Correlation Between Ophthalmologic and Neuroradiologic Findings in Type 1 Neurofibromatosis

Neurofibromatosis Type 1 (NF-1) is a genetic disease affecting the eye, and ocular findings such as Lisch nodules (LN) or optic pathway gliomas (OPGs) are a part of its diagnostic criteria. Recent imaging technologies such as infrared (IR) imaging and optical coherence tomography (OCT) have highlighted the visualization of choroidal focal abnormalities in these patients, even in the absence of other ocular lesions. This study aimed to establish a mor- phological multimodal evaluation of choroidal findings in patients with NF-1, correlating them with central nervous system (CNS) findings.

Original Contribution Section Editors: Clare Fraser, MD Susan Mollan, MD Correlation Between Ophthalmologic and Neuroradiologic Findings in Type 1 Neurofibromatosis Gonçalo Godinho, MD, João Esteves-Leandro, MD, Gonçalo Alves, MD, Carolina Madeira, MD, Olinda Faria, MD, Elisete Brandão, MD, Augusto Magalhães, MD, Fernando Falcão-Reis, PhD, Susana Penas, MD Background: Neurofibromatosis Type 1 (NF-1) is a genetic disease affecting the eye, and ocular findings such as Lisch nodules (LN) or optic pathway gliomas (OPGs) are a part of its diagnostic criteria. Recent imaging technologies such as infrared (IR) imaging and optical coherence tomography (OCT) have highlighted the visualization of choroidal focal abnormalities in these patients, even in the absence of other ocular lesions. This study aimed to establish a morphological multimodal evaluation of choroidal findings in patients with NF-1, correlating them with central nervous system (CNS) findings. Methods: This retrospective study included 44 eyes from 22 patients with NF-1. Central 30° IR imaging was obtained, and the number and total area of detectable lesions were calculated. Both macular and optic disc scanning with OCT were performed, with and without the enhanced depth imaging technique, to assess the presence of choroidal focal hyperreflective lesions. Central macular thickness, ganglion cell layer, and outer nuclear layer thickness were assessed, as well as subfoveal choroidal thickness. The peripapillary retinal nerve fiber layer (RNFL) thickness was also assessed. Patients’ magnetic resonance images (MRI) were reviewed and categorized by a neuroradiology specialist, determining the presence of OPGs and CNS hamartomas. Correlations between the ophthalmological and neuroradiological findings were established. Results: Patients’ mean age was 16.4 ± 7.3 years and 59.1% were women. On the MRI, 86.4% of the patients had CNS hamartomas, and 34.1% of the eyes had OPGs. LN were described in 29.5% of the eyes, whereas a total of 63.4% of the eyes presented the characteristic hyperreflective lesions in IR imaging, all of them matching the underlying choroidal lesions. A mean of 2.9 ± 3.3 lesions per eye and a median total lesion area of 1.52 mm2 were found. Ophthalmology Department (GG, JE-L, CM, OF, EB, AM, FF-R, SP), and Neuroradiology Department (GA), Centro Hospitalar Universitário São João, Porto, Portugal; and Surgery and Physiology Department (FF-R, SP), Faculty of Medicine, University of Porto, Portugal. The authors report no conflicts of interest. Address correspondence to Gonçalo Godinho, MD, Alameda Professor Hernâni Monteiro, 4200-319 Porto, Portugal; E-mail: [e-mail] Godinho et al: J Neuro-Ophthalmol 2022; 42: 101-107 The presence of OPGs was correlated with a greater number (P = 0.004) and a larger area (P = 0.006) of IR lesions. For a cut-off of 3.5 lesions per eye, the sensitivity and specificity for the presence of OPGs were 75% and 80%, respectively. For a total lesion area of 2.77 mm2, the sensitivity and specificity for the presence of OPGs were 69.2% and 93.1%, respectively. Eyes with OPGs presented a significant reduction in the temporal RNFL (P = 0.018) thickness, as well as a reduction in subfoveal choroid thickness (P = 0.04). No relations were found between CNS hamartomas and ophthalmological findings. Conclusions: This study suggests that focal choroidal abnormalities are correlated with the presence of CNS lesions as OPGs in patients with NF-1, and it might be a surrogate for the need for CNS imaging in these patients. Journal of