Abstract
Purpose :
To investigate whether analysis of sub-basal nerve plexus (SBNP) in a standardized anatomical region of interest can be used as an equivalent substitute to wide-area analysis of SBNP in a cohort including patients with type 2 diabetes.
Methods :
82 subjects including both healthy controls and patients with type 2 diabetes mellitus (T2DM) of different durations were included in the study. Wide-area mosaic images previously produced were analysed using a rectangular region of interest (ROI) sized1000 um x 650 um located in the central cornea. Parameters evaluated in the ROI included nerve density, branching and spacing. In entire mosaics, direction of nerve orientation was examined.
Results :
Wide-area mosaic images of mean size 6.0 mm2 were analysed. ROI-based nerve density (CNFL) underestimated mosaic CNFL (mCNFL) by 34% (3.73 ± 2.97 mm/mm2), and the difference was significant (median 14.3 mm/mm2 for mCNFL vs. 10.3 mm/mm2 for CNFL, P < 0.001). In subgroup analysis, the reduction in mCNFL in long-term T2DM versus healthy subjects was significant (difference 2.1 mm/mm2, P = 0.02), but the reduction in CNFL based on the ROI was not different (9.8 vs. 11.2 mm/mm2, difference 1.4 mm/mm2, P = 0.13). Nerve orientation analysis revealed that the left side of the SBNP had a significant reduction in nerve population in long-term T2DM, relative to healthy subjects.
Conclusions :
CNFL in an anatomically-defined central corneal region underestimated mCNFL 90% of the time and by a significant 34%. Further, the sub-basal nerve degeneration characterizing T2DM is not homogeneous across the SBNP but is sensitive to the specific region examined. Our results suggest that analysis of the central SBNP may not be the most sensitive or specific area for monitoring peripheral nerve degeneration. Rather, other anatomic regions of the SBNP may hold potentially improved diagnostic value.
This is a 2020 ARVO Annual Meeting abstract.