP Lena Jacobsen et al. e185
Table I: Grading of visual field defects monitored with Goldmann perimetry 0 1a 1b 2a 2b 3 4
Normal Defect for object I but not for object II ⁄ 4 in one hemifield Defect for object I but not for object II ⁄ 4 in both hemifields Defect for object II ⁄ 4 but not for object V ⁄ 4 in one hemifield Defect for object II ⁄ 4 but not for object V ⁄ 4 in both hemifields Defect for object V ⁄ 4 less than hemianopia Defect for object V ⁄ 4 total hemianopia
Defects for objects I and II, but not for object V ⁄ 4, are relative. Defects for object V ⁄ 4 are absolute defects.
experienced radiologists (AK, OF). The radiologists, blind to the results of the perimetry, noted whether the lesion involved the lateral geniculate body, the optic radiation, the visual cortex, or other parts of the occipital cortex.
RESULTS The children were divided into the following three groups depending on the pattern of brain damage: those with malfor-
mations (polymicrogyria, schizencephaly, and grey matter heterotopia, n=7), those with WMDI (periventricular leukomalacia or periventricular haemorrhagic infarction, n=13), and those with cortical–subcortical lesions (middle cerebral artery or anterior cerebral artery infarcts with or without basal ganglia or thalamus defects, n=9). Seven participants had bilateral brain lesions. The involvement of the visual system by the lesion of each participant and visual field outcome are shown in Table IIa–c. When assessed with Goldmann perimetry, 11 of 29 children had normal visual fields. A total of 12 out of 29 children had defects only with object I or objects I and II. More severe visual field defects were present in 6 of the 29 children (stages 3–4), all of whom were also identified with the confrontation test. Children with relative defects (stages 1–2) on Goldmann perimetry had a normal result when tested with the confrontation technique. Table IIa shows that the visual field may be normal in children with malformations, even if the lesion involves the posterior visual system. Table IIb shows that WMDI affecting
Table II: Involvement of the visual system and visual field outcome Participant no.
MACS level
Age at perimetry, y
(a) Children with malformations 1 II 15 2 II 9 3 II 9; 9 4 II 9 5 II 13 6 II 13 7 II 15; 16 Participant no.
MACS level
(b) Children with WMDI 8 II 9 II 10 I 11 II 12 II 13 I 14 II 15 II 16 I 17 II 18 I 19 II 20 II Participant no.
MACS level
Imaging
Type of brain damage
Side
LGN
OR
VC
OC
VF
MRI MRI CT MRI MRI MRI MRI
Polymicrogyria Polymicrogyria R Schizencephaly, bilateral Schizencephaly R and polymicrogyria L Schizencephaly Polymicrogyria and heterotopic grey matter Polymicrogyria
L R R>L R>L L L R
0 0 0 0 0 0 R
0 0 R R L 0 R
0 0 0 0 0 0 0
0 0 0 0 0 0 R
0 0 0 0 1b 1b 3
Age at perimetry, y
Imaging
Side
LGN
OR
VC
OC
VF
10 9 9 12 9; 11 8 12 13; 14 11 15 10 17 12
MRI MRI MRI MRI MRI MRI MRI MRI MRI MRI MRI MRI MRI
L R RL R>L L L R L L R>L RL 0 L 0 0 0 R R