Neuro-Ophthalmology 2022;42:101–107 doi: 10.1097/WNO.0000000000001241 © 2021 by North American Neuro-Ophthalmology Society N eurofibromatosis Type 1 (NF-1) is a genetic disease affecting 1 in 3,500 patients (1–4). It is an autosomal dominant disorder with incomplete penetrance (5), caused by a mutation in the NF-1 gene, located in 17q11.2 (1,3,6), which encodes the tumor suppressor neurofibromin (6). Approximately 50% of the reported cases are spontaneous mutations (6). The most common clinical manifestations are café au lait spots, Lisch nodules (LN), and neurofibromas (5). The frequency of most of the NF-1 signs increases with age (5). Approximately 50% of the patients are diagnosed in the first year of life, and at age 8 years, most of the patients are already identified (5). The diagnosis of NF-1 is clinical, based on 2 or more signs defined by the National Institute of Health (NIH) in 1997: 6 or more cafè au lait spots, axillary or inguinal freckling, 2 or more cutaneous neurofibromas, 1 plexiform neurofibroma, distinctive osseous lesions, optic glioma, 2 or more LN, and a first-degree relative with NF-1 (1,2,6,7). 101 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution LN are the most frequently reported ocular lesions so far (3,5,8), affecting 90% of the patients older than 16 years (5) but are present in only 43% of pediatric patients (9). An optic pathway glioma (OPG) is a benign space-occupying tumor, which causes compression and demyelination of the fibers of the optic nerve (ON) (2), affecting 15%–20% of the patients (2–5,10) and is the lesion with the greatest potential impact in visual function (2,3,10). Among the patients with OPG, 20%–50% present decreased visual acuity or endocrine pathology, depending on its location due to hypothalamic–pituitary axis changes (2). Unlike most of the NF-1 lesions, OPGs are usually present at the age of 6, and afterward, its onset or progression is rare (2,5). Magnetic resonance imaging (MRI) is the gold standard technique for OPG detection, helping to establish NF-1 diagnosis and to identify its complications. There is no consensus about its indication in this disease. Although some defend its annual implementation, others suggest its use only in case of decreased visual acuity, proptosis, or neurological changes (2,4), as treatment of OPG is mainly indicated when visual compromise occurs. Recently, infrared (IR) imaging findings allowed the detection of hyperreflective areas (3,6–8,11,12), undetectable by funduscopy or fluorescein angiography, corresponding anatomically to the choroidal nodules detectable with enhanced depth imaging spectral-domain optical coherence tomography (EDI SD-OCT) (1,3,6,8,11). SD-OCT allows high-quality images and easy sequential noninvasive acquisition (2,6). These lesions are rather common and are often the first ocular manifestation of the disease, encouraging its inclusion in the diagnostic criterion for NF-1 (1). However, although OCT and IR imaging are easy to perform examinations, it can be challenging to perform them in very young children or children with neurodevelopmental deficits. Nevertheless, the relation between CNS and ocular findings in NF-1 has not been established so far. This study aimed to establish a morphological multimodal evaluation of the ophthalmologic findings in patients with NF-1, correlating them with CNS lesions. METHODS Patients enrolled in this study were diagnosed with NF-1 based on NIH criteria at a tertiary referral center hospital and had at least one ophthalmologic observation since January 2016. Only patients with a brain MRI within 2 years from the last ophthalmologic appointment were considered. It was a retrospective study, conducted following the tenets of the Declaration of Helsinki, and approved by the hospital ethics committee. Data on patients’ file were registered. The best-corrected visual acuity measured using the Snellen chart was converted to the logarithm of the minimal angle of resolution. 