Previously, Neurolucida 360 was used to manually trace the fascicles of the vagus nerve. However, Vesselucida 360, despite being designed for vessels, provided a more seamless and less time-consuming tool for tracing and segmenting fascicles of the vagus nerve from their branching point till cervical level. This was the only software used in this study for tracing fascicles.

Despite only having fascicular, and not fiber, resolution of microCT as mentioned above, by performing this microCT and tracing method, the following can be determined, quantified, analyzed, and subsequently compared between individuals:

Routine histology and immunohistochemistry (IHC) are useful for identifying nerve anatomical features like nerve diameter, fascicle count and diameter, fiber count, and fiber diameter/type. IHC can further characterize nerves based on neurochemical/microtubule characteristics, distinguishing different fiber types. Trichrome staining uses 3 dyes to stain collagen, keratin, muscle fibers, cytoplasm, nuclei, and bone (O’Connor and Valle, 1982). By immunostaining with neurofilament (NF) and myelin basic protein (MBP) antibodies on one slide, accurate detection of nerve fiber sizes is possible, visualizing intermediate filaments (Yuan et al., 2012) and myelin sheaths (Deber and Reynolds, 1991) and thereby identifying both myelinated and non-myelinated fibers.

Due to limitations imposed during the Covid pandemic, timeframes with the human vagus nerve samples and entry rules into labs not within our department, some aspects of this study were affected. The original target was to perform both histology, including Trichrome and H&E, and IHC, including ChAT, and NF and MBP, on all sections from the ten nerves. Due to the timing and practical constraints, all except ChAT could be performed on the sections. ChAT could not be optimized within this timeframe for the use on human tissue and thus, was subsequently removed from the study. Although the remaining staining techniques were performed, if a cross section of nerve was missing from the section, this could not be corrected for prior to nerve tissue being returned to the facility. This work still, however, serves as a preliminary tracing and staining study for work to be done in the future. Ten vagus nerves (five left and five right) were successfully microCT scanned and reconstructed. Only three left and two right vagus nerves which have been fully traced from caudal to cranial level are presented in this manuscript.

The Evelyn Cambridge Surgical Training Centre, A Cambridge University Health Partners Facility, loaned human vagus nerve samples for this study. The Human Tissue Authority (HTA) Designated Individual (DI), namely Dr. C R Constant, approved the loan of human tissue from first-person, signed and witnessed consented (willed) donors for the use in this project. Dissection of vagus nerves from the donors was supervised by the HTA DI for the Centre’s license and all subsequent transport and work on the nerves was pre-approved by the HTA DI on their license and specified in a Material Transfer/Loan Agreement. Separate ethical approval was not required. Further information on the cadavers can be found in Supplementary Table 1.

Both left and right vagus nerves from 5 frozen human cadavers were dissected from the cervical region down to the branching regions of cardiac, recurrent laryngeal and pulmonary branches; all the branches left attached to the main trunk of the vagus nerve. The nerves were exposed using techniques that are similar to surgery for vagus nerve electrode placement. A linear incision through the skin and platysma muscle was made one centimeter above the clavicle extending cranially and parallel to the anterior border of the sternocleidomastoid muscle. This muscle was retracted laterally to expose the carotid sheath. The omohyoid muscle was then dissected from the carotid sheath and retracted cranially. The carotid sheath was then opened, and the vagus nerve was exposed between the common carotid artery and jugular vein. At this level (5 cm caudal from nodose ganglion), the vagus nerve was transected and exposed to follow it caudally toward any and all branches leading to the heart, lungs and larynx including the superior and inferior cardiac branches, pulmonary branches and the recurrent laryngeal branch. The vagus nerve was resected from below the last pulmonary branch to 2 cm cranial to the average VNS cuff placement and 5 cm caudal from the nodose ganglion, yielding samples about 25 cm in length. Sutures were placed around the vagus nerve prior to any branching region (i.e., the region where a branch leaves the main vagal trunk) to mark the branching point. Nerves were then placed in neutral buffered formalin (10%) for fixation, labeled accordingly and transported to UCL. All ten nerves were iodine stained, scanned by microCT, and had histology and IHC performed; five nerves proceeded to computerized segmentation to identify fascicle anatomy.

