Article Text
PDF
Abstract
Purpose Patients with non-proliferative macular telangiectasia type 2 (MacTel) have ganglion cell layer (GCL) and nerve fibre layer (NFL) loss, but it is unclear whether the thinning is progressive. We quantified the change in retinal layer thickness over time in MacTel with and without diabetes.
Methods In this retrospective, multicentre, comparative case series, subjects with MacTel with at least two optical coherence tomographic (OCT) scans separated by >9 months OCTs were segmented using the Iowa Reference Algorithms. Mean NFL and GCL thickness was computed across the total area of the early treatment diabetic retinopathy study grid and for the inner temporal region to determine the rate of thinning over time. Mixed effects models were fit to each layer and region to determine retinal thinning for each sublayer over time.
Results 115 patients with MacTel were included; 57 patients (50%) had diabetes and 21 (18%) had a history of carbonic anhydrase inhibitor (CAI) treatment. MacTel patients with and without diabetes had similar rates of thinning. In patients without diabetes and untreated with CAIs, the temporal parafoveal NFL thinned at a rate of −0.25±0.09 µm/year (95% CI [−0.42 to –0.09]; p=0.003). The GCL in subfield 4 thinned faster in the eyes treated with CAI (−1.23±0.21 µm/year; 95% CI [−1.64 to –0.82]) than in untreated eyes (−0.19±0.16; 95% CI [−0.50, 0.11]; p<0.001), an effect also seen for the inner nuclear layer. Progressive outer retinal thinning was observed.
Conclusions Patients with MacTel sustain progressive inner retinal neurodegeneration similar to those with diabetes without diabetic retinopathy. Further research is needed to understand the consequences of retinal thinning in MacTel.
- Retina
- Macula
- Imaging
- Degeneration
Data availability statement
All data relevant to the study are included in the article or uploaded as supplementary information.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
WHAT IS ALREADY KNOWN ON THIS TOPIC
The pathophysiology of macular telangiectasia type 2 (MacTel 2) is unclear but inner retinal loss, that is, neurodegeneration, has been reported, which is similar to what has been observed in diabetes mellitus without overt diabetic retinopathy.
A substantial portion of patients with MacTel 2 also have diabetes, but it has not been established whether a similar, increased or decreased level of neurodegeneration over time would be observed in MacTel 2.
WHAT THIS STUDY ADDS
Progressive thinning of the NFL/GCL layers occurs at a rate of 0.25 µm/year in patients with MacTel 2.
The rate of inner retinal loss is similar in those with Mactel 2 regardless of diabetes disease status.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
Further studies are needed to investigate the effects of inner retinal degeneration on visual function in Mactel 2.
Additional research is needed to determine whether therapies aimed at preserving inner retinal function might help reduce visual disability from MacTel 2.
Introduction
Macular telangiectasia type 2 (MacTel) is a bilateral, idiopathic condition that often results in chronic, progressive, central vision loss.1 2 Imaging findings among patients with the non-proliferative, that is without neovascularisation form of MacTel can consist of xanthophyll pigment redistribution on fundus autofluorescence, early leakage (especially temporally) of parafoveal capillaries on fluorescein angiography, as well as ellipsoid zone (EZ) loss and hyporeflective cavities in the retina on optical coherence tomography (OCT).1 3
While anti-angiogenic intravitreal injections have demonstrated efficacy in treating the neovascular complications of MacTel, there are no FDA-approved therapies for the non-proliferative form of MacTel, though early trial results with ciliary neurotrophic factor are encouraging.4 Anti-vascular endothelial growth factor (anti-VEGF) injections can reduce empty cystic cavities for the non-proliferative form of MacTel, but the clinical value is unclear and thus, the risk-benefit ratio of monthly intravitreal injections may not be favourable.5 6 We have shown that carbonic anhydrase inhibitors (CAIs)—which have a more favourable risk and treatment burden profile than repeated anti-VEGF injections—can also reduce macular cystic cavities and central macular thickness in non-proliferative MacTel, but analysis of the inner retinal layers, for example, nerve fibre layer (NFL) and ganglion cell layers (GCL), was not performed.7
Despite recent research elucidating some genetic underpinnings of a specific form of MacTel, its pathophysiology is still uncertain.8–10 Clinical exam and imaging findings suggest that both vascular and neurodegenerative components contribute to MacTel; Müller cell depletion has been demonstrated through postmortem histological analysis11 and OCT angiography has demonstrated abnormalities in the deep and superficial capillary plexi.12
Numerous studies have demonstrated a strong association of MacTel with diabetes mellitus, but the clinical features are distinct from diabetic retinopathy.13 14 Though the mechanistic relationship between MacTel and diabetes mellitus has yet to be established, both share a common feature whereby loss of the GCL and NFL can occur with no to minimal clinical manifestations of retinopathy.15–18 In diabetes without diabetic retinopathy, we found that the rate of NFL and GCL thinning was 0.25 µm/year and 0.29 µm/year respectively—extrapolated over 10–20 years, this is similar to the magnitude of inner retinal loss seen in severe glaucoma.18 19 Moreover, it is unknown whether the inner retinal loss in MacTel is progressive as it is in diabetes where it is increasingly recognised as an important component of the disease process.20 21 Establishing whether inner retinal neurodegeneration is progressive in MacTel would expand our understanding of disease mechanisms and pinpoint a potential target for future treatments. The purpose of the current study was to evaluate thinning in the NFL and GCL of non-proliferative MacTel patients and to evaluate how diabetes (with mild non-proliferative diabetic retinopathy at most) and CAI treatment may impact this thinning.
