Effects of aortic occlusion on regional spinal cord blood flow and somatosensory evoked potentials in sheep: Kaplan BJ, Friedman WA, Gravenstein N, et al. Neurosurgery 1987;21:668-75

Abstract

The precise mechanism underlying spinal cord injury resulting from aortic occlusion is not well understood; such injury remains a devastating complication associated with thoracoabdominal aortic repair. This elegant study documents the changes in somatosensory evoked potentials (SEPs) resulting from aortic clamping and correlates these changes with alterations in regional spinal cord blood flow (SCBF). Eleven sheep had left thoracotomy with elaborately monitored general anesthesia. The aorta was cross-clamped for 30 minutes just distal to the subclavian artery in nine animals; two additional animals served as nonclamped controls. After tibial nerve stimulation, SEPs were continuously recorded from nerve action potential at the hip, lumbar, cervical, and cortical montages. SCBF was determined at five specific points during the experiment: control before clamping, immediately after clamping, on initial change in SEP, 30 minutes after clamping, and 30 minutes after reperfusion. At each point, radioactive microspheres labeled with one of five different isotopes corresponding to the appropriate interval were injected into the left atrium. The two control animals received the five injections, each 15 minutes apart, without aortic clamping. After the final injection, animals were killed; the spinal cord was removed and divided into lumbar, thoracic, and cervical sections and the dorsal columns, lateral columns, and cortical gray matter separated. Regional SCBF was then calculated from the ratio of the radioactivity in the specimen relative to that of a reference organ and expressed as milliliters per 100 gm tissue per minute. SCBF was not affected by repetitive microsphere injections as demonstrated in the two control animals. In contrast, the nine experimental animals exhibited two distinct patterns of alterations with aortic clamping. Most animals (seven of nine) demonstrated total loss of SEPs within 15 minutes of aortic occlusion. Thirty minutes after reperfusion, evoked responses of diminished amplitude returned in only three animals and not at all in the remaining four. These seven animals also demonstrated profound reduction in thoracolumbar SCBF to both white and gray matter and subsequent reactive hyperemia after reperfusion. In two experimental animals SEPs were preserved, and the initial reduction in amplitude observed within 5 minutes of clamping resolved during the 30-minute period of ischemia. These two animals also developed a marked reduction in SCBF, which returned to 50% of baseline levels 30 minutes after clamping, corresponding to the recovery of SEP. Therefore spinal cord ischemia is associated with profound alternations in SEPs and the magnitude and duration of diminished evoked responses seem to correlate with similar reductions in SCBF. The derived data suggest that preservation of SEP is associated with maintenance of SCBF above 20 ml/100 gm tissue/min; SCBF less than 10 ml/100 gm tissue/min results in obliteration of SEP. The more peripheral the nerve tissue, the more resistance to ischemia. The data support the validity of SEPs in reflecting alterations in SCBF and underscore the value of this measure in further investigations of ischemic spinal cord injury.