Factors Associated With Lack of Vision Improvement in Children With Cortical Visual Impairment

Improvement in vision has been noted in children with cortical visual impairment (CVI), resulting from disparate types of brain injury. The purpose of our study was to determine the risk factors associated with poor recovery of vision in this group of patients. Case records of children who were born before 2010 with at least 4 follow-up visits for CVI were reviewed for underlying etiologies of CVI, visual acuity (VA), and associated neurological and ophthalmological disorders. VA was assessed in 6 qualitative grades. Changes in VA were recorded as the difference between the grades of VA at presentation and the last follow-up visit. The outcome was calculated as a ratio of actual improvement to potential improvement in grades of qualitative VA. Multiple linear regression determined factors associated with lack of vision improvement in all children and based on etiology. Fifty-three children with CVI were identified. The median age at presentation was 13.6 months (range: 2.9-76.4 months) and the median follow-up was 5.8 years (1.1-16.3 years). CVI resulted from central nervous system (CNS) malformation (9.4%), hypoxic/inflammatory injury (15.1%), seizures (24.5%), and combined causes (51.0%). Vision improvement was noted in 83% of children. Lack of VA improvement was associated with older age at presentation in all children with CVI and within each etiological group except CNS malformation. None of the other investigated variables were associated with poor recovery of VA. Most of the children with CVI showed improvement in vision. Older age at presentation, but not etiology of CVI, was associated with poor improvement in VA.

Original Contribution Factors Associated With Lack of Vision Improvement in Children With Cortical Visual Impairment Swati Handa, FRCS, MCI, Seyed E. Saffari, PhD, Mark Borchert, MD Background: Improvement in vision has been noted in children with cortical visual impairment (CVI), resulting from disparate types of brain injury. The purpose of our study was to determine the risk factors associated with poor recovery of vision in this group of patients. Methods: Case records of children who were born before 2010 with at least 4 follow-up visits for CVI were reviewed for underlying etiologies of CVI, visual acuity (VA), and associated neurological and ophthalmological disorders. VA was assessed in 6 qualitative grades. Changes in VA were recorded as the difference between the grades of VA at presentation and the last follow-up visit. The outcome was calculated as a ratio of actual improvement to potential improvement in grades of qualitative VA. Multiple linear regression determined factors associated with lack of vision improvement in all children and based on etiology. Results: Fifty-three children with CVI were identified. The median age at presentation was 13.6 months (range: 2.9- 76.4 months) and the median follow-up was 5.8 years (1.1- 16.3 years). CVI resulted from central nervous system (CNS) malformation (9.4%), hypoxic/inflammatory injury (15.1%), seizures (24.5%), and combined causes (51.0%). Vision improvement was noted in 83% of children. Lack of VA improvement was associated with older age at presentation in all children with CVI and within each etiological group except CNS malformation. None of the other investigated variables were associated with poor recovery of VA. Conclusions: Most of the children with CVI showed improvement in vision. Older age at presentation, but not etiology of CVI, was associated with poor improvement in VA. Journal of Neuro-Ophthalmology 2018;38:429-433 doi: 10.1097/WNO.0000000000000610 © 2017 by North American Neuro-Ophthalmology Society Pediatric Ophthalmology (SH), KK Women's and Children's Hospital, Singapore; Centre for Quantitative Medicine (SES), Duke-NUS Medical School, Singapore; and Children's Hospital Los Angeles and the Vision Center (MB), USC/Keck School of Medicine, Los Angeles, California. The authors report no conflicts of interest. Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the full text and PDF versions of this article on the journal's Web site (www.jneuro-ophthalmology.com). Address correspondence to Swati Handa, FRCS, MCI, Pediatric Ophthalmology, KK Women's and Children's Hospital, 100 Bukit Timah Road, Singapore 229899; E-mail: [e-mail] Handa et al: J Neuro-Ophthalmol 2018; 38: 429-433 C ortical visual impairment (CVI) is diagnosed when an insult to retrochiasmal visual pathways and/or visual cortex results in bilateral visual deficit in the absence of significant ocular pathology or anterior visual pathways disease (1). It commonly occurs with adverse cognitive, neurobehavioral, and psychological outcomes after early injury to the young brain (2-6). Hypoxic-ischemic insult at birth is reported as the most common cause (5,7-12), with 60% of these children developing CVI (13). Approximately 15% of CVI cases are caused by meningitis and/or encephalitis (3,6,9,12). One-third of children with hydrocephalus develop CVI (14). Head injuries account for approximately 4% of CVI (3,9) of which 50% result from nonaccidental injuries (10). Other causes of CVI include central nervous system (CNS) malformations and epilepsy (15). Typically, CVI results in reduced visual acuity (VA) that may be associated with visual field defects (7,8,16). Some residual visual function is believed to be present even in the most severe cases of CVI (17). Most individuals show some improvement in VA with age, although normal VA usually is not usually attained (7,10). CVI frequently is associated with severe neurological disorders (3,4,10) that limit the VA assessment. Superimposed optic atrophy and strabismus are common (1,3). Little is known about the factors associated with improvement in VA. These include etiology, timing, severity and type of the brain insult, and associated neurological conditions including epilepsy (7). Age of onset and age at primary insult have been found to be associated with the visual recovery (3,18). It is thought that the young brain recovers better if the primary insult occurs at an earlier age due to greater neuronal plasticity with activation of alternative pathways or by changes at a cellular level in the traumatized area (19). However, it also has been proposed that more global damage with long-term disability results if the primary insult occurs at a younger age (20). Our purpose was to assess the extent of improvement in qualitative VA in children with CVI, and to evaluate the factors associated with lack of improvement in VA in all children and in various etiological groups. 429 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution METHODS Children with CVI born before 2010 and evaluated between July 1992 and July 2011 with at least 4 followup visits to a single physician (M.B.) were identified from the Vision Center database at Children's Hospital Los Angeles. A diagnosis of CVI was made if a child older than 6 months presented with bilateral visual impairment with normal eye movements and normal pupillary responses without any associated ocular or anterior visual pathway pathology sufficient to explain the visual loss. Neuroimaging and electrophysiology were not required for the diagnosis. Patient charts were retrospectively reviewed. All children underwent a complete neuro-ophthalmologic examination including VA, cycloplegic refraction, and sensorimotor testing. All children treated with vigabatrin for seizures underwent ophthalmoscopy and electroretinography every 3 months to monitor for retinal toxicity. Spectacles were prescribed for children with significant refractive errors and the ability to follow targets. The amount of correction depended on the visual needs of the children (e.g., a child with severe cognitive impairment, confined to a wheelchair, and may have had myopia undercorrected by 1D). Significant refractive errors, if any, were fully corrected before visual assessment in children with ability to fixate targets. The Institutional Review Board of Children's Hospital of Los Angeles approved this study. Information was recorded on etiology of CVI (CNS malformations, hydrocephalus, seizures, hypoxic/inflammatory conditions, or combination of the above); age at presentation to neuro-ophthalmology clinic; age at primary insult; age of onset of seizures; duration and frequency of seizures at the last follow-up visit ($2 or ,2 seizures per week), VA at presentation, and followup examinations; duration of follow-up; associated neurological diseases such as cerebral palsy and developmental delay; and other eye conditions including strabismus, nystagmus, and optic atrophy. Qualitative VA was categorized into 6 grades (Table 1). When possible, quantitative acuity was measured using Snellen, HOTV, or Allen optotypes (Grade 1). Improvement in qualitative VA was calculated by recording the difference between the initial grades of VA and at the last follow-up visit. To avoid confounding amblyopia, the eye with better grade of qualitative VA was included. The outcome measure was calculated as a ratio of actual improvement to