You are viewing the site in preview mode

Skip to main content
Download PDF

Cerebrospinal fluid dynamics and subarachnoid space occlusion following traumatic spinal cord injury in the pig: an investigation using magnetic resonance imaging

Fluids and Barriers of the CNS volume 22, Article number: 6 (2025) Cite this article

Abstract

Background

Traumatic spinal cord injury (SCI) causes spinal cord swelling and occlusion of the subarachnoid space (SAS). SAS occlusion can change pulsatile cerebrospinal fluid (CSF) dynamics, which could have acute clinical management implications. This study aimed to characterise SAS occlusion and investigate CSF dynamics over 14 days post-SCI in the pig.

Methods

A thoracic contusion SCI was induced in female domestic pigs (22–29 kg) via a weight drop apparatus (N = 5, 10 cm; N = 5, 20 cm). Magnetic resonance imaging (MRI) was performed pre-SCI and 3, 7 and 14 days post-SCI. SAS occlusion length (cranial-caudal), and injury site SAS area (cross-sectional), were measured on T2-weighted MRI. CSF dynamics, specifically peak cranial/caudal mean velocity (cm/s), and the corresponding time to peak (% of cardiac cycle), were measured on cardiac gated, axial phase-contrast MRI obtained at C2/C3, T8/T9, T11/T12 and L1/L2. Linear-mixed effects models, with a significance level of α = 0.05, were developed to assess the effect of: (1) injury group and time point on SAS occlusion measures; and (2), time point and spinal level, adjusted by injury group, on CSF dynamics.

Results

For both injury groups, SAS occlusion length decreased from 3 to 7 days post-SCI, and 7 to 14 days post-SCI. The cross-sectional SAS area decreased after SCI, and increased to 14 days post-SCI, in both groups. At all spinal levels, peak cranial/caudal mean velocity and the time to peak caudal mean velocity decreased at day 3 post-SCI. From 3 to 14 days post-SCI, peak caudal mean velocity and the time to peak caudal mean velocity increased towards baseline values, at all spinal levels.

Conclusions

Spinal-level specific changes to CSF dynamics, with concurrent changes to SAS occlusion, occurred after SCI in the pig, suggesting that CSF pulsatility and craniospinal compliance were altered in the sub-acute post-traumatic period. These results suggest that PC-MRI derived CSF dynamics may provide a non-invasive method to investigate functional alterations to the spinal intrathecal space following traumatic SCI.

Introduction

Traumatic spinal cord injury (SCI) often causes permanent neurological impairment, but treatment development has been largely unsuccessful, partly due to complex underlying pathophysiology and heterogeneous injuries [1]. Oedema and haemorrhage within the parenchyma commence in the first few hours after injury [2, 3] and can cause the spinal cord to swell and occlude the subarachnoid space (SAS) [4], leading to spinal cord compression and reduced perfusion [5]. Timely surgical decompression of the spinal cord, to alleviate compression by the fractured or misaligned spinal column, is considered important for optimal neurological recovery [6]. However, patency of the SAS surrounding the injured spinal cord is not achieved in all patients, either due to spinal cord swelling or inadequate decompression [5, 7], and methods to monitor decompression adequacy remain limited [8].

Spinal cord compression and subsequent SAS occlusion following SCI can result in changes to the cerebrospinal fluid (CSF) system [9,10,11,12,13,14]. The mean and dynamic features of intrathecal pressure measured using invasive techniques in clinical studies of traumatic SCI can be altered locally (termed intraspinal pressure) [9, 12, 13] and remotely to the level of occlusion [9, 11, 14]. It has been suggested that detecting changes in intrathecal system dynamics may provide a method for monitoring decompression adequacy [8]. While structural anatomic magnetic resonance imaging (MRI) is the current standard to identify spinal cord compression and assist clinical decision making [15], intraspinal pressures can vary widely in cases where spinal cord compression is indicated on MRI [9]. This variability suggests that additional tools to monitor the injury site could better inform patient-specific treatment decisions.

Pulsatile CSF flow in the spinal SAS has been measured with phase-contrast MRI (PC-MRI) in healthy humans [16,17,18,19,20,21,22,23,24,25,26,27,28], non-human primates (NHP) [29], canines [30, 31], and domestic pigs [32]. The cardiopulmonary system [20,21,22, 27, 33], and posture [23, 24, 34] contribute to normal variability in CSF dynamics. The spatial origin of cardiovascular-derived CSF pulsations is unresolved: some studies show that CSF pulsations stem from intracerebral arteries where an increase in blood in the cranium causes CSF to flow into the spinal canal [17, 19, 28], others show that they are caused by local vascular pulsations within the spinal column [25, 26, 35]. There is evidence that respiration is also a major driver of CSF flow [20,21,22, 27, 33]. CSF pulsations propagate and attenuate caudally along the spinal cord axis. In healthy, awake, humans, CSF flow magnitude is greater and propagates faster than in anesthetised pigs [32] and NHP [29].

CSF dynamics, measured with PC-MRI, have diagnostic [36] and therapeutic [37, 38] potential. PC-MRI derived CSF dynamics in the cerebral aqueduct and at the cranio-cervical junction, are used to guide diagnosis and assess treatment outcomes in normal pressure hydrocephalus [39,40,41] and Chiari malformation [42,43,44,45,46,47]. Although pulsatile spinal CSF flow is relatively understudied compared to intracranial CSF flow, changes to the former have been observed in several clinical PC-MRI studies of obstructive, non-traumatic spinal pathologies: cervical myelopathy [48,49,50], lumbar spinal stenosis [51], and post-traumatic syringomyelia [52]. In addition, a small, heterogeneous cohort of traumatic and non-traumatic cervical SCI patients imaged once at 1 month to 6 years after injury, showed differences in CSF dynamics at and below the level of injury, compared to healthy controls [53]. However, it remains unclear whether changes to SAS morphology and spinal level-specific CSF dynamics occur with time after traumatic SCI.

Since the acute and subacute (≤ 14 days) period following SCI offers a potential treatment window to reduce the deleterious effects of spinal cord compression, an improved understanding of CSF dynamics during this timeframe is important. Monitoring CSF dynamics with PC-MRI may provide a non-invasive method to more comprehensively assess SAS patency following traumatic SCI and subsequent decompression efforts. Using MRI across 14 days following a thoracic SCI in a domestic pig model, the aims of this study were to (1) characterise SAS occlusion, and (2) investigate changes to CSF dynamics.

Methods

Animal ethics

This project was approved by the South Australian Health and Medical Research Institute Animal Ethics Committee (SAM 243 and SAM-22-031) and conducted in accordance with the Australian National Health and Medical Research Council Code of Care and Use of Animals for Scientific Purposes [54].

Study overview

Ten female Large White Landrace cross pigs (22 – 29 kg at baseline imaging) were randomly allocated to two injury groups: 10 cm injury (N = 5) and 20 cm injury (N = 5). Group sizes were determined a priori. Some of the animals in this study have been reported on previously (pre-SCI data [32]; and, 1 × 20 cm animal and 3 × 10 cm animals [55]), but the data presented herein is original.

All animals were acclimatised in a purpose-built facility for 7 − 10 days prior to the first procedure. Throughout the study, the animals had access to water, enrichment toys, and twice daily food rations. During acclimatisation they were administered once daily prophylactic antibiotics (Trimidine™ 0.15 mg/10kg; Sulfadimidine 430 mg/g, Trimethoprim 86 mg/g). Prior to surgery, the animals were trained for up to 1 h daily to transverse a 5 m long runway, using a target and clicker and a small amount of their normal food ration for reward. This was performed so the animals could successfully complete the Porcine Thoracic Behaviour Scale (PTIBS) motor function assessment task [56]. Central lines and intrathecal catheters (detailed below) were inserted pre-SCI at day-5 and day-3, respectively. MRI was performed at the following four time points: pre-SCI (day-5), and following SCI (day 0) at day-3, -7 and -14. Humane killing with intracardiac perfusion was performed on day-14 using 1 L of heparinised saline, followed by 2.5 L 10% neutral buffered formalin. One animal (P010; 20 cm group) was humanely killed prior to the experimental endpoint at day-10 post-SCI due to peritonitis of unknown cause; MRI and motor function data from this animal are included in this study up to day-8.

Anaesthetic protocol

Anaesthesia was administered for all surgeries and MRI. The animals were fasted overnight and transdermal fentanyl patches (slow release, 50 – 75 µg/hr) were applied approximately 12 h prior to the first surgery and were replaced every 3 days until 3 days post-SCI. Animals were pre-medicated with medetomidine (0.02 mg/kg) and ketamine (7.5 mg/kg); anaesthesia was inducted with ketamine (11 mg/kg) and propofol (2 mg/kg), and maintained with an intravenous protocol of ketamine (5.0 – 8.0 mg/kg/hr), propofol (2.0 – 6.0 mg/kg/hr), and fentanyl (8.0 – 15.0 mg/kg/hr). Midazolam (0.1 – 0.2 mg/kg/hr) was also administered for anaesthetic maintenance during post-SCI scans. A propofol bolus was administered when indicated to deepen anaesthesia. The animals were intubated and mechanically ventilated at 17 − 21 breaths/min with 270 − 400 mL/min of oxygen. Heart rate, oxygen saturation, and end-tidal carbon dioxide were recorded, and blood pressure were monitored throughout. Anaesthetics were titrated where necessary towards the end of the procedure or scan.

Catheter surgeries

Central lines using custom catheters (60 cm long x 2.5 mm outer diameter x 1.5 mm inner diameter polyethylene (PE) tubing) were placed 5 days before SCI, approximately 10 cm in the external jugular vein and the left carotid artery via surgical cut-down. These were used for arterial blood pressure monitoring, and administration of intravenous anaesthetics and drugs. Intrathecal catheters (25 cm long x 1.9 mm outer diameter x 1.4 mm inner diameter PE tubing) were inserted 3 days before SCI. The catheters were placed approximately 2 cm under the dura via laminectomy (< 1 cm) at spinal levels T7 and L1/L2. At the catheter-dura interface, cyanoacrylate glue (Loctite, Henkel, Düsseldorf, Germany) was applied to secure the catheter and prevent CSF leakage (confirmed visually until wound closure). These catheters were used for intrathecal pressure monitoring (not reported in this study).

