Steps Towards Predicting The Development Of Advanced Diabetic Eye Disease

Proliferative diabetic retinopathy (PDR) is a severe form of diabetic eye disease where blood vessels form in the back of the eye obscuring vision. Prompt laser treatment is usually needed to avoid severe vision loss. However, laser treatment has negative consequences, including reduced night vision.  Prevention of PDR could result in better vision for those with diabetic eye disease. At present we do not have the tools to predict who will progress from early to advanced diabetic eye disease to better target earlier treatment, or prevention of progression. Data from two groups summarized in this report suggest some approaches that could be used to predict this progression.

The University of Michigan group, led by Dr. Stein, has retrospectively examined 4617 recipients of eye care in a large managed-care network. Of the 4617 with diagnosed early eye disease, 7% went on to develop advanced disease between 2001 and 2009. Medical factors were identified that predicted this progression, including presence of non-healing ulcers and diabetic kidney disease. The ‘low risk’ group defined by these criteria had only a 5% chance of developing advanced disease within 5 years, while the high risk group had a 38% chance in the same timeframe.

Another study group examined patients enrolled in ranibizumab (brand name Lucentis) clinical trials for factors predicting progression from early to advanced stage eye disease. Factors predicting progression which can be assessed using fundus photographs or optical coherence tomography included baseline eye disease severity, presence of subretinal fluid, and loss of capillaries. Interestingly in this study HbA1c level at baseline and diabetes duration did not predict development of advanced disease, suggesting perhaps at this late stage of disease these factors have less of an influence on progression than other local factors in the eye, or perhaps genetic predisposition.

Collectively these studies suggest that it should be possible to better predict the onset of advanced diabetic eye disease.

Ramifications for Individuals with T1D:

Though some treatments for diabetic eye disease are available, these are not optimal and treatment at earlier stages of diabetic eye disease to prevent vision loss is not currently possible. One challenge to developing such therapies is the small proportion of those who would experience significant disease progression within a clinical trial of reasonable length (for example 3 years or less). If disease markers were available to better identify patients by their risk of progression, it might be possible to reduce this risk by targeting better diabetes management strategies and conducting more regular eye check-ups. In addition, such markers could enable conduction of clinical trials to assess the efficacy of both repurposed and novel therapies to treat early disease to prevent future vision loss.

JDRF Involvement:

JDRF did not fund these investigators but is assessing opportunities to develop better risk prediction.

Investigators and Institutions:

Harris, Stein and colleagues, University of Michigan and Ehrlich and colleagues, Genentech and University of Wisconsin


  1. Harris-Nwanyanwu K, Talwar N, and others. Predicting development of proliferative diabetic retinopathy. Diabetes Care. 2013 Jun;36(6):1562-8. doi: 10.2337/dc12-0790. Epub 2012 Dec 28.
  2. Ehrlich JS, Domalpally A, and others. Effects of Intranvitreal Ranibizumab on Diabetic Retinopathy Severity: 36 Month Data from the RISE and RIDE Phase III trials. Presentation at ARVO 2013 meeting. May 8th 2013.