Pre-processing and staining were performed in the same way as the pig vagus nerves in Thompson et al. (2023). After fixation, nerve samples were measured, sutures of 1 cm length were superglued to the vagal trunk in 4 cm intervals, and nerves cut into 4 cm lengths at the level of suture placement leaving half of the suture on the end of each section as a marker for subsequent co-registration. Two to three sections were placed into a tube of 50 mL Lugol’s solution (total iodine 1%; 0.74% KI, 0.37% I) (Sigma Aldrich L6141) for 5 days (120 h) prior to scanning to achieve maximum contrast between fascicles and the rest of the nerve tissue. On the day of the microCT scan, the nerve was removed from the tube and blotted dry on paper towel to remove any excess Lugol’s solution. The nerve sections were placed next to each other onto a piece of cling film (10 × 5 cm) (Tesco, United Kingdom) in order from cranial to caudal along the length of the nerve, with cranial ends at the top, and sealed with another piece of cling film to retain moisture during the scan as to avoid shrinkage of the nerve tissue. The sealed nerve samples were rolled around a cylinder of sponge (0.5 cm D × 4.5 cm) and wrapped in another two layers of cling film to form a tightly wound cylinder with a diameter of ∼1.5 cm to fit within the field of view at the required resolution. The wrapped cylinder was placed within a ∼2 cm sponge inside a 3D-printed mount (30 mm width × 60 mm height tube in a mount fitted to the Nikon XT H 225 scanner stage; stl files available in Supplementary files) filled with ∼2 cm layer of sponge around the edges, ensuring a tight fit and the ends sealed with tape (Transpore, 3M, United Kingdom).

A microCT scanner (Nikon XT H 225, Nikon Metrology, Tring, United Kingdom) was homed and then conditioned at 200 kVp for 10 min before scanning and the target changed to molybdenum. The scanner was calibrated as per Nikon XT H 225 protocols prior to scanning the nerves. In addition, correction images were obtained prior to each scan for normalization with the detector and any background noise during reconstruction. All nerves were scanned with the following scanning parameters: 35 kVp energy, 120 μA current, 7 W power, an exposure of 0.25 fps, optimized projections, and a resolution with isotropic voxel size of 7 μm. Scans were reconstructed in CT Pro 3D (Nikon’s software for reconstructing CT data generated by Nikon Metrology, Tring, United Kingdom). The center of rotation was calculated manually with dual slice selection. Beam hardening correction was performed with a preset of 2 and coefficient of 0.0. The reconstructions were saved as 16-bit volumes and triple TIFF 16-bit image stack files allowing for subsequent image analysis and segmentation in various software.

Image analysis, segmentation, and tracing were performed on reconstructed microCT datasets from five human vagus nerves (three left and two right). Reconstructed microCT scan images were analyzed in ImageJ (Schindelin et al., 2012) in the XY plane to view the cross-section of the nerve. The vertical alignment of the nerve was positioned so that the cross-sectional plane was viewed in the XY stack and the longitudinal plane in the XZ and YZ stacks. This allowed for validation of the scanning protocol, direction of the nerve, and visual analysis of the quality of the image and the distinguishability of the soft tissues—specifically the identification of the fascicles known to exist within the nerve. AVI files were created from ImageJ to enable stack slice evaluation, identification of suture positions and branching locations of the vagus nerve, and as a reference during segmentation. Image stacks (XY plane along the Z-axis) were loaded into Vesselucida 360 (Version 2021.1.3, MBF Bioscience LLC, Williston, VT, United States) and image histograms adjusted to optimize visualization of the fascicles when required. Fascicles of the three target organs/functions (namely, cardiac, recurrent laryngeal and pulmonary fascicular groups) were segmented from the rest of the nerve using Vessel mode from the Trace tools. Starting from identification within branches of the vagus nerve, the fascicles were traced through every slice of each scan proceeding cranially up the length of the nerve to the cervical region at the level of cuff placement. Points adding to the path of the vessel trace were place every 50–100 sections and the thickness of the vessel trace was altered to match the diameter of the fascicle being traced throughout its length. If fascicles merged with or split into others at a higher frequency, points of the vessel trace were placed at smaller intervals to ensure accurate tracing. The vessel trace was labeled according to the organ from which the fascicle originated. If the fascicle split into two or more fascicles, a bifurcating or trifurcating node was inserted and each branch exiting the node traced cranially up the nerve until either the fascicle merged with another fascicle, or another node needed to be created for splitting fascicles. If fascicles merged into one another, the vessel trace was ended and another trace started that was labeled as containing the respective merged fascicle types and therefore containing fibers from different target organs (i.e., pulmonary and recurrent laryngeal merged fascicle). To continue tracing across cut regions of the nerve, the superglued suture markers and distinct physiological regions or landmarks were used to align the cranial and distal ends of the cut nerves and tracing continued (see Supplementary Figure 1 for a visual depiction of how co-registration was performed). Tracing across the cut regions was successful for all nerves. Subsequent to tracing throughout the segments of nerve in each scan, the stack was viewed in the 3D Environment where the traces are visualized in 3D displaying the thickness, path and color (specified when labeling the fascicle traces) of the fascicles throughout the length of each scan. Due to software data size limitations, the overlapping, consecutive scans could not be physically stitched together to show the full length at once, but rather allowing each scan to be visualized separately.

