Purpose
To investigate the development and progression of retinal pigment epithelial and outer retinal atrophy (RORA) secondary to maternally inherited diabetes and deafness (MIDD).
Design
Retrospective observational case series.
Methods
Thirty-six eyes of 18 patients (age range, 22.4-71.6 years) with genetically proven MIDD and serial optical coherence tomography (OCT) images were included. As proposed reference standard to diagnose and stage atrophy, OCT images were longitudinally evaluated and analyzed for presence and precursors of RORA. RORA was defined as an area of (1) hypertransmission, (2) disruption of the retinal pigment epithelium, (3) photoreceptor degeneration, and (4) absence of other signs of a retinal pigment epithelial tear.
Results
The majority of patients revealed areas of RORA in a circular area around the fovea of between 5° and 15° eccentricity. Over the observation time (range, 0.5-8.5 years), evidence for a consistent sequence of OCT features from earlier disease stages to the end stage of RORA could be found, starting with loss of ellipsoid zone and subretinal deposits, followed by loss of external limiting membrane and loss of retinal pigment epithelium with hypertransmission of OCT signal into the choroid, and leading to loss of the outer nuclear layer bordered by hyporeflective wedges. Outer retinal tabulations seemed to develop in regions of coalescent areas of RORA.
Conclusions
The development and progression of RORA could be tracked in MIDD patients using OCT images, allowing potential definition of novel surrogate markers. Similarities to OCT features in age-related macular degeneration, where mitochondrial dysfunction has been implicated in the pathogenesis, support wide-ranging benefits from proof-of-concept studies in MIDD.
Highlights
- •
The development and progression of RORA could be traced using OCT images.
- •
Evidence for a consistent sequence of OCT-features to RORA was found.
- •
Described OCT features in maternally inherited diabetes and deafness (MIDD) revealed similarities to RORA development in AMD.
- •
Findings suggest MIDD to serve as initial testbed for mitochondrial therapies.
- •
Proof-of-concept studies in MIDD could have wide-ranging benefits.
Introduction
Maternally inherited diabetes and deafness (MIDD, Online Mendelian Inheritance in Man # 520000) is a mitochondriopathy, which was first described in 1992, and is primarily caused by mutation in the Mitochondrially Encoded TRNA Leucine 1 ( MTTL1 ) gene at position 3243. , The majority (50%-85%) of patients reveal a pattern like dystrophy and/or retinal pigment epithelial and outer retinal atrophy (RORA) with phenotypic similarities to age-related macular degeneration (AMD). ,
New promising therapeutic strategies focusing on mitochondrial dysfunction have recently been presented. Mitochondrial dysfunction also has been implicated in the pathogenesis of multifactorial disease, such as AMD, which is the leading cause of legal blindness in the industrialized world. Given the fact that multifactorial diseases are driven by heterogeneous risk factors, the effect of these specific strategies can more effectively be evaluated in a pure mitochondriopathy. In this context, MIDD is a model disease for ocular phenotypes caused by mitochondrial dysfunction.
To accelerate clinical testing, meaningful, validated clinical endpoints need to be developed. Most interventional trials currently rely on the progression of RORA, which depicts the end stage of many retinal diseases and an accepted endpoint by regulators for studies of atrophic AMD (called “geographic atrophy” [GA]). , However, the most effective upcoming therapeutic approach might be directed to earlier disease stages. Therefore, ideal surrogate markers should be easily captured, reflect the current disease severity, be predictive for long-term progression based on short-term changes, and identify early disease-associated alterations ideally before the hitherto unknown point of no return.
A consensus of worldwide retinal specialists has recommended that optical coherence tomography (OCT) should be the primary imaging method to define RORA. Earlier OCT alterations have been described and shown to be associated with the development of RORA in AMD, which have the potential to serve as surrogate markers for therapeutic approaches and even allow for automated analysis. However, there has not been a similar investigation of OCT features in patients with MIDD. Therefore, we systematically and longitudinally investigated OCT scans in MIDD. We aimed to define features that might be relevant to the development and progression of RORA in this specific mitochondrial disease. The results of this pilot study may lead to further investigations that analyze the reliability of these parameters and accuracy as predictive factors.