Methods
We carried out a University of Iowa Institutional Review Board-approved (#201703769) retrospective study of patients with MacTel presenting to two large, independent, geographically distinct retina practices (the University of Iowa and the Retina Consultants of Houston) with at least two Heidelberg Spectralis SD-OCT scans separated by 9 or more months. This study adhered to the tenets of the Declaration of Helsinki. Exclusion criteria included glaucoma, history of optic nerve disease, proliferative diabetic retinopathy, diabetic macular oedema, choroidal neovascularisation from any cause, vascular occlusion, prior plaque radiotherapy, anti-VEGF therapy or advanced age-related macular degeneration.
OCT imaging was performed with the Heidelberg Spectralis spectral-domain OCT (Heidelberg, Germany). All images were taken by an experienced certified ophthalmic photographer, centred on the fovea, registered volume scans acquired with Heidelberg’s TruTrak software to ensure direct comparison on longitudinal scans, and reviewed in real time to ensure they were of adequate quality. The scan protocol captured the central 20o × 15o with 19–61 sections and an automatic real-time averaging of nine frames.
OCT volumes were automatically segmented into 11 retinal layers using the Iowa Reference Algorithms (V.3.8.0). This consists of producing intraretinal surfaces then using a graph-based method to segment the retinal surfaces.22 23 The Iowa Reference Algorithms have been validated by us and others (eg,24) with reproducibility studies and yield a lower between measurements variation than the manufacturer supplied Spectralis software for 19 line scan OCTs in subjects with substantial fluid from diabetic macular oedema.25 A vertical intensity gradient obtains the cost function, and smoothness constraints are adopted between neighbouring A-scans.17 The retinal boundaries are then segmented by graph optimisation via solving the minimum s-t cut problem in the geometric graph used.
Segmented OCTs were reviewed and any image portions with inaccurate segmentation were removed manually. About 50% of scans required manual segmentation correction, mostly outside the central early treatment diabetic retinopathy study (ETDRS) regions. Corrections done within the ETDRS regions were necessary in <5% of scans. Regions containing empty-cystic cavities on OCT were also removed manually but the adjacent areas without empty-cystic cavities were included for analysis (online supplemental figure). As MacTel has a predilection for the temporal parafoveal region, we calculated thickness of the NFL and GCL (a) within ETDRS subfield 4 (temporal parafoveal) and (b) across all the area covered by the ETDRS grid (figure 1). For each visit of each person, we calculated years since baseline visit. As MacTel is known to affect Muller cells,11 26 we also performed thickness calculations for the inner plexiform layer and inner nuclear layer (INL), where the body of Muller cells reside.
Supplemental material
Regions of interest. Average thickness for NFL, GCL layers, INL, IPL and outer retina was computed within subfield 4 of the ETDRS grid (orange) and for the entire 6 mm diameter ETDRS circle (purple). Regions are aligned to an eye with MacTel in this Heidelberg Spectralis SLO image. MacTel, macular telangiectasia type 2; GCL, ganglion cell layer; NFL, nerve fibre layer; ETDRS, early treatment diabetic retinopathy study; IPL, inner plexiform layer; INL, inner nuclear layer; SLO, scanning laser ophthalmoscopy.
For each combination of layer and region of interest, three questions were asked1: What is the overall rate of NFL and GCL thinning in the eyes with MacTel?2 Does CAI use change the rate of inner retinal sublayer thinning in MacTel eyes?3 Does a diagnosis of diabetes mellitus change the rate of thinning in MacTel eyes? To answer these questions, we constructed a linear mixed effects model:
(For an explanation of each term in the model, see the Supplemental Methods.) We fit this model to the four-layer/subfield combinations using the lme function from the nlme package (V.3.1; [Pinheiro J, Bates D, DebRoy S, Sarkar D, R Core Team (2020). nlme: Linear and Nonlinear Mixed Effects Models. R package V.3.1–144, <URL: https://CRAN.R-project.org/package=nlme>]) for R (V.3.6.3; [R Core Team (2020). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL https://www.R-project.org/]). We set the ‘control’ parameter to be ‘lmeControl(opt=‘optim’).’ Since we fit 10 separate models, we changed our threshold for nominal significance from 0.05 to 0.005 by Bonferroni correction.
