Effect of Sevoflurane Concentration, Blood Pressure, Arterial Carbon Dioxide Tension, Temperature, and Stimulation on Spinal Cord Blood Flow in Patients Undergoing Spinal Surgery: An Exploratory Study

To the Editor: WE used laser speckle contrast imaging during spinal surgery to investigate how spinal cord blood flow responds to intraoperative factors relevant to anesthesia. Laser speckle contrast imaging is a noninvasive optical technique that generates high-resolution, full-field blood flow maps.1 The principle of laser speckle contrast imaging is illustrated in Supplemental Digital Content figure S1, A and B (https://links.lww.com/ALN/E358). We investigated the effect of anesthetic agent, arterial carbon dioxide tension (PaCO2), mean arterial pressure (MAP), and temperature on spinal cord blood flow. We also sought evidence of autoregulation and neurovascular coupling in the human spinal cord. This observational study followed the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) guidelines (see supplemental content https://links.lww.com/ALN/E364). Patients undergoing elective laminectomies were included, as well as patients recruited into Duroplasty for Injured cervical Spinal Cord with Uncontrolled Swelling (DISCUS; ClinicalTrials.gov NCT04936620) and Injured Spinal Cord Pressure Evaluation (ISCoPE; ClinicalTrials.gov NCT02721615), which are clinical studies recruiting patients with acute, severe traumatic spinal cord injuries. After surgical resection/decompression, we exploited various intraoperative protocols used in our department to address our questions. Spinal cord blood flow was assessed using a portable laser speckle contrast imaging machine. For details, see Supplemental Digital Content methods (https://links.lww.com/ALN/E363). This is an exploratory study. Patient cohorts were small, analyses were not predefined, and the aim is signal finding and parameter estimation for future studies rather than confirmation of mechanism. Thirty-five patients who underwent laminectomies, with or without dural opening, were included. We previously showed that an intact dura and cerebrospinal fluid have little influence on spinal cord blood flow measurements with laser speckle contrast imaging.1 Propofol and remifentanil were administered to 24 (68.6%) patients, while 11 (31.4%) patients received sevoflurane. Demographics are shown in Supplemental Digital Content tables S1 to S3 (https://links.lww.com/ALN/E365). Two patients undergoing craniotomy were used as controls to validate the technique, i.e., to investigate whether laser speckle contrast imaging detects increased cerebral blood flow in response to increasing sevoflurane minimum alveolar concentration (MAC) or increasing PaCO2. No patient experienced intraoperative awareness. The Supplemental Digital Content flowchart (https://links.lww.com/ALN/E362) shows patient numbers in each experimental protocol. In the two craniotomy patients, increasing sevoflurane MAC caused progressive increase in cerebral blood flow (Supplemental Digital Content fig. S1E, https://links.lww.com/ALN/E358), in agreement with earlier studies.2,3 In contrast, figure 1 shows that increasing sevoflurane concentration caused no significant change in spinal cord blood flow. During these assessments, MAP and PaCO2 were kept constant. Our findings suggest that human spinal cord vessels are less responsive to sevoflurane than cerebral vessels.Fig. 1.: Effect of sevoflurane minimum alveolar concentration (MAC), PaCO 2, and irrigation fluid temperature on spinal cord blood flow (SCBF). Shown are spinal cord vessels: photograph and corresponding laser speckle contrast imaging at sevoflurane (Sevo) MAC 0.7 versus 1.5, photograph and corresponding laser speckle contrast imaging at low (↓) versus high (↑) PaCO 2, and photograph and corresponding laser speckle contrast imaging at cold (↓) versus warm (↑) irrigation fluid temperature (temp). Also shown is spinal cord blood flow scale (arbitrary units [a.u.]): percent change in vessel blood flow (blue), tissue blood flow (black), and vessel diameter (dia) (purple) at sevoflurane MAC 1.0 or 1.5 versus 0.7, low (↓) or normal (→) versus high (↑) PaCO 2, and cold (↓) or room temperature (→) versus warm (↑) irrigation fluid. Mean ± 95% CI; paired Student’s t tests with Bonferroni corrections, *P < 0.05, ##P = 0.001; ns, not significant.In the two craniotomy patients, increasing PaCO2 caused a progressive increase in cerebral blood flow (Supplemental Digital Content fig. S1D, https://links.lww.com/ALN/E358). Figure 1 shows that increasing PaCO2 also causes a significant increase in spinal cord blood flow. This response of spinal cord blood flow to changes in PaCO2 was seen in the patients who received propofol and in those who received sevoflurane (Supplemental Digital Content table S4, https://links.lww.com/ALN/E365). Thus, carbon dioxide is a potent vasodilator in the spinal cord and increases spinal cord blood flow, analogous to its effect on cerebral vasculature.4 During these assessments, MAP and anesthetic concentration were kept constant. Interestingly, increasing sevoflurane MAC rendered the spinal cord blood flow more sensitive to PaCO2; i.e., the same change in PaCO2 produced a larger change in spinal cord blood flow (Supplemental Digital Content fig. S2, https://links.lww.com/ALN/E359). During these assessments, the MAP was kept constant. Heat causes vasodilation, and cold causes vasoconstriction.5Figure 1 shows that increasing the temperature of the irrigating saline caused little or no change in spinal cord blood flow, which suggests that spinal cord temperature during surgery is primarily