Methods
This retrospective observational case series was performed at Moorfields Eye Hospital, London, United Kingdom. The study was in adherence with the Declaration of Helsinki. Approval by the local research governance committee from the Research and Development Department, and patients’ informed consent was obtained for the study.
Patient Selection
Eligible participants were identified by text search of electronic medical records from dedicated Medical Retina clinics of the study team at the Moorfields Eye Hospital, London, United Kingdom, between 2004 and 2019. The search terms were maternally inherited diabetes and deafness , MIDD , and m.3243A>G .
The inclusion criteria were defined as (1) presence of the m.3243A>G variant in the MTTL1 gene, (2) a compatible phenotype consistent with the clinical diagnoses of MIDD, and (3) serial OCT scans with an interval of at least 6 months. Other macular pathology and/or OCT images inadequate to evaluate (eg, insufficient image quality) led to exclusion. If both eyes met the inclusion criteria, both were included.
Imaging
All participants underwent a complete ophthalmic examination including best-corrected visual acuity testing, slit-lamp examination, and indirect ophthalmoscopy in routine clinical settings. Before multimodal imaging, pupils were dilated. The imaging protocol included spectral-domain OCT (Spectralis HRA+OCT, Heidelberg Engineering, Heidelberg, Germany) and short-wavelength fundus autofluorescence (AF) imaging (488-nm excitation and 500-700-nm emission) using 2 different generations of a confocal scanning laser ophthalmoscopy system (HRA classic, HRA 2, or Spectralis HRA+OCT; Heidelberg Engineering) according to the availability at the study center at the time of visit. The image resolution was 512 × 512 pixels for the HRA classic, whereas HRA 2 and Spectralis HRA+OCT recorded images with 768 × 768 pixels (high-speed mode) or 1,536 × 1,536 pixels (high-resolution mode). Using the device-inbuilt tracking feature, up to 100 images (centered on the fovea) were automatically aligned and averaged to optimize the signal-to-noise ratio.
Image Analysis
OCT scans were analyzed using the manufacturer’s software (Heidelberg Eye Explorer; Heidelberg Engineering). In accordance with the previously described practice, presence of RORA was defined as (1) a region of hypertransmission, (2) a zone of attenuation or disruption of the retinal pigment epithelium, (3) evidence of overlying photoreceptor degeneration, and (4) absence of scrolled retinal pigment epithelium or other signs of a retinal pigment epithelial tear. After defining the exact location of RORA, present and preceding OCT scans were investigated for alterations associated with it. Subsequent images were consecutively evaluated in the follow-up mode to ensure spatial reliability. In multimodal comparisons, landmarks such as vessel bifurcations were used for localization and alignment.
Results
Cohort Characteristics
Thirty-six eyes of 18 genetically confirmed MIDD patients (11 females, 7 males) were evaluated in the study ( Table ). The mean age (±SD, years) was 53.4 ± 11.6 (range, 22.4-71.6), the mean age of onset (ie, first reported appearance of visual symptoms or retinal alterations) of 47.3 ± 8.4 years (range, 21.0-62.0), and the mean retinal disease duration was 6.1 ± 6.2 years (range, 0.0-20.7) at baseline (ie, first recorded visit with OCT imaging). The age of onset appeared to describe a Gaussian distribution with one outlier that was extensively reported before, who showed subjective symptoms significantly earlier (21.0 years) than the other patients (≥38.5 years). All MIDD patients had been diagnosed with diabetes mellitus. However, only 4 eyes of 2 patients revealed diabetic retinopathy (graded as mild nonproliferative diabetic retinopathy). Cataract or pseudophakia was reported in all patients older than 50 years (not specified in more detail in the records). Two patients were described with cardiovascular complications (cardiomyopathy), 2 eyes with glaucoma, and 1 eye with Fuchs endothelial dystrophy.
