Neuro-Ophthalmology•

Macular exudate in idiopathic intracranial hypertension affects outer retina and visual acuity

Abstract

Background Optical coherence tomography (OCT) is suggested as a potential tool for retinal biomarkers in idiopathic intracranial hypertension (IIH). We explored how macular exudate (ME) affects retinal structure in IIH and investigated its relationship with their clinical features.

Methods Patients diagnosed with IIH and matched controls were enrolled. ME detection was done on fundus photography; swept-source OCT was used to image and measure the retinal sublayer thicknesses, including the retinal nerve fibre layer, ganglion cell-inner plexiform layer (GCIPL), inner nuclear layer (INL) and outer retinal layer (ORL). IIH patients underwent lumbar puncture where intracranial pressure (ICP) was assessed.

Results 195 eyes from 98 IIH patients (42 eyes had ME) and 224 eyes from 112 controls were included. IIH patients had thicker INL and ORL compared with controls (both p<0.001) while IIH eyes with ME had thicker INL and ORL thicknesses compared with eyes without ME (both p<0.05). In IIH patients, the retinal sublayer thicknesses correlated with their ICP levels, and GCIPL thickness correlated with visual acuity (VA). Furthermore, ME was associated with higher ICP, worse papilledema and lower VA (all p<0.001).

Conclusion ME affects retinal thickness in IIH patients and is associated with more severe clinical features in IIH. OCT may provide biomarkers informative of clinical changes in IIH. Further longitudinal studies are needed to explore the evolution of ME and its relationship to VA and retinal structure.

What is already known on this topic

  • Retinal structural changes in optical coherence tomography (OCT) are suggested as potential indicators for idiopathic intracranial hypertension (IIH), but the status of patients and method analysis caused inconsistent results among published studies.

What this study adds

  • Macular exudate (ME) affects retinal structural thickness and is associated with increased intracranial pressure and decreased visual acuity in IIH patients.

How this study might affect research, practice or policy

  • ME can be detected by OCT analysis in IIH and used in future clinical services and research.

Introduction

Idiopathic intracranial hypertension (IIH) is a clinical syndrome characterised by symptoms and signs of increased intracranial pressure (ICP).1 Several clinical biomarkers, such as papilledema and optic neuropathy, strongly suggest that increased ICP affects the optic nerve (ON), an observation confirmed by numerous reports.2 3

Although the ON is the main ocular site of pathological features in IIH, retinal changes outside of the ON may occur. Macular exudate (ME) is one of the most common retinal manifestations of IIH.4 Recent reports showed evidence that ME in IIH is related to alterations in the outer retinal structural integrity5 6; moreover, ME is increasingly acknowledged to be linked with visual abnormalities in IIH patients.7 8 In this regard, retinal optical coherence tomography (OCT) can image and measure the retinal structural integrity non-invasively and has been explored as a possible neuroimaging technique for finding biomarkers in most neurological diseases.9 Previous studies using OCT demonstrated changes of macular retinal nerve fibre layer (RNFL) and ganglion cell-inner plexiform layer (GCIPL) thickness in IIH patients compared with controls.10 11 However, the evaluation of the outer retina metrics has been underexplored.

The purpose of this study is to evaluate how ME affects the retinal structural thicknesses in IIH patients and its association with clinical features. Visual field abnormalities are part of the cardinal symptoms of IIH and clinical trials are making efforts to include visual abnormalities as clinical endpoints to test the efficacy of therapies.12 Thus, finding sensitive and reliable biomarkers for visual outcomes is of keen interest in IIH.

Materials and methods

Patient and public involvement

Patients or the public were not involved in the design, conduct, reporting or dissemination of this research.

Participants

This is a prospective cohort study that consecutively recruited patients diagnosed with IIH in the neurology department of West China Hospital, Sichuan University, from April 2021 to December 2023. IIH was diagnosed according to internationally accepted diagnostic criteria based on neurological examination, neuroimaging (magnetic resonance imaging or digital subtraction angiography), lumbar puncture and cerebrospinal fluid (CSF) examination.13 The exclusion criteria were: (1) neurological diseases, such as strokes, neuromyelitis optica spectrum disorders, multiple sclerosis and myelin oligodendrocyte glycoprotein antibody disease; (2) concomitant corticosteroid or immunosuppressant therapy; (3) poor OCT image quality. Controls included individuals age-matched and sex-matched who attended the hospital for medical check-ups without neurological or ophthalmologic disorders. Demographic information and vascular risk factors including hypertension, diabetes, dyslipidaemia, smoking status and drinking status were recorded for all participants.

