Slosh Simulation in a Computer Model of Canine Syringomyelia

Abstract

The exact pathogenesis of syringomyelia is unknown. Epidural venous distention during raised intrathoracic pressure (Valsalva) may cause impulsive movement of fluid ("slosh") within the syrinx. Such a slosh mechanism is a proposed cause of syrinx dissection into spinal cord parenchyma resulting in craniocaudal propagation of the cavity. We sought to test the "slosh" hypothesis by epidural excitation of CSF pulse in a computer model of canine syringomyelia. Our previously developed canine syringomyelia computer model was modified to include an epidural pressure pulse. Simulations were run for: cord free of cavities; cord with small syringes at different locations; and cord with a syrinx that was progressively expanding caudally. If small syringes are present, there are peaks of stress at those locations. This effect is most pronounced at the locations at which syringes initially form. When a syrinx is expanding caudally, the peak stress is typically at the caudal end of the syrinx. However, when the syrinx reaches the lumbar region; the stress becomes moderate. The findings support the "slosh" hypothesis, suggesting that small cervical syringes may propagate caudally. However, when the syrinx is large, there is less focal stress, which may explain why a syrinx can rapidly expand but then remain unchanged in shape over years.

Keywords: Bernard Williams hypothesis; Chiari malformation; Valsalva maneuver; biomechanics; cavalier King Charles spaniel; cerebrospinal fluid; pathophysiology syringomyelia.

Figure 1

The finite element model of the canine spinal cavity (mid-sagittal view). The arrows denote the area at which the epidural pressure was applied. Detail A shows the anatomical layers included in the model, as well as 1 cm long sections (21–24) of the syrinx.

Figure 2

Examples of model configurations used in the study. (a) Spinal cord free of syringes. (b) One small (10 mm long) syrinx starting at 70 mm from the cranial end of the cord; syrinx radius is 70% of the cord radius (S₈). (c) An 80 mm long syrinx stretching from the cranial end (S_1–8); syrinx radius is determined from MRI data. (d) Full length of the syrinx (S_1–28).; radii determined from MRI data.

Figure 3

Processing of the data for mechanical stress in the spinal cord. (a) traces of the von Mises stress in two elements in the spinal cords. The symbols indicate the peak values recorded over the duration of the simulated event. (b) The scatter gives the peak values of stress for all the elements of the spinal cord. The thick black line gives the median values of peak stress for 1 mm thick slices of the spinal cord.

Figure 4

Input pressure waveform (thick line), pressure trace in the SAS near a small syrinx (thin line), and pressure trace in the syrinx (symbol). The results were obtained for model configuration with a single small syrinx (S₈). Time is measured from the onset of the pressure excitation in the epidural space.

Figure 5

The distribution of axial (caudal-to-cranial) velocity in the SAS, spinal cord, syrinx for the cranial portion of the model. The displayed results are for 0.08 s after the onset of the pressure pulse in the epidural space. (a) The model configuration with one small syrinx stretching from x = 20 mm to x = 30 mm (S₃). (b) The model configuration with an enlarging syrinx stretching from x = 0 to x = 30 mm (S_1–3).

Figure 6

Peak von Mises stress in the spinal cord recorded in each element over the duration of the simulated event, for three configurations with a small syrinx, and one with an expanding syrinx (lateral view). (a) Small syrinx stretching from x = 0 to x = 10 mm (S₁). (b) Small syrinx stretching from x = 70 mm to x = 80 mm (S₈). (c) Small syrinx stretching from x = 240 mm to x = 250 mm (S₂₅). (d) An expanding syrinx stretching from x = 0 to x = 40 mm (S_1–4).

Figure 7

Distribution of the stress along the length of the spinal cord. Coordinate x corresponds to the axial position measured in the cranio-caudal direction (see Figure 1). The stress is quantified via the median values of the peak Von Mises stress recorded in 1 mm-thick slices over the duration of the simulated event. Shading gives the stress distribution for the spinal cord free of syringes. Thick black lines display results for the following four model configurations: (a) S₁, (b) S₈, (c) S₁₅, and (d) S₂₅. Short vertical lines on the horizontal axis denote cranial and caudal ends of the syrinx. Vertical broken lines divide the cord into cervical (C), thoracic (T), and lumbar (L) segment.

Figure 8

Peak stresses, and stress increases from the baseline values for the 28 positions of the small initial syringes. (a) Peak value of the median Von Mises stress (σ_max). (b) Increase of the median Von Mises stress from the baseline values (Δσ). Vertical broken lines divide the cord into cervical (C), thoracic (T), and lumbar (L) segment.

Figure 9

Distribution of von Mises stress along the length of the spinal cord as the syrinx is expanding from the cranial toward the caudal end in 10 mm increments. The x coordinate corresponds to the axial position measured in the cranio-caudal direction (see Figure 1). The stress was quantified via the median value of the peak von Mises stress recorded in 1 mm thick slices over the duration of the simulated event. Shading gives the stress distribution for the spine free of syringes. The lines display results for the following four scenarios: (a) syrinx expanding from S₁ to S_1–4, (b) syrinx expanding from S_1–5 to S_1–10, (c) syrinx expanding from S_1–11 to S_1–18, and (d) syrinx expanding from S_1–19 to S_1–28. Thick black lines correspond to the results for S_1–4, S_1–10, S_1–18, and S_1–28. Short vertical lines on the horizontal axis denote crania and caudal ends of the syrinx. Vertical broken lines divide the cord into cervical (C), thoracic (T), and lumbar (L) segment.

Figure 10

T2-weighted mid-sagittal MRI of the brain and cranial cervical spinal cord in CKCS with Chiari-like malformation and developing syringomyelia: (a) 8 months and (b) 3 years old. In (a), there is a (black) fluid signal-void sign within the center of the developing syrinx suggesting pulsatile or turbulent flow (arrow). In (b), the (white) syrinx has progressed to a large fluid-filled cavity within the spinal cord (red asterisk). There is less (black) fluid void sign suggesting less turbulent flow in this established syrinx.