Paraplegia Prevention in Thoracic and Thoracoabdominal Aortic Aneurysm Repair

The overlapping effects of aortic occlusion, collateral blood flow, metabolism, oxygen supply and demand, surgical technique, and response to ischemia underlie the unpredictability of spinal cord ischemia and subsequent paralysis in the surgical treatment of thoracoabdominal aortic aneurysms (TAAAs). However, experimentally validated interventions that enhance collateral circulation, increase ischemic tolerance, and add neurochemical protection have reduced the paraplegia risk by 80% to 90% since the 1980s ( Figure 1 ).

Open full size image

O/E ratios (ratio of observed to expected number of deficits) for paralysis have declined significantly since 1985 owing to the introduction of experimentally validated strategies that optimize collateral blood flow and increase ischemic tolerance of the spinal cord. Direct intercostal reimplantation and the use of assisted circulation alone have not improved O/E ratios.

Experimental studies of aortic occlusion show a decrease in spinal cord blood flow of 20% to 50% from baseline in animals not becoming paralyzed and a 60% to 90% decrease in those experiencing paralysis. With reperfusion, blood flow returns to baseline or slightly above in intact animals. However, paralyzed animals show a hyperemic response, with reperfusion blood flows two to three times greater than baseline. This reperfusion hyperemia correlates with the severity of ischemic insult to the spinal cord. Interventions or events causing small positive or negative changes in spinal cord blood flow or oxygen demand and supply relative to the critical threshold necessary to maintain viability have a profound impact on clinical outcome (i.e., paralysis).

The most important radicular artery that supplements the anterior spinal artery is considered to be the greater radicular artery, originating from the artery of Adamkiewicz, usually between vertebral levels T8 and L2. Like the baboon’s, the anterior spinal artery in humans is continuous but has functional zones determined by size that cause resistance to cephalad flow at the level of the greater radicular artery. This resistance to flow creates the vulnerable zone where spinal cord infarction occurs.

An alternative paradigm, validated by Griepp and Etz, hypothesizes that there is an axial collateral network of small arteries in the spinal canal, perivertebral tissues, and paraspinous muscles that receives inputs from the subclavian, mammary, and hypogastric arteries. These small arteries anastomose with each other and with the anterior and posterior spinal arteries that provide blood flow to the cord. The reduction in mean arterial pressure and blood flow in the collateral network that occurs with disruption of intercostal and lumbar arteries with aortic replacement is transient, and experimentally mean pressure and blood flow return to normal within 96 hours. The axial collateral network may be especially vulnerable to hypotension from hemorrhage, sepsis, dialysis, or myocardial infarction, resulting in delayed paraplegia. The concept of collateral circulation explains how maintaining a high arterial pressure and cardiac index reduces spinal cord ischemia and infarction during and after TAAA surgery. It also explains why most patients are not paralyzed after TAAA surgery when no or few intercostal arteries are revascularized.

The most important factors affecting spinal cord blood flow and injury during aortic occlusion, which have been confirmed clinically and experimentally, are hypotension, anemia (rupture and shock), the length of aorta replaced (Crawford types) correlating with quantitative disruption of intercostal blood flow, temperature, cerebrospinal fluid (CSF) dynamics, neurotransmitters, cardiac function, and aortic occlusion time.

These factors have been assessed before, during, and after aortic occlusion, and multivariate modeling has been used to stratify them based on statistical significance to quantify their relative importance in reducing paralysis when acting simultaneously ( Table 1 ). Extent of aortic replacement, acuity, and cardiac function are more important than temperature and occlusion time, which suggests that deleterious changes in other critical factors, such as CSF pressure, cardiac function, and oxygen delivery after aortic occlusion, are the underlying problem with longer aortic occlusion times.

Strategies to reduce spinal cord ischemia have fallen into four areas of clinical investigation and intervention: direct anatomic perfusion and reconstruction of spinal cord circulation, manipulation of cord metabolism with drugs and hypothermia to prolong ischemic tolerance and reduce reperfusion injury, augmenting collateral blood flow by controlling hemodynamics, controlling CSF pressure, and combinations of these therapies. To compare clinical reports of paralysis outcomes for thoracoabdominal aortic replacement, we developed an accurate mathematical model of paralysis risk, which accounted for 97% of the variability in paralysis rates in clinical reports up to that time ( Table 2 ). This model confirmed, as do most clinical reports, that paralysis risk is a function of the amount of aorta replaced and clinical presentation and that paralysis risk is also technique dependent. The ratio of observed to expected number of deficits (O/E ratio) is a quantitative measure of the effectiveness of different spinal cord protective strategies. The model makes it possible to compare different techniques such as simple aortic occlusion with intercostal reimplantation with and without assisted circulation, which have O/E ratios of 1 (1.15 and 0.98) (see Table 2 ).