LONG-TERM PROGRESSION OF ELLIPSOID ZONE LOSS AND ASSOCIATED FEATURES ON OPTICAL COHERENCE TOMOGRAPHY IN MACULAR TELANGIECTASIA TYPE 2.

28/11/2025

Ophthalmology

Authors: Kristin Raming, Lukas Goerdt, Eva Begemann, Nele Steffens BsC, Jose Luis Rodriguez Garcia, Cathy Egan, Clemens Lange, Simone Tzaridis, Martin Friedlander, Tjebo Heeren, Peter Charbel Issa, Frank G Holz, Maximilian Pfau, Kristina Pfau

PURPOSE: To characterize the detailed natural history of ellipsoid zone (EZ) loss progression in Macular Telangiectasia Type 2 (MacTel) and to identify structural predictors of disease acceleration using demographic and OCT-based parameters.

DESIGN: Retrospective longitudinal cohort study.

SUBJECTS: Patients with MacTel registered for the Natural History Observation Registry (NHOR) study at the University Hospital Bonn, Germany.

METHODS: The EZ loss area was assessed on en-face maps deriving from ultradense OCT volume scans. The EZ loss was graded as either no EZ loss (Pattern 0), a sharply demarcated hyporeflective space in the outer retina with preserved retinal architecture (Pattern 1), a subsidence of retinal layers internal to the EZ into the EZ loss and a loss of retinal architecture (Pattern 2), or a mix of these morphologies (Pattern 3). A random forest model was trained to predict future progression rates using the explanatory variables age, baseline EZ loss area, and Pattern. Hierarchical Bayesian models adjusted for onset variability were applied to find an 'overall´ slope of disease progression.

MAIN OUTCOME MEASURES: EZ loss progression (mm/year) dependent on Pattern, baseline EZ loss area, age adjusting for the nested data of two eyes within one patient.

RESULTS: The study included 606 eyes from 327 patients (mean age, 59.8 ± 9.6 years; median follow-up 5.41 years [4.16-8.11], single visit in 210 patients). Median EZ loss progression was 0.08 mm/year. Lesion morphology was significantly associated with progression: Pattern 1 lesions progressed slowly (0.02 mm/year), while Pattern 2 and Pattern 3 lesions progressed more rapidly (0.08 and 0.09 mm/year, respectively; p < 0.001). Bayesian modeling revealed a sigmoid progression trajectory with an inflection point at approximately 0.5 mm of EZ loss and an estimated disease duration of ∼30-40 years. EZ loss type independently predicted acceleration beyond this threshold.

CONCLUSIONS: EZ loss in MacTel follows a characteristic sigmoid trajectory, with distinct morphologic types predicting progression velocity. While Pattern 1 lesions remain stable for longer periods, Pattern 2 and 3 indicate fast progression. These findings provide structural benchmarks that may help guide clinical management and treatment timing, particularly in light of emerging implant-based therapies such as Revakinagene taroretcel Encelto®.

PMID: 41320134