potential improvement in grades of qualitative VA. For example, an actual improvement in qualitative VA from Grade 5 to 3 (2 grades) represents 50% (0.5) of the potential improvement from Grade 5 to Grade 1 (4 grades). The ratio ranged from 0 to 1, where 0 indicated no improvement and 1 indicated maximum possible improvement in the qualitative vision. Those few children in whom qualitative VA was Grade 1 at presentation (thus not improvable) were excluded from the analysis and are described separately. The factors associated with lack of VA improvement were analyzed for the entire cohort and by etiology of CVI. Combined cases were included in the respective specific etiologies for analysis. Statistical Analysis Statistical analysis was performed using SAS version 9.4 for Windows (SAS, Inc, Cary, NC). Statistical significance was set at P # 0.05. Demographic and clinical variables were reported for various etiological groups and the entire cohort using frequencies (percentages) for categorical variables and median (range) and mean (standard deviation) for continuous variables. Multiple linear regression analysis was used to assess the association between the demographic variables and the ratio of actual improvement to possible improvement in grades of qualitative VA for each etiological group adjusted for all variables that achieved the P , 0.2 significance level in the initial univariate analysis. RESULTS Patient Cohort Characteristics TABLE 1. Visual acuity assessment in children with cortical visual impairment Measurement of Qualitative Visual Acuity Categories Grade Grade Grade Grade 6 5 4 3 Grade 2 Grade 1 430 Definition No visual behavior Behavioral response to light Behavioral response to motion Poor fixation with ability to follow face or large object Fixation and pursuit of a 6-inch toy at 1 foot Fixation and pursuit of a 1-inch toy at 1 foot or ability to measure optotype acuity (better than or equal to 20/400) Demographics of the entire cohort and various etiological groups are summarized in Supplemental Digital Content 1 (see Table E1, http://links.lww.com/WNO/A284). More than half of the children (51%) had combinations of various causes including CNS malformations in 29.6%; hydrocephalus in 44.4%; hypoxic or inflammatory injuries in 55.5%; and seizures in 96.3%. No child had hydrocephalus as stand-alone etiology. Of 53 children, improvement of at least 1 grade of VA was noted in 44 (83%). The median potential improvement in qualitative VA for all children with CVI was 4 grades, and the median actual improvement was 2 grades. On simple linear regression, lack of VA improvement was associated with older age (P = 0.172), cerebral palsy (P = 0.096), and strabismus (P = 0.098). Handa et al: J Neuro-Ophthalmol 2018; 38: 429-433 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution Characteristics by Etiology Twenty-three children had CVI attributed only to CNS malformations (n = 5, 22%) including microcephaly, ventriculomegaly, neural tube defects, lipomyelomeningocele, polygyria, or to a combination of CNS malformations with hydrocephalus and/or seizures (n = 18, 78%). CVI was ascribed to hypoxic/inflammatory insults in 23 children. Of 4 children with inflammatory etiologies, 2 had Group B streptococcal meningitis with perinatal stroke; 1 had meningitis with hydrocephalus and seizures; and 1 had bacterial meningitis alone. Among 39 children in the seizure group, 13 children had only seizures (33%) of which 10 (77%) had infantile spasms. Two had neonatal-onset seizures, and 1 had Lennox-Gastaut syndrome. Seizures developed at median age 6 months (range: 0.03 months-28 months) and lasted for 6.5 years (range: 1.2 years-15.7 years). Seizures were well controlled in 51% (n = 20). On simple linear regression, older age at presentation (P = 0.018), older age at primary insult (P = 0.089), and shorter duration of followup period (P = 0.106) were associated with lack of improvement in VA in this group. Age at presentation and age at primary insult were significantly related (P = 0.002) on Pearson correlation test. Adjusting for age at presentation and duration of follow-up period on multiple regression, there was a trend toward lack of improvement in VA in children who had longer duration of seizures (20.039 ± 0.021; P = 0.078). Combined etiologies were seen in 27 children. All (n = 26, 96.3%) but 1 child had seizures. Older age at presentation and primary insult (P = 0.021 and P = 0.093, respectively) was noted to have lack of improvement in VA on simple regression. VA did not improve in 3 (13%), 5 (22%), 6 (15.4%), and 4 (14.8%) children in CNS malformation, hypoxic/inflammatory, seizure, and combined group, respectively. Supplemental Digital Content 2 (see Table E2, http:// links.lww.com/WNO/A285) presents the results of multiple linear regression in all children and various etiological groups adjusted for possible confounders noted on simple linear regression (P , 0.2) and those of clinical relevance. Supplemental Digital Content 3 (see Table E3, http:// links.lww.com/WNO/A286) summarizes the demographic and clinical features of children with CVI whose VA could be assessed quantitatively. We did not detect evidence of retinal damage in our patients taking vigabatrin. DISCUSSION Consistent with previous reports, vision improved in most of our patients with CVI. Depending on different etiologies and distinctive methods to evaluate VA, some improvement in VA has been noted in 32%-73% children with CVI (3,4,9,12,18). In our study, 83% children had at least 1 grade improvement in qualitative VA, and the median Handa et al: J Neuro-Ophthalmol 2018; 38: 429-433 actual improvement was 50% of the potential improvement. It is uncertain if this improvement is related to any therapeutic intervention. All patients with CVI were referred for early vision services through either the school district or regional center. However, data on the type or extent of interventions were not available. Previous studies of CVI have identified perinatal hypoxic/ischemic/inflammatory injuries as the leading cause, usually associated with cerebral palsy (7,9-11). Although this etiology was prevalent in our population, only 15% had hypoxic/inflammatory injury as a standalone etiology. Nearly half of our cohort (49%) had superimposed seizures. This is consistent with previous studies that documented seizures as the most common neurological association in children with CVI, although as a single etiology it is reported only in 4%-10% children (3,9). By contrast, seizures were the most common single etiology (n = 13; 24.5%) for CVI in our cohort. Most children with seizures in our study started as infantile spasms (77%) without associated cerebral malformations on MRI. Seizures, especially infantile spasms, may cause CVI by damaging the optic radiations and/or visual cortex (3,21) and, in patients treated with vigabatrin, have been associated with a poor visual prognosis (22). Yet, none of our patients on vigabatrin had detectable signs of retinal toxicity to account for lack of visual improvement. It is uncertain whether the high prevalence of seizures in our cohort was due to a skewed sample at our pediatric medical center or whether this represents a change in those available for vision monitoring due to improving developmental outcomes in children with infantile spasms (23). Visual function improves with better control of seizures (7). Wong et al (4) noted poor prognosis in children who had uncontrolled seizures for 3 months after an initial neurological insult. Children who had poor functional VA were more likely to have seizures at follow-up visits (83.5% vs 69%, P , 0.01) compared with those who had better functional VA (12). Surprisingly, in our study neither better control of seizures nor the age at onset was significantly associated with vision improvement. A trend toward lack of VA improvement was seen with longer duration of seizures (P = 0.078), but it did not reach statistical significance, possibly due to inadequate sample size. No single etiology of CVI was associated with lack of VA improvement. Our results are similar to Huo et al (3) who reported no association of etiology with functional levels of VA. We determined VA qualitatively but quantified the improvement by using the ratio of actual improvement to potential improvement in grades of VA to assess the relationship of etiology with improvement in VA. Our results are in agreement with Watson et al (24) who assessed VA in children with CVI using sweep visual evoked potential and reported no association with etiology (24). By contrast, a few previous reports do find a relationship between etiology and visual outcome (7). An improvement in vision 431 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution ranging from minimal in children with structural brain damage to greater improvement in children with meningitis and hypoxic-ischemic encephalopathy has been reported in patients with CVI (9). An insult to the optic radiations is associated with not only poorer visual function (13) but also poorer visual recovery and prognosis, especially in the presence of periventricular