Analyzing Retinal Microanatomy in Retinopathy of Prematurity to Improve Care

Overview

Retinopathy of prematurity (ROP) is a disorder of development of the retina and its vasculature that can impact vision in vulnerable preterm neonates for a lifetime. A major barrier to improving ROP outcomes is the lack of easy access and low stress means to obtain objective measures of ROP disease severity across the retina in these infants. The long-term goal of this program is to provide information which will improve preterm infant health and vision via objective bedside imaging and analysis that characterizes retina-wide ROP level of disease, its response to treatment and development, and to rapidly translate this for better early intervention and improved future vision care.

Retinopathy of prematurity (ROP) remains the leading cause of childhood blindness in the US and other developed countries, blinding 150,000-200,000 children worldwide annually and leaving many more visually impaired. This vision loss is lifelong and impacts neurodevelopment. While current treatments including the addition of anti-VEGF therapies decrease the likelihood of severe vision loss from ROP, these benefits have come with a burden of monitoring for recurrence, added to the widespread task of monitoring for onset of treatment requiring ROP. This is a global issue, as the improved survival of younger preterm infants increases the burden of care in a world where there is a worsening shortage of experts for bedside ROP exams. A major barrier to improving ROP management and outcomes is the lack of easy access to incisive, objective measures of ROP disease severity, especially at critical junctures for referral or treatment, response to treatment and reactivation without causing infant stress in the nursery. ROP monitoring by standard care exam and handheld widefield photographs (e.g., RetCam by Natus) use white light, are stressful to the infant, and are repeated more often in more premature infants and with more severe ROP; their induced stress can contribute to poorer neurodevelopment. Drawing/scores of an exam are subjective, and Retcam photos are limited by loss of view from areas of shadow, worse in eyes with dark pigmentation or vitreous haze, and from avoidance response in infants, especially after term age when ROP retreatment decisions must be made. These may contribute to expert disagreement on ROP disease severity based on such photos. AI-generated models are often based on curated sets of good quality images which do not reflect real-world, lower-quality photos. With the newest generation of handheld OCTs, a field-of-view wider than ROP photographs is possible, as the researchers have established for the proposed renewal. In the current grant period, they have established that with high speed swept-source OCT imaging of a smaller field-of-view, they can obtain: OCT images of good contrast that are agnostic to fundus pigmentation, less infant stress due to lack of visible light, vascular views further into the margin of the imaging field due to coherence gating inherent to OCT, and reproducible depth-resolved retinal measures of risk (e.g. choroid) for treatment requiring disease and of neovascularization, regression and traction at the vascular-avascular junction. Multiple research groups have studied OCT imaging in preterm infants, but rigorous studies to classify ROP on OCT relative to standard care classification are rare. Without such translation, there is a risk that with the use of OCT images physicians will overtreat features, rather than treating ROP at severity levels based on clinical trial evidence. The proposed study will provide clinician-scientists with more objective and precise OCT-based measures of level of severity within (and relative to) current ROP classification. Having completed the necessary groundwork, the investigators are prepared to test our hypotheses and validate OCT-based measures for referral warranted (RW) and treatment requiring (TR)-ROP. The researchers will demonstrate through measures of infant stress, scalable grading and AI tools, the basis for OCT imaging use to improve ROP care.