Subsequent to scanning, stained nerves were placed back into neutral buffered formalin for at least a week which allowed for the Lugol’s solution to be soaked out. The formalin was refreshed weekly prior to histology. Histology and immunohistochemistry (IHC) were performed on all 10 nerves. From each human vagus nerve, 0.5-cm-long segments were excised at predefined anatomical locations: at the level of cuff placement, 2 cm distal to the cuff, and at 4-cm intervals thereafter along the nerve trunk. Additional segments were collected immediately cranial to major branching points and from 0.5 to 1 cm within each branch. This sampling strategy yielded 6–7 segments from the nerve trunk at regular intervals, 3–4 segments cranial to branching regions, and 3–5 segments from the branches, resulting in 9–11 trunk segments and a total of 12–16 segments per nerve for histological and IHC analysis. Segments from the main vagal trunk were placed into a cassette with the order and placement of the segments noted. In another cassette, the segments prior to branching region were placed in order in one row, and those from the branches were placed in order in another row. These nerve samples were embedded in paraffin and six sequential sections cut every 4 μm from each block; two were placed on uncoated slides and three on coated slides. One section of each block on an uncoated slide was stained with Hematoxylin and Eosin (H&E, a routine stain used to demonstrate the general morphology of tissue) (Sheehan and Hrapchak, 1987), the second section on a coated slide stained with Trichrome, and the third section on a coated slide immunostained for both NF (anti-neurofilament heavy polypeptide antibody ab8135 1:1,000) and MBP (anti-myelin basic protein antibody ab7349 1:4,000) with a secondary biotin labeled antibody (donkey anti-rabbit IgG H&L (Biotin) ab6801 1:1,000) (see Supplementary Information). The remaining two sections, one coated and one uncoated, were extra. For H&E staining, a Gemini AS Automated Slide Stainer (Epredia, Portsmouth, New Hampshire, United States) was used. A Leica BOND RX (Leica Biosystems, Milton Keynes, United Kingdom) machine was used for automated IHC. All slides were imaged with light microscopy using a NanaZoomer S360 (Hamamatsu Photonics, Hamamatsu City, Japan).

Identification of histopathological features was performed, and the images of the histology and immunohistochemistry cross-sections were then compared to the corresponding slice in the microCT scan of the same nerve for comparison and validation. The presence and number of fascicles visualized in the golden standard of histology were compared to those identified during segmentation of microCT scans to confirm that the segmentation results correlated (Supplementary Figures 5, 6 in Thompson et al., 2023). In addition, fiber sizes within organ-specific fascicles were observed.

NDP.view 2 (Hamamatsu Photonics, Hamamatsu City, Japan) was used to view the images of the histology and IHC cross-sections on the multiple slides. Using NDP.view 2 annotation tools, namely “ruler” and “freehand region,” nerve dimensions were quantified from the H&E slides at the mid-cervical level (Supplementary Tables 3, 4, 6). As the nerves exhibit non-uniform or non-circular cross-sectional morphology, both the short-axis and long-axis diameters were measured; the reported diameter represents the mean of these two values. Cross-sectional area and circumference were measured using the freehand region tool, which allowed precise tracing of the nerve boundary and accounted for irregular nerve geometry. This approach avoided assumptions of circularity and minimized geometric bias that would arise from estimating area or circumference based solely on diameter measurements. These were then averaged for the ten nerves, and for the left and right nerves, respectively.

Intra- and inter-observer repeatability of morphometric measurements was assessed using repeated measurements of a single human nerve cross-section. Three independent researchers each performed nine repeated measurements of nerve circumference, cross-sectional area, short diameter, long diameter, and mean diameter. Researchers were blinded to each other’s measurements during analysis. Intra-observer repeatability was high, as demonstrated by very low within-researcher standard deviations and variance, narrow 95% confidence intervals, and low coefficients of variation across all morphometric parameters (Supplementary Table 2). Inter-observer agreement was assessed using one-way ANOVA followed by Tukey’s corrected pairwise comparisons between researchers. No statistically significant differences were observed between any researcher pairs for any morphometric parameter (all adjusted p > 0.05, Supplementary Table 2). Collectively, these findings demonstrate excellent intra- and inter-observer repeatability, indicating that the morphometric measurements are robust, reproducible, and not dependent on the individual researcher. This supports the reliability of all nerve morphometric measurements reported in this study. Fascicle counts was also consistent between researchers.

