Multimodal Imaging Characteristics and Diagnostic Value of Choroidal Nodules in Patients with Neurofibromatosis Type 1

Neurofibromatosis type 1 (NF-1) is an autosomal dominant disease caused by deletions or mutations in the neurofibromin gene located on chromosome 17p11.2.¹ It has a reported incidence of approximately 1/3000 and a prevalence between 1/4000 and 1/5000.² NF-1 is characterized by a range of findings, including nerve tumors that can develop in various parts of the body, cutaneous pigmentation changes (cafe au lait spots, axillary and inguinal freckling), vascular abnormalities, and bone lesions (pseudoarthrosis, sphenoidal wing hypoplasia), and is also commonly associated with ocular involvement.^{3, 4, 5}

The NF-1 diagnostic criteria determined by the National Institutes of Health include the presence of two or more Lisch nodules on the iris and optic nerve glioma as ophthalmological findings.⁶ In 2021, Legius et al.⁷ suggested that the presence of at least two choroidal abnormalities described as “bright, patchy nodules” on optical coherence tomography (OCT) or infrared reflectance (IR) imaging should also be used as a diagnostic criterion for NF-1. These lesions, which occur at the level of the choroid, are hamartomatous nodules also referred to as “Yasunari nodules”.⁸ Their reported prevalence in NF-1 cases ranges from 28% to 100%.^{8, 9, 10, 11}

The aim of this study was to evaluate the incidence of choroidal nodules in patients with NF-1, their multimodal imaging characteristics, and their place in the diagnosis of NF-1.

This retrospective study was conducted within the framework of the principles of the Declaration of Helsinki and approved by the Dokuz Eylül University Local Ethics Committee (approval number: 2023/10-01, date: 29.03.2023). We retrospectively analyzed the medical records and recorded OCT, enhanced depth imaging OCT (EDI-OCT), IR, OCT angiography (OCTA), and color fundus images of patients who were examined in the Dokuz Eylül University Faculty of Medicine, Department of Ophthalmology between January 2022 and December 2023 and had at least two of the National Institutes of Health diagnostic criteria for NF-1.⁶ Exclusion criteria were optic media opacity that could interfere with posterior segment imaging, comorbid systemic or ocular disease other than NF-1, and a history of systemic drug use.

All patients included in the study underwent a comprehensive ophthalmological assessment including anterior and posterior segment examination. Spectral-domain OCT images (7-mm fovea-centered radial B-scan sections with a30-degree lens, ART: 9), EDI-OCT images (9-mm horizontal sections passing through the fovea) and IR images (fovea-centered 55-degree area) were obtained with a standard protocol using the Heidelberg Spectralis (Heidelberg Engineering, Heidelberg, Germany) device. OCTA images covering a 12x12-mm area centered on the fovea and/or optic disc were obtained using the DRI OCT Triton (TOPCON, Tokyo, Japan) device, and central 45-degree color fundus images including the optic disc and macula were obtained using the VISUCAM 500 (Carl Zeiss Meditec, Jena, Germany) or DRI OCT Triton device. We recorded the patients’ age, gender, presence of cafe au lait spots, and cranial magnetic resonance imaging (MRI) findings for each patient; central macular thickness (CMT) and presence of Lisch nodules, eyelid plexiform neurofibroma, optic glioma, and Yasunari nodules in each eye; and subfoveal choroidal thickness (SFCT) for eyes with EDI-OCT images passing through the fovea.

Using the device software, CMT was determined by manually measuring the shortest vertical distance between the internal limiting membrane and Bruch’s membrane in OCT images passing through the fovea, and SFCT was determined by manually measuring the shortest vertical distance between the hyperreflective outer border of the retinal pigment epithelium (RPE) and the choroid-sclera junction in EDI-OCT images passing through the fovea.

The presence of optic glioma was evaluated by examining contrast-enhanced orbital MRI.

Lisch nodules were defined as yellowish brown dome-shaped solid hamartomatous lesions 1-2 mm in diameter on the iris surface on slit-lamp examination.¹⁰

Choroidal nodules were identified on IR images as domed, placoid, or patchy choroidal lesions with relatively irregular borders and high reflectivity.⁸ The presence of these nodules was also evaluated on OCT images with image quality ≥18 and OCTA images with image quality ≥40 for patients with recorded images.

Informed consent was not obtained because the study was retrospective and used no identifying data or images.

Statistical Analysis

Statistical analyses were performed using the SPSS statistical program (IBM Corp, Armonk, NY, USA). Descriptive statistics were used to summarize the data. Categorical variables are presented as numbers and percentages, and quantitative variables are presented as mean ± standard deviation.

The study included a total of 54 eyes of 27 patients. Descriptive statistics of the study group are given in Table 1; detailed demographic and clinical characteristics are shown in Table 2.

Fourteen (51.9%) of the patients were male and 13 (48.1%) were female. The mean age of the patients was 22.33±14.62 years (range, 8-57 years). All patients (100%) had at least 6 cafe au lait spots.

Lisch nodules were observed on slit-lamp examination in 40 (74.1%) of the 54 eyes. Neurofibroma on the upper eyelid was noted in 1 eye (1.9%) (patient 12, left eye).

