Thoracic endovascular aortic repair — descending thoracic aneurysm & Type B dissection stent-graft planning
Thoracic endovascular aortic repair begins long before the catheter lab: with a fine-cut, ECG-gated computed tomography angiogram (CTA) from the thoracic inlet to the femoral heads. This single study defines the disease (degenerative aneurysm vs. acute or chronic Type B dissection), measures the proximal and distal landing zones, characterizes access vessels, and identifies every branch vessel that could be jeopardized by graft coverage.
Descending thoracic aortic disease presents in two principal forms that TEVAR addresses differently:
Degenerative (atherosclerotic/fusiform) aneurysm: • Progressive dilation of all three wall layers beyond 1.5× normal diameter (>4 cm thoracic) • Risk of rupture rises sharply above 6 cm (annual rupture risk 6.9%/yr) — elective repair typically offered at 5.5–6.0 cm, or growth >0.5 cm/6 months • Often fusiform (circumferential) but may be saccular, which carries disproportionate rupture risk at smaller diameters
Type B aortic dissection (Stanford classification — tear distal to the left subclavian artery): • An intimal tear allows blood to dissect the media, creating a true lumen (compressed, carries antegrade flow) and false lumen (often larger, may thrombose or remain patent) • Uncomplicated Type B: managed medically first-line (blood pressure and heart rate control) — TEVAR reserved for complicated cases • Complicated Type B: malperfusion (visceral, renal, limb), rupture, refractory pain, or rapid false-lumen expansion — TEVAR is the guideline first-line intervention • High-risk uncomplicated features (large aortic diameter >4 cm, false lumen >22 mm, patent false lumen with entry tear) increasingly favor early TEVAR to prevent late aneurysmal degeneration
Successful TEVAR depends entirely on a durable seal proximal and distal to the pathology. CTA-based 3D center-lumen-line reconstruction (not axial slices, which overestimate diameter on an oblique aorta) is used to measure:
• Proximal landing zone: a segment of relatively normal, non-diseased, non-angulated aorta at least 20–25 mm long, sized to the graft with 10–20% oversizing • Distal landing zone: similarly sized healthy segment distal to the pathology, ideally proximal to the celiac axis to preserve visceral branches • Aortic diameter, angulation (arch and thoracoabdominal), tortuosity, and mural thrombus burden — all influence conformability and risk of endoleak or bird-beaking • Access vessel diameter and calcification (femoral/iliac) — sheath sizes for thoracic devices run 20–25 Fr outer diameter, requiring iliofemoral vessels typically ≥7–8 mm
The proximal landing zone in the aortic arch is precisely what the Ishimaru zone classification (Stage 2) formalizes — because the closer the pathology sits to the great vessels, the more the procedure must account for cerebral and upper-extremity perfusion.
Introduced by Ishimaru in 1999 and now the universal planning language of arch and thoracic endovascular repair, the zone classification divides the aorta from the aortic valve to the celiac axis into numbered landing segments. Where the required proximal seal falls dictates whether the left subclavian artery — or the carotid and innominate arteries as well — must be intentionally covered, and therefore whether debranching or revascularization is needed before graft deployment.
The classification runs proximal (zone 0) to distal (zone 5):
• Zone 0 — proximal edge at the origin of the innominate (brachiocephalic) artery; a graft landing here must cover the origins of all three arch branches unless debranched • Zone 1 — between the innominate and left common carotid artery origins; coverage here sacrifices the left carotid and left subclavian unless revascularized • Zone 2 — between the left common carotid and left subclavian artery origins; coverage here sacrifices only the LSA — the most common "elective coverage" scenario • Zone 3 — the first 2 cm of aorta distal to the LSA origin; a short proximal cuff is available without covering any branch • Zone 4 — remainder of the proximal-to-mid descending thoracic aorta • Zone 5 — distal descending thoracic aorta, from roughly the mid-descending segment to the celiac axis origin
Zones 0–2 require some form of branch management (debranching bypass or in-situ/branched device) because the graft must intentionally cross a great vessel origin to achieve a competent seal. Zones 3–5 generally allow simple LSA preservation.
