Comparison of the retinal vascular network and structure in patients with optic neuritis associated with myelin oligodendrocyte glycoprotein or aquaporin-4 antibodies: an optical coherence tomography angiography study

Optical coherence tomography angiography measurements in neuromyelitis optica spectrum disorder and myelin oligodendrocyte glycoprotein antibody disease: A systematic review and meta-analysis

PURPOSE

Neuromyelitis optica spectrum disorder (NMOSD) and myelin oligodendrocyte glycoprotein antibody disease (MOGAD) are immune-mediated disorders that can often manifest with optic neuritis (ON) among other symptoms. Optical coherence tomography angiography (OCTA) is an emerging diagnostic method that can quantify retinal capillary blood flow and vessel density (VD), which have been shown to be affected in NMOSD and MOGAD. Hence, we aimed to systematically review the studies addressing retinal microvasculature using OCTA in these diseases.

DESIGN

Systematic review and meta-analysis.

METHODS

PubMed, EMBASE, and Web of Sciences were systematically searched to identify articles addressing OCTA measurements in patients with NMOSD or MOGAD. Following the data extraction, a meta-analysis was performed on the study population and OCTA types amongst at least two homogenous studies.

RESULTS

Twenty-two studies on NMOSD, MOGAD, or both were included. Parafoveal superficial retinal capillary plexus (SRCP) VD and radial peripapillary capillary (RPC) VD were diminished in NMOSD ON+ and NMOSD ON- groups compared to healthy controls (HCs). In addition, both the SRCP VD and RPC VD were significantly reduced in NMOSD ON+ compared to NMOSD ON-. However, meta-analysis for deep retinal capillary plexus (DRCP) did not show a significant difference between NMOSD patients and HCs, or among ON+ and ON- patients. Furthermore, there was no significant difference in foveal avascular zone (FAZ) area size between NMOSD patients and HCs. Regarding MOGAD, the meta-analysis showed decreased parafoveal SRCP VD and RPC VD in MOGAD ON+ patients compared to HCs. Comparing NMOSD ON+ and MOGAD ON+, a meta-analysis was conducted for RPC VD, which showed no significant difference between the two groups.

CONCLUSIONS

This systematic review and meta-analysis confirmed reduced VD in the macular and peripapillary areas in NMOSD and MOGAD eyes, particularly in the parafoveal SRCP and RPC, which is further impacted by prior ON.

Optical coherence tomography angiography of peripapillary vessel density in non-arteritic anterior ischemic optic neuropathy and demyelinating optic neuritis

Background

In cases of optic disc edema or a pale optic disc, distinguishing an episode of optic neuritis (ON) from that of non-arteritic anterior ischemic optic neuropathy (NAION) during a clinical examination is challenging. Optical coherence tomography angiography (OCTA) can reveal differences in peripapillary vascular network structures and provide biomarkers for differential diagnosis.

Methods

A total of 23 eyes with NAION, 22 eyes with demyelinating ON (DON), and 27 eyes from healthy participants were imaged using OCTA to observe the radial peripapillary capillaries (RPCs). Optical coherence tomography was used to measure peripapillary retinal nerve fiber layer (RNFL) thickness and the macular ganglion cell complex (mGCC). Data for all patients were recorded at 2-3 weeks and more than 3 months after the symptom onset.

Results

A total of 23 affected eyes from 23 patients with NAION (average age 52.17 ± 7.92 years), 22 eyes from 22 patients with demyelinating optic neuritis (DON) (average age 47.88 ± 19.24 years), and 27 eyes from 27 healthy individuals (average age 46.43 ± 14.08 years) were included in the study. There were no significant differences in sex, age, and eye laterality between any two groups (F = 0.968, 0.475, 0.870; p > 0.05). Throughout the course of NAION and DON, the superior RPC, superior mGCC, and peripapillary RNFL decreased with time (p < 0.05). In contrast, the inferotemporal RPC and inferior mGCC did not decrease from the acute to chronic stage in NAION (t = 1.639, 0.834, p = 0.117, 0.413). Compared with the normal group, patients with NAION and DON exhibited a sharp reduction in the average RPC, RNFL, and GCC from the acute to the chronic stage (p < 0.05). Patients with DON exhibited a significant decrease in the inferotemporal RPC and inferior mGCC compared with the patients with NAION (p < 0.05). In contrast, there were no significant differences in the inferior mGCC at the chronic stage between the patients with NAION and those with ON (t = 2.547, p = 0.093).