Contrast-enhanced cranial MRI was performed in 23 patients (85.2%). Hamartoma was detected in 12 patients (52.2%); non-specific hyperintense foci, venous anomaly, and ventricular dilatation were each detected in 2 patients (8.7%); and thinning of the internal carotid artery, encephalomalacia, sphenoid bone dysplasia, plexiform neurofibroma, glial tumor, Chiari malformation type I, expansile lesion in the pons, and arachnoid cyst were each detected in 1 patient (4.3%). Three patients (12.9%) had no pathological findings.

Of 14 patients (51.8%) who underwent contrast-enhanced orbital MRI, 9 (33.3%) had optic glioma (unilateral in 8 patients [57.1%] and bilateral in 1 patient [7.1%]).

The mean CMT for all eyes was 244.31±25.17 µm (range, 181-380 µm), while the mean SFCT for 47 eyes with EDI-OCT images passing through the fovea was 320.49±63.63 µm (range, 139-469 µm).

IR imaging revealed at least one choroidal nodule in 52 eyes (96.3%). One pediatric patient (3.7%) had bilateral Lisch nodules but no detectable choroidal nodule in either eye (patient 11, 11 years old). Adequate quality OCTA images were available for 43 eyes (79.6%). In 31 (72.1%) of these eyes, choroidal nodules were observed as areas of low reflectivity in the choriocapillaris layer.

Two (3.7%) of the 54 eyes included in the study had no choroidal nodules despite the presence of Lisch nodules. In contrast, 16 eyes (29.6%) had no Lisch nodules despite the presence of choroidal nodules. In the remaining 36 eyes (66.7%), both Lisch nodules and choroidal nodules were detected.

The multimodal imaging characteristics of patient 2 are shown in Figures 1 and 2, and those of patient 18 are presented in Figures 3 and 4 as examples.

Histopathological studies have revealed that the choroidal nodules observed in NF-1 are hamartomatous in character, consisting of ovoid bodies that contain Schwann cells which proliferate in an annular structure around axons, and may also include neural crest-derived melanocytes and ganglion cells.^{1, 12, 13} These nodules are reported to impact choroidal blood flow, causing compression and thinning of the choriocapillaris and subsequently leading to choroidal and retinal thinning.^{14, 15} These lesions also show structural similarities to the cutaneous neurofibromas and Lisch nodules observed in the iris.¹⁶

Due to the high wavelength light used, IR imaging is a suitable modality for evaluating the ocular structures that lie beyond the RPE and is useful in detecting changes at the choroidal level.¹⁵ Although the choroidal nodules of NF-1 patients are not visible on fundoscopic examination, fundus autofluorescence imaging, or fluorescein angiography, they appear as bright, patchy areas in IR imaging and as hypocyanescent, patchy areas in the early phases of indocyanine green angiography.^{8, 13} In addition, the bright, patchy areas observed on IR imaging were found to be areas of high flow in the deep choroidal segment of OCTA.¹⁷ In our study, choroidal nodules were observed in 31 (72.1%) of 43 eyes with adequate quality OCTA images, appearing as areas of flow deficit in the choriocapillaris layer. Although it is generally accepted that these nodules cannot be observed on fundoscopic examination, some authors have stated that the degree of pigmentation may vary and as a result, areas of hyperpigmentation in the fundus might be detectable in some cases.^{11, 13}

Studies in the literature that used IR imaging to examine choroidal nodules in NF-1 are summarized in Table 3.^{8, 11, 13, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32}

There are publications reporting that choroidal abnormalities, which can be detected up to 100% of NF-1 patients, are much more common than cafe au lait spots (98%) and Lisch nodules (41-68%) (classically considered the most common findings of the disease) and may actually be the most common sign of NF-1.^{13, 33} Similar to cutaneous findings and Lisch nodules, choroidal findings are observed more frequently with older age, but they occur earlier than Lisch nodules. The prevalence of choroidal findings and Lisch nodules in pediatric cases has been reported to be 64-95% and 41-52%, respectively.^{19, 28}

Moreover, choroidal findings without Lisch nodules have been observed in 14-37% of NF-1 cases, while Lisch nodules without choroidal findings were observed in 2.5-16%.^{28, 31, 33} In our series, choroidal nodules were observed in 29.6% of eyes without Lisch nodules, whereas 3.7% of the eyes had only Lisch nodules.

Study Limitations

The main limitations of this study are that it was retrospective and single-centered with a relatively small number of patients. In addition, assessments that can indicate functional status, such as visual acuity and visual field, were not included in our study. However, the reason for this is that most previous publications examining choroidal nodules in NF cases have stated that these findings are asymptomatic and do not cause functional problems. Furthermore, as the OCTA images in our study were recorded for the purpose of evaluating for choroidal nodules, 12x12-mm images were obtained in order to evaluate the largest possible area. However, due to the features of the OCTA device we used, we could not obtain data on quantitative parameters such as vascular density and the foveal avascular zone with 12x12-mm images. Considering that choroidal nodules can lead to vascular alterations, especially at the choriocapillaris level, this can also be considered a limitation.

As NF-1 can cause ophthalmological findings, patients often present or are referred to ophthalmology clinics, both through outpatient clinic admissions and consultations by other medical disciplines. Therefore, it is important for ophthalmologists to recognize the ocular findings of NF-1. Although the presence of Lisch nodules can be evaluated by biomicroscopic examination, choroidal nodules can be observed much more frequently in these patients, as in our series, and can be easily detected using IR imaging, which is a rapid and non-invasive method. The presence of choroidal nodules before the appearance of Lisch nodules may facilitate early diagnosis in these patients.