Spinal cord injury and post-injury care

A comprehensive description of the surgical procedures, and post-surgical animal monitoring, has been provided previously [55]. In brief, spinal level T10 was located with C-arm fluoroscopy (Veradius, Philips, Netherlands) and a T9 – T13 dorsal midline incision was made. A T10 laminectomy (3.5 cm x 1.2 cm) was performed to expose the dura. Haemostasis of exposed trabecular bone was achieved with bonewax (W31G, Ethicon, NJ, USA), and of exposed vessels in the surgical site with Surgicel (1953, Ethicon, NJ, USA) and/or cautery. A custom weight drop device was secured to the spinal column at T12 and T13, and a 50 g impactor was released onto the exposed dura from 10 or 20 cm, and an additional 100 g weight was added for 5 min. A vacuum wound drain (Mini-Redon 50, Primed, Germany) was placed in the surgical site, the wound was closed in layers, and local anesthetic (Marcain 0.5% bupivacaine) was administered.

Control of the injury device and acquisition of synchronised data was performed with a custom controller and a data acquisition system that has previously been described [55]. Data were post-processed to determine the impactor velocity immediately prior to contact with the dura, and the peak impact force, impulse, and dura/cord displacement during impact [55].

A durotomy and immediate dural repair (designed as a surgical control for a separate study, unpublished) at the T10 dorsal midline was performed on three animals with a 20 cm injury (P001, P002 and P004), at 1 h post-SCI. A midline dural incision (2.5 cm long, centred at T10) was performed, and then immediately closed with a simple continuous stitch (5-0 polypropylene suture, EPH9702H, Ethicon) with minimal alteration of dural diameter. Axial ultrasound image sequences were obtained at T10 using a handheld linear array probe (L14-5/38, Ultrasonix RP, Ultrasonix Medical Co., Richmond, BC, Canada) with image resolution of 0.063 or 0.083 mm/pixel. Using the method previously reported [55], the dural diameter (dorsal-ventral) was measured on the image in the sequence corresponding to peak radial distension of the dura during the cardiac cycle. The mean reduction in dural diameter following durotomy was 5% (P001: 2%, P002: 6%, P004: 7%) (Additional file 1: Table 1). These animals were included in the 20 cm injury group.

All animals were provided continuous 24-hour care for 3 days post-SCI. During this time, they were administered analgesic (Meloxicam; 0.4 mg/kg) daily, antibiotics (Temgesic; 0.005–0.01 mg/kg buprenorphine) as required, and oxygen saturation, body temperature, food/water intake, and urine/faeces output were continuously monitored. From 3 to 14 days post-SCI, the animals were monitored at least three times a day and administered analgesics and antibiotics under veterinary guidance.

Spinal cord injury assessment: motor function and spinal cord lesion

SCI severity was assessed by hind-limb motor function, and histological analyses of spinal cord cumulative percentage lesion area and lesion extent. At day-8 and -13 post-SCI, two blinded investigators scored motor function of the animals from 1 (least) to 10 (normal) functional ability of the hind limbs using the PTIBS assessment scale [56]. A detailed description of histology procedures and image analysis has been reported [55]. In brief, spinal cord tissue was harvested following perfusion at day-14 post-SCI, processed and embedded in paraffin, and serial sectioned (axial, 5 μm thickness). Sections selected 500 μm apart were stained with luxol fast blue and hematoxylin and eosin. Lesion and spinal cord area were manually segmented on each section with NDP.view 2 software (Version 2.7.39, Hamamatsu Photonics), and the following parameters were obtained: cumulative percentage lesion area, calculated by summing the percentage of lesion (lesion area divided by the spinal cord area) on each section across the entire lesion; and, lesion extent, which was the total cranial-caudal length of the lesion.

Magnetic resonance imaging protocol

MRI was performed on a 3T scanner (Siemens, Magnetom Skyra, Germany). Animals were positioned in left lateral recumbency with two 18-channel body coils (Siemens, Germany) wrapped over the torso and neck. The following scans were performed: T2-weighted turbo spin echo (TSE), which was used to prescribe anatomical locations and orientations for PC-MRI planning; T2-weighted three-dimensional SPACE (sampling perfection with application optimised contrasts using different flip angle evolution), which was used for SAS occlusion measurements (described below); and, PC-MRI. Image and sequence parameters are available in Table 1.

Table 1 Magnetic resonance imaging (MRI) parameters for all sequences used in this study. TSE: turbo spin echo, SPACE: sampling perfection with application optimised contrasts using different flip angle evolution. PC: phase contrast. NRQ: not required. TR: repetition time. TE: echo time
Full size table

PC-MRI acquisition was performed with retrospective cardiac gating via a pulse oximeter attached to the tail. At spinal levels C2/C3, T8/T9, T11/T12 and L1/L2, single, axial slices orthogonal to the spinal cord in the sagittal and coronal planes were acquired (Fig. 1A-F). The images were aligned with the centre of the adjacent intervertebral disc. PC-MRI were obtained with anterior-posterior phase encoding; cranially directed flow was encoded positively, and caudally directed flow encoded negatively. The number of acquired cardiac phases was selected according to the target R-R interval (R-R interval of the cardiac cycle, divided by the repetition time) and ranged from 14 to 33. For the first seven animals, encoding velocities (VENC) of 4 cm/s and 6 cm/s were selected based on a pilot study. Following completion of the first seven animals, aliasing was detected in animals at C2/C3 in the VENC = 6 cm/s scans; therefore, for the last three animals (P008 – P010), VENC of 6 cm/s, 8 cm/s, 10 cm/s were selected.

Fig. 1
figure 1

Pre-injury anatomical T2 turbo spin echo (TSE) mid-sagittal magnetic resonance images (MRI) of the (A) cervical to thoracic and (B) thoracic to lumbar spine, illustrating the planes of phase-contrast MRI. Magnitude image and the corresponding phase images at peak cranial and caudal flow from one animal (P005), pre-injury, with region of interest in red segmented on the border of flow signal (separately into dorsal (d) and ventral (v) regions), at (C) C2/C3 (with an additional dorsal region because of the two distinct compartments of flow signal in the magnitude scan), (D) T8/T9, (E) T11/T12, and (F) L1/L2. (G) Exemplar pre-injury, post-processed cerebrospinal fluid (CSF) mean velocity over one cardiac cycle, (*) illustrates peak cranial and peak caudal mean velocity for C2/C3 and L1/L2

Full size image

Subarachnoid space occlusion

SAS occlusion measurements were performed using Materialise Mimics Software (Version 22.0, Materialise, NV). SAS occlusion length was defined as the cranial-caudal length with no apparent CSF signal on the mid-sagittal slice. It was selected in either the dorsal or ventral region exhibiting the longest occlusion. SAS occlusion length was measured along the midline of the spinal cord, normalised to the total distance between T8/T9 and T11/T12, and expressed as a percentage (Fig. 2A). Cross-sectional SAS area was measured on a single axial slice exhibiting maximum occlusion in post-SCI images, and a slice centred on T10 in pre-SCI images. It was calculated as the total SAS cross-sectional area (i.e., the CSF and spinal cord signal since MRI resolution precludes visualisation of the meninges) minus the spinal cord area (Fig. 2B-C). When there was no visible CSF signal, SAS cross-sectional area = 0 mm2.

Fig. 2
figure 2

Representative T2-weighted magnetic resonance imaging (MRI) illustrating subarachnoid space (SAS) occlusion measurements. (A) A T10 spinal cord injury (SCI) on a mid-sagittal slice. The dotted arrow illustrates SAS occlusion length, and the solid line arrow illustrates the distance between T8/T9 and T11/T12. (B) An axial slice centred on T10 in a pre-SCI MRI, with the same image in (C) illustrating the spinal cord and SAS cross-sectional area measurements

Full size image

Post-processing PC-MRI

For all spinal levels, regions of interest (ROI) were manually segmented using Segment software (Version 3.2, Medviso, Lund, Sweden) [57]. The ROI delineated the border of pixels, within the SAS, whose signal intensity was above the background (Fig. 1C-F). This was performed on the magnitude image that displayed the greatest contrast between flow and the background signal. Flow signal was not uniform in the SAS (e.g., flow signal detectable in the ventral region but not the dorsal region, and vice versa), so the SAS was segmented into dorsal and ventral regions (reported in Additional File 2: Table 2). When two distinct compartments were observed within one region, two ROIs were segmented (Fig. 1C) to ensure that only pixels with flow signal were included in the ROI. The ROIs were propagated to all temporal images since no motion of the spinal canal was detected. The following data were obtained for each ROI: mean velocity (cm/s) across the ROI for each cardiac image; CSF flow (mL/s) across the ROI for each cardiac image; stroke volume (mL/cycle), which is the sum of the absolute negative and positive flow volume across one cardiac cycle (mL/cycle); maximum cranial and caudal velocity (single pixel; cm/s) across one cardiac cycle.

All images were assessed for aliasing artefact (regions where the flow velocity is higher than the VENC, causing phase-wrapping of the signal and thereby display of velocity in the opposite direction). The VENC selected for flow analyses was chosen to obtain the optimal velocity resolution while minimising aliasing artefact. Aliasing was corrected with phase-unwrapping in Segment software in 73% of scans at C2/C3, 3% at T8/T9, 6% at T11/T12, and 9% at L1/L2, with detected flow. At most six consecutive cardiac phases were corrected (mean: 3 ± 1), and at most 14% of the total number of pixels were corrected at C2/C2, 3% at T8/T9, 2% at T11/T12, and 11% at L1/L2.

