June 2023
Volume 64, Issue 8
Open Access
ARVO Annual Meeting Abstract  |   June 2023
Comparison of artificial intelligence-based OCT and manual fundus autofluorescence measurements of geographic atrophy in the phase III OAKS/DERBY trials
Author Affiliations & Notes
  • Julia Mai
    Medizinische Universitat Wien, Wien, Wien, Austria
  • Gregor Sebastian Reiter
    Medizinische Universitat Wien, Wien, Wien, Austria
  • Wolf-Dieter Vogl
    RetInSight GmbH, Vienna, Vienna, Austria
  • Amir Sadeghipour
    RetInSight GmbH, Vienna, Vienna, Austria
  • Emma Foos
    Apellis Pharmaceuticals Inc, Waltham, Massachusetts, United States
  • Hrvoje Bogunovic
    Medizinische Universitat Wien, Wien, Wien, Austria
    Christian Doppler Laboratory for Artificial Intelligence in Retina, Vienna, Vienna, Austria
  • Ursula Schmidt-Erfurth
    Medizinische Universitat Wien, Wien, Wien, Austria
  • Footnotes
    Commercial Relationships   Julia Mai None; Gregor Reiter None; Wolf-Dieter Vogl None; Amir Sadeghipour None; Emma Foos None; Hrvoje Bogunovic Apellis Pharmaceuticals, Code F (Financial Support), Apellis Pharmaceuticals, Code R (Recipient); Ursula Schmidt-Erfurth Apellis Pharmaceuticals, Code C (Consultant/Contractor), Apellis Pharmaceuticals, Code F (Financial Support)
  • Footnotes
    Support  None
Investigative Ophthalmology & Visual Science June 2023, Vol.64, 1875. doi:
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Julia Mai, Gregor Sebastian Reiter, Wolf-Dieter Vogl, Amir Sadeghipour, Emma Foos, Hrvoje Bogunovic, Ursula Schmidt-Erfurth; Comparison of artificial intelligence-based OCT and manual fundus autofluorescence measurements of geographic atrophy in the phase III OAKS/DERBY trials. Invest. Ophthalmol. Vis. Sci. 2023;64(8):1875.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

Purpose : To perform an artificial intelligence (AI)-based analysis of retinal pigment epithelium (RPE) loss on optical coherence tomography (OCT) and compare it with the manual geographic atrophy (GA) measurements on fundus autofluorescence (FAF) in the prospective phase III OAKS and DERBY trials.

Methods : OCT (Spectralis, Heidelberg Engineering, Heidelberg, Germany) images from the OAKS (NCT03525613) and DERBY (NCT03525600) trials, two phase III, randomized controlled clinical studies of GA secondary to age-related macular degeneration with intravitreal pegcetacoplan were analyzed. RPE loss was measured on OCT in a fully automated way by a previously validated deep learning-based algorithm. The RPE loss was compared to the manual GA measurements on FAF at baseline and every two months up to month 24 for subjects pooled across all study arms.

Results : Data are presented as mean ± standard deviation (SD) in change from baseline (CFB) GA lesion size. At month 12, the mean difference in CFB between GA area on FAF and RPE loss on OCT was -0.169 ± 0.776 mm2 in the OAKS trial and -0.195 ± 0.881 mm2 in the DERBY trial. At month 24, the mean difference in CFB between FAF and OCT was -0.367 ± 1.035 mm2 in the OAKS trial and -0.285 ± 1.250 mm2 in the DERBY trial. While CFB in GA lesion size measured on FAF and CFB of the area of RPE loss measured on OCT is weakly to moderately correlated at early timepoints, the CFB measure by FAF and OCT were overall highly correlated: r = 0.774 in OAKS and r = 0.713 in DERBY at month 12 and r = 0.863 in OAKS and r = 0.813 in DERBY at month 24.

Conclusions : Automated OCT analysis was generally consistent for GA measurements compared to FAF when assessing change in lesion size. This suggests that AI-based assessment of GA on OCT can be used in future clinical trials and in routine clinical practice, upon the wider availability of automated tools.

This abstract was presented at the 2023 ARVO Annual Meeting, held in New Orleans, LA, April 23-27, 2023.

 

Pearson correlation coefficient between change from baseline GA lesion size on FAF and change from baseline RPE loss on OCT in mm2 from baseline through month 24 for study participants pooled across all arms (treatment and sham) in OAKS (blue) and DERBY (black).

Pearson correlation coefficient between change from baseline GA lesion size on FAF and change from baseline RPE loss on OCT in mm2 from baseline through month 24 for study participants pooled across all arms (treatment and sham) in OAKS (blue) and DERBY (black).

×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×