102 Slit-lamp examination and indirect binocular fundus ophthalmoscopy findings were registered. IR images centered on the macula with 30° were evaluated using the Spectralis HRA + OCT 5.1.2.0 (Heidelberg Engineering, Heidelberg, Germany; excitation light, 488 nm, barrier filter, 500 nm). Hyperreflective lesions were identified and manually segmented, and their area was calculated (Fig. 1). Optic disc and macular SD-OCT images were obtained with and without the EDI technique. Both linear horizontal scans crossing the fovea and macular 30 · 25° raster cube scanning were used. Peripapillary retinal nerve fiber layer (RNFL), central macular thickness (CMT), and central ganglion cell layer (GCL) and outer nuclear layer (ONL) thickness were automatically performed. For subfoveal choroidal thickness, a manually performed measurement using a caliper positioned from the retinal pigment epithelium external limit to the choroidal–scleral interface was performed. Using the confocal method, a correlation was sought between the lesions found in IR images and changes in OCT choroidal reflectivity (Fig. 1). The process was performed by a masked reader (G.G.) and reviewed by an OCT experient retinal specialist (S.P.). For the statistical analysis, added data on total number/area of hyperreflective lesions from both eyes were used. Findings in the brain MRI were reviewed by a neuroradiologist (G.A.), and the presence and location of OPGs and cerebral hamartomas were registered. Hamartomas were diagnosed as focal areas of T2 hyperintensity with no T1 changes or modification after gadolinium. Unidentified bright objects (UBO) and focal areas of signal intensity (FASI) were included. In patients with chiasmal or tract gliomas, both eyes were considered as having OPGs. Statistical analyses were performed using the SPSS software version 23.0 (SPSS, Chicago, IL). Categorical variables were reported as proportions and percentages. The normal distribution of continuous variables was verified by the Shapiro–Wilk test. Continuous variables were represented as mean ± SD, when normally distributed, or median (minimum–maximum) for variables with nonnormal distribution. When normality was defined, the t test was performed. If normality was denied, the Mann–Whitney test was applied. Correlation of the area of the lesions between eyes was calculated using the Spearman rank correlation coefficients. To evaluate correlations, the Pearson or Spearman test was used, according to the normality of the sample. To evaluate the strength of the correlation, Evan’s guide was used, classifying it as “very weak,” “weak,” “moderate,” “strong,” or “very strong.” The sensitivity and specificity for detecting the presence of OPG in one eye were evaluated for a wide range of the number and the total area of lesions per eye. The cut-off point was defined as the point closest to the upper left corner of the operational characteristics curve. Godinho et al: J Neuro-Ophthalmol 2022; 42: 101-107 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution FIG. 1. Multimodal confocal analysis of choroidal abnormalities. On the left image: manual segmentation of 2 hyperreflective lesions (HLs) using infrared images (outline) and their respective area; on the right image: an enhanced depth imaging spectral-domain optical coherence tomography horizontal scan presenting an hyperreflective choroidal lesion (outline) with a perfect anatomic correlation with the lower HLs. RESULTS Forty-four eyes of 22 patients with NF-1 were included, with a mean age of 16.4 ± 7.3 years. Most of the patients were women (n = 13; 59.1%). Thirteen eyes (29.5%) had LN described in the anterior segment evaluation. A total of 19 patients (86.4%) had identifiable CNS hamartomas on the MRI. OPGs were present in 15 eyes (34.1%) (Table 1). HL were detected in IR images of 28 eyes (63.4%). The number of HL detected in each affected eye was 2.9 ± 3.3, with a median of the total area of 1.52 (0.00–20.95) mm2. A moderate correlation was found between the patients’ age and the total number of HL (P = 0.017; R = 0.505) (Fig. 2). Although no significant correlation was initially found between patients’ age and HL total area (P = 0.067), when 2 outliers older than 30 years were removed from the analysis, the correlation turned to be strongly positive (P = 0.001; R = 0.685) (Fig. 3). The area of HL was highly correlated between both eyes (P = 0.001; R = 0.873). The OCT-assessed parameters such as mean CMT, ONL, GCL, RNFL, and subfoveal