Published in Research

How prevalent are hyper-reflective foci in DME?

This is editorially independent content
4 min read

A new study published in Investigative Ophthalmology and Visual Science investigated the baseline distribution of hyper-reflective foci (HRF) in diabetic macular edema (DME) and their co-localization with cystoid intraretinal fluid (IRF).

Give me some background first.

Stemming from diabetic retinopathy (DR), DME is characterized by swelling in the macula and may lead to vision problems and blindness.

HRF, on the other hand, are small, distinct objects that generate a highly reflective signal on spectral-domain optical coherence tomography (SD-OCT). 

They have been proposed as a biomarker for disease severity and progression in DME and are present in the following retinal diseases:

  • DME
  • Neovascular age-related macular degeneration
  • Retinal vein occlusion
  • Uveitic macular edema

What does HRF have to do with DME?

Previous studies have shown inconsistent results as to whether the presence or number of HRF are indicators of the visual outcome following treatment—however, the investigators noted that a greater proportion of these studies showed that HRF at baseline predicted a poor visual outcome.

This led them to hypothesize whether a link exists between the baseline distribution of HRF within the retina and disease severity.

Now, talk about the study.

Investigators segmented 1,527 baseline SD-OCT volume scans from phase 3 clinical trials using a deep learning-based algorithm to detect HRF.

HRF count and volume were then assessed.

The clinical trials included:

  • YOSEMITE (NCT03622580) and RHINE (NCT03622593)
    • Both evaluated the efficacy, safety, and pharmacokinetics of Faricimab in participants with DME

How was the study conducted?

HRF distributions were analyzed regarding:

  • Best-corrected visual acuity (BCVA)
  • Central subfield thickness (CST)
  • IRF volume in quartiles
  • Diabetic Retinopathy Severity Scores (DRSS) in groups

Findings?

The researchers found that:

  • HRF were present in most patients (up to 99.7%)
  • HRF co-localized with IRF in the en face projection
  • Median HRF count was 64.0
    • Interquartile range (IQR) = 96.0
  • Median IQR HRF volume within the 3-mm diameter Early Treatment Diabetic Retinopathy Study (ETDRS) ring: 1964.3 (3325.2) pL

Go on …

Median HRF volumes were greater with:

  • Decreasing BCVA
    • nominal P = 0.0109
  • Increasing CST
    • nominal P < 0.0001
  • IRF
    • nominal P < 0.0001
  • DRSS up to very severe nonproliferative diabetic retinopathy (NPDR)
    • nominal P < 0.0001

Limitations?

The use of variable HRF definitions were not consistently reported, which made comparing past studies difficult.

The authors also noted that inconsistencies in the literature regarding HRF presence or count as a predictor for vision outcomes may have been due to biases introduced by manual counting.

Additionally, the algorithm used in this study could not track individual objects in a full three-dimensional segmentation. For a full overview of the limitations, see the study.

Expert opinion?

The researchers recommended that future studies use automated approaches to analyze full SD-OCT volume scans to minimize bias.

Take home.

This study showed that HRF were nearly universally present within the retina of DME patients and may be indicators of disease severity.


How would you rate the quality of this content?