To understand the role of IOM , it is useful to consider the surgical procedure in stages and consider the strategy for use of IOM so as to reduce the risk of paralysis [19]. Initially, proximal cross-clamping of the aorta is done. At this point, all spinal cord perfusion comes from the vertebral arteries. This cross-clamp also causes a critical elevation of blood pressure proximal to the clamp with accompanying elevations in cerebrospinal fluid (CSF) pressure in the brain and spinal cord. This increased CSF pressure can further reduce spinal cord perfusion pressure (SCPP) [SCPP = spinal arterial pressure − spinal CSF pressure] compounding the reduction of spinal cord blood flow [2]. In order to minimize hypoperfusion, many surgeons try to limit the aortic cross-clamp to no more than 30–40 min.

Since perfusion is often not adequate, a surgeon may place a shunt or bypass pump between the proximal and distal aorta to extend the safe operative time. This bypass reduces the proximal hypertension and the associated increased CSFP . It provides better distal spinal cord perfusion for patients who are dependent on the caudal vessels [8]. Although a bypass perfusion pressure of 60–70 mmHg is generally accepted as adequate, the pressure necessary to effectively perfuse critical vessels is not known in any given patient without functional testing provided by IOM. Adequate pressure where IOM studies maintain pre-cross-clamp amplitude and latency is as high as 90–110 mmHg [20, 21]. Use of these distal perfusion techniques has been reported to reduce the risk of paralysis to about 10 % from the risk of 30–50 % with cross-clamping alone [14].

In some patients, distal perfusion alone is adequate to provide spinal cord perfusion. In other patients, perfusion from the critical radicular arteries is also required to prevent paraplegia. In more than half of the patients with aneurysms, a rich collateral network surrounding the spinal cord prevents ischemia, even when otherwise critical intercostal arteries are sacrificed [22]. But a group of patients with inadequate collateral networks may exist, in which reimplantation of intercostal arteries may be crucial for maintaining adequate spinal cord blood flow. These patients can be identified using IOM testing during segmental cross-clamping of the aorta.

Several methods are available to provide reperfusion of critical arteries once they are identified. One approach is to preserve the back wall of the aneurysm where the intercostal arteries arise and use it in the aortic reconstruction thus restoring critical perfusion [2]. Since this technique may not be feasible, other techniques have been developed to identify and implant the arteries in the repaired aorta segment. Because of its importance, some surgeons attempt to identify the ARM preoperatively to ensure that it is reimplanted into the aortic graft. Simply reimplanting the ARM further reduces the risk of paralysis from 10 % with distal perfusion to 5–6 % [14, 23]. If it cannot be identified, some surgeons reimplant all intercostals in the T8–T12 range [2].

Intraoperative monitoring can also be useful to help assess if other techniques employed to reduce the risk of spinal cord ischemia are useful. Of particular note is the use of CSF drainage to maintain a low CSFP and improve the net perfusion pressure (CSFPP). Intraoperative monitoring can assist in identifying the critical CSFP that improves perfusion [24, 25].

In short, the goal of the surgical technique is to optimize the perfusion and reduce the overall ischemic time of the spinal cord. Hence, the surgeon must work expeditiously during the cross-clamp time. The value of IOM is to promptly identify spinal cord regions that are ischemic while the insult is reversible and then guide the interventions to correct the ischemia and reduce the risk. IOM has been applied successfully to these surgeries and has reduced the risk of intraoperative paralysis.

Attempts to increase the allowable ischemic time in TAA repair by using hypothermia have not been universally applied because it can lead to adverse systemic effects of prolonged cardiopulmonary bypass and profound hypothermia (e.g., coagulopathy, endothelial dysfunction, systemic inflammation) when applied systemically [26]. However, since deep hypothermia (14.1–20.0 °C) can reliably protect the spinal cord from ischemic damage, techniques using regional hypothermia are being explored. The results in early reports of TAA repair under regional spinal cord hypothermia using a custom-designed epidural catheter were excellent [26, 27]. Unfortunately, an ideal agent for pharmacologic neuroprotection has yet to be found [27].

Since SSEP monitoring has been available since the 1980s, the most extensive clinical experience has been with SSEP monitoring. During aortic surgery, use of this technique can identify ischemia of the neural tract including peripheral nerve, white matter tracts of the dorsal columns, and ischemia in the brainstem or cerebral cortex.

A slow onset of cortical SSEP change (>15 min) usually indicates peripheral nerve ischemia. Isolated events, such as the femoral artery bypass cannula occluding flow to the leg, is one of the most frequent causes of peripheral nerve ischemia and unilateral loss of SSEP. If a bilateral change occurs within 15 min, it has been assumed to be of spinal cord origin, either from inadequate distal aorta perfusion or loss of critical intercostals. Thus, clinicians have used the SSEP to determine if bypass was necessary, if the pressure of bypass was adequate, and if it was safe to exclude a specific radicular artery.