Measurement of ICP

Lumbar puncture was performed in the left lateral decubitus position. After obtaining CSF, a spinal fluid manometer was connected to measure CSF pressure (in mm H2O). To avoid artificially elevated CSF pressure, the patients were relaxed with their legs extended and their necks in a neutral position. An assessment of the CSF pressure was conducted after waiting for 5–10 min and then the CSF was removed. A second measurement of CSF pressure was taken immediately after the CSF had been removed. CSF pressure was the height of the lowest part of the meniscus at the top of the fluid column.

Ophthalmic examination and OCT imaging

Participants had retinal photographs taken (head-centred images of the ON and macula images from both eyes). Evaluation of ME and papilledema was based on colour fundus photography and OCT images by two ophthalmologists blinded to their clinical information. ME was observed as hard exudates in colour fundus photography and hyper-reflective foci around the inner nuclear layer (INL) and outer retinal layer (ORL) in OCT images (figure 1).14 The modified Frisen scale was used to grade the papilledema of each eye based on colour fundus imaging from 0 (normal) to 5 (severe papilledema).15 The inter-rater kappa coefficient was 0.87 for ME and 0.92 for Frisen scores.

Figure 1
Request permissionrequest permission icon
Figure 1

The manifestation of control eyes and idiopathic intracranial hypertension eyes with macular exudate (ME) or without ME in fundus photograph, confocal scanning laser ophthalmoscope (cSSO) and optical coherence tomography (OCT) imaging. GCIPL, ganglion cell-inner plexiform layer; INL, inner nuclear layer; ORL, outer retinal layer; RNFL, retinal nerve fibre layer.

Patients with ophthalmic diseases, such as retinal vascular occlusion, retinal haemorrhages and cotton wool spots, were excluded from our study. Visual acuity (VA) was measured using a Snellen chart and the VA of both eyes was converted to a logarithm of the minimum angle of resolution for statistical purposes.

Swept-source OCT (SS-OCT) imaging

The SS-OCT (VG200S; SVision Imaging; V.2.1.016) was used for retinal structural imaging. Our previous study provided detailed specifications of the OCT tool.16 OCT images covered an area of 6×6 mm2 centred on the fovea with a raster scan protocol of 512 horizontal B scans and each B-scan contained 512 A scans.

Structural imaging

The OCT tool segmented the retinal structure automatically. The automated segmented layers were inspected for errors. Retinal imaging was done in both eyes. Here, we assessed the RNFL, GCIPL, INL and ORL, as shown in figure 1. OCT images with ophthalmic disorders, such as age-related macular degeneration, severe cataracts, diabetic retinopathy and glaucoma, were not included; eyes with optic atrophy were also excluded. If a participant had any of these disorders in one eye, the other eye was used; if both eyes had the aforementioned ophthalmic disorders, the participant was excluded from the study. OCT images with signal quality ≤7 were excluded from our data analysis. In our study, the mean sublayer thicknesses and thickness in each sector (superior, temporal, nasal and inferior) were analysed by the OCT tool and used in our data analysis. OCT data displayed in our study followed the OSCAR-IB quality criteria and APOSTEL recommendation.17 18

Statistics analysis

Baseline characteristics of participants were described as mean±SD for continuous variables and frequencies with percentages for categorical variables. The t-test, Kruskal-Wallis test and Fisher’s exact test were used to compare clinical characteristics between IIH and control groups. A generalised estimating equation (GEE) was applied to the OCT metrics between IIH patients and controls in mean thickness or thickness in the four sectors. Multivariable linear regression was used to compare OCT metrics among control eyes, IIH eyes with ME and IIH eyes without ME. Linear regression was also used to evaluate the correlation between OCT metrics and clinical features (ICP and VA). GEEs and linear regression covariates were age, gender and vascular risk factors. The forest plot was performed to display the correlation between clinical features and retinal thickness in control and IIH eyes with or without ME. Logistic regression was used for the comparison of disease features (ICP, VA and papilledema severity) between eyes with ME and eyes without ME. P values less than 0.05 (p<0.05) were considered statistically significant. All statistical analysis and plotting were conducted in R V.4.2.3.

