The impact of spinal cord nerve roots and denticulate ligaments on cerebrospinal fluid dynamics in the cervical spine

Abstract

Cerebrospinal fluid (CSF) dynamics in the spinal subarachnoid space (SSS) have been thought to play an important pathophysiological role in syringomyelia, Chiari I malformation (CM), and a role in intrathecal drug delivery. Yet, the impact that fine anatomical structures, including nerve roots and denticulate ligaments (NRDL), have on SSS CSF dynamics is not clear. In the present study we assessed the impact of NRDL on CSF dynamics in the cervical SSS. The 3D geometry of the cervical SSS was reconstructed based on manual segmentation of MRI images of a healthy volunteer and a patient with CM. Idealized NRDL were designed and added to each of the geometries based on in vivo measurments in the literature and confirmation by a neuroanatomist. CFD simulations were performed for the healthy and patient case with and without NRDL included. Our results showed that the NRDL had an important impact on CSF dynamics in terms of velocity field and flow patterns. However, pressure distribution was not altered greatly although the NRDL cases required a larger pressure gradient to maintain the same flow. Also, the NRDL did not alter CSF dynamics to a great degree in the SSS from the foramen magnum to the C1 level for the healthy subject and CM patient with mild tonsillar herniation (∼ 6 mm). Overall, the NRDL increased fluid mixing phenomena and resulted in a more complex flow field. Comparison of the streamlines of CSF flow revealed that the presence of NRDL lead to the formation of vortical structures and remarkably increased the local mixing of the CSF throughout the SSS.

Figure 1. Rendering of the 3D geometry of cervical SSS in the healthy subject containing idealized nerve roots and fine structures.

The dimensions of these structures are based on anatomical measurements in Table 1. Meshwork delineates the dural surface and the top portion is transparent to better visualize the anatomy. Inset represents the extracted dimension from the medical literature (θ = median descending angle, t = nerve root thickness, RL = radicular line length).

Figure 2. 3D geometry of the cervical SSS and the selected axial planes (violet) for the healthy (left) and patient (right) case.

Spinal cord NRDL are shown in green (FM = foramen magnum, C2M = middle of 2^nd cervical vertebra, C2P = junction of C2/C3 vertebra).

Figure 3. Plots of normalized pressure along paths through SSS (showed in the right inset) as calculated with different grid refinements (Table 2) at peak systole.

Green and blue areas denote nerve roots and denticulate ligaments, respectively.

Figure 4. Peak systolic and diastolic CSF velocity for the Healthy (top) and Chiari I malformation patient along the spine with and without NRDL.

Figure 5. Velocity magnitude contours at different locations along cervical SSS plotted at the time corresponding to the peak systole for the healthy case (A) and patient diagnosed with Chiari I malformation (B).

In each set of contours, the left and right column represent the results for cases without and with NRDL, respectively.

Figure 6. Sagittal view for the healthy (top) and patient (bottom) subject showing the location of flow jets between dorsal and ventral nerve roots.

Figure 7. Streamline plots for healthy case with fine structures (left) and without fine structures (right).

Streamlines are calculated at t = 0.56 s.

Figure 8. Secondary flow parameter (SFP) during cardiac cycle with and without NRDL for the healthy (top) and patient cases (bottom).

Gray area highlights the duration of bidirectional flow in the presence of the NRDL for each case. Note that SFP is largest due to the change in flow direction.

Figure 9. Unsteady pressure gradient comparison with and without NRDL in the healthy (top) and patient (bottom) subject.