Patients in the CAI group were either on oral methazolamide or acetazolamide for off-label use at the discretion of the retinal physician when there were empty-cystic cavities on the OCT. Patients had to have been on the CAI continuously for the duration of the scans separated by >9 months to be included in this group.
SEs and CIs for the means of the diabetic and CAI groups were calculated from the SEs of the baseline group and the difference. The ± symbol indicates SE of the estimated mean.
Results
A total of 134 subjects with MacTel that had at least two visits with OCT scans were collected. 14 subjects were excluded due to having less than 9 months between imaging studies; two were excluded due to retinal vein occlusion; two were excluded due to choroidal neovascularisation; one was excluded due to glaucoma in both eyes. A total of 115 subjects (224 eyes) were included in the study (86 female; 29 male). The mean age at baseline was 63 years old (range: 36–over 80 years). Each patient was seen an average of 7.1 (range: 2–38) times over an average of 3.5 years (range: 0.8–8.1); see online supplemental figure for histogram of the number of patients seen per years of follow-up. A total of 57/115 subjects (50%) presented with diabetes with 7 subjects having any history of mild non-proliferative diabetic retinopathy (NPDR) consistent with a low reported rate of NPDR in MacTel,27 28 and 21/115 (18%) were treated with CAIs. A total of 1585 OCT volume scans were automatically segmented and manually reviewed for segmentation errors.
We computed thickness of the NFL and GCL in ETDRS subfield 4 (temporal parafoveal) and globally across the entire ETDRS grid (figure 1). Using these thicknesses, we fit mixed effects models for all the eyes in the dataset (figure 2). No statistically significant differences in thicknesses were identified in those with diabetes or treatment with CAIs compared with the MacTel only group at baseline (table 1). Considering all ETDRS subfields together, there was no significant change in NFL or GCL thickness over time (table 2) in those with MacTel only, those on CAI and those with diabetes. However, in MacTel patients without diabetes and untreated with CAIs, the NFL in subfield 4 thinned at rate of −0.25±0.09 µm/year (95% CI [−0.42 to –0.09]; p=0.003; table 2). In subfield 4, the rate of thinning observed in the eyes treated with CAI (−1.23±0.21 µm/year) was statistically significantly faster than in untreated eyes for GCL (−0.19±0.16 µm/year; p<0.001; table 2) and INL (−0.88±0.15 µm/year; p<0.001). The rate of NFL and GCL thinning in those with diabetes was not significantly different compared with patients without diabetes. There was progressive thinning of the outer retina in all ETDRS regions and subfield 4 in those with MacTel only (table 2).
Fitted rates of inner retinal thickness over time for all samples in 10 regions (A, nerve fibre layer (NFL) in ETDRS subfield 4; B, NFL of entire grid; C, ganglion cell layer (GCL) in ETDRS subfield 4; D, GCL of entire ETDRS grid; E, inner plexiform layer (IPL) in ETDRS subfield 4; F, IPL of entire grid; G, inner nuclear layer (INL) in ETDRS subfield 4; H, INL of entire grid; I, outer retina in ETDRS subfield 4; J, outer retina of entire grid). Each point represents thickness for each optical coherence tomographic (OCT) (ie, each scan per each visit) and each line represents the model fit for each eye which demonstrates a trend towards gradual thinning over time in these subgroups and layers. ETDRS, early treatment diabetic retinopathy study.
Retinal thickness (in μm) by region at baseline visit for those with MacTel only, those on carbonic anhydrase inhibitors (CAI) and those with diabetes mellitus. There was no statistically significant difference in average baseline thickness of any layer/region analysed with or without CAI treatment or diabetes
Rates of change (in μm/year) in retinal thickness by region for nerve fibre layer (NFL), ganglion cell layer (GCL), inner plexiform layer (IPL), inner nuclear layer (INL) and outer retina (OR). There was a significantly increased rate of NFL and OR thinning of ETDRS subfield 4 in those with MacTel and GCL and INL thinning in subjects who used CAIs
Discussion
MacTel remains a clinical diagnosis of uncertain pathophysiology. While previous reports have sought to illustrate the alterations in retinal layer thickness between patients with and without MacTel, these studies have not measured changes of individual layers within the inner retina (ie, NFL and GCL) longitudinally. In this study, we report the thickness of inner retinal layers at baseline and the rate of inner retinal thinning in MacTel, and further elucidate how diabetes and CAI treatment modify these rates.