Results

A total of 115 IIH patients and 113 controls were initially enrolled in our study. 17 participants were excluded due to carotid artery stenosis, corticosteroid treatment, macular pathologic findings or poor-quality OCT scans as shown in online supplemental figure 1. Our final data analysis included 195 eyes from 98 IIH patients (mean age=35.57±11.99 years; 44.79% men) and 224 eyes from 112 controls (mean age=34.65±12.00 years; 38.39% man). IIH patients showed reduced VA compared with controls. Of the 195 eyes from IIH patients, 42 eyes (21.54%) had ME; 164 eyes (84.10%) presented with papilledema. 6 patients (6.12%) had ME in unilateral eye and 18 patients (18.95%) had ME in bilateral eyes; besides, 6 patients had papilledema in unilateral eye and 79 patients (80.61%) had papilledema in bilateral eyes. Compared with the control group, IIH patients had thicker INL and ORL thicknesses (all p<0.001). However, there was no significant difference in RNFL (p=0.358) and GCIPL (p=0.612) thicknesses when both groups were evaluated. Table 1 displays the demographics and clinical characteristics of our study participants.

Table 1
|
Clinical features, retinal sign and OCT parameters for participants

Compared with controls, IIH patients showed thicker INL and ORL thicknesses in four sectors (all p<0.05, figure 2 and online supplemental table 1). The nasal sector (p=0.013) of the GCIPL was thicker, while the temporal (p=0.009) sector was thinner in IIH patients than in controls. No significant difference (p>0.05) was seen in the RNFL thickness when both groups were evaluated. When we divided IIH eyes into eyes with and without ME, we showed that IIH eyes with ME had increased RNFL, INL and ORL thickness compared with control eyes and IIH eyes without ME (all p<0.05, online supplemental table 1). Eyes without ME also showed thicker INL thickness than control eyes (p=0.008, online supplemental table 2).

Figure 2
Request permissionrequest permission icon
Figure 2

Comparison of optical coherence tomograph parameters between controls and idiopathic intracranial hypertension (IIH) patients in four sectors (superior (S), temporal (T), nasal (N) and inferior (I)). GCIPL, ganglion cell-inner plexiform layer; INL, inner nuclear layer; ORL, outer retinal layer; RNFL, retinal nerve fibre layer.

The correlation between OCT metrics and clinical features in IIH patients is shown in figure 3 and online supplemental table 3. A significant correlation between ICP and thickness of retinal sublayers was observed in IIH patients (p<0.001 for GCIPL and INL; p=0.016 for RNFL and p=0.011 for ORL). When IIH eyes were divided into eyes with ME and without ME, a significant correlation was found between ICP and thickness of INL and ORL in IIH eyes without ME (all p<0.001); contrarily, no correlation was observed between ICP and retinal sublayer thicknesses (RNFL, INL, ORL) in IIH eyes with ME. A significant correlation was seen between VA and RNFL and GCIPL thicknesses in eyes with and without ME (all p<0.001).

Figure 3
Request permissionrequest permission icon
Figure 3

Correlations between clinical features and optical coherence tomograph (OCT) parameters. (A) Correlation between intracranial pressure (ICP) and OCT parameters. (B) Correlation between visual acuity (VA) and OCT parameters. GCIPL, ganglion cell-inner plexiform layer; INL, inner nuclear layer; ME, macular exudate; ORL, outer retinal layer; RNFL, retinal nerve fibre layer.

Table 2 compares clinical features between IIH patients with ME and those without ME. IIH patients with ME had higher ICP levels, more severe papilledema and impaired VA (all p<0.001) compared with IIH eyes without ME.

Table 2
|
Clinical features compared between eyes with ME and eyes without ME

Discussion

In this study, we showed that IIH patients had thicker INL and ORL thicknesses than controls. When IIH eyes were stratified into eyes with and without ME, we showed that IIH eyes with ME had thicker INL and ORL thicknesses than eyes without ME and controls, respectively. In addition, IIH patients’ retinal sublayer thicknesses correlated with their ICP levels. Besides, ME in IIH patients was associated with higher ICP levels, more severe papilledema (measured by Frisen scores) and reduced VA.

The clinical use of subretinal thicknesses as a surrogate marker for IIH remains inconclusive, given the inconsistent findings between groups and various limitations associated with patients in OCT-based studies.19–22 These limitations include the difficulty for patients with papilledema and/or ME to cooperate with study protocols and the persistent challenge of diagnosing IIH clinically without ambiguity. Deviations in the reported results of macular RNFL and GCIPL in patients with IIH arise from intrinsic OCT apparatus variability including the eye tracking system, length of examination time, and resolution for adequately visualising the retina.16 23 Contrary to our previous report, we did not observe any significant differences in macular RNFL and GCIPL between IIH patients and controls due to different scanning areas.16 However, in our current study, when stratified into IIH eyes with or without ME, we showed that IIH eyes with ME had thicker RNFL thickness than controls or IIH eyes without ME, respectively. Here, we suggest that RNFL thickening in IIH eyes with ME may reflect axonal swelling due to capillary leakage and axoplasmic flow stasis, as previously reported.22