Zone assignment is not academic — it dictates the entire procedural plan:
• Zone 0–1: extensive coverage requires total or partial arch debranching (carotid-carotid and carotid-subclavian bypass, or innominate bypass) via sternotomy or, increasingly, branched/fenestrated arch devices in specialized centers • Zone 2: the pivotal decision zone — LSA can be covered with or without prior revascularization depending on collateral anatomy (Stage 3) • Zone 3–5: standard TEVAR without branch vessel management; the main determinant of risk shifts to overall coverage length and spinal cord perfusion (Stage 5)
Accurate zone measurement from centerline CTA reconstruction — not visual estimation — is essential, since a seal that appears to fall in zone 3 on an axial image can, once the true curvature of the arch is accounted for, actually require zone 2 coverage.
The left subclavian artery supplies the left vertebral artery (a major contributor to the posterior cerebral circulation and, critically, to spinal cord collateral flow), the left arm, and in many patients a left internal mammary artery available for future coronary bypass. Deciding whether to sacrifice, preserve, or proactively revascularize the LSA is one of the highest-stakes decisions in thoracic endograft planning.
Revascularization — most commonly a carotid-subclavian bypass (synthetic graft tunneled from the common carotid to the subclavian artery) or subclavian-to-carotid transposition — is recommended when:
• The left vertebral artery is dominant, hypoplastic/absent on the right, or terminates directly into the posterior inferior cerebellar artery (PICA) without a complete circle of Willis — covering the LSA here risks posterior circulation stroke • A patent left internal mammary artery (LIMA) graft supplies coronary circulation — LSA coverage would cause acute myocardial ischemia • A functioning left upper extremity arteriovenous fistula/graft for hemodialysis is present • Long segment coverage is planned (extending the risk to spinal cord perfusion, since the left vertebral artery is a major collateral feeder — Stage 5) • The patient has had prior infrarenal aortic surgery or hypogastric artery occlusion, reducing collateral reserve elsewhere
In emergent, ruptured, or malperfusion presentations, coverage often proceeds first to control the acute threat, with staged or same-setting revascularization as anatomy and time permit — accepting a small transient stroke/ischemia risk to address the immediate life threat.
Three technical approaches accomplish LSA revascularization:
• Carotid-subclavian bypass: a prosthetic (typically 8mm PTFE or Dacron) conduit tunneled subcutaneously from the common carotid artery to the subclavian artery distal to the vertebral origin — preserves antegrade flow to both vertebral and internal mammary arteries, durable patency • Subclavian-to-carotid transposition: the subclavian artery is divided and directly reimplanted into the common carotid artery — avoids prosthetic material and has excellent long-term patency, but is technically more demanding • Endovascular in-situ fenestration or branched/parallel graft (chimney/snorkel) techniques: preserve LSA flow through the endograft itself without open surgery — evolving options in specialized centers, avoiding a cervical incision entirely
When revascularization is deferred and the LSA is simply covered, roughly 5–10% of patients develop arm claudication and a smaller fraction require delayed bypass for symptomatic subclavian steal or digital ischemia.
Deployment is the irreversible step: once the constraining sheath is withdrawn, the self-expanding nitinol stent-graft springs open against the aortic wall and cannot be easily repositioned. Every prior step — imaging, zone planning, and subclavian management — exists to make this moment as safe and precise as possible.
1. Bilateral femoral (or iliac conduit) access is obtained; a stiff guidewire (e.g., Lunderquist) is advanced under fluoroscopy across the arch into the ascending aorta to provide a stable rail 2. Pre-deployment arteriography defines the target vessel origins and confirms planned landing zones against the CTA measurements 3. The delivery system, pre-loaded with the compressed nitinol/polyester (or ePTFE) endograft, is advanced over the wire and rotated/positioned under fluoroscopic roadmap so the proximal marker sits precisely at the planned landing zone 4. Systemic heparinization is confirmed (ACT >250–300s) before manipulation across the pathology 5. Rapid ventricular pacing (or adenosine-induced asystole in select cases) is initiated to transiently drop cardiac output and aortic pulsatility 6. The outer sheath is withdrawn in a single controlled motion, allowing the self-expanding graft to appose the aortic wall precisely at the intended zone 7. Pacing is stopped, blood pressure is allowed to recover, and completion arteriography confirms graft position, patency of preserved branches, and absence of endoleak 8. Proximal and distal seal zones are balloon-molded if needed to eliminate residual folds or type Ia/Ib endoleak
The single greatest technical risk during deployment is graft migration or malposition caused by the "windsock effect": before full apposition, systolic ejection can catch the leading edge of the still-unsheathing graft like a parachute, displacing it distally or proximally by centimeters — potentially covering (or failing to cover) the intended branch vessel.