The number of fascicles (and fascicle bundles—defined as the number of fascicles separated by thin layer of perineurium from but within the same thicker layer of perineurium as other fascicles) were counted at regular intervals along the trunk of the nerve and within the branches for the ten human vagus nerves from the histology images. Specifically, the Trichrome slides were used for this for clear distinguishability of the fascicles from the rest of the nerve tissue. In the cases where a nerve cross-section was not visible on the slides, H&E slides were used. This was compared between nerves (Supplementary Tables 4, 5). The distinguishability of the fascicles and fascicle bundles was visible upon analysis of the microCT scans, when visualizing the movement of the fascicles in the Z direction; however, for unambiguous identification of this, the Trichrome histology slides were used. The distance between branches was calculated for all dissected nerves and averaged to provide general gross vagal anatomical information and compared between individuals (Supplementary Table 6). Additionally, the cervical vagus nerve average diameters, areas and fascicle counts were compared to histological studies (Table 1; Hammer et al., 2015; Hammer et al., 2018; Pelot et al., 2020; Stakenborg et al., 2020; Verlinden et al., 2016).

Fascicles present in the cardiac, pulmonary and recurrent laryngeal peripheral branches were preserved after entry into the main vagus trunk (Figures 1–4 and Supplementary Figures 2–6) (n = 5). The organization of these organ-specific fascicles was partially preserved for approximately 1–1.5 cm. Detailed morphology of the nerve and its branches can be seen in the “Analysis of nerve morphology” section. In the left vagus nerves, approximately 1–1.5 cm cranial to branch entry, cardiac fascicles merged with some pulmonary fascicles. The remaining cardiac fascicles merged 2.5 cm cranial to branch entry point. The recurrent laryngeal fascicles remained on one half of the nerve for at least 2–3 cm and then eventually all fascicles of the recurrent laryngeal merged by 6–7 cm cranially from the branch entry point. In the right vagus nerves, the inferior cardiac not only predominantly merged with the pulmonary fascicles (forming a common, pink fascicle, Supplementary Figures 2–6) as in the left vagus nerves, but in both right vagus nerve samples segmented, the inferior cardiac fascicles merged with a recurrent laryngeal fascicle (forming a common, yellow fascicle, Figure 4 and Supplementary Figures 5, 6) at 1.7 and 2.1 cm, respectively, from point of entry in the nerve till merge point. This difference of merging pattern may be attributed to the differing branching point of recurrent laryngeal branch between left and right vagus nerves. At the point where all three organ-specific fascicle types had merged, microCT has a limitation of not allowing for the visualization and thus, the tracing of the fibers of the nerves. Therefore, the organization of the organ-specific fibers within these fascicles remains unknown. The vagus nerve structure became plexiform after all the three organ-specific fascicles merged together (Figure 3). Fascicles were observed to merge or split at an approximate rate of one to two fascicles every 0.5 mm. This measurement is based on fixed, Lugol’s-stained nerve tissue and so, shrinkage should be accounted for if translating to fresh nerve tissue. Adjusting this measurement to account for an estimated cumulative 5–15% tissue shrinkage arising from formalin fixation (approximately 5–10% for non-embedded tissue) and subsequent Lugol’s iodine staining (∼5%), the adjusted nerve length is estimated to be 0.53–0.59 mm. Further up the nerve, below the cervical level, a superior cardiac branch entered the vagus nerve and remained separated from the other fascicles for a short length (2–3 cm) which is 0.5–1 cm caudal to the VNS stimulator placement. In summary, so far, organization and separation between the organ-specific fascicular groups exists at approximately clavicular level and the superior cardiac fascicles are separate from the other vagal fascicles at a position close to VNS cuff placement.

Histology and immunohistochemistry were performed on all nerves including H&E, Trichrome, NF with MBP (Supplementary Figure 7). NF with MBP successfully stained fibers simultaneously, allowing clear visualization of fiber caliber heterogeneity within the cross-sections with ease (Figures 5, 6). The large, myelinated, efferent fibers in the recurrent laryngeal fascicles (Figures 5E, 6D) are evident in the cross-sections from the branch as well as in the vagal trunk just cranial to the entry of the fascicles. At the cervical level (Figures 5A, 6A), however, these large fibers appear to be mixed with smaller fibers dispersed around the cross-section and only small clusters of large fibers observed. Despite hypotheses that the clusters of fibers will remain organ-specific, this cannot be deciphered without tracing of the fibers throughout the nerve.

