June 2022
Volume 63, Issue 7
Open Access
ARVO Annual Meeting Abstract  |   June 2022
Characterizing early residual fluid in neovascular age-related macular degeneration using a machine learning algorithm in routine clinical practice
Author Affiliations & Notes
  • Anna K Wu
    Center for Ophthalmic Bioinformatics, Cleveland Clinic Cole Eye Institute, Cleveland, Ohio, United States
    Case Western Reserve University School of Medicine, Cleveland, Ohio, United States
  • Scott W Perkins
    Cleveland Clinic Lerner College of Medicine of Case Western Reserve University, Cleveland, Ohio, United States
  • Rishi P Singh
    Center for Ophthalmic Bioinformatics, Cleveland Clinic Cole Eye Institute, Cleveland, Ohio, United States
  • Footnotes
    Commercial Relationships   Anna Wu None; Scott Perkins None; Rishi Singh Apellis, Aerie, Graybug, Code F (Financial Support), Novartis, Genentech, Regeneron, Alcon, Bausch and Lomb, 41 Gyroscope, Code I (Personal Financial Interest)
  • Footnotes
    Support  This study was supported in part by the NIH-NEI P30 Core Grant (IP30EY025585), Unrestricted Grants from The Research to Prevent Blindness, Inc., and Cleveland Eye Bank Foundation awarded to the Cole Eye Institute.
Investigative Ophthalmology & Visual Science June 2022, Vol.63, 321 – F0152. doi:
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Anna K Wu, Scott W Perkins, Rishi P Singh; Characterizing early residual fluid in neovascular age-related macular degeneration using a machine learning algorithm in routine clinical practice. Invest. Ophthalmol. Vis. Sci. 2022;63(7):321 – F0152.

      Download citation file:


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

      ×
  • Supplements
Abstract

Purpose : Treating neovascular age-related macular degeneration (nAMD) often starts with intensive therapy with anti-vascular endothelial growth factor (VEGF) injections. In prior post-hoc studies, early residual fluid (ERF), defined as intraretinal fluid (IRF) and/ or subretinal fluid (SRF) after the loading dose, was associated with poorer long-term visual acuity. This retrospective study examined the impact of ERF on visual acuity using a machine learning (ML) algorithm to quantify fluid on optical coherence tomography (OCT) in routine clinical practice.

Methods : 286 treatment naïve patients with nAMD were identified from a retrospective analysis. Only one eye was included per patient. Best visual acuity (BVA) (ETDRS letters) and OCT data were collected every 3 months from baseline to 12 months. OCTs were analyzed by the Notal OCT analyzer ML algorithm, quantifying IRF, SRF, and total retinal fluid (TRF). The ERF group was defined as those with fluid at week (W) 12. Fluid type at W12 was stratified and volumes were divided into quartiles to see their predictive value for W52 BVA. Paired t-tests and ANOVA compared BVA and fluid changes within and between fluid subgroups.

Results : At W12, 58.4% of patients had ERF. The breakdown of subgroups at W12 were no fluid (41.6%), IRF-only (21.6%), SRF-only (45.5%), and IRF & SRF (32.9%). When comparing long term BVA outcomes according to W12 fluid status, ERF and ERF-free groups had similar mean BVA gains from baseline to W52 (+5.70±15.44 vs. +4.89±17.95; p= 0.69). There was also no significant difference in mean BVA gains from baseline to W52 when comparing the W12 fluid subgroups of no fluid (+4.89±17.95), IRF-only (+4.56±16.36 ), SRF-only (+5.55±12.49), and IRF & SRF (+6.61±18.45), p=0.93. Quartile analysis of W12 IRF, SRF, and TRF quantities revealed no predictive pattern for W52 BVA.

Conclusions : These results from routine clinical practice diverge from prior post-hoc studies, since there was no significant difference in long-term BVA gains between W12 ERF and ERF-free cohorts, as well as between the various fluid subgroups. In addition, fluid quantifications of ERF did not appear clinically relevant when predicting long term BVA. Thus, the clinical utility of a ML algorithm for predicting visual outcomes of ERF cohorts in nAMD warrants further exploration.

This abstract was presented at the 2022 ARVO Annual Meeting, held in Denver, CO, May 1-4, 2022, and virtually.

×
×

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.

×