Evaluation of Corneal Confocal Microscopy for the Identification and Prediction of Neuropathy in Type 1 Diabetes

The diffuse injury to peripheral nerves (diabetic neuropathy) is exceptionally common in type
1 diabetes, but there is a lack of an objective surrogate marker to identify early
subclinical stages when treatments might be most effective, prior to late-stage progression
to troublesome and costly foot infection, ulceration, and limb amputation. In contrast to the
ability to objectively measure disease-specific surrogate markers for retinopathy and
nephropathy, this lack of a diabetic neuropathy surrogate marker has seriously impeded the
development of specific interventions in clinical research trials. Representing 5 independent
research groups that have together created a consortium of investigators dedicated to the
development of a surrogate marker for early diabetic neuropathy, we have focused on using the
eye as a window to non-invasively image by a method of in-vivo corneal confocal microscopy
(CCM) the small nerve fibres that innervate the cornea. We have demonstrated that changes in
these nerve fibre endings occur early in the development of neuropathy, reflect well the
changes seen in other peripheral nerves by invasive skin biopsy evaluation, and that their
measurement is feasible and reproducible. As a multinational consortium, we have the benefit
in this proposal of pooling multiple cohorts to apply the most valid study methods in
biomarker development. First, we aim to determine in the analysis of an existing pooled
dataset of 516 type 1 and 524 type 2 diabetes subjects the exact levels of CCM measurement
that can identify the presence of diabetic neuropathy. Secondly, we propose over three years
to re-examine at least 70% of this cohort, which will provide 5- to 7-year follow-up data to
determine which type and level of CCM measurement can predict the future onset of neuropathy,
as well as its progression in those who had neuropathy at baseline. Finally, we will evaluate
the role of time- and cost-saving automated image analysis software. By virtue of large
sample size from data pooling, we are uniquely afforded the methodological power to confirm
our objectives by way of separate derivation and validation analysis sets. Through a unique
and unprecedented multinational pooled dataset approach for diabetic neuropathy, this work
will derive and validate specific CCM parameter thresholds for the identification of
neuropathy, and - more importantly - the identification of individuals at future risk. These
results will permit application in clinical practice and intervention trials for neuropathy
risk stratification. Evaluation of automated image analysis will influence likelihood of
successful knowledge translation of this surrogate biomarker into clinical practice - in
which the procedure could be harmonized with annual retinal examinations - and into
intervention trials.

Inclusion Criteria:

- Individuals of any gender or race aged 18 or above

- Type 1 diabetes mellitus or type 2 diabetes mellitus as defined by the American
Diabetes Association guidelines (2014) of any duration

- Availability of the initial CCM examination performed two to eight years ago

- Ability to understand and cooperate with study procedures

Exclusion Criteria:

- Confirmed to have neuropathy owing to non-diabetic causes (such as familial,
alcoholic, nutritional, uremic)

- Current eye infection, corneal damage, or severe movement disorders which could
preclude a safe CCM exam

- Allergy to proparacaine (the ocular topical anaesthetic used for the CCM exam)