The Relationship between Choroidal Abnormalities and Visual Outcomes in Pediatric Patients with NF1-associated Optic Pathway Gliomas

Neurofibromatosis type 1 (NF1) is an autosomal dominant disorder that occurs in approximately 1:3,500 births. ^{1, 2} The disease can affect nearly all organ systems in the body and is particularly important for ophthalmologists since the eyes, ocular adnexa, and visual pathways may be involved with the potential for irreversible injury to visual function from compressive optic neuropathy and glaucoma. The diagnostic criteria for NF1 include ocular findings such as Lisch nodules, optic pathway gliomas, and, more recently added, choroidal abnormalities (CAs) visualized on near-infrared reflectance (NIR) imaging.^{3, 4}

CAs were first described in histopathologic reports as ovoid bodies in the choroid, consisting of hyperplastic Schwann cells, melanocytes, and ganglion cells.^{5, 6} In vivo, these lesions are undetectable by conventional ophthalmoscopy, but with advances in ophthalmic imaging and the widespread use of spectral domain optical coherence tomography (SD-OCT) with near-infrared reflectance (NIR) imaging of the fundus, CAs were identified as an objective biomarker in NF1.^7–9 These lesions appear as bright patchy lesions on SD-OCT in NIR mode and are commonly identified in patients with NF1. Flores et al. demonstrated that CAs are more prevalent and may be detected even earlier than Lisch nodules in pediatric patients with NF1.⁷ Previous studies have shown that CAs can progress in number and size over time.^{10, 11} More recently, Godinho et al. reported a significant association between the number and area of CAs and the presence of optic pathway glioma (OPG).¹² Interestingly, the impact of CAs on visual function in children with NF1-OPG has not been reported.

The purpose of this study is to evaluate the relationship between CAs and visual acuity in children with NF1-OPGs. We also aimed to investigate if changes in the number and size of the lesions are associated with visual structure and function in a multicenter cohort of children with NF1-OPGs followed longitudinally.

METHODS

This was a prospective observational cohort study of pediatric patients with NF1-OPGs from Boston Children’s Hospital, Children’s Hospital of Philadelphia, and The Hospital for Sick Children. The Institutional Review Board of each site approved the study. A parent or guardian of each participant gave written informed consent, and children also provided written assent when applicable as determined by the local Institutional Review Board. The study adhered to the tenets of the Declaration of Helsinki for research involving human subjects and were conducted in accordance with the regulations of the Health Insurance Portability and Accountability Act.

The study included children 18 years and younger with formal diagnostic of NF1 based on National Institute of Health diagnostic criteria,⁴ and with radiological diagnosis of an OPG. The clinical information from baseline and the last follow-up visit, including age, gender, race, best-corrected visual acuity (BCVA), OPG location, spectral-domain optical coherence tomography (SD-OCT) scans, and NIR images (Heidelberg Engineering, GmbH, Dossenheim, Germany) were extracted. The visual acuity assessment was performed using age-appropriate visual function tests¹³ following a standardized protocol and BCVA was converted to logMAR (Logarithm of the Minimum Angle of Resolution). Eligible subjects for this study were required to have good quality SD-OCT scans of the optic nerve and macula and at least 12 months follow-up.

Optical Coherence Tomography Parameters

At baseline and follow-up visits, patients underwent OCT scans centered on the optic nerve to measure the circumpapillary RNFL thickness and the macula to measure the ganglion cell layer (GCL) and inner plexiform layer (IPL) combined. After automated segmentation of each scan location, the same investigator (AG) would manually inspect and correct all segmentation errors. The retinal thickness map consists of three concentric rings with diameters of 1, 3, and 6 mm, with the 3- and 6-mm rings divided into quadrants. The ganglion cell-inner plexiform layer (GCIPL) thickness was calculated as an average of the four 3 mm subfields and four 6 mm subfields of combined GCL and IPL (the sum of GCL and IPL thickness, termed ganglion cell complex [GCC] hereafter) excluding the central area (1 mm radius) that corresponded to the foveola, as previously described.¹⁴

Assessment of Choroidal abnormalities

The NIR images centered at the optic nerve and macula of each eye at the baseline visit and at the last follow-up were automatically recorded using HRA + OCT 5.1.2.0 (Heidelberg Engineering, Heidelberg, Germany; excitation light, 488 nm, barrier filter, 500 nm). Two trained neuro-ophthalmologists evaluated each image for presence, size, and number of CAs and lesions were included by consensus. The borders of the hyperreflective lesions were manually delimitated using the OCT area tools (Figure 1), and the area of each lesion was automatically calculated. The total area of CAs per eye was calculated as the sum of the area of the lesions.

Figure 1.

Near-infrared reflectance images of the optic nerve (a) and macula (b) of children with Neurofibromatosis type 1-associated optic pathway gliomas.

Statistical Analysis

Demographic and clinical characteristics were summarized by standard descriptive summaries (e.g., means and standard deviations for continuous variables such as age and percentages for categorical variables such as gender). Cross-sectional associations of the presence, number, and total area of CAs with visual acuity, RNFL, and GCC at baseline and the last follow-up were evaluated by generalized regression models with adjustment by age and gender, and inter-eye correlation was account for by using generalized estimating equations. All statistical analyses were performed using SAS v9.4 (SAS Institute Inc., Cary, NC), and two-sided p-value<0.05 was considered statistically significant.

RESULTS

A total of 41 children (82 eyes) with NF1-OPGs from 3 centers were included in the study. The mean ± standard deviation (SD) age of the children was 10.2 ± 3.3 years at baseline and 11.8 ± 3.3 years at the last visit with mean follow-up period of 20.4 ± 7.2 months. Twenty-three (56.1%) children were female. The mean visual acuity in logMAR was 0.08 ± 0.31 at baseline and 0.08 ± 0.29 at the last follow-up visit. The mean RNFL was 84.0 ± 24.0 microns (ranging from 34 to 160 microns) at baseline and 83.1 ± 24.1 (ranging from 30 to 157 microns) at the last follow-up. The demographic and clinical characteristics of the children included in the study are described in Table 1.

Table 1.

Clinical features at baseline and last follow-up of the subjects included in the study

Demographic and Clinical feature	Baseline	Last Follow-up
Subjects	41	41
Age (years), mean (SD)	10.1 (3.3)	11.8 (3.3)
Gender (female)	23 (56.1%)
Follow-up (months), mean (SD)	20.4 (7.2)
Eyes	82	82
Presence of CAs, yes (%)	38 (46.3%)	39 (47.6%)
Number of CAs, median (IQR)**	1.0 (1.0, 3.0)	1.0 (1.0, 3.0)
CAs total area (mm2), Median (IQR)**	1.4 (0.9, 2.4)	1.6 (1.0, 2.7)
Visual acuity (logMAR)	N = 78	N = 80
Mean (SD)	0.08 (0.31)	0.08 (0.29)
Median (IQR)	0.00 [−0.10, 0.20]	0.00 [−0.10, 0.10]
RNFL (microns)	N = 77	N = 78
Mean (SD)	84.0 (24.0)	83.1 (24.1)
Median (IQR)	89.0 [67.0, 100.0]	87.0 [65.5, 98.0]
GCC Global (6mm+3mm) (microns)	N = 73	N= 70
Mean (SD)	70.8 (13.9)	69.6 (13.0)
Median (IQR)	74.5 [61.2, 79.0]	71.6 [58.6, 78.5]

CAs= choroidal abnormalities; IQR= interquartile range; GCC= Ganglion cell complex; Max= maximum; Min= minimum; RNFL= Retinal nerve fiver layer; SD= Standard deviation

CAs were identified in 38 eyes (46.3%) at baseline. Among the eyes with CAs, the number of CA per eye ranged from 1 to 7 lesions with mean of 2.1 ± 1.7 (median 1.0, [IQR] 1.0, 3.0) lesions and 22 (26.8%) eyes with only one CA. At the last follow-up visit, CAs were identified in 39 eyes (47.6%) with a mean of 2.2 ± 1.8 (median 1.0, [IQR] 1.0, 3.0) lesions per eye. The lesions had a mean total area of 2.0 ± 1.7 mm² (median 1.4, [IQR] 0.9, 2.4) at baseline and 2.3 ± 2.1 mm² (median 1.6, [IQR] 1.0, 2.7) at the last follow-up. Figure 2 shows a representative example of one eye with a single CA at baseline (a) and the last follow-up (b). In this eye, there was an increase of 0.20 mm² over 14 months interval.

Figure 2.

Near-infrared reflectance images of a child with Neurofibromatosis type 1-associated optic pathway glioma. Figure (a) shows a hyperreflective lesion with area of 0.89 mm², and figure (b) shows the same lesion 14 months later, measuring 1.09 mm².

In the regression analyses, the visual acuity was not significantly associated with the presence, total number, and area of CAs at baseline (P = 0.15, P = 0.27, P = 0.34, respectively) or last follow-up (P = 0.18, P = 0.09, P = 0.09, respectively). Neither RNFL nor GCIPL was associated with the presence, number, or total area of CAs (P > 0.08 in all comparisons, Table 2).

Table 2.

Regression analysis for evaluating the cross-sectional association between presence, number and area of choroidal abnormalities with visual acuity and optic coherence tomography parameters at baseline and last follow-up, adjusted by age and gender.

