Epidural versus subdural spinal cord cooling: Cerebrospinal fluid temperature and pressure changes

Background. Regional spinal cord cooling can increase the tolerable duratio n for spinal cord ischemia resulting from aortic clamping. We compared the efficacy of epidural and subdural cooling and the effect of the resulting c erebrospinal fluid-pressure (CSF) increases on spinal cord motor neuron fun ction. Methods. In 8 pigs, CSF temperature and pressure were assessed in the subdu ral space at L4, T15, and T7. Saline was infused at 333, 666, and 999 ml/h at four consecutive locations: L4 subdural, L4 epidural, T15 subdural, and T15 epidural. First, the influence of CSF-pressure increases during normoth ermic infusion on transcranial motor evoked potentials (tc-MEPs) was assess ed. Then, hypothermic infusion (4 degrees C) was performed to assess CSF-te mperature changes. Results. During normothermic infusion, baseline CSF pressures increased uni formly from 6 +/- 4 mm Hg to 34 +/- 18, 42 +/- 17, and 50 +/- 18 mm Hg with increasing infusion rates (p < 0.001), and did not differ between epidural or subdural infusion. Tc-MEPs indicated spinal cord ischemia in 6 animals when CSF pressures reached 65 +/- 11 mm Hg. During hypothermic infusion, CS F temperatures decreased from 37 degrees to 35 +/- 1.2 degrees, 31 +/- 2.2 degrees, and 28 +/- 2.8 degrees C, but increasing CSF-temperature gradients were observed between the infusion location and distant segments. Subdural cooling resulted in lower CSF temperatures (p < 0.001), but caused larger CSF-pressure increases (p < 0.001). Conclusions. Subdural and epidural infusion cooling produce localized spina l cord hypothermia in pigs. The concurrent pressure increases, however, are uniformly distributed and can result in tc-MEP evidence of ischemia. (Ann Thorac Surg 2000;70:222-8) (C) 2000 by The Society of Thoracic Surgeons.