Magnetic resonance imaging of continuous spinal anesthesia with hyperbaric lidocaine: Root or white matter lesion?

Introduction: Persistent neurologic deficits in patients after continuous spinal anesthesia with hyperbaric lidocaine may have been erroneously characterized as a cauda equina syndrome (1,2). First, clinical reports of symptoms rarely include pain (typical for cauda equina lesions) but most often consist of saddle anesthesia, paraplegia and sphincter dysfunction (typical for conus medullaris lesions). Secondly, spinal cord damage in dogs after intrathecal procaine (3) or lidocaine (4) is confined only to the myelin sheath beneath the pia mater and not nerve roots. In this study we use diffusion-weighted magnetic resonance imaging (DWI) to further characterize hyperbaric lidocaine-induced spinal cord damage. DWI measures diffusion of free protons and is a very sensitive detector of white or gray matter injury (5). Normal spinal cord gray and white matter as well as nerve roots are clearly visualized by DWI because the proton diffusion coefficient (DC) - the critical parameter measured with DWI - within these tissues differs. The effect of hyperbaric 5% lidocaine on the spinal cord diffusion coefficient will be compared with those found in ischémie conditions. Methods: The study was approved by the Duke University Animal Care Institutional Committee. Nine female adult Fisher rats (weighing 150-190g) were anesthetized with isoflurane, intubated and mechanically ventilated. Catheters were inserted into the right external jugular vein and carotid artery. Heart rate, arterial blood pressure and body temperature was monitored continuously. Diffusion-weighted MR images (DWI) were acquired at the level of C6 in five rats before and after ischemia induced by cardiac arrest. In another four rats catheters were first placed intrathecally via the atlanto-occipital membrane. Subsequently, DWI's were acquired at the level of C6 before and during 3 hr continuous intrathecal perfusion with 5% hyperbaric lidocaine. Hemodynamic stability during lidocaine administration was maintained with i.v. hydration, atropine and ephedrine as needed. All imaging was done on a 7 T magnet using a standard spin echo diffusion pulse sequence (5). Results: Ischemia reduced the diffusion coefficient (DC) of white matter by 25%, gray matter by 45% and nerve roots/dorsal root ganglia by 35%. Figure 1 shows that one hour of continuous intrathecal administration of 5% hyperbaric lidocaine reduces the DC of white matter by 15% and that the changes are not as profound as those found during ischemia (results are presented as mean ± SD). Figure 2 shows that in contrast to white matter, DCs of adjacent nerve roots are unaffected by 5 % hyperbaric lidocaine. Discussion: The present study has led to the following conclusion: (1) diffusion-weighted imaging of spinal cord in rats with indwelling intrathecal catheters is possible in vivo, (2) the diffusion coefficient of white matter is reduced by 15% after 1 hr of 5% hyperbaric lidocaine indicating onset of cytotoxic edema, (3) the lidocaine-induced changes in the white matter DC do not reach ischémie levels, (4) adjacent gray matter as well as nerve roots are unaffected. Ongoing work will provide information as to the exact pathophysiological sequence of events which lead to irreversible spinal cord damage during continuous spinal anesthesia with high dose hyperbaric 5% lidocaine. It is our goal that this research will be applicable to local anesthesia toxicity studies and useful as a future neurotoxicity screening tool for new intralhecal drugs designed for both anesthesia and analgesia.

Duke Scholars

Author G. Allan Johnson Radiology