Macular Telangiectasia Type 2: A Classification System Using MultiModal Imaging MacTel Project Report Number 10

FROM OUTER RETINAL NEOVACULARIZATION TO EXUDATIVE SUBRETINAL NEOVASCULARIZATION IN MACULAR TELANGIECTASIA TYPE 2

PURPOSE

To describe the progression from outer retinal neovascularization (ORNV) to exudative subretinal new vessels (SRNVs) in idiopathic macular telangiectasia type 2.

METHODS

A total of 135 patients (270 eyes) imaged with optical coherence tomography angiography were included.

MAIN OUTCOME MEASURES

Ellipsoid zone loss, outer retinal hyperreflectivity, ORNV, and SRNVs. Outer retinal neovascularization was defined as a flow signal passing through the outer plexiform layer, with or without vertical linear outer retinal hyperreflectivity on the optical coherence tomography B-scan. Subretinal new vessels were defined as an abnormal capillary network with a peripheral anastomotic arcade seen on en face optical coherence tomography angiography and a convex hyperreflectivity at the retinal pigment epithelium.

RESULTS

Subretinal new vessels were observed in 38/270 eyes (14%). Subretinal new vessels were at a fibrotic stage in 24/38 eyes and at an exudative stage in 6/38 eyes, and a progression from ORNV to SRNVs was documented in 8/38 eyes. All cases showed an ellipsoid zone loss. In seven eyes (2.5%), SRNVs were also associated with subepithelial neovascularization. No retinochoroidal anastomosis was detected. The visual acuity dropped when SRNVs were present.

CONCLUSION

In this case series, SRNVs were found in 14% of eyes. In all cases, they were associated with an ellipsoid zone loss and with outer retinal hyperreflectivity. A progression from ORNV to SRNVs was observed.

Quantification of vascular changes in macular telangiectasia type 2 with AngioTool software

PURPOSE

To compare AngioTool (AT) vascular parameters (VP) between MacTel2 eyes and normal eyes. Secondary outcome measures were to correlate VP with BCVA and to analyze VP between various grades of Simple MacTel Classification.

METHODS

This is a retrospective study. SD OCTA images of the superficial vascular complex (SVC) and deep capillary complex (DVC) were exported into Image J and AT. The explant area (EA), vessel area (VA), vessel percentage area (VPA), total number of junctions (TNJ), junction density (JD), total vessel length (TVL), average vessel length (AVL), total number of endpoints (TNE) and mean E lacunarity (MEL) were studied.

RESULTS

Group 1 had 120 MacTel2 eyes. Group 2 had 60 age-matched normal eyes. All VP were significantly different between the two groups except EA and TNE in both complexes. None of the VP had a correlation with BCVA. Interquadrant analysis (IQA) in SVC and DVC showed statistical significance in VPA, AVL and JD and in AVL, TNE, JD, VPA respectively. Post hoc analysis in SVC and DVC showed statistical significance in TNJ, JD, TVL and AVL between grade 1 and grade 3, and in VA, VPA, TNJ, JD, TVL and MEL between grade 0 and grade 3 respectively.

CONCLUSION

VP were affected in MacTel2 eyes. VP did not correlate with BCVA. Occurrence of pigmentation is an important event in the progression of disease. AT may provide quantitative markers to measure disease progression.

Self-Supervised Learning for Improved Optical Coherence Tomography Detection of Macular Telangiectasia Type 2

Importance

Deep learning image analysis often depends on large, labeled datasets, which are difficult to obtain for rare diseases.

Objective

To develop a self-supervised approach for automated classification of macular telangiectasia type 2 (MacTel) on optical coherence tomography (OCT) with limited labeled data.

Design, Setting, and Participants

This was a retrospective comparative study. OCT images from May 2014 to May 2019 were collected by the Lowy Medical Research Institute, La Jolla, California, and the University of Washington, Seattle, from January 2016 to October 2022. Clinical diagnoses of patients with and without MacTel were confirmed by retina specialists. Data were analyzed from January to September 2023.

Exposures

Two convolutional neural networks were pretrained using the Bootstrap Your Own Latent algorithm on unlabeled training data and fine-tuned with labeled training data to predict MacTel (self-supervised method). ResNet18 and ResNet50 models were also trained using all labeled data (supervised method).

Main Outcomes and Measures

The ground truth yes vs no MacTel diagnosis is determined by retinal specialists based on spectral-domain OCT. The models' predictions were compared against human graders using accuracy, sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), area under precision recall curve (AUPRC), and area under the receiver operating characteristic curve (AUROC). Uniform manifold approximation and projection was performed for dimension reduction and GradCAM visualizations for supervised and self-supervised methods.

Results

A total of 2636 OCT scans from 780 patients with MacTel and 131 patients without MacTel were included from the MacTel Project (mean [SD] age, 60.8 [11.7] years; 63.8% female), and another 2564 from 1769 patients without MacTel from the University of Washington (mean [SD] age, 61.2 [18.1] years; 53.4% female). The self-supervised approach fine-tuned on 100% of the labeled training data with ResNet50 as the feature extractor performed the best, achieving an AUPRC of 0.971 (95% CI, 0.969-0.972), an AUROC of 0.970 (95% CI, 0.970-0.973), accuracy of 0.898%, sensitivity of 0.898, specificity of 0.949, PPV of 0.935, and NPV of 0.919. With only 419 OCT volumes (185 MacTel patients in 10% of labeled training dataset), the ResNet18 self-supervised model achieved comparable performance, with an AUPRC of 0.958 (95% CI, 0.957-0.960), an AUROC of 0.966 (95% CI, 0.964-0.967), and accuracy, sensitivity, specificity, PPV, and NPV of 90.2%, 0.884, 0.916, 0.896, and 0.906, respectively. The self-supervised models showed better agreement with the more experienced human expert graders.