choroidal thickness for the whole sample are summarized in Table 2. Eyes with OPG presented a significantly higher number of HL than eyes without OPG (P = 0.004) (Table 3). Considering the ROC curve, for a cut-off of 3.5 lesions per eye, the sensitivity and specificity for detecting OPG on the MRI were 75% and 80%, respectively (ROC = 0.78). Eyes with OPG also had significantly greater HL areas than eyes without OPG (P = 0.006). Considering the ROC Godinho et al: J Neuro-Ophthalmol 2022; 42: 101-107 curve, for a cut-off of 2.77 mm2 of HL, the sensitivity and specificity for the presence of OPG were 69.2% and 93.1%, respectively (ROC = 0.763). Eyes of patients with OPG presented a significantly impaired VA (P , 0.001). Comparing patients with and without CNS hamartomas, no differences were found related to the number (P = TABLE 1. Demographic and clinical data Patients (n) Eyes (n) Age, yrs Sex—female/male, n (%) VA—logMAR (mean ± SD) Spherical equivalent—D (median [range]) Patients with CNS hamartomas, n (%) OPG location, n (%) Optic nerve Chiasm Optic tract Eyes with Lisch nodules, n (%) Eyes with OPG, n (%) Eyes with HL, n (%) No. of lesions/eye (mean ± SD) Lesion area/eye—mm2 (median [range]) 22 44 16.4 ± 7.3 13 (59.1)/9 (41.9) 0.0 (0.0–2.0) 0.0 (23.8; +7.0) 19 (86.4) 5 (50.0) 3 (30.0) 2 (20.0) 13 (29.5) 15 (34.1) 28 (63.4) 2.9 ± 3.3 1.52 (0.00–20.95) CNS, central nervous system; HL, hyperreflective lesions; logMAR, logarithm of the minimal angle of resolution; OPG, optic pathway glioma; VA, visual acuity. 103 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution FIG. 2. Graph illustrating the relation between the age of the patients and the total number of the hyperreflective lesions in both eyes using infrared reflectance. FIG. 3. Graph illustrating the relation between the age of the patients and the total area of the hyperreflective lesions in both eyes using infrared reflectance. 0.080) or area of HL (P = 0.067) (Table 4). No correlations were detected for the presence of CNS hamartomas and OPG (P = 1.0). When correlating SD-OCT thickness parameters with the HL, a negative moderate correlation was found between the mean HL area and the mean RNFL (P , 0.001; R = 20.565). No further correlation was found, namely with GCL (P = 0.134), ONL (P = 0.846), subfoveal choroidal (P = 0.185), and CMT (P = 0.879). Furthermore, no relation was found between the number and area of HL and the VA (P = 0.398 and P = 0.951, respectively). Eyes with OPG showed an increased subfoveal choroidal thickness (P , 0.001). These eyes also had a decreased temporal RNFL (P = 0.018). However, no other associations were found in relation to mean RNFL (P = 0.067), GCL (P = 0.196), ONL (P = 0.318), and CMT (P = 0.944) (Table 3). The presence of CNS hamartomas was neither correlated with the choroidal or the retinal layers’ thickness nor with the VA (P = 0.787). (Table 4). This number may be underestimated because these data were retrospectively collected from patients’ files. HL seems to be far more frequent than LN in patients with NF-1, with reported frequencies ranging from 71% to 100% (1,3,6,8). We described HL on IR reflectance in more than twice the patients with LN. Histopathologically, they seem to correspond to choroidal hamartomas and are expected to be increased in older patients (12). We found a moderate correlation between patients’ age and the total number of HL, as well as a strong correlation (removing 2 outliers) between age and the area of HL, which suggests a progressive increase in the number and growth of these lesions over time. We also found OPGs in 34.1% of the eyes, a higher percentage than described in the literature (15%–25%) (5,7,10). Nevertheless, as we only included patients who performed a brain MRI, our sample probably had a selection bias. The patients probably had a higher clinical TABLE 2. SD-OCT thickness CONCLUSIONS We found a significant correlation between CNS and ophthalmological findings in patients with NF-1, and to the best of our knowledge, this is the first study relating CNS findings with choroidal lesions in patients with NF-1. Recent developments in multimodal imaging in ophthalmology have promoted the search for biomarkers aiding in the diagnosis and follow-up decisions. LN and OPG are major ophthalmological diagnostic criteria for NF-1. The frequency of LN tends to increase with age and has been reported in 43% of the pediatric population and 90% of patients older than 16 years (5). In our study, LN were described in only 29.5% of the patients. 