determined by the core temperature rather than the temperature of the irrigation fluid. During these assessments, MAP, PaCO2, and anesthetic concentration were kept constant. Next, we sought evidence of autoregulation, i.e., whether spinal cord blood flow stays constant despite changes in MAP. A sudden increase in MAP caused a transient increase in spinal cord blood flow in patient 5 with return to baseline spinal cord blood flow (fig. 2) but a persistent increase in spinal cord blood flow in patient 30 recorded from within the injured cord (Supplemental Digital Content fig. S3, https://links.lww.com/ALN/E360), indicating intact and impaired autoregulation, respectively. Supplemental Digital Content table S5 (https://links.lww.com/ALN/E365) shows that, in patients without traumatic spinal cord injuries, autoregulation was generally preserved. In contrast, autoregulation was impaired in many patients with traumatic spinal cord injuries. Thus, the human spinal cord autoregulates, akin to brain autoregulation.6 As with brain autoregulation,7 spinal autoregulation is preserved in patients receiving propofol or sevoflurane. The existence of spinal cord autoregulation and its disruption in traumatic spinal cord injury support the concept of optimum spinal cord perfusion pressure, defined as the perfusion pressure that optimizes autoregulation,8 which can be targeted to rationalize the management of traumatic spinal cord injuries by avoiding cord hypo- and hyperperfusion.Fig. 2.: Normal spinal cord autoregulation. Shown are a spinal cord vessel photograph and corresponding laser speckle contrast imaging with mean arterial pressure (MAP) at each time point shown. Also shown is a plot of mean arterial pressure (black), spinal cord blood flow (SCBF) in an artery (red), and tissue spinal cord blood flow (blue) versus time for patient 5. a.u., arbitrary units.In the brain, increased neuronal activity elicits a localized increase in cerebral blood flow, thus coupling energy demand and energy supply.9 In patients 34 and 35, direct electrical stimulation of one dorsal column was associated with an increase in spinal cord blood flow bilaterally that peaked within a second of stimulation onset and disappeared within a second of cessation of stimulation (Supplemental Digital Content fig. S4, https://links.lww.com/ALN/E361; Supplemental Digital Content video, https://links.lww.com/ALN/E366). This suggests that the spinal cord also exhibits neurovascular coupling. When one side of the cord is activated, commissural interneurons transmit information contralaterally to facilitate left–right upper limb coordination for object manipulation and left–right lower limb coordination for walking; this may account for the bilateral increase in dorsal column spinal cord blood flow in response to unilateral dorsal column stimulation. Our findings directly validate the previously reported functional magnetic resonance imaging findings of bilateral dorsal spinal cord blood flow increase during unilateral vibratory dermatomal stimulation.10 Our study has limitations. First, the number of patients is small. Second, only some patients were studied in each protocol, based on practical constraints, which may introduce selection bias. Third, the patients are heterogeneous. Fourth, laser speckle contrast imaging penetrates to a depth of 2 to 3 mm from the cord surface,1 which monitors only the dorsal columns. Fifth, spinal cord oxygen demand was not measured to determine whether sevoflurane uncouples flow–metabolism matching.7 Sixth, in each patient, autoregulation was studied as response of spinal cord blood flow to a single MAP change. Despite the limitations, our study offers clinically important insight into human spinal cord blood flow physiology. For example, based on our findings, in patients with cord compression, hypocarbia may be detrimental by causing hypoperfusion at the site of compression. In conclusion, our data from this small patient cohort suggest that, in patients receiving sevoflurane or propofol, spinal cord blood flow exhibits autoregulation, carbon dioxide reactivity, and neurovascular coupling. Sevoflurane MAC and the temperature of irrigating saline had no significant effect on spinal cord blood flow. Future studies with larger, protocol-specific cohorts will be essential to confirm these observations and better define their clinical implications. Research Support This work was funded by the Wings for Life Spinal Cord Research Foundation (London, United Kingdom; research grant to Dr. Saadoun and Dr. Papadopoulos), the National Institute of Health Research (NIHR) Efficacy and Mechanism Evaluation Program (EME, grant No. NIHR130048), and the Neurosciences Research Foundation (London, United Kingdom; McKissock fellowship to Dr. Asif). Dr. Asif’s salary was funded by the EME, NIHR130048, and Dr. Boseta is partly funded by the NIHR Clinical Research Network. Competing Interests The authors declare no competing interests. Supplemental Digital Content Supplemental Figure S1. Intraoperative laser speckle contrast imaging, https://links.lww.com/ALN/E358 Supplemental Figure S2. Effect of sevoflurane MAC on spinal cord blood flow CO2 reactivity, https://links.lww.com/ALN/E359 Supplemental Figure S3. Impaired spinal cord autoregulation, https://links.lww.com/ALN/E360 Supplemental Figure S4. Neurovascular coupling, https://links.lww.com/ALN/E361 Flowchart, https://links.lww.com/ALN/E362 Supplemental methods, https://links.lww.com/ALN/E363 STROBE checklist, https://links.lww.com/ALN/E364 Supplemental tables, https://links.lww.com/ALN/E365 Supplemental video, https://links.lww.com/ALN/E366