Early outcomes after endovascular management of acute, complicated type B aortic dissection
Article Outline
Objectives
Surgical management of acute, complicated type B aortic dissection is associated with significant morbidity and mortality. This study examined the feasibility and safety of endovascular treatment of this pathology.
Methods
We reviewed a prospectively maintained thoracic endovascular database and medical records at a single institution from 2005 to 2007. The study group comprised of acute, complicated type B dissections, defined as duration of symptoms ≤14 days and involving either false lumen rupture, malperfusion, intractable pain, or uncontrolled hypertension. All repairs were performed using the TAG device (W. L. Gore and Associates, Flagstaff, Ariz). Select 30-day or in-hospital outcomes were reported.
Results
Of the 216 thoracic endovascular aortic aneurysm repairs performed during the study period, 33 (15%) were for acute, complicated type B dissections. There were eight women (24%). The mean age was 61 ± 15 years. The average duration of symptoms was 2.9 ± 4.1 (median, 1) days. The indications for repair included rupture in 15 patients (46%) and mesenteric/renal/lower extremity malperfusion in 11 (33%). Mean fluoroscopy time and contrast volume were 30 ± 16 minutes and 176 ± 55 mL, respectively. Eight (73%) of 11 patients with malperfusion required branch vessel stenting. The 30-day in-hospital mortality was 21% (7 of 33). Causes of death included cardiac arrest in 3, progressive multisystem organ failure in 2, rupture in 1 and unknown in 1. At least one major complication occurred in 76% of the patients, including respiratory failure in 11 (33%), permanent spinal cord ischemia in 5 (15%), renal failure requiring dialysis in 4 (12%), and stroke in 4 (12%). The mean postoperative length of stay was 17.2 ± 16.5 days, and only 14 (42%) were discharged to home.
Conclusions
Emergency endovascular repair of acute, complicated type B dissection is associated with significant mortality and morbidity. The overall role of this therapy in the treatment of this lethal problem should be better defined and compared with other surgical or interventional options before being generally adopted.
Aortic dissection is a relatively rare entity, with an incidence of 2.9 cases per 100,000 per year.1 Current management of uncomplicated type B dissections is intensive medical therapy with blood pressure and pulse rate reduction using nitrates and β-blockers, which has been associated with a 3-year actuarial survival of 78%.2 However, urgent intervention is indicated when an acute dissection is complicated by malperfusion or rupture.3, 4 Treatment options have included direct aortic replacement, extra-anatomic revascularization (eg, axillary-femoral bypass) for lower extremity ischemia, and percutaneous fenestrations for visceral malperfusion.5, 6
More recently, endovascular stent graft treatment of acute aortic dissections has been used as a potentially minimally invasive alternative to conventional aortic surgery.7 The principles of therapy involve restoration of true lumen flow with closure of the primary tear and restoration of visceral/lower extremity perfusion. Although no commercially available device in the United States is currently approved for this indication, the TAG endograft (W. L. Gore and Associates, Flagstaff, Ariz) has been used off-label in these cases. In this study, we review the early outcomes of endovascular stent graft repair of acute, complicated acute aortic dissections at a single institution.
Methods
We retrospectively reviewed our prospectively maintained database of thoracic endovascular aortic repairs (TEVAR), electronic medical records, and preoperative and postoperative imaging studies at a single tertiary-care university medical center. During the 27 months between September 2005 and December 2007, 216 TEVARs were performed. Fig 1 illustrates the distribution of underlying pathologies that were treated. Of these, 33 (15%) were performed for acute (duration of symptoms ≤14 days), complicated DeBakey type III/Stanford type B dissections, defined as a primary tear distal to the left subclavian artery. Specific indications included rupture in 15 (46%); malperfusion in 11 (33%), comprising mesenteric in 3, renal in 3, and lower extremity in 10; intractable pain in 3 (9%), limited retrograde extension of the dissection in 2 (6%), intramural hematoma in 1 (3%), and severe hemolysis in 1 (3%; Fig 2). All repairs were performed using the TAG device (off-label).
Fig 1.
Distribution of pathologies treated with thoracic endovascular aneurysm repair during the study period. AD, Acute, complicated type B dissections; APU, aortic penetrating ulcers; BTT, blunt traumatic transections; CD, chronic dissections; dTAA, descending thoracic aortic aneurysms repaired with TAG and other devices under an investigational device exemption clinical trial; hybrid, includes both abdominal visceral and thoracic arch debranching procedures.
Fig 2.
Rupture from an acute type B dissection. Note the large left hemothorax, which was drained by tube thoracostomy after endovascular repair.
