Perioperative CSF Drains

Historically, the incidence of paraplegia following thoracic aortic aneurysm (TAA) surgery was on the order of 40%, although more recent estimates are below 20% for open TAA surgery and 8% for thoracic endovascular aneurysm repair (TEVAR). Possible mechanisms of injury include: (1) Decrease in spinal cord perfusion pressure during use of the aortic cross-clamp, (2) Increase in intrathecal pressure (ITP) transmitted from elevated ICP and hypercarbia, (3) Ligation of collateral vessels, (4) Thrombosis, (5) Hematoma, (6) Repefusing injury causing cord edema, (7) Predisposing conditions (peripheral vascular disease, diabetes, prior aortic surgery).

Intrathecal drains have been used perioperatively to decrease the chances of perioperative spinal cord ischemia by improving SC blood flow. The equation for spinal cord perfusion pressure (SCPP) is SCPP = MAP = ITP. Ideally, SCPP should be above 50-60 mm Hg. In addition to providing adequate BP (which may increase infra-clamp pressures by collaterals), the ITP should be kept below 15 by draining CSF. This may be achieved by passive drainage at a rate of no more than 10-15 cc/hr (although rates as high as 20 cc/hr may be used if necessary to improve SCPP). Higher rates of drainage risk stretching and tearing of dural bridging veins (may be seen as blood coming from drain).

There is a controversy of whether the drain should be kept open and allowed to drain continuously or used for continuous monitoring with intermittent drainage. The advantage of continuous drainage is avoidance of excessive ITP. The disadvantage is greater chance of bleeding. If continuous drainage were to be used, it is recommended to keep the height of the reservoir no less than 10 cm above the drain to avoid an excessive decrease in ITP. The advantage of continuous monitoring is detection of obstruction by disappearance of the normal ITP wave and avoidance of excess drainage. The disadvantage is possible excess accumulation of CSF (unlikely with vigilant monitoring). Due to the concern of intracranial hemorrhage, some authors recommend continuous monitoring.

There is a paucity of quality human data on the efficacy of spinal drains. Most studies are small and do not control for all variables. The largest study by Coselli of 145 patients showed an impressive 80% reduction in postoperative neurologic deficits using spinal drains in combination with hypothermia, left-heart partial bypass, and selective intercostal reimplantation. A general impression is that use of CSF drainage alone is ineffective intraoperatively. Postoperatively, drainage has been shown to be effective for delayed onset paraplegia.

As mentioned, spinal drainage is part of a multifaceted approach to spinal cord protection that may include distal perfusion techniques (eg, extracorporeal bypass, reimplantation of intercostals), epidural cooling, deep circulatory arrest, monitoring with SSEP/MEP, pharmacologic protection (although most agents have not beeen proven effective, eg, steroids, barbituates, calcium entry blockers, NMDA antagonists, naloxone, papaverine, allopurinol, zirconitide, activated protein C). The use of naloxone is especially intriguing because intrathecal opioids have been shown to worsen neurologic outcome, possibley through mu- and delta-receptor, but not kappa-receptor, agonism.

Complications of intrathecal catheters include infections (0.08% meningitis), CSF leak (caution should be used during blood patch placement to avoid excess increases in ITP), intracranial hemorrhage (best seen by T2-weighted MRI), neuraxial hematoma, and fractured catheters. The overall mortality associated with spinal drains is estimated with by 0.6% (40% if hemorrhage occurs).