Eddy current offset correction (to reduce gradient field distortions) was applied to each scan. This was performed by iteratively adjusting the magnitude threshold in Segment software to achieve as close to zero flow as possible, across all the ROIs, using a method previously reported [32]. The mean offset magnitude applied at C2/C3 was 0.36 ± 0.11 cm/s, T8/T9 was 0.27 ± 0.04 cm/s, T11/T12 was 0.28 ± 0.04 cm/s, and L1/L2 was 0.28 ± 0.05 cm/s.

Data were further processed using a custom MATLAB program (Version R2022a, Mathworks Inc, Natick, MA). Due to sample size of the study, all ROIs within one scan were combined during post-processing: mean velocity data from each ROI were averaged, and flow data from each ROI was summed. To normalise for heart rate across animals and acquisition session, flow and mean velocity versus time data for each cycle were interpolated to 100 points using a cubic spline (“pchip” MATLAB function which preserves maxima and minima) and expressed as a percentage of the cardiac cycle (Fig. 1G). The following parameters were obtained: peak cranial (positive) and caudal (negative) mean velocity (cm/s) and flow (mL/s), which were selected in the half of the cardiac cycle which corresponds to the systolic and/or diastolic pulse in healthy animals; and, the corresponding time to peak cranial and caudal mean velocity (% of cardiac cycle, from t = 0 in the cardiac cycle).

Statistics

All statistical analyses were performed with SPSS (version 28, IBM Corporation, Armonk, NY). Mann-Whitney U tests were used to compare impact velocity, peak force, motor function, cumulative percentage lesion area, and lesion extent (significance level of α = 0.05; two sided) between injury groups. Normality and homogeneity of variance were assessed for all remaining outcomes using Shapiro-Wilk and Levene’s tests. Separate linear mixed-effects models (LMM), with animal number as a random effect, were developed to assess the following: (1) effect of injury group, time point, and their interaction, on occlusion length and cross-sectional SAS area; (2) effect of each scanning session (time point and spinal level), on heart rate, respiratory rate, and end-tidal carbon dioxide; and, (3) effect of time point and spinal level, adjusted for injury group, on peak mean CSF velocity, peak CSF flow, time to peak mean velocity, maximum velocity, and stroke volume. The interaction of time and spinal level was not significant in any model and was removed from the final models. Post-hoc pairwise comparisons of time point at each spinal level were performed, with Bonferroni correction, for categorical factors with three or more levels (significance level of α = 0.05). Estimated marginal mean (EMM) differences and 95% confidence interval (CI) are presented in text. Pairwise comparisons and estimated marginal means are available in Additional files 3–7.

Results

Assessment of the biomechanical, functional, and histological data indicated that SCI was induced in all of the animals in this study. The two weight drop heights induced stratified injuries, except for lesion extent (Table 2). All mechanical outputs and injury severity assessments are available in Additional file 8: Table 8. Percentage lesion area data are available in Additional file 9: Fig. 1.

Table 2 Assessment of spinal cord injury (SCI) across the 20 cm and 10 cm injury groups. Data presented as median and range, and p-values correspond to Mann-Whitney U tests. PTIBS: Porcine Thoracic Behaviour Scale
Full size table

Subarachnoid space occlusion

SAS occlusion length, normalised to the distance between T8/T9 and T11/T12, was longer in the 20 cm animals than the 10 cm animals at day-3 (EMM difference: 36.7%, 95% CI [17.3, 54.2], p = 0.002), day-7 (EMM difference: 35.8%, 95% CI [17.3, 54.2], p = 0.002) and day-14 (EMM difference: 28.2%, 95% CI [9.6, 46.8], p = 0.007). SAS occlusion length decreased from day-3 to day-7 in the 20 cm group (EMM difference: 11.4%, 95% CI [5.3, 17.5], p < 0.001) and in the 10 cm group (EMM difference: 11.4%, 95% CI [5.3, 17.6], p < 0.001). SAS occlusion length decreased from day-7 to day-14 in the 20 cm group (EMM difference: 14.4%, 95% CI [7.8, 21.1], p < 0.002) and in the 10 cm group (EMM difference: 6.9%, 95% CI [0.7, 21.1], p = 0.028) (Fig. 3A-B, C).

Fig. 3
figure 3

Representative mid-sagittal T2-weighted magnetic resonance imaging (MRI) of subarachnoid space (SAS) occlusion following a T10 spinal cord injury (SCI) from (A) one 20 cm injury animal (P003) and (B) one 10 cm animal (P006), from pre-SCI to day-3, -7 and -14 post-SCI. The dashed arrow between the dashed lines illustrates SAS occlusion length between T8/T9 and T11/T12. Axial T2-weighted MRI at the injury epicentre, with labelled dorsal (d) and ventral (v) directions, and illustrative cross-sectional measurements in the panels below; blue indicates spinal cord cross-sectional area, and red indicates SAS cross-sectional area. (C) SAS occlusion length between T8/T9 and T11/T12 (%) for day-3, -7 and -14 post-SCI for each animal (circles) with mean (solid line) and standard deviation (shaded area); 20 cm injury group in black and 10 cm injury group in grey. (D) SAS cross-sectional area at T10 for pre-SCI, and at the injury epicentre for day-3, -7 and -14 post-SCI, for each animal (circles) with mean (solid line) and standard deviation (shaded area); 20 cm injury group in black and 10 cm injury group in grey. Brackets indicate statistical significance, which correspond to linear mixed effects models with Bonferroni post-hoc pairwise comparisons

Full size image

There was no difference in cross-sectional SAS area between the 20 cm and 10 cm animals pre-SCI (EMM difference: 1.88 mm2, 95% CI [-4.21, 7.98], p = 0.52). At day-3 following SCI, the SAS appeared fully occluded (SAS cross-sectional area = 0 mm2) in all of the 20 cm animals, and in two of five 10 cm animals. The 20 cm animals had lower cross-sectional SAS area at the injury epicentre than the 10 cm animals at day-3 (EMM difference: -6.88 mm2, 95% CI [-12.98, -0.78], p = 0.03), day-7 (EMM difference: -14.09 mm2, 95% CI [-20.19, -8.00], p < 0.001) and day-14 (EMM difference: -18.23 mm2, 95% CI [-24.43, -12.04], p < 0.001). Cross-sectional SAS area decreased from pre-SCI to day-3 in the 20 cm group (EMM difference: 27.78 mm2, 95% CI [24.70, 30.87], p < 0.001) and in the 10 cm group (EMM difference: 22.79 mm2, 95% CI [19.70, 25.87], p < 0.001). Cross-sectional SAS area increased from day-3 to day-7 in the 20 cm group (EMM difference: 9.12 mm2, 95% CI [-12.21, -6.04], p < 0.001) and in the 10 cm group (EMM difference: 16.33 mm2, 95% CI [-19.42, -13.25], p < 0.001), and increased from day-7 to day-14 in the 20 cm group (EMM difference: 6.31 mm2, 95% CI [-9.63, -2.98], p < 0.001) and in the 10 cm group (EMM difference: -10.45 mm2, 95% CI [-13.54, -7.36], p < 0.001) (Figure A-B, D).

Cerebrospinal fluid flow detection

Across all spinal levels and animals, a total of 34/40 pre-SCI and 97/116 post-SCI PC-MRI scans detected CSF flow. Data were excluded from analyses when no flow was detected since it is possible that there was flow below a detectable threshold determined by the amount of phase noise present. No flow was detected in five scans at T8/T9 day-3 (Additional file 2: Table 2). The animal with the greatest length of SAS occlusion (P002; 99.5% occlusion between T8/T9 and T11/T12 at day-3 post-SCI) had no flow signal at T8/T9 and T11/T12 pre-SCI and at day-3. At T11/T12, two animals at day-3, and three animals at day-7 had no detected flow in the dorsal SAS, but flow in the ventral SAS. Abnormal cardiac gating, determined by incomplete capture of a full CSF pulsation, was identified in one pre-SCI (P009), and one day-3 (P003) scan. These data were excluded from temporal analyses (Additional file 10: Fig. 2). For P009, data were retained for analyses of peak mean velocity and flow (with peak data allocated to the correct CSF flow direction), since the maxima and minima were within one standard deviation, or marginally outside the range, of the other data, and for stroke volume since complete bi-directional pulses were captured. For P003, data were retained for analyses of peak caudal mean velocity and flow at C2/C3, and peak cranial mean velocity and flow at T8/T9 – L1/L2, but not for stroke volume, since only a portion of the cardiac cycle was captured. Two C2/C3 scans at day-7 (P001 and P007), and four at day-14 post-SCI (P004, P005, P006, and P007), were removed from analyses due to severe aliasing (more than 50% of the ROI were aliased). One L1/L2 scan at day-14 was excluded because only nine cardiac phases were acquired (P003).

Heart rate did not change from pre-SCI to day-3 (p = 0.372) and day-3 to day-7 (p = 1.000), but heart rate was slower at day-14 than at day-3 (p = 0.002) and day-7 (p = 0.044) (Table 3). Respiratory rate was faster pre-SCI than at day-7 (p = 0.023) and day-14 (p = 0.019). There was no change in end-tidal carbon dioxide between each time point (p = 0.721).

Table 3 Mean physiological recordings at each time point pre- and post-spinal cord injury (SCI)
Full size table

CSF dynamics after spinal cord injury

Spinal level-specific changes to CSF dynamics, adjusted for injury group, were assessed over time after SCI. Injury group (10 cm vs. 20 cm) was not a significant covariate for any of the CSF dynamics measures, except peak cranial mean velocity (EMM difference: 0.23 cm/s, 95% CI [-0.41, -0.05], p = 0.021) and stroke volume (EMM difference: -0.03 mL/cycle, 95% CI [-0.07, 0.00], p = 0.049).

