Nerve spectroscopy: understanding peripheral nerve autofluorescence through photodynamics

Background Being able to accurately identify sensory and motor nerves is crucial during surgical procedures to prevent nerve injury. We aimed to (1) evaluate the feasibility of performing peripheral human nerve visualization utilizing nerves' own autofluorescence in an ex-vivo model; (2) compare the effect of three different nerve fiber fixation methods on the intensity of fluorescence, indicated as the intensity ratio; and (3) similarly compare three different excitation ranges. Methods Samples from various human peripheral nerves were selected postoperatively. Nerve fibers were divided into three groups: Group A nerve fibers were washed with a physiologic solution; Group B nerve fibers were fixated with formaldehyde for 6 h first, and then washed with a physiologic solution; Group C nerve fibers were fixated with formaldehyde for six hours, but not washed afterwards. An Olympus IX83 inverted microscope was used for close-up image evaluation. Nerve fibers were exposed to white-light wavelength spectrums for a specific time frame prior to visualization under three different filters-Filter 1-LF405-B-OMF Semrock; Filter 2-U-MGFP; Filter 3-U-MRFPHQ Olympus, with excitation ranges of 390-440, 460-480, and 535-555, respectively. The fluorescence intensity of all images was subsequently analyzed using Image-J Software, and results compared by analysis of variance (ANOVA). Results The intensity ratios observed with Filter 1 failed to distinguish the different nerve fiber groups (p = 0.39). Conversely, the intensity ratios seen under Filters 2 and 3 varied significantly between the three nerve-fiber groups (p = 0.021, p = 0.030, respectively). The overall intensity of measurements was greater with Filter 1 than Filter 3 (p < 0.05); however, all nerves were well visualized by all filters. Conclusion The current results on ex vivo peripheral nerve fiber autofluorescence suggest that peripheral nerve fiber autofluorescence intensity does not greatly depend upon the excitation wavelength or fixation methods used in an ex vivo setting. Implications for future nerve-sparing surgery are discussed.