Rapid pacing at 160–180 bpm reduces stroke volume and pulse pressure by inducing a state resembling ventricular fibrillation hemodynamically (without true fibrillation), dropping systolic pressure transiently below 50–70 mmHg. This "aortic standstill" gives the operator a brief, hemodynamically quiet window to release the graft with millimeter precision. Adenosine-induced transient asystole is a useful alternative or adjunct, particularly when a temporary pacing wire is not already in place.
Once deployed, the graft is oversized 10–20% relative to native aortic diameter specifically so that the radial force of the nitinol stent skeleton maintains apposition against arterial pressure even before any endothelialization or fibrous encapsulation occurs over subsequent weeks.
Spinal cord ischemia (SCI), manifesting as paraparesis or paraplegia, is the complication most feared after TEVAR because it is often permanent and devastating to quality of life despite a technically perfect graft. Unlike open thoracoabdominal repair, TEVAR causes no direct intercostal artery ligation — yet SCI still occurs because endograft coverage abruptly excludes a length of aorta supplying segmental (intercostal and lumbar) arteries that feed a rich collateral network sustaining the cord.
The spinal cord is not supplied by a single vessel but by a redundant "collateral network" — a term formalized by Griepp and Griepp (2007) that reframed decades of thoracoabdominal surgical thinking. The anterior spinal artery runs the length of the cord, fed segmentally by radiculomedullary branches from intercostal and lumbar arteries (the largest often called the artery of Adamkiewicz, typically arising T8–L1), and cross-connected to the subclavian (via vertebral arteries), internal iliac (hypogastric), and paraspinous muscular collateral beds.
Acute coverage of a long segment of thoracic aorta abruptly removes many segmental artery inputs simultaneously. Spinal cord perfusion pressure (SCPP) — the net driving pressure for cord blood flow — falls when the collateral network loses too many contributing vessels at once, even though no single artery is thought to be indispensable. SCPP is approximately mean arterial pressure minus the higher of intrathecal (CSF) pressure or central venous pressure — which is exactly why both blood pressure augmentation and CSF drainage are the two levers available to protect the cord.
Risk rises with the extent to which the collateral network is disrupted, not simply with aneurysm size:
• Coverage length: the single strongest predictor — long-segment coverage (>15–20 cm, or >8 vertebral body levels) sharply increases risk by removing more segmental artery origins • Concurrent LSA coverage without revascularization: removes a major collateral inflow (via the vertebral artery) at the same time as thoracic segmental arteries are excluded • Prior or planned infrarenal aortic repair / hypogastric artery occlusion: removes distal collateral inflow, "sandwiching" the cord between two disrupted beds • Perioperative hypotension: even brief episodes can drop SCPP below the ischemic threshold in an already-compromised collateral bed • Extent of prior aortic coverage (staged repairs, prior open or endovascular surgery)
Prevention strategy combines several complementary measures: 1. Prophylactic lumbar CSF drainage — a spinal drain lowers intrathecal pressure, directly increasing spinal cord perfusion pressure (SCPP = MAP − max(CSF pressure, CVP)); drains are typically placed for high-risk cases (long coverage, prior aortic surgery, planned LSA coverage) and kept at CSF pressure <10 mmHg for 24–72 hours 2. Permissive/induced hypertension — maintaining mean arterial pressure above baseline (often MAP >90 mmHg or SBP >140–160 mmHg) in the immediate postoperative period directly raises SCPP 3. Staged repair — when feasible, covering the aorta in two sequential procedures allows the collateral network time to remodel and augment between stages, rather than losing all segmental inputs simultaneously 4. Minimizing total coverage length and preserving the LSA and hypogastric arteries whenever the landing zone and pathology allow 5. Neuromonitoring (motor evoked potentials) in high-risk cases to detect ischemia in real time and trigger immediate blood pressure augmentation or CSF drainage adjustment 6. Vigilant serial neurologic examination for 24–48 hours, since roughly half of SCI events are delayed in onset, appearing after an initial normal exam — reflecting evolving collateral insufficiency rather than an intraoperative injury.