104 Central macular thickness, mm Macular ONL, mm Macular GCL, mm Peripapillary RNFL, mm Median average Nasal superior Nasal Nasal inferior Temporal inferior Temporal Temporal superior Subfoveal choroid 267 (201–308) 87.9 ± 4.0 15.1 ± 4.3 92.2 ± 21.7 103.0 ± 31.0 67.2 ± 21.5 132.1 ± 34.9 131.6 ± 31.1 66.6 ± 18.3 103.9 ± 33.1 351.3 ± 69.8 Values represented as mean ± SD or median (range). GCL, ganglion cell layer; ONL, outer nuclear layer; RNFL, retinal nerve fiber layer. Godinho et al: J Neuro-Ophthalmol 2022; 42: 101-107 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution TABLE 3. Clinical, near-IR, and OCT findings in eyes with and without OPG Number of HL Median area of HL, mm2 VA (logMAR) SD-OCT thickness, mm Central macular retina Macular ONL Macular GCL Peripapillary RNFL Median Nasal superior Nasal Nasal inferior Temporal inferior Temporal Temporal superior Subfoveal choroid With OPG* N = 15 Without OPG N = 29 P 10 (0–16) 4.03 (0.00–20.95) 0.2 (0.0–2.0) 2 (0–13) 0.85 (0.00–9.01) 0.0 (0.0–2.0) 0.004†‡ 0.006†‡ ,0.001†‡ 266 (201–308) 88.9 ± 4.9 13.4 ± 5.4 268 (238–293) 87.5 ± 3.6 15.7 ± 3.6 0.944 0.318 0.196 80.5 ± 29.2 94.9 ± 35.7 58.2 ± 21.8 91.2 ± 42.1 114.0 ± 46.7 56.8 ± 21.5 114.9 ± 40.9 385.7 ± 78.1 97.5 ± 15.2 106.6 ± 28.6 71.2 ± 20.4 109.7 ± 27.2 140.2 ± 25.0 71.0 ± 15.1 139.2 ± 22.5 336.5 ± 61.7 0.067 0.260 0.670 0.166 0.076 0.018†§ 0.0620 0.040†§ Variables represented as mean ± SD or as median (range). *Patients with chiasmal or tract gliomas, both eyes were considered as having OPG. † P , 0.05: statistically significant. ‡ Comparison made with the Mann–Whitney test. § Comparison made with the t test. GCL, ganglion cell layer; HL, hyperreflective lesions; IR, infrared; logMAR, logarithm of the minimal angle of resolution; OCT, optical coherence tomography; ONL, outer nuclear layer; OPG, optic pathway glioma; RNFL, retinal nerve fiber layer; VA, visual acuity. suspicion related to the presence of OPGs, such as visual function impairment, neurological changes, or proptosis, explaining the higher frequency of this lesion in our sample. We also found an association between the presence of OPGs and the number and the total area of HL. We cannot explain our finding, but we speculate that a common specific genetic predisposition for the development of tumors might be present in some of these patients. Given its specificities and sensitivities, we found that 3.5 HL per eye and a total HL area of 2.77 mm2 are values that could be used as guides for further brain MRI study in asymptomatic patients. We must acknowledge that the TABLE 4. Clinical, near-IR, and OCT findings in eyes of patients with and without CNS hamartomas Number of HL Median HL area, mm2 VA (logMAR) OCT thickness, mm Central macular retina Macular ONL Macular GCL Peripapillary RNFL Median Nasal superior Nasal Nasal inferior Temporal inferior Temporal Temporal superior Subfoveal choroid With Hamartomas N = 38 Without Hamartomas N=6 3 (0–16) 3.55 (0.00–20.95) 0.0 (0.0–2.0) 0 (0–3) 0.23 (0.00–1.56) 0.0 (0.0–2.0) 0.081 0.670 0.412 266 (201–308) 87.6 ± 3.9 14.8 ± 4.4 269 (240–277) 90.4 ± 4.8 17.0 ± 2.7 0.832 0.146 0.282 92.7 ± 21.6 106.3 ± 30.2 67.7 ± 22.1 105.9 ± 32.9 131.3 ± 34.3 64.7 ± 17.5 133.5 ± 30.8 357.5 ± 70.4 89.5 ± 24.0 82.8 ± 30.9 64.2 ± 19.1 92.3 ± 35.1 137.2 ± 41.7 78.0 ± 20.7 120.7 ± 33.5 301.2 ± 27.8 0.745 0.086 0.715 0.360 0.706 0.100 0.357 0.086 P Variables represented as mean ± SD or as median (range) GCL, ganglion cell layer; HL, hyperreflective lesions; IR, infrared; logMAR, logarithm of the minimal angle of resolution; OCT, optical coherence tomography; ONL, outer nuclear layer; OPG, optic pathway glioma; RNFL, retinal nerve fiber layer; VA, visual acuity. Godinho et al: J Neuro-Ophthalmol 2022; 42: 101-107 105 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution mean age of patients included is higher than the age of patients in whom OPG screening is more important, which does not allow further