Abstract
Purpose :
To assess the impact of DR severity on the risk of DR progression to clinically significant macular edema (CSME) or proliferative DR (PDR) in patients with diabetes mellitus (DM) screened at primary care centers in the US; and the association of ≥2-step DR worsening with race or DM type.
Methods :
Eyes of 22,109 patients with DM were analyzed based on DR severity data collected by masked professional reading center graders from 1999–2016 (Inoveon Corporation, OK). DR severity was graded using the Early Treatment Diabetic Retinopathy Study (ETDRS) DR Severity Scale (DRSS). Eyes with valid baseline (BL) and ≥1 post-BL DRSS value (42,011 eyes) were analyzed. Rate of ≥2-step DR worsening was assessed in the overall population; the development of CSME (using ETDRS criteria) or PDR was analyzed among eyes with no CSME and no PDR at BL, respectively. Association analyses were performed in a subset where patients reported race and DM type.
Results :
For all eyes evaluated, time to first ≥2-step DR worsening was 2.7% at year 2 and 7.1% at year 4 (Kaplan-Meier analysis). Rate of ≥2-step DR worsening was greatest among eyes with BL DRSS 43–53. Analysis of time to first development of PDR or CSME in all eyes showed the presence of 3 distinct clinical subtypes: increased risk of CSME, increased risk of PDR, and increased risk of CSME and PDR (smallest subset). This trend was not dependent on BL DRSS. Only 4058 (18%) and 11,334 (51%) patients reported race and DM type, respectively. Of these, >95% patients had BL DRSS 10–35; in this subset, a higher rate of ≥2-step DR worsening was associated with Black and White races (Fig) and type 1 DM (T1D).
Conclusions :
Eyes with BL DRSS 43–53 were at greatest risk of progression to CSME or PDR. Existence of 3 distinct clinical DR subtypes, which were not driven by BL DR severity, was also observed. The association analyses, conducted in subsets where race and/or DM type was reported (>95% with BL DRSS 10–35), showed a trend of faster DR progression in Black and White patients and those with T1D; more research is needed as our data may not accurately reflect DR dynamics in underrepresented minorities. Study limitations were: small sample size of the Black and T1D patient subsets; majority of patients reporting race and/or DM type were from OK or FL; and more thorough targeted screening conducted for patients with DM in the Native American population.
This is a 2021 ARVO Annual Meeting abstract.