One of the earliest human studies using the SSEP was done by Cunningham and Laschinger et al. [28]. They observed four patterns of change in the SSEP depending on the mechanism of ischemia. A type I pattern was where the SSEP was altered in 3–4 min following proximal cross-clamping of the aorta with complete loss in 8–9 min. This pattern was thought to be indicative of inadequate perfusion of the spinal cord from the distal arterial supply. This emphasized the value of bypass perfusion [29]. A type II pattern was where the SSEP did not change after aortic cross-clamping, suggesting adequate spinal perfusion was present from cephalad vessels.

A type III pattern was where the patient was critically dependent on a radicular artery and changes in the SSEP were noted when it was occluded. This points out the use of the SSEP to locate the critical intercostals during aortic cross-clamping. The SSEP was shown to be restored when the critical intercostal was reimplanted or unclamped. Other studies report similar changes; rapid reperfusion of critical intercostal vessels restores the cortical SSEP and appears to reduce the risk of paraplegia [28, 30, 31]. A type IV pattern, where a gradual “fade out” reduction in amplitude but not latency of the SSEP over 30–50 min, is considered characteristic of lower extremity hypoperfusion [32]. This may signal the need to reorient the bypass cannula.

Use of the SSEP has not been completely reliable for always being able to predict paralysis since it does not monitor the motor tracts and instead monitors proprioception and vibration pathways in the posterior spinal cord [20]. In some cases, an SSEP loss has been correlated with motor injury. This is seen most often when an SSEP loss occurred rapidly after cross-clamping (within 3–5 min) or when the duration of an SSEP loss was prolonged (40–60 min) [29, 33–35].

With the advent of reliable motor-evoked potential monitoring, clinical inquiry turned toward the use of transcranial motor-evoked potential (MEP) monitoring. This form of monitoring differs from SSEP monitoring in that the responses can detect ischemia in the anterolateral white matter tracts, the spinal grey matter, and the peripheral nerve and muscles. The use of MEPs is of particular interest since aortic occlusion in dogs demonstrated that the predominant injury from spinal cord ischemia is grey matter necrosis [36, 37].

Although false negatives with MEP monitoring have occurred (i.e., immediate postoperative paralysis with preserved intraoperative MEP), most studies show an excellent correlation of outcome with MEP findings [8, 13, 20, 21, 25, 32, 38]. All studies, both clinical and experimental, have reported a rapid change in MEP, within 2–4 min, to ischemic conditions where the ischemia includes the spinal grey matter serving the muscles monitored. This rapid response provides more useful feedback in the intraoperative environment where time is critical in determining outcome. These clinical studies suggest that MEP monitoring is more sensitive to ischemia than SSEP monitoring and provides significant additional reductions in paraplegia.

Clinical studies have found that when comparing SSEP and MEP findings, there is a relatively long delay (7–30 min) between the onset of ischemia and the disappearance of SSEP, whereas the MEP generally changes comparatively quickly within 2–5 min [21, 32, 39–41]. The most important finding was that SSEP changes were not always associated with MEP changes.

Clearly, the effectiveness of the monitoring and some of the differences between the SSEP, MEP, and epidural recorded responses relate to the specific neural tracts being monitored and their susceptibility to ischemia. Based on the available data, the time to electrical failure of various neural tissues is shown in Table 40.1. The cerebral cortex is the most sensitive to ischemia, with a loss of electroencephalographic (EEG) activity within 20 s after inadequate perfusion. In the spinal cord, the grey matter is most sensitive to ischemia, with a loss of synaptic activity within 1–2 min. Conduction in sensory and motor white matter tracts (axons) demonstrates alterations in activity within 3–6 min (SSEP tract) or 11 min (motor tract) and a loss of conduction at between 7 and 18 min (SSEP) or 11 and 17 min (MEP). Hence, the white matter tracts of the sensory and motor systems are about equal to each other in terms of sensitivity to ischemia [20, 32, 42, 43]. Isolated peripheral nerve ischemia results in the loss of SSEP and MEP conduction after 20–30 min.

The second issue comparing SSEP to MEP pertains to the prediction of outcome. When focusing on motor outcome, the MEP has the greater predictive power because it actually measures function in the cortical-spinal tract (CST). However, since only 5 % of the CST fibers are involved in the response, the correlation is not exact [44]. Further, a variety of other descending tracts are needed for coordinated motor function, so losses in associated tracts may hamper motor function despite maintenance of adequate motor cell function in the CST [44]. Use of SSEPs also has the drawback that ischemia in the anterior spinal region that supplies the motor tracts may not be reflected in the posteriorly located SSEP tracts. Thus, the differences in arterial blood supply, AntSA versus PostSAs, produce a differential vulnerability to ischemia. Hence, the loss of the SSEP and MEP acts as a surrogate for the entire functional aspect of the spinal cord.