Qualitatively, bi-directional CSF velocity waveforms were usually observed at all spinal levels and time points pre- and post-SCI. At day-3 post-SCI, T8/T9 lacked a pronounced cranial-caudal pulse in all animals with detected flow (Fig. 4). In one animal (P007), cranial flow at T8/T9 and T11/T12 at both day-7 and day-14 occurred in the first half of the cardiac cycle, with a second cranial pulse in the second half of the cycle (Additional file 11: Fig. 3).The standard deviation at all time points were consistent across the cardiac cycle at the thoracolumbar levels, and the standard deviation at C2/C3 was larger than the thoracolumbar levels, at every time point (Additional file 12: Fig. 4).

Fig. 4
figure 4

Averaged mean velocity waveforms over one cardiac cycle for four spinal levels pre-SCI, and at day-3, -7 and -14 post-spinal cord injury (SCI) for the (A) 20 cm group and (B) 10 cm group

Full size image

Peak cranial and caudal mean velocity changed over time after SCI (Fig. 5). At all spinal levels, peak cranial mean velocity decreased from pre-SCI to day-3 (EMM difference: 0.33 cm/s, 95% CI [0.08, 0.59], p = 0.004) and day-7 (EMM difference: 0.26 cm/s, 95% CI [0.02, 0.51], p = 0.029) post-SCI. There was no change in peak cranial mean velocity from pre-SCI to day-14 (EMM difference: 0.09 cm/s, 95% CI [-0.17, 0.34], p = 1.000). Peak caudal mean velocity decreased from pre-SCI to day-3 (EMM difference: -0.35 cm/s, 95% CI [-0.62, -0.08], p = 0.004), and increased from day-3 to day-14 (EMM difference: 0.32 cm/s, 95% CI [0.04, 0.61], p = 0.018). There was no difference in peak caudal mean velocity between pre-SCI and day-14 (EMM difference: -0.04 cm/s, 95% CI [-0.30, 0.23], p = 1.000).

Fig. 5
figure 5

Peak mean velocity (cranial/caudal) pre-spinal cord injury (SCI), and day-3, -7 and -14 post-SCI in animals with a 20 cm (black) or 10 cm (grey) weight-drop height SCI. Significant differences between time points correspond to linear mixed effects models, adjusted for injury group, with Bonferroni-corrected post-hoc pairwise comparisons. Each data point (circle) is from one animal, and error bar is mean ± one standard deviation

Full size image

Time to peak caudal mean velocity changed over time after SCI, but there was no difference in time to peak cranial mean velocity over time (p = 0.375) (Fig. 6). At all spinal levels, time to peak caudal mean velocity decreased from pre-SCI to day-3 (EMM difference: 11.10%, 95% CI [3.35 18.84], p = 0.001) and day-7 (EMM difference: 9.56%, 95% CI [2.19 16.94], p = 0.004), and increased from day-3 to day-14 (EMM difference: -8.20%, 95% CI [-16.01 -0.38], p = 0.034).

Fig. 6
figure 6

Time to peak mean velocity (cranial-caudal) pre-spinal cord injury (SCI), and at day-3, -7 and -14 post-SCI in animals with a 20 (black) cm or 10 cm (grey) weight-drop height SCI. Significant differences between time points correspond to linear mixed effects models, adjusted for injury group, with Bonferroni-corrected post-hoc pairwise comparisons. Each data point (circle) is from one animal, and error bar is mean ± one standard deviation. Note: the caudal/cranial pulse is in the opposite portion of the cardiac cycle at C2/C3 than at the thoracolumbar levels

Full size image

Maximum cranial and caudal velocity changed over time after SCI. At all spinal levels, from day-3 to day-14, maximum cranial velocity (EMM difference: -1.41 cm/s, 95% CI [-2.61, -0.20], p = 0.013) and maximum caudal velocity increased (EMM difference: 1.53 cm/s, 95% CI [0.35, 2.71], p = 0.004); however, from pre-SCI to post-SCI time points there was no change in maximum cranial velocity (day-3: EMM difference: 0.29 cm/s, 95% CI [-0.86, 1.45], p = 1.000; day-7: EMM difference: -0.18 cm/s, 95% CI [-1.29, 0.93], p = 1.000; day-14: EMM difference: -1.11 cm/s, 95% CI [-2.29, 0.07], p = 0.076) and caudal velocity (day-3: EMM difference: -0.58 cm/s, 95% CI [-1.70, 0.55], p = 1.000; day-7: EMM difference: 0.08 cm/s, 95% CI [-1.00, 1.16], p = 1.000; day-14: EMM difference: 0.95 cm/s, 95% CI [-0.20, 2.11], p = 0.170). There was no change to peak cranial flow (p = 0.138), peak caudal flow (p = 0.358), and stroke volume (p = 0.074) over time after SCI.

Discussion

SAS occlusion and pulsatile CSF dynamics were measured using MRI across 14 days after a T10 contusion SCI in the pig. The SAS appeared fully occluded at the injury epicentre in all of the 20 cm animals, and in two of five 10 cm animals at day 3 post-SCI. The occlusion decreased in length, and the cross-sectional SAS area increased, from day-3 to day-14 post-SCI. Decreased peak mean velocity and the timing of caudal CSF pulsations at day-3 post-SCI, with return towards baseline at day-14, were observed.

The magnitude of SAS occlusion, measured as occlusion length and cross-sectional SAS area on MRI, was greatest at day-3 and reduced over time up to day-14 following SCI. In the current study, a higher weight drop was associated with larger cumulative percentage lesion area at 14 days post-SCI, greater longitudinal extent of SAS occlusion, reduced cross-sectional SAS area at the injury epicentre, and worse hind limb dysfunction across two weeks post-SCI (Table 2). Longitudinal extent of SAS occlusion decreased from day-3 to -14 post-SCI in 9 of 10 animals. In addition, radial SAS occlusion did not develop or resolve uniformly, leading to apparent occlusion on the dorsal side, but not on the ventral side, and vice versa (example in Fig. 3B). Oedema and haemorrhage development can cause the spinal cord to swell, press against dural tissue, and occlude the SAS [5, 9, 13], relative to the initial injury severity [4]. There is limited data addressing the temporal progression of SAS occlusion reduction after SCI; one clinical study showed that increased length of SAS occlusion was associated with worse American Spinal Injuries Association Impairment Scale (AIS) score on admission, and that occlusion length decreased approximately exponentially from the first scan to ∼50 days post-SCI [5]. In the current study, SAS occlusion post-SCI was caused by apparent spinal cord swelling in 65% of all post-SCI scans, and by epidural haematoma formation at the laminectomy site (i.e., extradural compression dorsally) in 35% of post-SCI scans. The presence of this extradural compression has previously been reported, and shown on MRI, for some of the animals in this study [55]. Clinically, haematomas may occur asymptomatically following spinal decompression in up to 50% of patients with various obstructive spinal pathologies [58].

Peak cranial and caudal mean CSF velocity were reduced at each monitored level above and below the injury site at day-3 after SCI in this model. Similar reductions to peak mean CSF velocity, above and below an occlusion, have been observed clinically in other SAS obstructive pathologies. In cervical myelopathy with confirmed occlusion (defined as dural compression), peak cranial/caudal mean CSF velocity were lower at spinal levels above and below the occlusion in patients with worse functional and sensory scores [49]. Peak caudal velocity was lower at spinal levels above and below the occlusion in cervical myelopathy patients, compared to healthy controls, and returned to normal following decompression surgery [50]. Although spinal morphology and SAS occlusion were not described, lower peak cranial/caudal mean CSF velocity were observed at, and above, the level of the syrinx (T3 – L1) in post-traumatic syringomyelia [52], and lower peak cranial mean CSF velocity was observed below the injury in traumatic and non-traumatic cervical SCI (C2 – C7) from 1 month to 6 years after injury [53], compared to healthy adults. In these cervical SCI patients, peak caudal mean CSF velocity was higher at the level of maximum occlusion (or level of SCI), compared to the C4/C5 level in healthy controls [53]. In the current study, the studied spinal levels were selected because they reliably had a patent SAS based on a pilot study of 20 cm injury animals. Future studies could acquire flow data at the injury epicentre following occlusion resolution in all of the 10 cm animals, and a subset of the 20 cm animals.

In this SCI model, peak caudal mean CSF velocity returned towards pre-SCI baseline between day-3 and day-14 post-SCI, and peak cranial mean CSF velocity showed a similar (non-significant) tendency towards baseline. In the current study, substantial aliasing could not be corrected at C2/C3, for 2 of 10 animals at day-7, and 4 of 10 animals at day-14, leading to exclusion of these data from the statistical analyses. Inclusion of these higher velocity data would have elevated the C2/C3 peak caudal and cranial mean velocity, and amplified the recovery-to-baseline observed at day-7 and -14 post-SCI at this level. These changes to peak mean CSF velocity, and the concurrent changes to SAS occlusion, over time post-SCI indicates a potential relationship between the two. To our knowledge, there are no longitudinal clinical PC-MRI studies of pathology-associated CSF flow; however, clinical studies have shown that a worse stenosis is associated with lower peak mean velocity in lumbar spinal stenosis [51] and cervical myelopathy [49, 50]. In cervical myelopathy, pulsatile CSF waveforms were absent in patients with “severely” compressed dura. These suggest that the degree of SAS occlusion, and its subsequent resolution, likely contributed to the observed changes in peak caudal mean velocity over time, although the current study was not designed to test the effect of occlusion on CSF dynamics. It is possible that the changes to CSF dynamics observed in the current study may also be due, at least in part, to SCI-related changes to the mechanical properties of the spinal cord and surrounding tissues (i.e., as a result of oedema and/or haemorrhage) and/or systemic physiology. To assess the effect of SAS occlusion on CSF dynamics in isolation, future studies could occlude the SAS without SCI (e.g., via extradural ligature or intradural balloon).

