Type of Proposal

Visual Presentation (Poster, Installation, Demonstration)


Schulich School of Medicine Windsor

Faculty Sponsor

Dr. Bruce A Perkins ( Associate Professor and Clinician Scientist Division of Endocrinology and Metabolism, University of Toronto Institute of Health Policy, Management and Evaluation, University of Toronto, Lunenfeld Tanenbaum Research Institute)

Start Date

24-3-2015 11:00 AM

End Date

24-3-2015 11:50 AM

Importance of the Project

Diabetic neuropathy has major negative impacts on one's morbidity, mortality, and quality of life. It is highly costly to the healthcare system. The lack of an early biomarker for nerve injury of neuropathy in clinical practice impedes the benefit of early identification, early management, prevention of neuropathy-related sequelae, and the accurate evaluation of future therapies in clinical trials. The current gold standard test for small nerve injury uses skin biopsies to examine the morphological changes of intra-epidermal nerve fibres, a method that is both invasive and costly. Alternatively, morphological examination of the small nerve fibres of the cornea by in-vivo corneal confocal microscopy is a potential early non-invasive method for neuropathy detection. The corneal nerve fibres are thought to closely represent those that are involved in the length-dependent process of neuropathy. It is therefore important to identify any variables that may influence the corneal nerve morphology, and interfere with its validity as a screening tool of diabetic neuropathy. For example, if there were relationships between corneal nerve fibre length and age, we would need to establish age-stratified thresholds for screening protocols of neuropathy. Although there exists research studying the associations of variables associated with corneal nerve morphology, our study is the first to examine variables specifically associated with corneal nerve fibre length.

Existing State of Knowledge

Much work has been accomplished regarding the concurrent validity and reproducibility of this technology. Corneal nerve fibre length obtained from corneal nerve microscopy has been shown to be highly reproducible and to have a high level of diagnostic validity for the identification of neuropathy. In terms of clinical variables associated with corneal nerve morphology, previous studies have showed conflicting results. Some studies showed that there is no correlation between corneal nerve density and age, others showed increasing age is associated with decreased corneal nerve density. In terms of the effect of contact lens wear on corneal nerve morphology, previous work has suggested that it has no effect. However, the parameters of corneal nerve morphology in this research did not look at nerve fibre length, but rather parameters such as hepatocyte density, nerve fibre width, and nerve fibber density. Our study is the first to examine variables specifically associated with corneal nerve fibre length.

Research Question

Are there any variables such as age or contact lens wear that may influence corneal nerve fibber length, and thus interfere with its validity as a biomarker of diabetic neuropathy?


We collected clinical variables, electrophysiological exam results, as well as eye history from 64 healthy volunteers. We performed corneal confocal microscopy on these patients, manually traced the nerves on the captured images, and used an image analysis software(CCM Image Analysis tool v1.1 which was provided by Drs R. Malik and M. Dabbah, University of Manchester, Manchester, UK ) to determine their corneal nerve fibre length. We then used univariate and multivariate linear regression analysis to determine if there are any variables associated with corneal nerve fibre length.

Your Findings

We found that the corneal nerve fibre length actually is quite variable in healthy volunteers. From the univariate regression analysis, we observed that even though HbA1c(a marker of glycemic control), age and status of contact lens wear were all associated with corneal fibber length. However, HbA1c was the only variable associated with corneal fibre length in the multivariate regression analysis. Since HbA1c is a well-known marker of glycemic control in those living with diabetes, its strong inverse association with corneal nerve fibber length (CNFL) indicates that CNFL is a sensitive measure of nerve damage in non-diabetic individuals. Those who had shorter CNFL tend to have HbA1c in the pre-diabetic range. This helps to support the use of corneal confocal microscopy as a tool to detect early nerve injury.

We could not determine the effect of refractive surgery on corneal nerve fibre length because there were only 2 subjects who had undergone LASIK surgeries. We only selected 1 image per eye for determination of corneal nerve fibber length. A higher number of images analyzed may improve the precision of the measurements. There is a need of large-scale longitudinal studies to determine the validity of corneal nerve fibre length in predicting future onset of diabetic neuropathy. The manual nerve tracing method used in this study proved to be a time-consuming process. Currently there is a newly developed software(ACCMetrics), which allows fully automated image analysis. Research efforts are needed to assess the concurrent and predictive validity of the fully automated image analysis method in comparison to the manual image analysis method.


Mar 24th, 11:00 AM Mar 24th, 11:50 AM

Corneal Confocal Microscopy, a promising new clinical tool for diabetic neuropathy screening.