The progression of spinal cord ischemia after TAA interventions can frequently be identified by IOM and arrested before irreversible infarction results [27]. This spinal cord rescue depends on the early detection and immediate multimodal intervention to maximize spinal cord oxygen supply. The routine application of IOM in TAA repair remains controversial, although many operative teams have successfully integrated it into their practice [15]. Some studies show a significant role of MEPs during TAA repair and the subsequent surgical strategies necessary to maintain or regain spinal cord function, leading to significantly reduced paraplegia rates [45]. The recent American College of Cardiology/American Heart Association Guidelines for the Diagnosis and Management of Patients with Thoracic Aortic Disease indicate that motor- or somatosensory-evoked potential monitoring can be useful when the data will help to guide therapy (CLASS IIa, Level of Evidence: B) [16]. Protocols to guide therapy include those mentioned above (e.g., CSF drainage, increase in mean arterial and distal aortic pressure, reattachment of intercostal arteries) [45].

Although open surgical repair is the superior technique, thoracic endovascular aneurysm repair (TEVAR) avoids subjecting patients whose comorbid conditions predispose them to significant morbidity or mortality to the open procedure (especially a thoracotomy) [46]. Endovascular grafting may be of particular value in patients with significant comorbid conditions (older age, substantial cardiac, pulmonary, and renal dysfunction) who would be considered poor or unacceptable candidates for open surgery [16]. Recommendations from the American College of Cardiology/American Heart Association Guidelines for the Diagnosis and Management of Patients with Thoracic Aortic Disease are that in patients with degenerative or traumatic aneurysms of the descending thoracic aorta exceeding 5.5 cm, saccular aneurysms, or postoperative pseudoaneurysms, endovascular stent grafting should be strongly considered when feasible (CLASS I, Level of Evidence: B) [3, 16].

The potential advantages of endovascular grafting include the absence of a thoracotomy incision and the need for partial or total extracorporeal circulatory support and clamping of the aorta. This procedure also has the advantage of a shorter ischemic time for distal tissues. In addition, several factors increase the blood flow reserve of the spinal cord during the perioperative period. These include (1) an increased cardiac output, (2) increased blood pressure, (3) avoidance of hypoxemia and anemia, (4) preservation of the critical intercostal arteries, and (5) reduced CSF pressure [21, 47].

The clinical advantages of endovascular stent placement have been stated to include shorter intensive care unit and hospital stay, lower 30-day mortality, fewer pulmonary complications, reduced pain, and reduced transfusion requirements [46, 48]. For example, 50 % of patients have chronic obstructive pulmonary disease, and the reduced pulmonary complications associated with stent placement improve the likelihood of hospital discharge [46].

However, endovascular treatment of TAAs is significantly more complex due to the need to incorporate visceral arteries in the repair at the same time that the aneurysm is excluded. One major problem with the endovascular techniques is the limited ability to provide vascular supply to blood vessels originating from the aorta in the region of the stent. This has led to stents with branches or fenestrations to provide a means to supply vessels such as the renal or mesenteric arteries [46]. The stents have not, however, been well developed for providing flow to the ARM or critical radicular arteries, such that spinal cord ischemia could occur if these vessels are compromised and are essential. These stents may require 4–6 weeks of manufacturing time, making them less useful in an emergency or for an urgent repair [15].

Several factors may limit the ability to use TEVAR in an individual patient. These include the absence of suitable “landing zones” above and below the aneurysm (usually 2–3 cm of normal diameter aorta without circumferential thrombus for the stent to attach), an aortic width at the landing zones that exceeds the recommended width for the largest available endovascular grafts (generally 10–15 % larger than the width of the aorta), a lack of vascular access sites, and severe aortic atherosclerosis [3, 16].

The disadvantages of TEVAR include late complications, which are uncommon with the open technique (graft migration, stent fracture, endovascular leaks, and stent failure necessitating surgical correction) [48]. In addition, TEVAR is associated with increased risk of atheroembolic stroke and retrograde type A dissection (especially in patients with fragile aortic wall) [49].

Although the complications of the open portion of the procedure are reduced, there is evidence that endovascular treatment of TAAs, similarly to open repair, carries the same risks of paraplegia, renal failure, stroke, and death [15]. High-risk factors of spinal cord injury during endovascular aortic repair are (1) coverage of the left subclavian artery, (2) extensive coverage of long segments of the thoracic aorta, (3) prior downstream aortic repair, (4) compromising important intercostal (T8–L1), vertebral, pelvic, and hypogastric collaterals, and (5) an aorta prone to give off atheromatous emboli [47].