AI-Powered OCT Analysis Enhances Understanding of Geographic Atrophy Progression and Treatment Response
核心洞察
AI-driven analysis of OCT images offers a more precise and automated method for assessing geographic atrophy (搜索) (GA) progression compared to traditional methods.
Post-hoc analysis of the FILLY trial demonstrates AI's superior accuracy in predicting GA growth rates using OCT data compared to expert ophthalmologists.
Studies of the OAKS and DERBY trials reveal that AI-quantified changes in retinal pigment epithelium (RPE (搜索)) and ellipsoid zone (EZ (搜索)) loss correlate with the therapeutic effects of pegcetacoplan.
Researchers are leveraging artificial intelligence (AI) to deepen the understanding of geographic atrophy (搜索) (GA), a late-stage form of age-related macular degeneration (搜索) (AMD). The focus is on early intervention and prevention of GA progression, with AI simplifying imaging analysis and providing more precise retinal details.
AI Analysis in the FILLY Study
Ursula Schmidt-Erfurth, MD, presented a post hoc analysis of an optical coherence tomography (OCT)-based AI study from the FILLY randomized clinical trial (NCT02503332) at the 2024 American Society of Retina Specialists Annual Meeting. The study analyzed the natural progression of GA in participants treated with pegcetacoplan therapy (Syfovre; Apellis Pharmaceuticals). The AI algorithm predicted progression based solely on baseline OCT images (Spectralis; Heidelberg Engineering).
The study included 134 eyes of 134 patients from the phase 2 clinical trial, with 2880 gradings performed by four ophthalmologists. The main outcomes were accuracy, weighted κ, and concordance index (c-index). Human experts achieved accuracy values of 0.37, 0.43, and 0.41 on fundus autofluorescence (FAF), near-infrared reflectance (NIR)+OCT, and FAF+NIR+OCT, respectively. Their c-indices were 0.62, 0.59, and 0.60. In comparison, the automated AI-based analysis reached an accuracy of 0.48, a κ value of 0.23, and a c-index of 0.69 using only OCT imaging.
Schmidt-Erfurth noted that AI performed in an automated manner, proving superior and faster than analyses performed by ophthalmologists. The advantages of AI include full automation, reliable analysis of routine OCT, and accessibility via the cloud.
Insights from the OAKS and DERBY Trials
Another report highlighted AI's contribution to understanding GA. Schmidt-Erfurth led a post hoc OCT-based AI analysis to identify changes in the mean area of retinal pigment epithelial (RPE (搜索)) loss and ellipsoid zone (EZ (搜索)) loss over time. The results showed that OCT-based AI analysis objectively identified and quantified degeneration of the photoreceptor and RPE in patients with GA secondary to AMD.
The investigators quantified morphologic changes in photoreceptors and RPE (搜索) layers in GA patients treated with pegcetacoplan. These patients had participated in the OAKS (NCT03525613) and DERBY (NCT03525600) phase 3 clinical trials. Spectral-domain OCT images were analyzed over 24 months for changes in the mean area of RPE loss and EZ (搜索) loss in the pooled sham arms and the monthly (PM)/every other month (PEOM) treatment arms.
Key Findings from the Analysis
The analysis included 897 eyes of 897 patients. At 24 months, the analysis showed a therapeutic reduction of RPE (搜索) loss growth by 22% and 20% in the OAKS trial and 27% and 21% in DERBY for PM/PEOM compared with the sham arm, respectively. The reduction on the EZ (搜索) level was significantly higher with 53% and 46% in the OAKS trial and 47% and 46% in the DERBY trial for PM/PEOM compared with sham at 24 months.
"The therapeutic benefit for RPE (搜索) loss growth increased with larger EZ (搜索)-RPE difference quartiles from 21.9%, 23.1%, 23.9% to 33.6% for PM vs sham (P < .01 for all comparisons) and from 13.6% (P = .11), 23.8%, 23.8% to 20.0% for PEOM vs sham (P < .01 for all comparisons) in quartiles 1, 2, 3, and 4, respectively, at 24 months," the authors stated. "Regarding maintenance of the EZ layer, the therapeutic reduction of loss increased from 14.8% (P = 0.09), 33.3%, 46.6% to 77.8% (P < .0001) between PM and sham and from 15.9% (P =.08), 33.8%, 52.0% to 64.9% (P < .0001) between PEOM and sham for quartiles 1 to 4 at 24 months."
The authors concluded that OCT-based AI analysis objectively identifies and quantifies photoreceptor and RPE (搜索) degeneration in GA.
Schmidt-Erfurth and colleagues anticipate that AI-based clinical tools will become widely available via cloud-based technology, potentially benefiting providers, healthcare systems, and patients.