conclusions. However, we believe that further studies with a larger and younger subset of patients with NF-1 could clarify if the evaluation of HL could allow a more careful use of brain MRI, as it is an expensive, invasive, and time-consuming procedure that frequently requires sedation in younger or more neurologically affected patients (4). Although depending on patients’ cooperation, IR and OCT acquisition are cheaper, easier, and quicker to perform (6). Although it might be challenging to perform these examinations in particular cases as in younger children or patients with major neurodevelopmental deficits, an IR frame is even easier to perform than an OCT and in those particular cases, the information provided may allow for a better management of the disease. Our experience proved that with the help of a patient and experient OCT performer these examinations were often possible in more challenging cases, even in patients who did not collaborate for a visual acuity assessment. Nevertheless, we have to acknowledge that a lower HL number/area does not exclude the presence of OPG, and the decision to perform a brain MRI must be always equated by the patient’s physician. Several studies reported a decrease of peripapillary RNFL and ONL thickness in eyes with OPG (4–6,13). Chang et al (2) described a decrease in peripapillary and macular RNFL thickness in patients with NF-1 and OPG, reporting similar measures between patients with NF-1 without OPG and controls. In our sample, no differences were found in eyes with OPG related to CMT, GCL, or ONL thickness. However, the peripapillary RNFL was decreased, presumably because of an optic atrophy caused by an OPG. We also found that patients with OPG presented thicker choroids, probably because of the presence of the choroidal nodules that could alter the choroidal vascular anatomy. Curiously, a choroidal thinning in patients with NF-1 have been described when compared with healthy controls, and the authors suggested this could be due to an abnormal choroidal flow secondary to the local presence of these focal abnormalities (6). We found that eyes with OPG presented worse visual acuity than eyes without OPG. Nevertheless, we found no correlation between the visual acuity and the total number/ area of HL. Once a visual decline is determinant for OPG treatment decision, we have to be cautious when presuming that the imaging of choroidal lesions is enough for OPG screening because it does not replace a clinical evaluation. Notably, the presence of CNS hamartomas was not correlated with any of the ocular findings. This could mean that either they are unrelated findings or that our sample was not large enough for a significant correlation. Further studies with larger samples could provide more certainties on this matter. Recently, the importance of updating the NF-1 diagnostic criteria has been brought to light, considering the 106 recent imaging findings. Choroidal lesions are being proposed as new diagnostic criteria (14,15), which could increase the sensitivity and specificity of the current criteria and could provide an earlier diagnosis in younger patients. However, we have to acknowledge that our study presents some limitations. It is an observational, retrospective study, with a small sample. Despite a maximum of 2 years between brain MRI and IR imaging, choroidal and brain lesions could have changed during that time. Furthermore, the choroidal lesions were manually segmented, both in IR and SD-OCT images and therefore dependent on the observer. Furthermore, we only evaluated HL visible on the IR 30° central field, thus underestimating the total number and area of choroidal lesions that could be present in a more peripheral location. We also did not perform a transverse OCT imaging (mode C) to evaluate the thickness of the choroidal nodules that could give more detailed information on the anatomy of these lesions and potential correlation to CNS findings. This study concluded that choroidal abnormalities correlated with the presence of CNS lesions as OPGs in patients with NF-1, suggesting, for the first time, the use of choroidal findings as biomarkers for CNS abnormalities in this group of patients. However, further studies including