leukomalacia (18). The only variable in our study consistently noted with lack of improvement in VA was age at presentation to the neuro-ophthalmologist. Older children were less likely to have improvement in VA, except those in the congenital CNS malformation group. Previous studies have shown that older children (presenting after age 3 years) were less likely to have improvement in VA (3,12). However, Lambert et al (18) studied only children with hypoxic insult at a documented age and found worse visual outcome with younger age. In our study, the mean age at primary insult was 2.1 months (median: 0 months), but the mean age at presentation was 1.53 years (median: 1.13 years). There was no association of age at presentation with VA at presentation (Pearson correlation = 0.19; P = 0.17). There is no way of knowing if there was an improvement before presentation or if potential for improvement declines after a certain age, at least with the intervention services provided in our locale. It remains to be seen whether or not new, more frequent, or more prolonged forms of vision services will impact this outcome (25). Even without evidence that early interventional services are beneficial, our data suggest that delay in diagnosis and referral for services is not helpful, and may be detrimental. There are several limitations of our study. Coming from a major pediatric medical center, the cases may be skewed toward more severe conditions with poorer outcomes. Despite long-term follow-up and more older children compared with previous studies, few of our patients had quantifiable VA due to poor cognitive or communication skills. No quantitative acuity tests have been validated for children with CVI. Teller acuity cards have only been validated in children with cataracts, amblyopia, strabismus, or ptosis (26). Teller acuity cards have been used to assess vision in studies of CVI, but results can only be categorized as normal, impaired, or blind (27). Inconsistent Teller acuity and visually evoked potential scores in nearly half of children with neurologic conditions raise questions about their utility in CVI (28). We have found preferential looking paradigms such as Teller acuity cards to be of limited value in assessing acuity of children with CVI due to fluctuations in visual attention, variable saccadic latencies with dysmetria, aversive head and eye movements, and frequent visual field deficits (29,30). Our system for grading vision was not linear and did not provide true quantification of improvement. This deficiency was largely, but not completely, ameliorated by the use of a ratio of improvement to potential improvement. Also, grading of VA based on ability to fixate and pursue visual objects does not incorporate analysis of many of the abnor432 mal features of vision noted in patients with CVI, such as latency of fixation or inability to deal with visual complexity (25). Finally, there is an impact of confounding variables such as nystagmus (28%) and optic atrophy (20%) on VA, which was, however, largely alleviated by performing multiple linear regressions. In conclusion, most children with CVI show an improvement in VA. Older age at presentation is associated with lack of vision improvement. Concerted efforts by multidisciplinary teams of pediatricians, neurologists, and pediatric ophthalmologists in early diagnosis and referral of children with CVI will be important in developing studies to test the effectiveness of various interventions on visual performance. STATEMENT OF AUTHORSHIP Category 1: a. Conception and design: M. Borchert and S. Handa; b. Acquisition of data: S. Handa; and c. Analysis and interpretation of data: S. Ehsan and S. Handa. Category 2: a. Drafting the manuscript: S. Handa and b. Revising it for intellectual content: M. Borchert and S. Ehsan. Category 3: a. Final approval of the completed manuscript: S. Handa, S. Ehsan, and M. Borchert. REFERENCES 1. Binder NR, Kruglyakova J, Borchert MS. Strabismus in patients with cortical visual impairment: outcomes of surgery and observations of spontaneous resolution. J AAPOS. 2016;20:121-125. 2. Jan JE, Groenveld M, Sykanda AM, Hoyt CS. Behavioural characteristics of children with permanent cortical visual impairment. Dev Med Child Neurol. 1987;29:571-576. 3. Huo R, Burden SK, Hoyt CS, Good WV. Chronic cortical visual impairment in children: aetiology, prognosis, and associated neurological deficits. Br J Ophthalmol. 1999;83:670-675. 4. Wong VC. Cortical blindness in children: a study of etiology and prognosis. Pediatr Neurol. 1991;7:178-185. 