Acute Spinal Cord Injury: Correlations and Causal Relations Between Intraspinal Pressure, Spinal Cord Perfusion Pressure, Lactate-to-Pyruvate Ratio, and Limb Power

Abstract

Background/objective: We have recently developed monitoring from the injury site in patients with acute, severe traumatic spinal cord injuries to facilitate their management in the intensive care unit. This is analogous to monitoring from the brain in patients with traumatic brain injuries. This study aims to determine whether, after traumatic spinal cord injury, fluctuations in the monitored physiological, and metabolic parameters at the injury site are causally linked to changes in limb power.

Methods: This is an observational study of a cohort of adult patients with motor-incomplete spinal cord injuries, i.e., grade C American spinal injuries association Impairment Scale. A pressure probe and a microdialysis catheter were placed intradurally at the injury site. For up to a week after surgery, we monitored limb power, intraspinal pressure, spinal cord perfusion pressure, and tissue lactate-to-pyruvate ratio. We established correlations between these variables and performed Granger causality analysis.

Results: Nineteen patients, aged 22-70 years, were recruited. Motor score versus intraspinal pressure had exponential decay relation (intraspinal pressure rise to 20 mmHg was associated with drop of 11 motor points, but little drop in motor points as intraspinal pressure rose further, R2 = 0.98). Motor score versus spinal cord perfusion pressure (up to 110 mmHg) had linear relation (1.4 motor point rise/10 mmHg rise in spinal cord perfusion pressure, R2 = 0.96). Motor score versus lactate-to-pyruvate ratio (greater than 20) also had linear relation (0.8 motor score drop/10-point rise in lactate-to-pyruvate ratio, R2 = 0.92). Increased intraspinal pressure Granger-caused increase in lactate-to-pyruvate ratio, decrease in spinal cord perfusion, and decrease in motor score. Increased spinal cord perfusion Granger-caused decrease in lactate-to-pyruvate ratio and increase in motor score. Increased lactate-to-pyruvate ratio Granger-caused increase in intraspinal pressure, decrease in spinal cord perfusion, and decrease in motor score. Causality analysis also revealed multiple vicious cycles that amplify insults to the cord thus exacerbating cord damage.

Conclusion: Monitoring intraspinal pressure, spinal cord perfusion pressure, lactate-to-pyruvate ratio, and intervening to normalize these parameters are likely to improve limb power.

Keywords: Blood pressure; Intraspinal pressure; LPR; Management; Microdialysis; Monitoring; Spinal cord injury.

Conflict of interest statement

All authors have no conflicts of interest.

Figures

Fig. 1

Monitoring setup. a Intraspinal pressure probe and microdialysis catheter inserted intradurally to monitor from injured cord. b Preoperative sagittal T2 MRI of patient no. 52 with spinal cord injury at C6. c. Postoperative axial CT with intraspinal pressure probe and microdialysis catheter in situ. d Multi-modality monitoring of motor score (yellow), intraspinal pressure (blue), mean arterial pressure (green), spinal cord perfusion pressure (red) as well as tissue glucose (pink), lactate (orange), pyruvate (cyan), and lactate-to-pyruvate ratio (purple) (Color figure online)

Fig. 2

Standardized motor score correlates with injury site physiology. Standardized motor score versus a intraspinal pressure, b spinal cord perfusion pressure and c mean arterial pressure. Mean ± standard error. Trends (dotted gray line) modeled as exponential decay (intraspinal pressure, R2 = 0.98, P < 0.0005), linear (spinal cord perfusion pressure in the range < 50 to 110 mmHg, R2 = 0.96, P < 0.0005), and bounded exponential (Mean arterial pressure, R2 = 0.73, P < 0.05)

Fig. 3

Standardized motor score correlates with injury site metabolism. a Relation between standardized motor score and lactate-to-pyruvate ratio. b Glucose and c Lactate + Pyruvate versus lactate-to-pyruvate ratio. Mean ± standard error. In a, the dotted line is the best fit straight line for lactate-to-pyruvate ratio in the range 20 to > 70 mmHg, R2 = 0.92, P < 0.005

Fig. 4

Intraspinal pressure correlates with injury site metabolism. a Tissue glucose, b tissue lactate-to-pyruvate ratio and c tissue pyruvate (black) + lactate (gray), versus Intraspinal pressure. Mean ± standard error. In a, best fit straight line R2 = 0.57, P < 0.05

Fig. 5

Spinal cord perfusion pressure correlates with injury site metabolism. a Tissue glucose. b Tissue lactate-to-pyruvate ratio and c. tissue pyruvate (black) + lactate (gray), versus ISP. Mean ± standard error. In a, best fit straight line R2 = 0.83, P < 0.005

Fig. 6

Granger causality relations. Each arrow indicates the direction of information flow, i.e., causal influence, with corresponding F and P values. ‘ + ’ or ‘–’ indicate the correlation between the variables. Causality arrows are shown if P < 0.05. Analysis is shown for lag = 1, but also holds for lag = 2 or 3. For details, see supplement