The overall incidence of SCI (i.e., paresis or paralysis) after TEVAR ranges from to 2–10%. Other studies demonstrated a 0–10.3% SCI range, with an average of 4.5% . The onset and severity of injury after TEVAR depends on the ability of the collateral network to supply the marginally vascularized area in the critical zone of the spinal cord, known as the watershed area in the grey matter of the spinal cord [5, 6]. These data show the critical nature of intercostal circle and its determining role in establishing the temporal way of onset of neurological complications. Risk factors for development of SCI after TEVAR include coverage of the left subclavian or hypogastric artery, embolization during intervention, renal failure, perioperative hypotension, prior abdominal aortic aneurysm repair, more than 3 aortic stent grafts used in the procedure, a length of total aorta to be covered of more than 20 cm .
The spinal cord perfusion is a terminal circulation and, as a consequence, the pathophysiology of SCI essentially follows an ischemia-infarction pattern. Aortic pathologies may alter this vascular network making it proner to hemodynamic instability. The concept of a collateral network was proposed recently , suggesting that multiple factors contribute to SCI, in addition to occlusion of some critical intercostal arteries, as Adamkiewitz affirmed some decades ago.
Segmental inputs from intercostal arteries are assumed to be essential for maintaining adequate flow to the anterior spinal artery, but only at certain levels do the anterior and posterior radicular-medullary arteries, which represents intercostal arteries’ last divisions, do cross the dura and reach the medulla. In fact, only a few of these segmental branches remain in adults .
The left subclavian artery and the internal iliac arteries also contribute to the spinal cord perfusion by delivering branches that feed the radiculo-medullary vessels.
The mechanism of SCI may involve global hypoperfusion, as with aortic cross clamping or systemic hypotension or hemodynamic instability; selective ischemia from ligation of dominant segmental vessels; or secondary insult, such as with reperfusion injury. In such patients the atherosclerotic disease and endovascular procedure may occlude many segmental arteries and promote collateral vessels enlargement, altering significantly the normal patterns of blood supply to the spinal cord, which are moreover extremely variable among individuals.
However other factors, among which extravascular components also, may play a significant role in the selection of patients at risk of developing neurological complication.
Neurologic deficits may also be related to degenerative changes of the spinal column resulting in narrowing of the spinal canal, myelopathy, or to spinal cord compression in general. Stenotic damage to the spinal cord is thought to be the result of two processes: direct compression of the neural elements and ischemic disruption of arterial and venous structures surrounding the spinal cord .
Several studies have shown that the radicular venous system proved to be particularly affected by the compression since the veins in the region were reduced in number or collapsed, and showed a grossly visible congestion proximal to the lesion. Congestion of the venous system, vulnerable to compression, decreases perfusion in capillaries directly feeding nerve roots, producing ischemia and consequent intra-radicular edema caused by breakdown of the blood-nerve barrier .
Our patient preoperative spinal MRA showed severe stenosis of lumbar canal.
It is possible that in poor circulating condition the degree of narrowing of the spinal cord could contribute to decrease of the grade of tolerance to any ischemic injury, determining a state of basal suffering which makes it more sensitive to hypoperfusion and to development of hypoxic damage. This reserve, similarly to coronary arteries, is a delicate balance between ischemic injury and capacity to meet physiological needs through activation of collateral networks and neurohormonal blood flow regulatory mechanism.
According to James et al. , a LSS may therefore increase the perfusion pressure required to maintain tissue perfusion to the spinal cord at the level of the stenosis, such as in this case. The fall in spinal cord perfusion pressure following loss of many contributing segmental vessels and blocked anterior spinal artery could then easily tip the balance to drop the perfusion pressure below critical level at the level of the stenosis promoting steal phenomena.
Spinal cord perfusion pressure is the difference between mean arterial blood pressure (MAP) and CSF (or central venous pressure, whichever is greater). According to general guidelines for minimizing SCI, it is advisable to increase MAP (i.e., > 90 mmHg) and drain CSF (≤ 10 mmHg) to maintain spinal cord perfusion pressure at levels above 80 mmHg . However, many studies have shown that the patients with lumbar spinal stenosis have an increased epidural pressure at the level of the stenosis . It may be hypothesized that patients with spinal canal disorders require an even higher blood pressure value than those proposed by standard prevention protocols, suggesting that the development of a uniform multimodal preventive treatment protocol can be elusive.
Other considerations could be done on the timing of the surgical procedure. In our case the second step was performed as an emergency by implantation of a four branched custom made endoprosthesis due to substantial rise of diameter and early sign of rupture. The onset of neurological complications led us to defer the second surgical step; an attitude of observation could be justified in order to both obtain the maximum benefits from the phenomenon of collateralization and guaranteeing a personalized treatment for the patient, an adequate preoperative planning, even if the risks deriving from postponing the treatment of a pathology that may rapidly evolve still remain a concern.
Selective reimplantation of intercostal/lumbar arteries represent an open surgical alternative performed in high-volume centers for the prevention of spinal cord ischemia. James et al.  documented a case in which, despite having reimplanted two pairs of intercostal arteries in a patient with post-dissecative thoracoabdominal aneurysm and concomitant stenosis of the medullary canal, irreversible medullary ischemia developed anyway. Despite to comorbidities, an open repair with selective reimplantation of the intercostal-lumbar arteries would have been rational in our case. Unfortunately, the most sensible tract of the spinal cord is the stenotic one, as theorized by some authors [10, 11] and such a surgical procedure, technically difficult and not always feasible, may not prevent SCI anyway.
In conclusion, it is likely that stenotic pathology of the medullary canal is a comorbidity that may be investigated in candidates for TAAA endovascular repair. This comorbidity may constitute an additional anatomic risk factor in those patients currently recognized as prognostically associated to development of perioperative neurological complications. The preoperative use of spinal cord MRA could be useful not only in the assessment of collateral circles but also in both the diagnosis of concomitant medullary associated diseases and the evaluation of their relative weight on the spinal injury risk. It cannot be excluded that these patients with spinal canal disorder may benefit from a personalized SCI prevention protocol.