Conclusions and Relevance

The findings suggest that self-supervised learning may improve the accuracy of automated MacTel vs non-MacTel binary classification on OCT with limited labeled training data, and these approaches may be applicable to other rare diseases, although further research is warranted.

Developing a Continuous Severity Scale for Macular Telangiectasia Type 2 Using Deep Learning and Implications for Disease Grading

PURPOSE

Deep learning (DL) models have achieved state-of-the-art medical diagnosis classification accuracy. Current models are limited by discrete diagnosis labels, but could yield more information with diagnosis in a continuous scale. We developed a novel continuous severity scaling system for macular telangiectasia (MacTel) type 2 by combining a DL classification model with uniform manifold approximation and projection (UMAP).

DESIGN

We used a DL network to learn a feature representation of MacTel severity from discrete severity labels and applied UMAP to embed this feature representation into 2 dimensions, thereby creating a continuous MacTel severity scale.

PARTICIPANTS

A total of 2003 OCT volumes were analyzed from 1089 MacTel Project participants.

METHODS

We trained a multiview DL classifier using multiple B-scans from OCT volumes to learn a previously published discrete 7-step MacTel severity scale. The classifiers' last feature layer was extracted as input for UMAP, which embedded these features into a continuous 2-dimensional manifold. The DL classifier was assessed in terms of test accuracy. Rank correlation for the continuous UMAP scale against the previously published scale was calculated. Additionally, the UMAP scale was assessed in the κ agreement against 5 clinical experts on 100 pairs of patient volumes. For each pair of patient volumes, clinical experts were asked to select the volume with more severe MacTel disease and to compare them against the UMAP scale.

MAIN OUTCOME MEASURES

Classification accuracy for the DL classifier and κ agreement versus clinical experts for UMAP.

RESULTS

The multiview DL classifier achieved top 1 accuracy of 63.3% (186/294) on held-out test OCT volumes. The UMAP metric showed a clear continuous gradation of MacTel severity with a Spearman rank correlation of 0.84 with the previously published scale. Furthermore, the continuous UMAP metric achieved κ agreements of 0.56 to 0.63 with 5 clinical experts, which was comparable with interobserver κ values.

CONCLUSIONS

Our UMAP embedding generated a continuous MacTel severity scale, without requiring continuous training labels. This technique can be applied to other diseases and may lead to more accurate diagnosis, improved understanding of disease progression, and key imaging features for pathologic characteristics.

FINANCIAL DISCLOSURE(S)

Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.

Unraveling the mysteries of macular telangiectasia 2: the intersection of philanthropy, multimodal imaging and molecular genetics. The 2022 founders lecture of the pan American vitreoretinal society

PURPOSE

Offer a personal perspective on the scientific advances on macular telangiectasia type 2 (MacTel2) since the launch of the MacTel Project in 2005.

DESIGN

Literature review and personal perspective.

METHODS

Critical review of the peer-reviewed literature and personal perspective.

RESULTS

Generous financial support from the Lowy Medical Research Institute laid the foundations of the MacTel Project. MacTel Project investigators used state of the art multimodal retinal imaging and advanced modern biological methods to unravel many of the mysteries surrounding MacTel2. Major accomplishments includes elucidation of the pathogenic role that low serine levels, elevated 1-deoxysphingolipids and other mechanisms induce mitochondrial dysfunction which lead to Müller cell and photoreceptor degeneration; the use of objective measures of retinal structures such as the area of ellipsoid zone disruption as an outcome measure in clinical trials; the demonstration that the ciliary neurotrophic factor slows down retinal degeneration and the development of a new severity scale classification based on multimodal imaging findings.

CONCLUSIONS

MacTel2 is a predominantly metabolic disease characterized by defects in energy metabolism. Despite relatively good visual acuities, MacTel2 patients experience significant visual disability. The Mac Tel Project has been instrumental in advancing MacTel2 knowledge in the past two decades.

Longitudinal anatomical and visual outcome of macular telangiectasia type 2 in Asian patients

Limited information regarding the anatomical and visual prognosis of macular telangiectasia (MacTel) type 2 in the Asian population is currently available. Herein, we conducted a retrospective longitudinal analysis of Japanese patients diagnosed with MacTel type 2. Disease progression was evaluated using the Simple MacTel Classification developed by Chew EY et al. in 2023, and its association with visual changes was analyzed. Sixteen eyes of eight Japanese patients were included in the study, with an average follow-up period of 8.2 ± 3.9 years (range, 2.2-14.0). At the initial visit, 7 (44%) and 5 (31%) eyes were classified as Grade 2 (central ellipsoid zone break) and Grade 3 (noncentral pigment), respectively. The proportion of eyes that progressed by 1 or 2-steps in grade after 1, 3, 5, 8, and 12 years was 0%, 14%, 43%, 70%, and 100%, or 0%, 7%, 7%, 30%, and 75%, respectively. The visual acuity significantly deteriorated during the follow-up period, particularly in the two eyes with full-thickness macular holes (FTMH). Three out of 7 patients exhibited low serum serine concentrations, although no apparent correlation with anatomical or visual outcomes was observed. Overall, this cohort demonstrated chronic disease progression, both anatomically and functionally, in eyes with MacTel type 2, with FTMH potentially associated with greater visual loss.