A shift in the temporal characteristics of pulsatile CSF flow was detected after SCI in the pig, which may indicate a change in craniospinal compliance. In the current study, peak caudal mean velocity occurred earlier in the cardiac cycle at day-3 post-SCI, and returned towards baseline timing at day-14 post-SCI, at each measured spinal level (C2/C3 − L1/L2). Caudal CSF pulsations can also occur earlier in the cardiac cycle in patients with Chiari malformation (without a syrinx) [45] with CSF flow obstruction out of the cranium, and post-traumatic syringomyelia [52] with SAS obstruction surrounding the syrinx, than in healthy humans. Following decompression surgery in Chiari malformation, the timing of the caudal pulse occurred later in the cardiac cycle [44, 46], and although compliance of the spinal compartment was not measured in the study, the compliance of the intracranial compartment increased compared to pre-operative values [47]. While obstructive spinal pathologies may change compliance of the system due to a pathological increase in the volume of the intradural contents and/or a decrease in the local volume available for CSF, pathologic tissue changes that are independent of volumetric changes may also alter spinal compliance. Changes to the timing of CSF pulsations have been measured along the spine in amyotrophic lateral sclerosis patients [59], a pathology not marked by SAS occlusion. These observed changes to CSF dynamics suggest that compliance of the tissues within and surrounding the spinal canal, and changes to their biomechanical properties with pathology (i.e., oedema and haemorrhage), may also contribute to changes in the temporal features of CSF flow; however, the extent of this effect is unclear.

In the current study, variable physiological parameters were observed despite a standardised anaesthetic protocol. The effect of heart rate, respiratory rate, and arterial carbon dioxide, on CSF flow magnitude in the pig, has not been reported. However, there were no marked changes in the physiological recordings within a single scanning session, and no indication that vital signs became unstable during scanning at pre-SCI and all days post-SCI. Heart rate decreased at day-14 post-SCI, but current evidence for the effect of heart rate on pulsatile CSF flow is limited and conflicting. Change in heart rate has little effect on pulsatile CSF flow in rats [33], whereas in a three dimensional computational fluid dynamics model, increased heart rate caused increased bidirectional CSF velocity [60]. In the current study, additional LMMs were developed to assess the effect of time point and spinal level on CSF dynamics while adjusting for heart rate or respiratory rate. In these models, heart rate was not a significant covariate, except for time to peak caudal mean velocity (Additional file 13). Time to peak caudal mean velocity occurred earlier in the cardiac cycle at day-3 post-SCI than pre-SCI, but there was no change from day-3 to day-14 post-SCI. In addition, although evidence shows that respiration is a major driver of CSF flow [20,21,22, 33], and respiratory rate (under ventilation) was slower at day-7 and -14 compared to pre-SCI, respiratory rate was not a significant covariate in any of the models testing CSF dynamics over time after SCI. These analyses suggest that respiratory rate (and heart rate) were not the primary drivers of changes to CSF dynamics in the current study. However, the potential effect of other systemic physiological changes associated with SCI on CSF dynamics is not clear, and should be measured in future studies. For example, we did not measure arterial pressure or flow concurrently with image acquisition; doing so in future studies may provide further information on the coupling of arterial and CSF pulsations.

The complex interplay between cardiopulmonary conditions and spinal vasculature on CSF dynamics needs to be better understood within and across species to help bridge the gap between pre-clinical models and clinical studies of CSF flow. Respiratory conditions (spontaneous breathing vs. positive pressure ventilation) can affect the magnitude of pulsatile CSF flow in the spinal SAS [33]. Intrathoracic and intra-abdominal pressures drive CSF flow with respiration, which are likely due to changes in blood volume within epidural veins [22]. In pigs, the ventral vertebral venous plexus (epidural plexus in humans) dominates drainage of blood from the brain (in lateral recumbency) compared to the internal jugular vein in humans (in supine position) [61]. This may suggest a difference in the volume of blood (or size of these veins) relative to the spinal canal in pigs, than in humans; however, the effect of this on CSF dynamics remains unknown. In addition, although the SAS appeared fully occluded in the current study (i.e., no visible CSF signal at the injury epicentre on anatomical MRI), pulsatile CSF flow was detected below the occlusion (at T11/T12) in 8 of 9 animals. This may suggest that the SAS was not completely occluded in this study (image spatial resolution and partial volume averaging effects preclude absolute confirmation), and/or that local spinal sources for CSF pulsations in the pig are different than in humans. These indicate that further study is warranted.

Altered or abnormal CSF dynamics, measured with PC-MRI, may be a viable non-invasive biomarker of morphological and/or pathological alterations to spinal central nervous system. This may have implications for potential patient-specific treatment and management strategies following SCI, for example: adapting intrathecal drug delivery protocols based on CSF dynamics to optimise drug delivery to target tissue [37], expansion duraplasty to target dural compression [13], and CSF drainage to reduce mean CSF pressure [62]. Although continuous measurement of CSF pressure pulsations can be achieved with pressure monitoring, invasive surgical placement of indwelling sensors is required for injury-site monitoring, which increases the risk of further damage to the spinal cord [8] or other surgical complications. Remote CSF pressure dynamics are also of current interest to assess spinal cord compression: CSF pressure pulsations were absent [11], or altered [14], below a confirmed occlusion in SCI patients. While the current study demonstrates that PC-MRI can detect changes to CSF dynamics in 14 days following a contusion SCI in a pig, further studies are needed to understand the concordance between CSF pressure measures and PC-MRI flow measures.

This study is limited with respect to sample size, but similar numbers have been used in other large animal studies of CSF flow [29,30,31] and survival SCI [63,64,65]. Injury group was a priori designated a covariate rather than a main effect in LMMs that examined whether spinal-level specific changes occurred to CSF dynamics over time after SCI. Future studies could increase group size to test the effect of injury group, and the interaction of spinal level and time point, on CSF dynamics. In the current study, intrathecal catheters were present at T7 and L2 in all post-SCI scans. CSF leaks can occur with intrathecal catheters, as shown in patients undergoing intrathecal drug therapy [66]. While CSF leaks were not observed immediately following insertion in the current study, and there was no evidence of major CSF leaks indicated on MRI (as extradural CSF signal on T2-weighted STIR fat suppression sequence; spinal levels imaged: T4 – L1; in-plane resolution: 0.8125 × 0.8.125 mm/pixel; slice thickness: 0.800 mm) at each time point, it is possible that CSF leaks developed with no indication on MRI, similar to that observed clinically [67]. In addition, a study of CSF flow in NHP showed that intrathecal catheters of similar outer diameter (1.98 mm, cf. 1.9 mm in this study), inserted from L5 to T12/L1 or C5, can influence hydraulic resistance within the spinal SAS and reduce cranial/caudal CSF flow pulsations [68]. However, in the current study, the intrathecal catheters were inserted under the dura the distance of approximately one spinal level, and the pig SAS has larger cross-sectional area (28.7 ± 2.6 mm2) than that of the NHP (∼ 20 – 25 mm2, estimated from graphical data [68]). Furthermore, CSF mean velocity and time to peak mean velocity returned towards pre-SCI (and pre-intrathecal catheter insertion) baseline values at day 14, suggesting that the laminectomy surgery and the presence of intrathecal catheters, did not dominate the responses observed post-SCI. Although 3 of 5 animals in the 20 cm injury group received a durotomy, the mean dural diameter at T10 immediately following durotomy (6.96 ± 0.24 mm) was only slightly lower than one standard deviation from the mean for pre-SCI pigs (7.47 ± 0.39 mm) and less than 7% lower than the animal-specific pre-injury SAS diameter. At each post-SCI time point in the durotomy animals, local SAS volume remained reduced because of extradural compression or continued spinal cord swelling, likely minimising any potential effects of the durotomy. In addition, the cumulative percentage lesion area, lesion extent, and functional data appeared similar across the 20 cm animals with and without the durotomy. While it is possible that the durotomy had an effect on CSF dynamics along the spine, the temporal changes for the durotomy animals were broadly consistent with that of the non-durotomy animals (Additional file 14: Fig. 5).

Although cardiac gated PC-MRI have demonstrated increasing clinical utility, there are some limitations of the method. Two-dimensional PC-MRI cannot identify complex flow patterns, such as flow in the in-plane direction, since it encodes flow in the through-plane direction. Complex flow patterns have been shown at the level of the obstruction in Chiari malformations [43]. Four-dimensional PC-MRI has been used to investigate CSF flow in the spine [43, 69, 70] and although this technique can identify these complex flow patterns, the markedly longer scan duration may present challenges in animal models and clinical implementation. The best practice for acquiring and post-processing PC-MRI in the spine has not been standardised: imaging parameters, ROI segmentation method, eddy current compensation, and temporal adjustments appear to vary across research groups. In the current study, ROIs were segmented according to CSF flow signal on the magnitude scan, rather than necessarily encompassing the full SAS area. A limitation of this approach is that ROI area varied between time points for the same animal. Mean ROI area increased from pre-SCI to day-14 post-SCI in the dorsal and ventral regions. In addition, CSF flow within the ROI could be different to the total CSF flow within the SAS at that spinal level. Together these could potentially influence flow outcomes reported. Some studies have used semi-automated methods to segment ROIs based on pixel intensity [28]; however, manual segmentation of CSF flow signal in healthy pigs, by the same investigator as in this study (MAB), has previously demonstrated excellent intra-rater reliability [32]. In some animals at C2/C3, substantial aliasing was detected at day-7 and -14 post-SCI (i.e. the CSF velocities were higher than could be reliably measured), and the exclusion of these data likely reduced the group peak mean velocity values reported at those time points. At day-14 post-SCI, 3 of 4 of these exclusions were in the 10 cm injury group. In this study, the number of cardiac phases acquired was adjusted for heart rate variability to maintain adequate temporal resolution and signal-to-noise ratio. As a result, temporal resolution varied (15 – 57 ms), which may influence the temporal outcomes reported [71]. The minimum number of phases collected were relatively low compared to other studies, and future studies could increase the minimum number of phases acquired. While technical failures and/or insufficient CSF flow have been reported in clinical PC-MRI studies [20, 25, 59, 72], the cause of no detected flow in some of the pre-SCI scans, and abnormal cardiac gating in two scans, remains unresolved. Future investigations into this may clarify the cause of no detected flow at post-SCI time points.