	Coefficient (SE)*	P-value	Coefficient (SE)*	P-value
CAs	Baseline VA N= 78		Last follow-up VA N= 82
Presence		0.19		0.16
No	0.13 (0.06)		0.12 (0.06)	0.16
Yes	0.04 (0.04)		0.04 (0.03)
Number (per unit increase) **	−0.02 (0.02)	0.40	−0.02 (0.01)	0.12
Total area (per unit increase) **	−0.02 (0.02)	0.43	−0.02 (0.01)	0.08
CAs	Baseline RNFL N= 77		Last follow-up RNFL N= 78
Presence		0.71		0.55
No	82.8 (5.2)		81.2 (5.2)
Yes	84.9 (3.8)		84.6 (3.7)
Number (per unit increase) **	0.21 (1.9)	0.91	0.83 (1.7)	0.62
Total area (per unit increase) **	0.89 (1.8)	0.61	1.13(1.3)	0.42
CAs	Baseline GCC (Global) N= 73		Last follow-up GCC (Global) N= 70
Presence		0.87		0.56
No	70.5 (3.2)		67.8 (3.1)
Yes	71.1 (2.0)		69.8 (1.9)
Number (per unit increase) **	−0.09 (1.1)	0.94	−0.22 (1.2)	0.85
Total area (per unit increase) **	0.22 (1.1)	0.84	0.04 (1.0)	0.97

CAs= choroidal abnormalities; SD= standard deviation; IR= interquartile; GCC= ganglion cell complex

^*Adjusted by age and gender

^**As a continues value

DISCUSSION

To the best of our knowledge, this is the first prospective study to examine the relationship between CAs and parameters of visual function and structure in children with NF1-OPGs. We did not find a statistically significant association of CAs (presence, number, or total area) with visual acuity and OCT parameters at baseline nor at the last follow-up. These data suggest that CAs are not associated with structure and visual function in pediatric patients with NF1-OPGs.

CAs are a new ocular biomarker for NF1, but their relationship with structure and visual function still needs to be clarified. These nodules are located in the deep layers of the choroid below the choriocapillaris. It has been suggested that their presence may compress the overlying choroidal vessels causing some degree of choroidal ischemia.¹⁵ Touzé et al.¹⁶ hypothesized that CAs could affect the electric activity of the RPE in subjects with NF1, but did not find a statistically significant difference in electrooculogram (EOG) values measures between NF1 patients with and without CAs, and no significant correlation between surface and number of CAs with measures of EOG. Godinho et al.¹² also assessed the correlation between CAs abnormalities and RNFL in a cohort of patients with NFI and found no significant correlation between RNFL and CAs. In contrast to these studies that evaluated the patients at one point in time and included NF1 patients with and without OPG, we evaluated the association of CAs parameters with visual structure and function in patients with NF1-OPG who were evaluated at baseline and also after a mean follow-up of 20 months. We aimed to understand if the presence, size, and number of these lesions were associated with visual acuity, the RNFL and GCIPL. This is essential information for clinicians since a decline in visual parameters affects the management and treatment decisions in these patients. However, in agreement with previous studies, we also did not find a significant relationship of CAs parameters with visual acuity, RNFL or GCIPL suggesting that if there are any associated vascular changes, they may not be sufficient to affect retina function.

In our study, CAs were present in 46% of eyes at baseline and 48% of the eyes at the last follow-up. Only one eye developed a new CAs during the follow-up period. The total area of CAs was fairly stable with median of 1.4 mm² at baseline and 1.6 mm² at the last follow-up. Interestingly, the prevalence of CAs in our cohort is smaller than the reported in the pediatric population with NF1, which varies between 64 −78.9 %.^{3, 7, 17, 18} It may be partly explained by the lower age of the children included in our study, with a mean age of 11.9 ± 3.1 years at the last follow-up.

Our study had limitations. The CAs were manually measured, which may have interfered with the accuracy of the measurements of the CAs, however, this was unlikely to significantly impact the results. Also, the NIR images were analyzed by a trained neuro-ophthalmologist and by a dual-trained pediatric and neuro-ophthalmologist to reduce observer bias. As another limitation, we only evaluated OCT NIR images 30^o field posterior pole and around the optic nerve; thus, we did not account for CAs present in a more peripheral location. Nevertheless, there is no standardized protocol to identify CAs, and most choroidal nodules are located in the posterior pole. ^{3, 19} Also, it could be questioned whether enhanced depth imaging (EDI) OCT would be helpful in identifying CAs. Although EDI OCT visualizes the depth extent of the choroid, the identification of CAs is based on the NIR images.

This study found that although prevalent, CAs are not ubiquitous in children with NF1-OPGs. Our results suggest that the presence, number, and size of CAs are not associated with RNFL and GCC thickness or visual acuity. However, further studies are necessary to evaluate the best imaging method to identify and determine the extent of these lesions, and longer follow-up may be required to confirm our findings.

Abbreviations:

BCVA

best corrected visual acuity

CAs

choroidal abnormalities

EDI

enhanced depth imaging

GCL

ganglion cell layer

GCIPL

ganglion cell-inner plexiform layer

IPL

inner plexiform layer

IQR

interquartile range

logMAR

logarithm of the minimum angle of resolution

MD

mean deviation

NF1

neurofibromatosis type 1

NF1-OPGs

NF1-associated optic pathway gliomas

NIR

near-infrared imaging

OCT

optical coherence tomography

OPG

optic pathway glioma

RNFL

retinal nerve fiber layer thickness

SD

standard deviation

SD-OCT

spectral domain optical coherence tomography

VA

visual acuity