RIPL count is a reproducible OCT biomarker for vision-threatening diabetic retinopathy, validated against ultra-widefield fluorescein angiography, with each additional RIPL increasing the odds of VTDR by 22%.
Key Findings
Results
Retinal ischemic perivascular lesions (RIPLs) were highly reproducible between two independent graders assessing diabetic retinopathy severity.
RIPLs were identified on a per-lesion basis using 61 serial B-scans from Heidelberg SPECTRALIS OCT with a 30° × 25° field of view.
RIPLs were defined as thinning of the inner nuclear layer, disruption of the outer plexiform layer, and elevation of the outer nuclear layer.
The study included 76 patients (97 eyes) with diabetic retinopathy who underwent same-day UWF-FA and OCT.
Results
Higher RIPL counts were significantly associated with vision-threatening diabetic retinopathy (VTDR).
Median RIPL counts were 0 in non-VTDR eyes and 4 in VTDR eyes.
Each additional RIPL increased the odds of having VTDR by 22% (95% CI, 1.09–1.37; P = 0.001).
Associations were evaluated using generalized estimating equation models clustered by patient.
53 eyes were graded as VTDR (54.6%) and 44 eyes as non-VTDR (45.4%).
Results
ROC analysis demonstrated good diagnostic performance of RIPL count for detecting VTDR.
Area under the curve (AUC) was 0.813 (95% CI, 0.730–0.897).
An optimal threshold of 2.5 RIPLs yielded a sensitivity of 70% and specificity of 84% for detecting VTDR.
DR severity was graded on UWF-FA, which served as the reference standard.
The retrospective cohort included 97 eyes from Northwestern Memorial Hospital.
Conclusions
RIPLs were validated as an OCT biomarker for VTDR using ultra-widefield fluorescein angiography as the grading reference standard.
This is a retrospective cohort study using same-day UWF-FA and OCT imaging.
Low-density OCT scans (61 serial B-scans) were sufficient for RIPL identification, suggesting practical clinical applicability.
The study used UWF-FA-based DR severity grading, providing a more comprehensive retinal assessment than standard field photography.
The authors describe RIPLs as offering 'a quantitative and practical OCT-based biomarker to identify diabetic patients at high risk for vision loss.'
What This Means
This research suggests that small structural changes visible on a common eye imaging test called optical coherence tomography (OCT) can help identify diabetic patients who are at high risk of serious vision loss. These changes, called retinal ischemic perivascular lesions (RIPLs), appear as subtle damage to specific layers of the retina and are thought to result from reduced blood flow to the eye in people with diabetes. The study compared OCT findings in 97 eyes of diabetic patients against a more detailed imaging technique (ultra-widefield fluorescein angiography, or UWF-FA) to confirm whether RIPLs reliably signal advanced, vision-threatening diabetic retinopathy.
The researchers found that eyes with vision-threatening diabetic retinopathy had a median of 4 RIPLs, compared to 0 in eyes without vision-threatening disease. Each additional RIPL detected on OCT raised the likelihood of vision-threatening retinopathy by 22%. When using a cutoff of approximately 2.5 RIPLs, the test correctly identified vision-threatening disease 70% of the time (sensitivity) and correctly ruled it out 84% of the time (specificity). Two independent graders counting RIPLs showed very high agreement with each other, suggesting this measurement is reliable and reproducible.
This research suggests that counting RIPLs on standard OCT scans — a widely available, non-invasive imaging tool already used in many eye clinics — could offer a practical way to flag diabetic patients who need closer monitoring or more aggressive treatment before they suffer significant vision loss. This could be particularly valuable in settings where more complex imaging like fluorescein angiography is not readily accessible, potentially improving early detection and management of diabetic eye disease.