Conclusion

Spinal-level specific changes to CSF dynamics, with concurrent reductions in SAS occlusion, were observed over 14 days following a thoracic contusion SCI in the pig. Reduced peak mean CSF velocity indicates decreased pulsatility of the CSF, and earlier onset of peak caudal mean velocity suggests a change in craniospinal compliance at day-3 post-SCI. This study provides evidence that further study using PC-MRI in this SCI model may assist understanding of alterations to CSF dynamics and their relationship with SAS occlusion, spinal cord compression, and pathological changes to spinal cord tissue following SCI.

Data availability

The datasets supporting the conclusions of this article are available in the Figshare repository, 10.25909/25623033.

Abbreviations

AP:

Anterior posterior

CI:

Confidence interval

CSF:

Cerebrospinal fluid

C2/C3:

Cervical spine between levels 2 and 3

LMM:

Linear mixed effects model

L1/L2:

Lumbar spine between levels 1 and 2

MRI:

Magnetic resonance imaging

NHP:

Non-human primates

PC-MRI:

Phase contract magnetic resonance imaging

ROI:

Region of interest

SAS:

Subarachnoid space

SCI:

Spinal cord injury

SD:

Standard deviation

TR:

Repetition time

TE:

Echo time

T8/T9:

Thoracic spine between levels 8 and 9

T11/T12:

Thoracic spine between levels 11 and 12

References

  1. Ahuja CS, Wilson JR, Nori S, Kotter MRN, Druschel C, Curt A, Fehlings MG. Traumatic spinal cord injury. Nat Reviews Disease Primers. 2017;3(1):17018.

    Article  PubMed  Google Scholar 

  2. Bilgen M, Abbe R, Liu S-J, Narayana PA. Spatial and temporal evolution of hemorrhage in the hyperacute phase of experimental spinal cord injury: in vivo magnetic resonance imaging. Magn Reson Med. 2000;43(4):594–600.

    Article  PubMed  CAS  Google Scholar 

  3. Fujii H, Yone K, Sakou T. Magnetic resonance imaging study of experimental Acute spinal cord Injury. Spine. 1993;18(14):2030–4.

    Article  PubMed  CAS  Google Scholar 

  4. Jones CF, Cripton PA, Kwon BK. Gross morphological changes of the spinal cord immediately after surgical decompression in a large animal model of traumatic spinal cord injury. Spine. 2012;37(15):E890–9.

    Article  PubMed  Google Scholar 

  5. Saadoun S, Werndle MC, Lopez de Heredia L, Papadopoulos MC. The dura causes spinal cord compression after spinal cord injury. Br J Neurosurg. 2016;30(5):582–4.

    Article  PubMed  Google Scholar 

  6. Badhiwala JH, Wilson JR, Witiw CD, Harrop JS, Vaccaro AR, Aarabi B, et al. The influence of timing of surgical decompression for acute spinal cord injury: a pooled analysis of individual patient data. Lancet Neurol. 2021;20(2):117–26.

    Article  PubMed  CAS  Google Scholar 

  7. Aarabi B, Olexa J, Chryssikos T, Galvagno SM, Hersh DS, Wessell A, et al. Extent of spinal cord decompression in Motor Complete (American Spinal Injury Association Impairment Scale Grades A and B) traumatic spinal cord Injury Patients: Post-operative Magnetic Resonance Imaging Analysis of Standard operative approaches. J Neurotrauma. 2019;36(6):862–76.

    Article  PubMed  PubMed Central  Google Scholar 

  8. Saadoun S, Papadopoulos MC. Spinal cord injury: is monitoring from the injury site the future? Crit Care. 2016;20(1):308.

    Article  PubMed  PubMed Central  Google Scholar 

  9. Werndle MC, Saadoun S, Phang I, Czosnyka M, Varsos GV, Czosnyka ZH, et al. Monitoring of spinal cord perfusion pressure in acute spinal cord injury: initial findings of the injured spinal cord pressure evaluation study*. Critcal Care Med. 2014;42(3):646–55.

    Article  Google Scholar 

  10. Leonard AV, Thornton E, Vink R. The relative contribution of edema and hemorrhage to raised intrathecal pressure after traumatic spinal cord injury. J Neurotrauma. 2015;32(6):397–402.

    Article  PubMed  Google Scholar 

  11. Kwon BK, Curt A, Belanger LM, Bernardo A, Chan D, Markez JA, et al. Intrathecal pressure monitoring and cerebrospinal fluid drainage in acute spinal cord injury: a prospective randomized trial: clinical article. J Neurosurgery: Spine SPI. 2009;10(3):181–93.

    Google Scholar 

  12. Saadoun S, Chen S, Papadopoulos MC. Intraspinal pressure and spinal cord perfusion pressure predict neurological outcome after traumatic spinal cord injury. J Neurol Neurosurg Psychiatry. 2017;88(5):452–3.

    Article  PubMed  Google Scholar 

  13. Phang I, Werndle MC, Saadoun S, Varsos G, Czosnyka M, Zoumprouli A, Papadopoulos MC. Expansion duroplasty improves intraspinal pressure, spinal cord perfusion pressure, and vascular pressure reactivity index in patients with traumatic spinal cord injury: injured spinal cord pressure evaluation study. J Neurotrauma. 2015;32(12):865–74.

    Article  PubMed  PubMed Central  Google Scholar 

  14. Kheram N, Boraschi A, Pfender N, Friedl S, Rasenack M, Fritz B, et al. Cerebrospinal Fluid Pressure Dynamics as a Bedside Test in traumatic spinal cord Injury to Assess Surgical spinal cord decompression: safety, feasibility, and Proof-of-Concept. Neurorehabilit Neural Repair. 2023;37(4):171–82.

    Article  Google Scholar 

  15. Fehlings MG, Martin AR, Tetreault LA, Aarabi B, Anderson P, Arnold PM, et al. A clinical practice Guideline for the management of patients with Acute spinal cord Injury: recommendations on the role of Baseline Magnetic Resonance Imaging in clinical decision making and Outcome Prediction. Global Spine J. 2017;7(3 Suppl):s221–30.

    Article  Google Scholar 

  16. Henry-Feugeas MC, Idy-Peretti I, Blanchet B, Hassine D, Zannoli G, Schouman-Claeys E. Temporal and spatial assessment of normal cerebrospinal fluid dynamics with MR imaging. Magn Reson Imaging. 1993;11(8):1107–18.

    Article  PubMed  CAS  Google Scholar 

  17. Enzmann D, Pelc N. Normal flow patterns of intracranial and spinal cerebrospinal fluid defined with phase-contrast cine MR imaging. Radiology. 1991;178:467–74.

    Article  PubMed  CAS  Google Scholar 

  18. Greitz D. Cerebrospinal fluid circulation and associated intracranial dynamics. A radiologic investigation using MR imaging and radionuclide cisternography. Acta Radiol Supplements. 1993;386:1–23.

    CAS  Google Scholar 

  19. Alperin N, Vikingstad EM, Gomez-Anson B, Levin DN. Hemodynamically independent analysis of cerebrospinal fluid and brain motion observed with dynamic phase contrast MRI. Magn Reson Med. 1996;35(5):741–54.

    Article  PubMed  CAS  Google Scholar 

  20. Dreha-Kulaczewski S, Konopka M, Joseph AA, Kollmeier J, Merboldt K-D, Ludwig H-C, et al. Respiration and the watershed of spinal CSF flow in humans. Sci Rep. 2018;8(1):5594.

    Article  PubMed  PubMed Central  Google Scholar 

  21. Dreha-Kulaczewski S, Joseph AA, Merboldt K-D, Ludwig H-C, Gärtner J, Frahm J. Identification of the Upward Movement of Human CSF in Vivo and its relation to the brain venous system. J Neuroscience: Official J Soc Neurosci. 2017;37(9):2395–402.

    Article  CAS  Google Scholar 

  22. Lloyd RA, Butler JE, Gandevia SC, Ball IK, Toson B, Stoodley MA, Bilston LE. Respiratory cerebrospinal fluid flow is driven by the thoracic and lumbar spinal pressures. J Physiol. 2020;598(24):5789–805.

    Article  PubMed  CAS  Google Scholar 

  23. Alperin N, Burman R, Lee SH. Role of the spinal canal compliance in regulating posture-related cerebrospinal fluid hydrodynamics in humans. J Magn Reson Imaging. 2021;54(1):206–14.

    Article  PubMed  Google Scholar 

  24. Muccio M, Chu D, Minkoff L, Kulkarni N, Damadian B, Damadian RV, Ge Y. Upright versus supine MRI: effects of body position on craniocervical CSF flow. Fluids Barriers CNS. 2021;18(1):61.

    Article  PubMed  PubMed Central  Google Scholar 

  25. Bert RJ, Settipalle N, Tiwana E, Muddasani D, Nath R, Wellman B, et al. The relationships among spinal CSF flows, spinal cord geometry, and vascular correlations: evidence of intrathecal sources and sinks. Am J Physiol Regul Integr Comp Physiol. 2019;317(3):R470–84.

    Article  PubMed  CAS  Google Scholar 

  26. Henry–Feugeas M-C, Idy–Peretti I, Baledent O, Poncelet–Didon A, Zannoli G, Bittoun J, Schouman–Claeys E. Origin of subarachnoid cerebrospinal fluid pulsations: a phase-contrast MR analysis. Magn Reson Imaging. 2000;18(4):387–95.

    Article  PubMed  Google Scholar 

  27. Aktas G, Kollmeier JM, Joseph AA, Merboldt K-D, Ludwig H-C, Gärtner J, et al. Spinal CSF flow in response to forced thoracic and abdominal respiration. Fluids Barriers CNS. 2019;16(1):10.