| Parameter | Subjects |
|---|---|
| Number of eyes (patients) | 36 (18) |
| Sex, female/male | 11/7 |
| Age at baseline, y a | 53.4 ± 11.6 |
| Age of onset, y a | 47.3 ± 8.4 |
| Disease duration at baseline, y a | 6.1 ± 6.2 |
| Best-corrected visual acuity at baseline, logMAR a | 0.19 ± 0.25 |
| Presence of RORA at baseline, eyes (patients) | 33 (17) |
| Size of RORA at baseline, mm 2 a | 15.9 ± 20.1 |
| Focality of RORA at baseline a | 2.8 ± 1.8 |
| Follow-up time, y a | 4.6 ± 2.8 |
RORA at Baseline
At baseline, 3 eyes of 2 patients did not show any area of RORA ( Figure 1 , A). Both patients were among the youngest included subjects (22.4 and 50.5 years of age). The other 33 eyes revealed atrophic patches, mostly in a circular distribution around the fovea between 5 and 15° eccentricity ( Figure 1 , B). If the area of RORA exceeded approximately 12 mm 2 , it has been ascertained to extend into the central 5°. Eight eyes presented with fovea-involving RORA at baseline, which was associated in the majority (7 eyes) with a best-corrected visual acuity of <1.0 logarithm of the minimum angle of resolution (LogMAR) (>20/200 Snellen equivalent). As previously reported, the size of atrophy was very variable in MIDD. It ranged from 0.06 to 86.91 mm 2 in the present cohort ( Table ). Peripapillary atrophy was a quite common finding and present to some extent in 28 eyes of 14 patients. Peripapillary atrophy was not confluent with the macular RORA except in the 3 eyes with the most extensive atrophic changes. In this subgroup, some form of delineation was identified that was characterized by an interrupted line of apparently partly preserved outer retinal layers in OCT ( Figure 1 , C).
Pre-RORA features
Over the observation period of 4.6 ± 2.8 years (range, 0.5-8.5 years), a variety of OCT features were identified that were particularly found to signify high risk of progression to RORA. In the following, we describe outer retinal abnormalities such as external limiting membrane (ELM) and ellipsoid zone (EZ) disruption, subretinal deposits, hypertransmission of the OCT signal into the choroid and sclera, outer retinal layer thinning associated with subsidence of inner retinal layers, hyporeflective wedges, outer retinal tubulations, or subretinal hyperreflective lesion.
EZ and ELM integrity loss and subretinal deposits
EZ and ELM represent hyperreflective bands in the outer retina. In our cohort, the loss of EZ was the first sign of RORA development and usually associated with subretinal deposits between Bruch membrane and retinal pigment epithelium that resemble cuticular drusen. The subretinal deposits usually showed overlying retinal pigment epithelium as a hyperreflective cap and homogenous internal structure of medium-low reflectivity. They were associated with hyperreflective dots in IR image and hyperautofluorescent dots in AF imaging. The course showed reflectivity changes in the overlaying OCT layers leading to loss of a delimitable ELM ( Figure 2 ).
Retinal pigment epithelial integrity loss and hypertransmission of the OCT signal
The retinal pigment epithelium displays as the most external hyperreflective OCT layer. Vanishing subretinal deposits led to a spatially related retinal pigment epithelial signal discontinuation that was associated with hypertransmission of the OCT signal into the choroid and loss of AF intensity ( Figure 2 ). Both retinal pigment epithelial loss and hypertransmission are part of the definition of RORA and were therefore found in the area of RORA. However, they were also found in some transient states (ie, areas of retinal pigment epithelial disruption with preserved but partly abnormal outer nuclear layer [ONL]; Figure 3 ).