5. Lanzi G, Fazzi E, Uggetti C, Cavallini A, Danova S, Egitto M, Ginevra F, Salati R, Bianchi P. Cerebral visual impairment in periventricular leukomalacia. Neuropediatrics. 1998;29:145-150. 6. Hoyt C. Visual function in the brain-damaged child. Eye (Lond). 2003;17:369-384. 7. Good WV, Jan JE, DeSa L, Barkovich AJ, Groenveld M, Hoyt CS. Cortical visual impairment in children. Surv Ophthalmol. 1994;38:351-364. 8. Good WV, Jan JE, Burden SK, Skoczenski A, Candy R. Recent advances in cortical visual impairment. Dev Med Child Neurol. 2001;43:56-60. 9. Khetpal V, Donahue SP. Cortical visual impairment: etiology, associated findings, and prognosis in a tertiary care setting. J AAPOS. 2007;11:235-239. 10. Groenveld M. Observations on the habilitation of children with cortical visual impairment. J Vis Impair Blind. 1990;84:11-15. 11. Flodmarkc O, Janc JE, Wongc PKH. Computed tomography of the brains of children with cortical visual impairment. Dev Med Child Neurol. 1990;32:611-620. 12. Matsuba CA, Jan JE. Long-term outcome of children with cortical visual impairment. Dev Med Child Neurol. 2006;48:508-512. 13. Cioni G, Fazzi B, Ipata AE, Canapicchi R. van Hof-van Duin J. Correction between cerebral visual impairment and magnetic resonance imaging in children with neonatal encephalopathy. Dev Med Child Neurol. 1996;38:120-132. 14. Persson EK, Anderson S, Wiklund L-M, Uvebrant P. Hydrocephalus in children born in 1999-2002: epidemiology, Handa et al: J Neuro-Ophthalmol 2018; 38: 429-433 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution 15. 16. 17. 18. 19. 20. 21. 22. outcome and ophthalmological findings. Childs Nerv Syst. 2007;23:1111-1118. Chen TC, Weinberg MH, Catalano RA, Simon JW, Wagle WA. Development of object vision in infants with permanent cortical visual impairment. Am J Ophthalmol. 1992;114:575-578. Afshari MA, Afshari NA, Fulton AB. Cortical visual impairment in infants and children. Int Ophthalmol Clin. 2001;41:159- 169. Philip SS, Dutton GN. Identifying and characterising cerebral visual impairment in children: a review. Clin Exp Optom. 2014;97:196-208. Lambert SR, Hoyt CS, Jan JE, Barkovich J, Flodmark O. Visual recovery from hypoxic cortical blindness during childhood computed tomographic and magnetic resonance imaging predictors. Arch Ophthalmol. 1987;105:1371-1377. Sabel BA. Restoration of vision I: neurobiological mechanisms of restoration and plasticity after brain damage a review. Restor Neurol Neurosci. 1999;15:177-200. Anderson V, Spencer-Smith M, Leventer R, Coleman L, Anderson P, Williams J, Greenham M, Jacobs R. Childhood brain insult: can age at insult help us predict outcome? Brain. 2009;132:45-56. Castano G, Lyons CJ, Jan JE, Connolly M. Cortical visual impairment in children with infantile spasms. J AAPOS. 2000;4:175-178. Wright T, Kumarappah A, Stavropoulos A, Reginald A, Buncic JR, Westall CA. Vigabatrin toxicity in infancy is associated with retinal defect in adolescence: a prospective study. Retina. 2017;37:858-866. Handa et al: J Neuro-Ophthalmol 2018; 38: 429-433 23. Carmant L. Vigabatrin therapy for infantile spasms: review of major trials in Europe, Canada, and the United States; and recommendations for dosing. Acta Neurol Scand. 2011;124:36-47. 24. Watson T, Orel-Bixler D, Haegerstrom-Portnoy G. Longitudinal quantitative assessment of vision function in children with cortical visual impairment. Optom Vis Sci. 2007;84:471-480. 25. Roman-Lantzy C. Cortical Visual Impairment: An Approach to Assessment and Intervention. New York, NY: American Foundation for the Blind Press, 2007. 26. Preston KL, McDonald M, Sebris SL, Dobson V, Teller DY. Validation of the acuity card procedure for assessment of infants with ocular disorders. Ophthalmology. 1987;94:644- 653. 27. Uggetti C, Egitto MG, Fazzi E, Bianchi PE, Bergamaschi R, Zappoli F, Sibilla L, Martelli A, Lanzi G. Cerebral visual impairment in periventricular leukomalacia: MR correlation. Am J Neuroradiol. 1996;17:979-985. 28. Westall CA, Ainsworth JR, Buncic JR. Which ocular and neurologic conditions cause disparate results in visual acuity scores recorded with visually evoked potential and teller acuity cards? J AAPOS. 2000;4:295-301. 29. Fazzi E, Signorini SG, Bova SM, La PR, Ondei P, Bertone C, Misefari W, Bianchi PE. Spectrum of visual disorders in children with cerebral visual impairment. J Child Neurol. 2007;22:294-301. 30. Good WV. Development of a quantitative method to measure vision in children with chronic cortical visual impairment. Trans Am Ophthalmol Soc. 2001;99:253-269. 433 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited.