Assessing Corneal Confocal Microscopy and Other Small Fiber Measures in Diabetic Polyneuropathy

Background & Objectives: Damage to small nerve fibers is common in diabetic polyneuropathy (DPN) and the diagnosis of DPN relies on subjective symptoms and signs in a combination with objective confirmatory tests, typically electrophysiology or intraepidermal nerve fiber density (IENFD) from skin biopsy. Corneal confocal microscopy (CCM) has been introduced as a tool to detect DPN. However, it is unclear if CCM can reliably be used to diagnose DPN and how the technique compares with other commonly used measures of small fiber damage, such as IENFD, cold detection threshold (CDT) and warm detection threshold (WDT). Therefore, we assessed and compared the use of CCM, IENFD, CDT and WDT in the diagnosis of DPN in patients with type 2 diabetes.

Methods: In this cohort study, the participants underwent detailed neurological examination, electrophysiology, quantification of IENFD, CCM, and quantitative sensory testing. Definition of DPN was made in accordance with the Toronto criteria for diabetic neuropathy (without relying on IENFD and thermal thresholds).

Results: A total of 214 patients with at least probable DPN, 63 patients without DPN and 97 controls without diabetes were included. Patients with DPN had lower CCM measures (corneal nerve fiber length (CNFL), nerve fiber density and branch density), IENFD, CDT, and WDT compared to patients without DPN (p=<0.001, <0.001, 0.002, p<0.001, p=0.003 and <0.005, respectively), whereas there was no difference between controls and diabetes patients without DPN. All three CCM measures showed a very low diagnostic sensitivity with CNFL showing the highest (14.4% (95% CI 9.8;18.4)) and a specificity of 95.7% (88.0;99.1). In comparison, the sensitivity of abnormal CDT and/or WDT was 30.5% (24.4;37.0) with a specificity of 84.9% (74.6;92.2). The sensitivity of abnormal IENFD was highest among all measures with a value of 51.1% (43.7;58.5) and a specificity of 90% (79.5;96.2). CCM measures did not correlate with IENFD, CDT/WDT or neuropathy severity in the group of patients with DPN.

Discussion: CCM measures showed the lowest sensitivity compared to other small fiber measures, in the diagnosis of DPN. This indicates that CCM is not a sensitive method to detect DPN in recently diagnosed type 2 diabetes.

Classification of Evidence: This study provides Class III evidence that CCM measures aid in the detection of diabetic polyneuropathy in recently diagnosed type 2 diabetics, but with a low sensitivity when compared to other small fiber measures.