    Article  PubMed  PubMed Central  Google Scholar 

  28. Baledent O, Henry-Feugeas M-C, Idy-Peretti I. Cerebrospinal Fluid Dynamics and Relation with Blood Flow: a magnetic resonance study with Semiautomated Cerebrospinal Fluid Segmentation. Invest Radiol. 2001;36(7):368–77.

    Article  PubMed  CAS  Google Scholar 

  29. Khani M, Lawrence BJ, Sass LR, Gibbs CP, Pluid JJ, Oshinski JN, et al. Characterization of intrathecal cerebrospinal fluid geometry and dynamics in cynomolgus monkeys (macaca fascicularis) by magnetic resonance imaging. PLoS ONE. 2019;14(2):e0212239.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  30. Christen MA, Schweizer-Gorgas D, Richter H, Joerger FB, Dennler M. Quantification of cerebrospinal fluid flow in dogs by cardiac-gated phase-contrast magnetic resonance imaging. J Vet Intern Med. 2021;35(1):333–40.

    Article  PubMed  Google Scholar 

  31. Cho H, Kim Y, Hong S, Choi H. Cerebrospinal fluid flow in normal beagle dogs analyzed using magnetic resonance imaging. J Vet Sci. 2021;22(1):e2–e.

    Article  PubMed  Google Scholar 

  32. Bessen MA, Gayen CD, Quarrington RD, Walls AC, Leonard AV, Kurtcuoglu V, Jones CF. Characterising spinal cerebrospinal fluid flow in the pig with phase-contrast magnetic resonance imaging. Fluids Barriers CNS. 2023;20(1):5.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  33. Liu S, Bilston LE, Flores Rodriguez N, Wright C, McMullan S, Lloyd R, et al. Changes in intrathoracic pressure, not arterial pulsations, exert the greatest effect on tracer influx in the spinal cord. Fluids Barriers CNS. 2022;19(1):14.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  34. Alperin N, Lee SH, Sivaramakrishnan A, Hushek SG. Quantifying the effect of posture on intracranial physiology in humans by MRI flow studies. J Magn Reson Imaging. 2005;22(5):591–6.

    Article  PubMed  Google Scholar 

  35. Nakamura K, Urayama K, Hoshino Y. Lumbar cerebrospinal fluid pulse wave rising from pulsations of both the spinal cord and the brain in humans. Spinal Cord. 1997;35(11):735–9.

    Article  PubMed  CAS  Google Scholar 

  36. Tawfik AM, Elsorogy L, Abdelghaffar R, Naby AA, Elmenshawi I, Phase-Contrast. MRI CSF Flow measurements for the diagnosis of normal-pressure Hydrocephalus: Observer Agreement of Velocity Versus volume parameters. Am J Roentgenol. 2017;208(4):838–43.

    Article  Google Scholar 

  37. Khani M, Burla GKR, Sass LR, Arters ON, Xing T, Wu H, Martin BA. Human in silico trials for parametric computational fluid dynamics investigation of cerebrospinal fluid drug delivery: impact of injection location, injection protocol, and physiology. Fluids Barriers CNS. 2022;19(1):8.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  38. Naseri Kouzehgarani G, Feldsien T, Engelhard HH, Mirakhur KK, Phipps C, Nimmrich V, et al. Harnessing cerebrospinal fluid circulation for drug delivery to brain tissues. Adv Drug Deliv Rev. 2021;173:20–59.

    Article  PubMed  CAS  Google Scholar 

  39. Bradley WG Jr., Scalzo D, Queralt J, Nitz WN, Atkinson DJ, Wong P. Normal-pressure hydrocephalus: evaluation with cerebrospinal fluid flow measurements at MR imaging. Radiology. 1996;198(2):523–9.

    Article  PubMed  Google Scholar 

  40. Sharma AK, Gaikwad S, Gupta V, Garg A, Mishra NK. Measurement of peak CSF flow velocity at cerebral aqueduct, before and after lumbar CSF drainage, by use of phase-contrast MRI: utility in the management of idiopathic normal pressure hydrocephalus. Clin Neurol Neurosurg. 2008;110(4):363–8.

    Article  PubMed  Google Scholar 

  41. Miyati T, Mase M, Kasai H, Hara M, Yamada K, Shibamoto Y, et al. Noninvasive MRI assessment of intracranial compliance in idiopathic normal pressure hydrocephalus. J Magn Reson Imaging. 2007;26(2):274–8.

    Article  PubMed  Google Scholar 

  42. Haughton VM, Korosec FR, Medow JE, Dolar MT, Iskandar BJ. Peak systolic and diastolic CSF velocity in the foramen magnum in adult patients with Chiari I malformations and in normal control participants. AJNR Am J Neuroradiol. 2003;24(2):169–76.

    PubMed  PubMed Central  Google Scholar 

  43. Bunck AC, Kroeger JR, Juettner A, Brentrup A, Fiedler B, Crelier GR, et al. Magnetic resonance 4D flow analysis of cerebrospinal fluid dynamics in Chiari I malformation with and without syringomyelia. Eur Radiol. 2012;22(9):1860–70.

    Article  PubMed  Google Scholar 

  44. Bhadelia RA, Bogdan AR, Wolpert SM, Lev S, Appignani BA, Heilman CB. Cerebrospinal fluid flow waveforms: analysis in patients with Chiari I malformation by means of gated phase-contrast MR imaging velocity measurements. Radiology. 1995;196(1):195–202.

    Article  PubMed  CAS  Google Scholar 

  45. Clarke EC, Stoodley MA, Bilston LE. Changes in temporal flow characteristics of CSF in Chiari malformation type I with and without syringomyelia: implications for theory of syrinx development: clinical article. J Neurosurg JNS. 2013;118(5):1135–40.

    Article  Google Scholar 

  46. Armonda RA, Citrin CM, Foley KT, Ellenbogen RG. Quantitative cine-mode magnetic resonance imaging of Chiari I malformations: an analysis of cerebrospinal fluid dynamics. Neurosurgery. 1994;35(2).

  47. Sivaramakrishnan A, Alperin N, Surapaneni S, Lichtor T. Evaluating the effect of decompression surgery on Cerebrospinal Fluid Flow and Intracranial Compliance in patients with Chiari Malformation with magnetic resonance imaging Flow studies. Neurosurgery. 2004;55(6):1344–51.

    Article  PubMed  Google Scholar 

  48. Bae YJ, Lee JW, Lee E, Yeom JS, Kim K-J, Kang HS. Cervical compressive myelopathy: flow analysis of cerebrospinal fluid using phase-contrast magnetic resonance imaging. Eur Spine J. 2017;26(1):40–8.

    Article  PubMed  Google Scholar 

  49. Shibuya R, Yonenobu K, Koizumi T, Kato Y, Mitta M, Yoshikawa H. Pulsatile Cerebrospinal Fluid Flow Measurement using phase-contrast magnetic resonance imaging in patients with cervical myelopathy. Spine. 2002;27(10):1087–93.

    Article  PubMed  Google Scholar 

  50. Tominaga T, Watabe N, Takahashi T, Shimizu H, Yoshimoto T. Quantitative Assessment of Surgical decompression of the cervical spine with Cine Phase Contrast Magnetic Resonance Imaging. Neurosurgery. 2002;50(4):791–6.

    Article  PubMed  Google Scholar 

  51. Kim H-J, Kim H, Kim Y-T, Sohn C-H, Kim K, Kim D-J. Cerebrospinal fluid dynamics correlate with neurogenic claudication in lumbar spinal stenosis. PLoS ONE. 2021;16(5):e0250742.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  52. Yeo J, Cheng S, Hemley S, Lee BB, Stoodley M, Bilston L. Characteristics of CSF velocity-time profile in posttraumatic syringomyelia. Am J Neuroradiol. 2017;38(9):1839–44.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  53. Kim SY, Shin MJ, Chang JH, Lee CH, Shin YI, Shin YB, Ko HY. Correlation of diffusion tensor imaging and phase-contrast MR with clinical parameters of cervical spinal cord injuries. Spinal Cord. 2015;53(8):608–14.

    Article  PubMed  Google Scholar 

  54. National Health and Medical Research Council. Australian code for the care and use of animals for scientific purposes. 8th Edition ed. Canberra: National Health and Medical Research Council 2013.

  55. Gayen C, Bessen MA, Dorrian R, Quarrington RD, Mulaibrahimovic A, O’Hare Doig R et al. A survival model of thoracic contusion spinal cord injury in the domestic pig. J Neurotrauma. 2022.

  56. Lee JHT, Jones CF, Okon EB, Anderson L, Tigchelaar S, Kooner P, et al. A Novel Porcine Model of traumatic thoracic spinal cord Injury. J Neurotrauma. 2013;30(3):142–59.

    Article  PubMed  Google Scholar 

  57. Heiberg E, Sjögren J, Ugander M, Carlsson M, Engblom H, Arheden H. Design and validation of segment–freely available software for cardiovascular image analysis. BMC Med Imaging. 2010;10:1.

    Article  PubMed  PubMed Central  Google Scholar 

  58. Shin HK, Jeong HJ, Kim E, Park JH, Park SJ, Cho Y. Should we check the routine postoperative MRI for hematoma in spinal decompression surgery? Clin Orthop Surg. 2017;9(2):184–9.

    Article  PubMed  PubMed Central  Google Scholar 

  59. Sass LR, Khani M, Romm J, Schmid Daners M, McCain K, Freeman T, et al. Non-invasive MRI quantification of cerebrospinal fluid dynamics in amyotrophic lateral sclerosis patients. Fluids Barriers CNS. 2020;17(1):4.

