Monitoring Spinal Cord Tissue Oxygen in Patients With Acute, Severe Traumatic Spinal Cord Injuries

OBJECTIVES: To determine the feasibility of monitoring tissue oxygen tension from the injury site (p(sct)o(2)) in patients with acute, severe traumatic spinal cord injuries. DESIGN: We inserted at the injury site a pressure probe, a microdialysis catheter, and an oxygen electrode to monitor for up to a week intraspinal pressure (ISP), spinal cord perfusion pressure (SCPP), tissue glucose, lactate/pyruvate ratio (LPR), and p(sct)o(2). We analyzed 2,213 hours of such data. Follow-up was 6-28 months postinjury. SETTING: Single-center neurosurgical and neurocritical care units. SUBJECTS: Twenty-six patients with traumatic spinal cord injuries, American spinal injury association Impairment Scale A-C. Probes were inserted within 72 hours of injury. INTERVENTIONS: Insertion of subarachnoid oxygen electrode (Licox; Integra LifeSciences, Sophia-Antipolis, France), pressure probe, and microdialysis catheter. MEASUREMENTS AND MAIN RESULTS: p(sct)o(2) was significantly influenced by ISP (p(sct)o(2) 26.7 +/- 0.3 mm Hg at ISP > 10 mmHg vs p(sct)o(2) 22.7 +/- 0.8 mm Hg at ISP <= 10 mm Hg), SCPP (p(sct)o(2) 26.8 +/- 0.3 mm Hg at SCPP < 90 mm Hg vs p(sct)o(2) 32.1 +/- 0.7 mm Hg at SCPP >= 90 mm Hg), tissue glucose (p(sct)o(2) 26.8 +/- 0.4 mm Hg at glucose < 6 mM vs 32.9 +/- 0.5 mm Hg at glucose >= 6 mM), tissue LPR (p(sct)o(2) 25.3 +/- 0.4 mm Hg at LPR > 30 vs p(sct)o(2) 31.3 +/- 0.3 mm Hg at LPR <= 30), and fever (p(sct)o(2) 28.8 +/- 0.5 mm Hg at cord temperature 37-38 degrees C vs p(sct)o(2) 28.7 +/- 0.8 mm Hg at cord temperature >= 39 degrees C). Tissue hypoxia also occurred independent of these factors. Increasing the Fio(2) by 0.48 increases p(sct)o(2) by 71.8% above baseline within 8.4 minutes. In patients with motor-incomplete injuries, fluctuations in p(sct)o(2) correlated with fluctuations in limb motor score. The injured cord spent 11% (39%) hours at p(sct)o(2) less than 5 mm Hg (< 20 mm Hg) in patients with motor-complete outcomes, compared with 1% (30%) hours at p(sct)o(2) less than 5 mm Hg (< 20 mm Hg) in patients with motor-incomplete outcomes. Complications were cerebrospinal fluid leak (5/26) and wound infection (1/26). CONCLUSIONS: This study lays the foundation for measuring and altering spinal cord oxygen at the injury site. Future studies are required to investigate whether this is an effective new therapy.