Histology and immunohistochemistry allowed for validation of microCT segmentation—confirming the presence of fascicles from that which was segmented from the scans. A nerve example of microCT compared with immunohistochemistry can be seen in Figures 7–9. Regions of the stack of XY slices of microCT data from where the histology slice was taken were compared to the histology slice of interest to find the matching slice for accurate cross-validation. For display purposes here, the closest already segmented and labeled cross-section was compared and therefore, there is slight mismatch. Recurrent laryngeal fascicles can be seen to contain more larger, myelinated fibers than the other fascicles in the nerve. The mixed (cardiac and pulmonary) fascicles contained a mix of fiber sizes. The superior cardiac (Figure 9) fascicle also contained a mixture of fiber sizes.

From the morphological data of the nerve (n = 10), the right human vagus nerves were 1.22 ± 0.1 SD times larger than the left vagus nerve in terms of its diameters, circumference, and area, with an average area of all 10 nerves of 3.63 mm² ± 1.43 SD (variance (var) = 1.84, 95% CI [2.61, 4.65]), an average area of the left vagus nerves of 3.08 mm² ± 1.77 SD [var = 2.51, 95% CI (0.88, 5.28)], and an average area of the right vagus nerves of 4.17 mm² ± 0.85 SD [var = 0.57, 95% CI (3.12, 5.22)]. In addition, the right vagus nerves contain 1.35 times more fascicles and fascicle bundles than the left: 10.80 ± 4.64 SD [var = 19.36, 95% CI (7.48, 14.12)], 9.60 ± 3.05 SD [var = 7.44, 95% CI (5.81, 13.39)], and 12.00 ± 5.96 SD [var = 28.40, 95% CI (5.00, 19.8)] for all, left and right fascicles, respectively (Supplementary Tables 3, 4 and Supplementary Figure 8). The left and right cervical vagus nerves consisted of 5.00 ± 1.58 SD [var = 2.00, 95% CI (3.04, 6.96)] and 7.20 ± 3.70 SD [var = 10.96, 95% CI (2.6, 11.8)] fascicle bundles, respectively.

The number of fascicles present in the cross-sections of the main trunk of the vagus nerve varied when progressing through the nerve at regular intervals (Supplementary Tables 3, 4). The least number of fascicles was 8.30 ± 3.43 and 8.67 ± 4.5 at 1 cm and 13 cm from the upper cervical level, respectively. At 9 cm from upper cervical level, there was the greatest average number of fascicles across the 10 nerves of 16.20 ± 7.28. Such changes in fascicle counts when progressing down the length of the nerve illustrate the anastomoses taking place. This coincides with the high standard deviations observed here such as a standard deviation of 10.19 for nerve 3. The mean fascicle count along the entire nerve length was 11.85 ± 5.75. This is greater than the mean at cervical level. On average between nerves, the superior cardiac branch contained 3.10 ± 1.45 fascicles, the inferior cardiac branch 2.33 ± 1.06, recurrent laryngeal branch 7.25 ± 1.98, and the pulmonary branch 6.14 ± 2.91 fascicles.

Unlike the pig vagus nerves, the human inferior cardiac branches exit the vagus nerve just above or at the same level as the recurrent laryngeal branches (Supplementary Table 5). The distance from cervical level on average to the inferior cardiac branches and recurrent laryngeal branch was 12.30 cm ± 1.89 and 12.30 cm ± 1.89, and 15.30 cm ± 1.44 and 15.00 cm ± 1.27 for the right and left nerves, respectively. The discrepancy of the branching point of the right and left recurrent laryngeal branches of the vagus nerve is expected. The superior cardiac branch exited the vagus nerve 6.38 cm ± 2.43 from the upper cervical level with a difference of 2.45 cm between the left and right nerves with the branch exiting at 5.15 cm and 7.60 cm, respectively. The pulmonary branches exited the vagus nerve from 18 to 20 cm from the upper cervical level. Four fifths of the right vagus nerves had three pulmonary branches and four fifths of the left vagus nerve has two pulmonary branches which coincides with the number of lobes the lungs consist of on either side. There was a difference of 1.2 cm ± 0.71 and 0.70 cm ± 0.27 between the first and second, and between the second and third pulmonary branches, respectively.

Compared to morphological data previously reported in literature, the effective diameters were comparable between studies (Table 1). In addition, the fascicle counts, when looking at fascicle bundles in this work, was comparable between studies too. The individual fascicle count reported here was larger; this may be indicative of the other studies not taking into account thin layers of perineurium. The nerve cross-sectional areas correlated with Pelot et al. (2020) but the other three studies reported greater areas of 1.5 times, albeit with overlapping ranges.