larger and younger samples are needed to confirm or refute our findings. STATEMENT OF AUTHORSHIP Category 1: a. Conception and design: G. Godinho, J. EstevesLeandro, G. Alves, C. Madeira, O. Faria, E. Brandão, A. Magalhães, F. Falcão-Reis, and S. Penas; b. Acquisition of data: G. Godinho, G. Alves, and S. Penas; c. Analysis and interpretation of data: G. Godinho, J. Esteves-Leandro, G. Alves, C. Madeira, O. Faria, E. Brandão, A. Magalhães, F. Falcão-Reis, and S. Penas. Category 2: a. Drafting the manuscript: G. Godinho, C. Mandeira, and S. Penas; b. Revising it for intellectual content: G. Godinho, J. Esteves-Leandro, G. Alves, C. Madeira, O. Faria, E. Brandão, A. Magalhães, F. Falcão-Reis, and S. Penas. Category 3: a. Final approval of the completed manuscript: G. Godinho, J. Esteves-Leandro, G. Alves, C. Madeira, O. Faria, E. Brandão, A. Magalhães, F. Falcão-Reis, and S. Penas. REFERENCES 1. Cassiman C, Casteels I, Stalmans P, Legius E, Jacob J. Optical coherence tomography angiography of retinal microvascular changes overlying choroidal nodules in neurofibromatosis type 1. Case Rep Ophthalmol. 2017;8:214–220. 2. Chang L, El-Dairi M, Frempong T, Burner E, Bhatti M, Young T, Leigh F. Optical coherence tomography in the evaluation of neurofibromatosis type 1 subjects with optic pathway gliomas. J AAPOS. 2010;14:511–517. 3. Moramarco A, Giustini S, Nofroni I, Mallone F, Miraglia E, Iacovino C, Calvieri S, Lambiase A. Near-infrared imaging: an in vivo, non-invasive diagnostic tool in neurofibromatosis type 1. Graefes Arch Clin Exp Ophthalmol. 2018;256:307–311. 4. Topcu-Yilmaz P, Kasim B, Kiratli H. Investigation of retinal nerve fiber layer thickness in patients with neurofibromatosis1. Jpn J Ophthalmol. 2014;58:172–176. 5. Jett K, Friedman J. Clinical and genetic aspects of neurofibromatosis 1. Genet Med. 2010;12:1–11. Godinho et al: J Neuro-Ophthalmol 2022; 42: 101-107 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution 6. Abdolrahimzadeh S, Felli L, Plateroti R, Plateroti A, Giustini S, Calvieri S, Recupero S. Morphologic and vasculature features of the choroid and associated choroid–retinal thickness alterations in neurofibromatosis type 1. Br J Ophthalmol. 2015;99:789–793. 7. Nakakura S, Shiraki K, Yasunari T, Hayashi Y, Ataka S, Kohno T. Quantification and anatomic distribution of choroidal abnormalities in patients with type I neurofibromatosis. Graefes Arch Clin Exp Ophthalmol. 2005;243:980–984. 8. Ueda-Consolvo T, Miyakoshi A, Ozaki H, Houki S, Hayashi A. Near-infrared fundus autofluorescence-visualized melanin in the choroidal abnormalities of neurofibromatosis type 1. Clin Ophthalmol. 2012;6:1191–1194. 9. Kinori M, Hodgson N, Zeid J. Ophthalmic manifestations in neurofibromatosis type 1. Surv Ophthalmol. 2018;63:518– 533. 10. Abdolrahimzadeh S, Felli L, Plateroti A, Perdicchi A, Contestabile M, Recupero S. Spectral domain optical coherence tomography evidence of retinal nerve fiber layer and ganglion cell loss in adult patients with neurofibromatosis type 1. Retina. 2016;36:75–81. Godinho et al: J Neuro-Ophthalmol 2022; 42: 101-107 11. Byun Y, Park Y. Indocyanine green angiographic findings of obscure choroidal abnormalities in neurofibromatosis. Korean J Ophthalmol. 2012;26:230–234. 12. Vagge A, Camicione P, Capris C, Sburlati C, Panarello S, Calevo M, Traverso C, Capris P. Choroidal abnormalities in neurofibromatosis type 1 detected by near-infrared reflectance imaging in paediatric population. Acta Ophthalmol. 2015;93:e667–e671. 13. Avery R, Cnaan A, Schuman J, Chen C, Glaug N, Packer R, Ishikawa H. Reproducibility of circumpapillary retinal nerve fiber layer measurements using handheld optical coherence tomography in sedated children. Am J Ophthalmol. 2014;18:780–787. 14. Viola F, Villani E, Natacci F, Selicorni A, Melloni G, Vezzola D, Barteselli G, Mapelli C, Pirondini C, Ratiglia R. Choroidal abnormalities detected by near-infrared reflectance imaging as a new diagnostic criterion for neurofibromatosis 1. Ophthalmology. 2012;119:369–375. 15. Karaconji T, Whist E, Jamieson R, Flaherty M, Grigg J. Neurofibromatosis type 1: review and update on emerging therapies. Asia Pac J Ophthalmol. 2019;8:62–72. 107 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited.