Silva PS, Stanton RC, Elmasry MA, Fleming A, Pellegrini E, van Hemert J, Tolls D, Tolson AM, Lewis D, Stainback J, Cavallerano JD, Sun JK, Aiello LP. Association of systemic comorbities with predominantly peripheral diabetic retinopathy lesions (PPL) identified on ultrawide field (UWF) retinal imaging. Poster presented at the 2019 ARVO Annual Meeting; April 28, 2019. Vancouver, Canada. [abstract] Invest Ophth Vis Sci. 2019 Jul 1; 60(9):4772.

PURPOSE: To evaluate the association of systemic comorbidities with presence of PPL on UWF imaging in eyes of patients with diabetes (DM).

METHODS: All Beetham Eye Institute UWF images and electronic medical records from 5/1/09-8/30/18 were reviewed. UWF images with no diabetic retinopathy (DR) or any level of nonproliferative DR (NPDR) were evaluated using an automated hemorrhage and/or microaneurysm (HMA) detection algorithm to determine HMA count and location within, and peripheral to, the ETDRS 7 fields. Algorithm performance AUC against manual MA identification was 0.90-0.95 for images with NPDR. PPL-HMA were defined as present when at least 1 field had greater HMAs number in the peripheral retina than within the ETDRS fields.

RESULTS: Images and records from 13,015 eyes with gradable UWF were studied. Mean age was 57.5±17.1 yrs, DM duration 21.0±10.6 yrs, 2-year mean HbA1c 8.2±1.6, 49.5% male, 71.6% white and 56.4% type 2. Baseline DR severity by UWF imaging was: no DR 6.9%, mild NPDR 48.2%, moderate 32.4% and severe 12.5%. PPL-HMA were present in 52.6% with mild and 59.9% with moderate or more severe NPDR. Presence of PPL-HMA was associated with longer DM duration (p<0.0001) and increasing DR severity (p<0.0001). Correcting for DM duration and DR severity, presence of PPL-HMA was associated with renal disease [↑serum creatinine, p=0.005; ↑urine albumin/creatinine ratio, p=0.02; ↓estimated glomerular filtration rate (eGFR), p=0.02] and anemia (↓hematocrit, p=0.001; ↓red blood cell count, p<0.0001; ↓mean corpuscular volume, p<0.0001). PPL were not associated with HbA1c or serum lipids when correcting for DM duration and DR severity. Presence of PPL-HMA was associated with severe chronic kidney disease (eGFR <15ml/min; present: 47.4% vs absent: 41.1%, p<0.0001).

CONCLUSIONS: PPL are associated with >4-fold increased risk of DR progression. In this cohort using a fully automated HMA detection algorithm, systemic associations of PPL-HMA with renal disease and anemia have been identified for the first time and are independent of DR severity and HbA1c. These findings are consistent with previously known associations between DR and nephropathy and anemia, and support studies to determine the potential use of automated HMA counts as a marker for systemic disease.