There is an association of diabetes mellitus and MacTel that was reported to be 28% in one natural history study of MacTel patients and 59% in another,27 with our study demonstrating 49% having concomitant diabetes mellitus.13 Both disorders can result in inner retinal loss without clinical manifestations of vasculopathy.18 29 Thus, we sought to investigate how the rate of inner retinal thinning in MacTel patients with and without diabetes might compare to the rate of inner retinal thinning in patients with diabetes without retinopathy. In 2016, we demonstrated progressive loss of the entire parafoveal NFL of 0.25 µm/year in patients with diabetes with no to minimally evident diabetic retinopathy.18 In our population, the temporal parafoveal NFL thinned by a statistically significant average of 0.25 µm/year for patients with MacTel and no diabetes. Additionally, there was no significant difference in our study between measured rates of thinning with and without diabetes. We did not analyse the entire parafoveal region in this study as the disease process in MacTel predominates in the temporal parafoveal region, but the near identical rate of NFL loss in those with MacTel and those with diabetes with no to mild NPDR highlights the similarities between the two disorders. Currently, it is unclear whether the NFL loss is due to dendritic changes, ganglion or Muller cell dysfunction, vascular abnormalities that have been observed on OCT angiography12 and/or a variety of other possible mechanisms, but further research is needed in this area to elucidate this and its functional consequences. It may be possible that NFL and GCL could be used as an additional reliable imaging-based biomarker to track disease progression or treatment response for a therapeutic trial similar to how EZ band has been used previously.30
While many treatments have been attempted in MacTel, CAIs have shown promise as a relatively benign treatment that may help decrease the presence of macular abnormalities.7 A goal of CAI treatment in MacTel is to reduce the presence of cystic cavities within the retina and, ultimately, prevent vision loss. In the current study, there was no statistically significant difference in average baseline thickness of any layer/region analysed between patients who had and had not undergone carbonic anhydrase inhibitor treatment. There was, however, a significantly increased rate of thinning of the temporal parafoveal GCL and INL in patients who were treated with CAIs compared with those who were not. These results are in concordance with a study by Chen et al which demonstrated reduced central macular thickness in patients treated with CAIs.7 The implications of this increased thinning are unclear but are specific to the temporal parafoveal region with no significant thinning found in the entire macula or NFL. This increased thinning may be secondary to the resolution of subclinical oedema (ie, axonal/nuclear swelling), the reduction of adjacent intraretinal cystic cavities or could be indicative of true retinal degeneration in this region. Further functional and histopathological studies are needed to better understand this finding in addition to the potential of CAI on long-term rates of visual complications, including the context of Muller cell loss that has been described in those with MacTel.11
Weaknesses of the study include its retrospective nature, and lack of histopathological and functional correlations. We did not perform correlations of visual acuity and neuroretinal loss as (1) our findings were focused on the NFL and GCL of the temporal parafovea; (2) no global thinning of the retina was found; and (3) visual acuity loss has not yet been associated with retinal diabetic neurodegeneration.16 29 Moreover, correlations of ellipsoid zone loss and visual acuity have been established for this disorder30 and these analyses are an area of on-going research by our group as we also observed progressive outer retinal thinning in those with MacTel. While having a control group would have reduced the possibility of test-retest variability as the cause of these observations, a p value <0.005, sizeable n, extensive follow-up from baseline and tight CIs that do not include 0 strongly suggest that the findings are likely real. Moreover, a similar rate of loss of NFL and GCL in diabetic neurodegeneration with the Iowa Reference Algorithms18 has been verified by many other groups worldwide suggesting that the progressive thinning observed in this study is biologically plausible.
Although this is the first study quantifying progressive inner retinal thinning in patients with MacTel, future studies are needed to investigate the effects of inner retinal degeneration on visual function, the effect of carbonic anhydrase inhibitors on MacTel-associated neurodegeneration and the role of carbonic anhydrase inhibitors in preventing secondary vision loss in MacTel including geographic atrophy and choroidal neovascularisation.
Supplemental material
Data availability statement
All data relevant to the study are included in the article or uploaded as supplementary information.
Ethics statements
Patient consent for publication
Ethics approval
This study was approved by the University of Iowa Institutional Review Board (#201703769).
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
X @elliottsohnmd
Contributors Design of the study (CCW and EHS), collection and management of the data (ALA, CW, CC, LJL and AMR), analysis and interpretation of the data (ALA, SSW, KL, MDA, CCW and EHS), manuscript preparation (ALA, SSW and EHS), review and approval of the manuscript (ALA, SSW, CW, CC, LJL, AMR, ICH, JCF, HCB, EMS, SRR, KL, MDA, CCW and EHS). ES, the guarantor of contributorship.
Funding This study was funded by National Eye Institute (R01EY035435), NEI (P30EY025580).
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.