Conclusion

Various structural and microvascular changes were observed in patients with NAION and ON, indicating distinct features of the optic nerve during the different stages of NAION and ON. Peripapillary vascular density, measured using spectral domain OCT (SD-OCT), may be a biomarker to distinguish NAION from ON.

MOG antibody-associated optic neuritis

Myelin oligodendrocyte glycoprotein (MOG) antibody-associated disease (MOGAD) is a demyelinating disorder, distinct from multiple sclerosis (MS) and neuromyelitis optica spectrum disorder (NMOSD). MOGAD most frequently presents with optic neuritis (MOG-ON), often with characteristic clinical and radiological features. Bilateral involvement, disc swelling clinically and radiologically, and longitudinally extensive optic nerve hyperintensity with associated optic perineuritis on MRI are key characteristics that can help distinguish MOG-ON from optic neuritis due to other aetiologies. The detection of serum MOG immunoglobulin G utilising a live cell-based assay in a patient with a compatible clinical phenotype is highly specific for the diagnosis of MOGAD. This review will highlight the key clinical and radiological features which expedite diagnosis, as well as ancillary investigations such as visual fields, visual evoked potentials and cerebrospinal fluid analysis, which may be less discriminatory. Optical coherence tomography can identify optic nerve swelling acutely, and atrophy chronically, and may transpire to have utility as a diagnostic and prognostic biomarker. MOG-ON appears to be largely responsive to corticosteroids, which are often the mainstay of acute management. However, relapses are common in patients in whom follow-up is prolonged, often in the context of early or rapid corticosteroid tapering. Establishing optimal acute therapy, the role of maintenance steroid-sparing immunotherapy for long-term relapse prevention, and identifying predictors of relapsing disease remain key research priorities in MOG-ON.

Development of a deep learning model to distinguish the cause of optic disc atrophy using retinal fundus photography

The differential diagnosis for optic atrophy can be challenging and requires expensive, time-consuming ancillary testing to determine the cause. While Leber's hereditary optic neuropathy (LHON) and optic neuritis (ON) are both clinically significant causes for optic atrophy, both relatively rare in the general population, contributing to limitations in obtaining large imaging datasets. This study therefore aims to develop a deep learning (DL) model based on small datasets that could distinguish the cause of optic disc atrophy using only fundus photography. We retrospectively reviewed fundus photographs of 120 normal eyes, 30 eyes (15 patients) with genetically-confirmed LHON, and 30 eyes (26 patients) with ON. Images were split into a training dataset and a test dataset and used for model training with ResNet-18. To visualize the critical regions in retinal photographs that are highly associated with disease prediction, Gradient-Weighted Class Activation Map (Grad-CAM) was used to generate image-level attention heat maps and to enhance the interpretability of the DL system. In the 3-class classification of normal, LHON, and ON, the area under the receiver operating characteristic curve (AUROC) was 1.0 for normal, 0.988 for LHON, and 0.990 for ON, clearly differentiating each class from the others with an overall total accuracy of 0.93. Specifically, when distinguishing between normal and disease cases, the precision, recall, and F1 scores were perfect at 1.0. Furthermore, in the differentiation of LHON from other conditions, ON from others, and between LHON and ON, we consistently observed precision, recall, and F1 scores of 0.8. The model performance was maintained until only 10% of the pixel values of the image, identified as important by Grad-CAM, were preserved and the rest were masked, followed by retraining and evaluation.

Retina thickness in clinically affected and unaffected eyes in patients with aquaporin-4 immunoglobulin G antibody seropositive neuromyelitis optica spectrum disorders: a systematic review and meta-analysis

BACKGROUND AND PURPOSE

Retina thickness has been studied in patients with neuromyelitis optica spectrum disorders (NMOSD) without distinguishing serostatus and limited data are available in unaffected eyes. We aimed to investigate retina thickness in eyes of aquaporin-4 immunoglobulin G antibody seropositive (AQP4-IgG⁺) NMOSD patients with optic neuritis (AQP4-ON) and without (AQP4-NON).

METHODS

Eligible studies were identified by searching PubMed and Embase. Mean difference (MD, μm) with corresponding 95% confidence interval (CI) was pooled with random-effect models. The primary measures were average thickness of peripapillar retinal nerve fiber layer (pRNFL) centered on optic disc and the combination of ganglion cell layer and inner plexiform layer (GCIPL) at macula.

