Objective: To quantify retinal ischemic perivascular lesions (RIPLs) and evaluate their potential as a biomarker for diabetic retinopathy (DR) severity. Design: Cross-sectional studyParticipants: Patients with diabetes mellitus (DM) and spectrum of DR severity, recruited from a tertiary referral center.Methods: RIPLs were quantified using 66 mm optical coherence tomography (OCT) scans based on established criteria: perivascular location, focal thinning of the inner nuclear layer, disruption of the outer plexiform layer, and elevation of Henle's fiber layer.Two independent graders assessed RIPL counts, and intergrader reliability was evaluated using the intraclass correlation coefficient.Main Outcome Measures: Association between RIPL count and DR severity using multinomial logistic regression.Receiver operating characteristic (ROC) curve analysis evaluated diagnostic performance of RIPL using area under the curve (AUC), sensitivity and specificity.Causal mediation analysis (CMA) estimated the direct and indirect effects of systemic comorbidities on DR severity.Results: 129 eyes were included: 51 with no DR (DMnoDR), 45 nonproliferative diabetic retinopathy (NPDR), and 33 proliferative diabetic retinopathy (PDR).RIPL counts increased with DR severity: 1.18 2.46 in DMnoDR, 3.69 4.91 in NPDR, and 10.10 10.44 in PDR.Eyes with higher RIPL counts had significantly greater odds of higher DR severity (NPDR vs. DMnoDR: odds ratio OR = 1.22,95% CI, 1.04-1.43,p = 0.0156; PDR vs. NPDR: OR = 1.11, 95% CI, 1.03-1.21,p = 0.0084; PDR vs. DMnoDR: OR = 1.36, 95% CI, 1.16-1.59,p = 0.0002).RIPLs were moderately predictive for vision-threatening DR, with an AUC of 0.779 (95%CI: 0.678-0.879);an optimized threshold of 3 RIPLs achieved sensitivity of 76% and specificity of 78%.CMA revealed that RIPLs mediated 30% of hypertension's effect on DR severity (p = 0.018), but not ischemic heart disease (p = 0.092). J o u r n a l P r e -p r o o fConclusions: RIPLs are discrete, quantitative OCT biomarkers associated with DR severity and partially explain the effect of hypertension on DR severity.RIPL may reflect the ischemic burden in the diabetic macula, and offer an additional marker of DR severity, bridging vascular and neural pathology, and capturing the impact of vascular comorbidities on DR.
Zhuang et al. (Sun,) studied this question.