Previous studies predominantly focused on the inner subretinal thicknesses (RNFL and GCIPL) with less attention on the deeper structural thicknesses of the retina. In our current study, we showed that IIH patients had thicker INL and ORL thicknesses compared with controls. ME, a common clinical finding in IIH patients, is suggested to be associated with ocular glymphatic dysfunction,24 which affects the INL and ORL thicknesses. Here, we suggest that the thickening of the INL and ORL in IIH patients may be due to axoplasmic flow stasis (due to dysfunction of the retinal glymphatic system/haemostasis) as previously reported.2 25 Besides, we showed that the thickening of INL and ORL was more substantial in IIH eyes with ME than in IIH eyes without ME. It is plausible to suggest that fluid accumulation (ME) leads to ORL thickening.

Regarding the clinical utility of OCT metrics as a biomarker of ICP levels in IIH patients, previous studies suggested that changes in the retinal structural thicknesses (RNFL and GCIPL) may reflect the ICP levels in IIH patients.10 12 20 In our current study, we showed that RNFL and GCIPL thicknesses correlated with ICP levels in IIH patients, which was in line with prior studies.12 16 We also showed that INL and ORL thicknesses correlated with their ICP levels. The pathophysiology explanation for this observed association between the deeper structural thicknesses and the ICP levels in IIH patients is intriguing. Increased ICP is associated with papilledema, ultimately resulting in ME in the retina. Since the INL and ORL are associated with retinal haemostasis,24 our findings underpinning that changes in these retinal sublayer thicknesses are associated with ICP levels in IIH patients suggest that these layers may reflect the dysfunction of homeostasis coupled with increased ICP levels in IIH patients. Additionally, there were significant correlations between ICP levels and INL and ORL thicknesses in IIH eyes without ME; however, in IIH eyes with ME, there was no significant correlation between ICP and the thickness of INL and ORL. Given that the INL and ORL are sensitive to glymphatic changes in the retina, we suggest that these structural changes in IIH patients may be associated with their ICP levels before the occurrence of ME. Therefore, we highlight the need for clinicians to pay attention to the assessment of the INL and ORL thicknesses in the assessment of the retina to prevent ME and further complications.

In the retina, ME is considered to be due to extravascular accumulation of proteins and lipids.26 27 In IIH patients, it is suggested that ME is a result of papilledema7; IIH eyes with ME had higher ICP levels and worse papilledema. Many of the retinal changes seem to be directly proportional to papilledema severity. Here, we suggest that changes in ICP levels may cause papilledema (venous congestion and capillary leakage) which may ultimately lead to ME in the retina. Of note, we showed that IIH eyes with ME had reduced VA compared with eyes without ME. Previous studies demonstrated that ME results in reduced VA, which is in line with our findings.26–28 Taken together, our findings highlight the need for clinicians to prioritise ME assessment for timely intervention to improve the prognosis for IIH patients.

The strengths of this study are the relatively large sample of patients with confirmed IIH and analyses of both eyes per participant with fundus photography and OCT. Detection of ME was confirmed by fundus photography and a comprehensive structural imaging protocol that generated intra-retinal thickness metrics describing each layer of the retina was done automatically by the OCT tool.

Our study has some limitations. The observational, cross-sectional design of this study does not allow for inferences regarding the causal mechanisms between retinal changes, VA and ME in IIH patients. To address this limitation, we have initiated a longitudinal study to explore the mechanisms underlying retinal changes associated with visual changes, retinopathy signs such as ME and ICP. We did not assess retinal vessels in our study. Moreover, participants with ocular abnormalities and patients with severe papilledema and ME (detachment of the ORL) were excluded, which might introduce selection bias. We did not perform visual field examinations for all IIH patients in our study. Analysis of visual field, ME and OCT features will be more precise to explore their relationship.

The results of this comprehensive study highlight those changes of retinal thickness (especially the deeper layers, INL and ORL) are correlated with ME. We also showed that in IIH patients, INL and ORL thickness was correlated with ICP levels. Importantly, we showed that ME was associated with higher ICP, worse papilledema and impaired VA. Patients with IIH may benefit from OCT for detecting ME and retinal structure changes. Future prospective longitudinal studies are needed to determine whether ME can be used as a surrogate marker for IIH patients.

Overview

Article metrics
Altmetric data not available for this article.
Dimensionsopen-url