    Article  PubMed  PubMed Central  Google Scholar 

  60. Linge SO, Mardal KA, Haughton V, Helgeland A, CSF Flow Dynamics in the normal and the Chiari I Subarachnoid Space during Rest and Exertion. Am J Neuroradiol. 2013;34(1):41–5.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  61. Habib CA, Utriainen D, Peduzzi-Nelson J, Dawe E, Mattei J, Latif Z, et al. MR imaging of the yucatan pig head and neck vasculature. J Magn Reson Imaging. 2013;38(3):641–9.

    Article  PubMed  Google Scholar 

  62. Theodore N, Martirosyan N, Hersh AM, Ehresman J, Ahmed AK, Danielson J, et al. Cerebrospinal fluid drainage in patients with Acute spinal cord Injury: a Multi-center Randomized Controlled Trial. World Neurosurg. 2023;177:e472–9.

    Article  Google Scholar 

  63. Neuropathological. Motor impairments after incomplete cervical spinal cord Injury in pigs. J Neurotrauma. 2021;38(21):2956–77.

    Article  Google Scholar 

  64. Züchner M, Escalona MJ, Teige LH, Balafas E, Zhang L, Kostomitsopoulos N, Boulland JL. How to generate graded spinal cord injuries in swine - tools and procedures. Dis Models Mech. 2021;14(8).

  65. Huang L, Lin X, Tang Y, Yang R, Li A, Jc Y, et al. Quantitative assessment of spinal cord perfusion by using contrast-enhanced ultrasound in a porcine model with acute spinal cord contusion. Spinal Cord. 2013;51(3):196–201.

    Article  PubMed  CAS  Google Scholar 

  66. Wilson C, Linczer J, Newman S, Weyhenmeyer J, Roper A, Miller J, et al. Intrathecal Baclofen and Opioid Therapy: cerebrospinal fluid leak and infection incidence, risk factors, and outcomes. World Neurosurg. 2023;171:e456–63.

    Article  PubMed  Google Scholar 

  67. Dobrocky T, Winklehner A, Breiding PS, Grunder L, Peschi G, Häni L, et al. Spine MRI in spontaneous intracranial hypotension for CSF leak detection: Nonsuperiority of Intrathecal Gadolinium to heavily T2-Weighted Fat-saturated sequences. AJNR Am J Neuroradiol. 2020;41(7):1309–15.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  68. Khani M, Fu AQ, Pluid J, Gibbs CP, Oshinski JN, Xing T, et al. Intrathecal catheter implantation decreases cerebrospinal fluid dynamics in cynomolgus monkeys. PLoS ONE. 2021;15(12):e0244090.

    Article  Google Scholar 

  69. Bunck AC, Kroger JR, Juttner A, Brentrup A, Fiedler B, Schaarschmidt F, et al. Magnetic resonance 4D flow characteristics of cerebrospinal fluid at the craniocervical junction and the cervical spinal canal. Eurpean Radiol. 2011;21(8):1788–96.

    Article  Google Scholar 

  70. Yiallourou TI, Kröger JR, Stergiopulos N, Maintz D, Martin BA, Bunck AC. Comparison of 4D phase-contrast MRI flow measurements to computational fluid dynamics simulations of cerebrospinal fluid motion in the cervical spine. PLoS ONE. 2012;7(12):e52284.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  71. Dorniak K, Heiberg E, Hellmann M, Rawicz-Zegrzda D, Wesierska M, Galaska R, et al. Required temporal resolution for accurate thoracic aortic pulse wave velocity measurements by phase-contrast magnetic resonance imaging and comparison with clinical standard applanation tonometry. BMC Cardiovasc Disord. 2016;16(1):110.

    Article  PubMed  PubMed Central  Google Scholar 

  72. Sonnabend K, Brinker G, Maintz D, Bunck AC, Weiss K. Cerebrospinal fluid pulse wave velocity measurements: in vitro and in vivo evaluation of a novel multiband cine phase-contrast MRI sequence. Magn Reson Med. 2021;85(1):197–208.

    Article  PubMed  CAS  Google Scholar 

Download references

Acknowledgements

The authors acknowledge the technical assistance of Dr. Chris Christou and the staff of the South Australian Health and Medical Research Institute (SAHMRI) Preclinical Imaging and Research Laboratories (PIRL). The authors acknowledge the facilities and scientific and technical assistance of the National Imaging Facility, a National Collaborative Research Infrastructure (NCRIS) capability, at the Large Animal Research and Imaging Facility (LARIF), SAHMRI. The authors thank Sandra Jenkner, Nikoo Soltan, and Keziah Skein for their assistance with animal care, and Jana Bednaz for statistical support.

Funding

This study was supported by a Basic Research Grant Award from the North American Spine Society (CFJ, AVL). MAB and RMD were supported by an Australian Government Research Training Program Scholarship (RTPS). CDG was supported by a The Hospital Research Foundation PhD Scholarship. ROD was supported by the Neil Sachse Centre for Spinal Cord Research (on behalf of it’s donors), the Lifetime Support Authority (GA00044 and GA00093) and the National Imaging Facility. RDQ was supported by an Australian Research Council Discovery Project during this study (DP190101209; CFJ). VK received support from the Swiss National Science Foundation (205321_182683). CFJ was partially supported by a National Health and Medical Research Council (Australia) Early Career Fellowship during this study (APP1072387).

Author information

Authors and Affiliations

  1. Adelaide Spinal Research Group & Centre for Orthopaedics and Trauma Research, Faculty of Health and Medical Sciences, The University of Adelaide, Level 7, Adelaide Health and Medical Sciences Building, North Terrace, Adelaide, SA, 5005, Australia

    Madeleine Amy Bessen, Christine Diana Gayen, Ryan David Quarrington, Adnan Mulaibrahimovic & Claire Frances Jones

  2. Adelaide Medical School, The University of Adelaide, Level 7, Adelaide Health and Medical Sciences Building, The University of Adelaide, North Terrace, Adelaide, SA, 5005, Australia

    Madeleine Amy Bessen, Ryan L. O’Hare Doig, Ryan David Quarrington & Adnan Mulaibrahimovic

  3. Neil Sachse Centre for Spinal Cord Research, Lifelong Health Theme, South Australian Health and Medical Research Institute, Level 7, South Australian Health and Medical Research Institute, North Terrace, Adelaide, SA, 5005, Australia

    Ryan L. O’Hare Doig

  4. Translational Neuropathology Laboratory, School of Biomedicine, The University of Adelaide, Level 2, Helen Mayo North Building, Frome Rd, Adelaide, SA, 5005, Australia

    Christine Diana Gayen, Ryan Michael Dorrian & Anna Victoria Leonard

  5. Institute of Physiology, University of Zurich, Winterthurerstrasse 190, Zurich, 8057, Switzerland

    Vartan Kurtcuoglu

  6. Zurich Center of Integrative Human Physiology, University of Zurich, Winterthurerstrasse 190, Zurich, 8057, Switzerland

    Vartan Kurtcuoglu

  7. Clinical and Research Imaging Centre, South Australian Health and Medical Research Institute, National Imaging Facility, Northern Pod, SAHMRI, North Terrace, Adelaide, SA, 5005, Australia

    Angela Catherine Walls

  8. School of Electrical and Mechanical Engineering, The University of Adelaide, Level 7, Adelaide Health and Medical Sciences Building, Adelaide, SA, 5005, Australia

    Claire Frances Jones

  9. Neuroscience Center Zurich, University of Zurich, Winterthurerstrasse 190, Zurich, 8057, Switzerland

    Vartan Kurtcuoglu

Authors
  1. Madeleine Amy Bessen
    View author publications

    You can also search for this author inPubMed Google Scholar

  2. Christine Diana Gayen
    View author publications

    You can also search for this author inPubMed Google Scholar

  3. Ryan L. O’Hare Doig
    View author publications

    You can also search for this author inPubMed Google Scholar

  4. Ryan Michael Dorrian
    View author publications

    You can also search for this author inPubMed Google Scholar

  5. Ryan David Quarrington
    View author publications

    You can also search for this author inPubMed Google Scholar

  6. Adnan Mulaibrahimovic
    View author publications

    You can also search for this author inPubMed Google Scholar

  7. Vartan Kurtcuoglu
    View author publications

    You can also search for this author inPubMed Google Scholar

  8. Angela Catherine Walls
    View author publications

    You can also search for this author inPubMed Google Scholar

  9. Anna Victoria Leonard
    View author publications

    You can also search for this author inPubMed Google Scholar

  10. Claire Frances Jones
    View author publications

    You can also search for this author inPubMed Google Scholar

Contributions

Conceptualization, MAB and CFJ. Methodology, MAB, CDG, RD, VK, AW and CFJ. Validation, MAB, RD, VK, AW and CFJ. Formal Analysis, MAB, CDG, RDQ, and CFJ. Investigation, MAB, CDG, RD, RMD, AM, and CFJ. Resources, MAB and CFJ. Data Curation, MAB, CDG, RD, RMD, RDQ, and CFJ. Writing – Original Draft, MAB and CFJ. Writing – Review & Editing, MAB, CDG, RD, RMD, AM, VK, AW, AVL and CFJ. Visualization, MAB and CFJ. Supervision, AVL and CFJ. Project Administration, MAB, CDG, RD, AVL, and CFJ. Funding Acquisition, AVL and CFJ.

Corresponding author

Correspondence to Claire Frances Jones.

Ethics declarations

Ethics approval and consent to participate

This project was approved by the South Australian Health and Medical Research Institute Animal Ethics Committee (SAM 243 and SAM-22-031) and conducted in accordance with the Australian National Health and Medical Research Council Code of Care and Use of Animals for Scientific Purposes [54].

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Additional Files

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Bessen, M.A., Gayen, C.D., Doig, R.L.O. et al. Cerebrospinal fluid dynamics and subarachnoid space occlusion following traumatic spinal cord injury in the pig: an investigation using magnetic resonance imaging. Fluids Barriers CNS 22, 6 (2025). https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12987-024-00595-9

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12987-024-00595-9

Keywords

Fluids and Barriers of the CNS

ISSN: 2045-8118