A recent study identified texture-based features in spectral domain optical coherence tomography that may serve as early biomarkers for intraocular inflammation in neovascular age-related macular degeneration patients undergoing anti-vascular endothelial growth factor treatment.
While the pathogenesis and adverse effects of intraocular inflammation (IOI) are well-documented, this research, published in Heliyon, sought to identify eyes at risk of developing IOI during anti-vascular endothelial growth factor (anti-VEGF) therapy. The investigators hypothesized that significant textural variations within the vitreous compartment, due to hyperreflective deposits and preretinal hyperreflective foci, may distinguish patients who developed inflammation from those who did not.
Using data from the phase 3 HAWK trial, which compared brolucizumab (Beovu) with aflibercept (Eylea) for neovascular age-related macular degeneration (nAMD) treatment, the study analyzed cases with IOI, endophthalmitis, and retinal vascular occlusion. The cohort included 67 eyes—34 with inflammation and 33 propensity-matched controls.
The researchers extracted 481 texture-based radiomics features from the vitreous compartment of spectral domain optical coherence tomography (SD-OCT) scans obtained prior to the onset of inflammation. Radiomics, an emerging field in medical imaging analysis, employs mathematical algorithms to extract quantitative data from medical images.
The most discriminating radiomic features distinguishing between control and safety groups were identified as:
- Median-Laws L3S3L3
- Skewness-Laws S3S3L3
- Median-Laws E3L3L3
- Median-Laws S3L3L3
- Kurtosis-Gabor XY-θ = 0.000, XZ-θ = 0.000, λ = 2.668, BW = 1
These features demonstrated potential in identifying eyes at higher risk of developing IOI during anti-VEGF treatment, even before clinical manifestation. The authors noted, the radiomics-based characterization of the vitreous compartment could identify early signals of adverse events, potentially aiding clinicians in assessing underlying posterior inflammation and informing decisions on subsequent anti-VEGF injections.
While promising, the investigators acknowledged the need for further research to validate these findings. The study's limitations included its retrospective nature and the potential for undetected IOI cases in the original trial. Full disclosures can be found in the study.