RESULTS

We included 21 studies enrolling 787 AQP4-IgG⁺ NMOSD patients. Compared with healthy control, pRNFL was thinner in eyes of AQP4-ON (- 32.78, 95% CI [- 36.24, - 29.33]) and AQP4-NON (- 2.76, 95% CI [- 3.94, - 1.58]), so was GICPL in AQP4-ON (-21.38, 95% CI [- 24.01, - 18.74]) and AQP4-NON (95% CI - 2.96, [- 3.91, - 2.00]). Compared with multiple sclerosis with ON, AQP4-ON had thinner pRNFL (- 13.56, 95%CI [- 16.51, - 10.60]) and GCIPL (- 9.12, 95% CI [- 11.88, - 6.36]). AQP4-ON and myelin oligodendrocyte glycoprotein antibody-associated demyelination with ON (MOG-ON) had similar pRNFL (0.59, 95% CI [- 6.61, 7.79]) and GCIPL thickness (- 0.55, 95% CI [- 2.92, 1.82]). AQP4-NON had similar pRNFL and GCIPL thickness to MOG-NON and multiple sclerosis without ON.

CONCLUSIONS

The average thickness of pRNFL and GICPL decreased both in AQP4-ON and AQP4-NON eyes. AQP4-ON eyes had a similar level of pRNFL and GICPL thinning to MOG-ON eyes, so did AQP4-NON to MOG-NON eyes.

Retinal structural and microvascular changes in myelin oligodendrocyte glycoprotein antibody disease and neuromyelitis optica spectrum disorder: An OCT/OCTA study

Purpose

To compare the optical coherence tomography (OCT)/OCT angiography (OCTA) measures in patients with neuromyelitis optica spectrum disorder (NMOSD) and myelin oligodendrocyte glycoprotein antibody disease (MOGAD).

Methods

Twenty-one MOG, 21 NMOSD, and 22 controls were enrolled in our study. The retinal structure [retinal nerve fiber layer (RNFL) and ganglion cell-inner plexiform layer (GCIPL)] was imaged and assessed with the OCT; OCTA was used to image the macula microvasculature [superficial vascular plexus (SVP), intermediate capillary plexus (ICP), and deep capillary plexus (DCP)]. Clinical information such as disease duration, visual acuity, and frequency of optic neuritis and disability was recorded for all patients.

Results

Compared with NMOSD patients, MOGAD patients showed significantly reduced SVP density (P = 0.023). No significant difference (P > 0.05) was seen in the microvasculature and structure when NMOSD-ON was compared with MOG-ON. In NMOSD patients, EDSS, disease duration, reduced visual acuity, and frequency of ON significantly correlated (P < 0.05) with SVP and ICP densities; in MOGAD patients, SVP correlated with EDSS, duration, reduced visual acuity, and frequency of ON (P < 0.05), while DCP density correlated with disease duration, visual acuity, and frequency of ON.

Conclusions

Distinct structural and microvascular changes were identified in MOGAD patients compared with NMOSD patients suggesting that the pathological mechanisms are different in NMOSD and MOGAD. Retinal imaging via the SS-OCT/OCTA might have the potential to be used as a clinical tool to evaluate the clinical features associated with NMOSD and MOGAD.

A comparative study of alteration in retinal layer segmentation alteration by SD-OCT in neuromyelitis optica spectrum disorders: A systematic review and meta-analysis

Background

To evaluate the feature of different retinal layer segmentation in neuromyelitis optica spectrum disorders (NMOSD) with spectral-domain optical coherence tomography (SD-OCT) and to compare it with that in multiple sclerosis (MS), healthy controls (HC), and idiopathic optic neuritis (ION).

Methods

We retrieved four electronic databases, including Pubmed, Embase, Cochrane Library, and Web of Science from inception to September 1st, 2021. A meta-analysis was performed to compare different retinal layer segmentation thicknesses between patients with or without a history of optic neuritis (ON) in NMOSD and the control group, including patients with MS, HC, and ION.

Results

Forty-two studies were included and the interval between the last ON onset and examination was greater than 3 months. Compared with that in HC eyes, the loss of retinal nerve fiber layer (RNFL) and macular ganglion cell and inner plexiform layer (GC-IPL) was serious in NMOSD eye especially after ON. Moreover, compared with that in ION eyes or MS-related-ON eyes, the injury to the peripapillary retinal nerve fiber layer (pRNFL) was severe in NMOSD-related-ON eyes. In addition, the correlation coefficient between pRNFL and prognostic visual acuity was 0.43. However, the one-arm study revealed the inner nuclear layer (INL) was thickened in NMOSD-related-ON eyes compared with HC eyes.

Conclusions

Inclusion of the RNFL and macular GC-IPL is recommended for monitoring disease progression and attention should be paid to changes in the INL.