The last three indications deserve additional clarification. In the two patients with “limited retrograde extension,” although the primary tear was distal to the left subclavian artery, the proximal extent of the dissection involved the arch but did not extend proximal to the innominate artery and the ascending aorta, which would technically make the condition a type A dissection. In these situations, coverage of the primary tear would theoretically depressurize the false lumen proximal as well as distal to the tear.8 Although patients with intramural hematoma are older, more commonly present with hypertension, and have radiographic features distinct from a classic aortic dissection,9 this condition was included in the study because it may represent a forme fruste of an aortic dissection,10 and its natural history can be accompanied by similar complications of intractable pain, progression, and rupture.11, 12
The case of severe hemolysis was an unusual indication. Extensive preoperative imaging, including color-flow Doppler transesophageal echocardiography (TEE), which demonstrated high-flow jets through several small fenestrations in the descending thoracic aorta, and consultations with hematology were obtained to exclude other potential etiologies before proceeding with treatment.
Anatomic suitability was based on a reference aortic (between the left common carotid and left subclavian artery) inner diameter (intimal) ≤37 mm, and a centerline distance from the distal margin of the left common carotid artery to the start of the primary tear ≥20 mm (Fig 3). Device oversizing was based on manufacturer's instructions for use. The side of insertion of the primary introducer sheath was based on the size of the access vessels and distal extension of the dissection. Intravascular ultrasound (IVUS) was used in 27 of 33 (82%) patients to verify the preoperative anatomy and true lumen passage of the principle guidewire. TEE was not routinely used.
Fig 3.
Multiplanar reformation of a computed tomography angiogram shows the measurements of the proximal landing zone that were used to determine anatomic suitability and endograft selection.
Endograft coverage typically started at the distal margin of the left common carotid artery origin and extended ≤20 mm of the celiac artery. Balloon molding was avoided accept for a significant endoleak at completion angiography, and only compliant balloons were used (Coda, Cook Medical, Bloomington, Ind).
A left carotid–subclavian bypass was performed in the setting of a dominant left vertebral artery or a patent left internal mammary to left anterior descending artery bypass graft. Adjunctive stenting was performed based on angiographic assessment of visceral or lower extremity vessel lumen and contrast flow after deployment of the thoracic endograft and closure of the primary tear.
Spinal drainage was used inconsistently either preoperatively or postoperatively based on a combination of operator preference or new-onset symptoms of spinal cord ischemia (SCI), or both. No patient presented with preoperative paraparesis or paraplegia. All patients were monitored postoperatively in the cardiac intensive care unit. For those patients who did not have spinal drainage, any evidence of lower extremity neurologic symptoms were managed with pharmacologic elevation of systolic blood pressure >160 mm Hg, fluid resuscitation, and prompt insertion of a spinal catheter ≤60 minutes by an on-call cardiac anesthesiologist. The spinal catheter was placed at 10 cm above the level of the heart and adjusted higher or lower depending on therapeutic effect and amount of spinal fluid drainage. Drainage was limited to <15 mL/h or 350 mL/24 h to avoid potential complications of subdural hematoma or cerebral herniation. The catheter was left to drain for 72 hours. If any degree of therapeutic effect was achieved, the catheter was clamped for another 24 hours in case symptoms returned, and then removed.
Reported outcomes included 30-day or in-hospital mortality, morbidity, intraoperative procedural measures, and discharge disposition from the index hospitalization. The study was approved by the Institutional Review Board.
Results
The mean age of the patients was 61 ± 15 years, and eight (24%) were women. The mean duration of symptoms was 2.9 ± 4.1 days (median, 1; range, 0-14 days). Immediately before repair, in the subset of patients with malperfusion, eight (73%) had an estimated glomerular filtration rate of <60 mL/min/m2 and 5 (45%) had a lactic acid level ≥1.5 mmol/L.
The mean diameter and length of the proximal landing zone were 30.5 ± 4.5 mm and 47.3 ± 45.2 mm, respectively, and the TAG devices were oversized at a mean of 17.3% ± 9.3%. The mean fractional aortic coverage (length of aorta covered/total length of thoracic aorta from left common carotid to celiac artery) was 87%. In 22 of 33 patients (67%), the proximal landing zone involved coverage of the left subclavian artery and extension to the left common carotid artery. A left carotid–subclavian bypass was required in three (14%) of these 22 patients and was performed either immediately before or after the endovascular repair based on radiographic assessments of their vertebrobasilar circulation by preoperative computed tomography angiography or intraoperative angiography, or both. No symptomatic left arm ischemia occurred in the remaining 19 patients. No patient required an iliac conduit for compromised access.
Mean fluoroscopy time was 30 ± 16 minutes, contrast volume was 176 ± 55 mL, overall procedure time was 130 ± 69 minutes, and estimated blood loss was 300 ± 127 mL. Right brachial catheterization was used in 10 of 33 (30%) for angiographic access when the dissection extended into the contralateral external iliac or femoral artery, or both, and precluded safe vascular entry into the true lumen. Adjunctive endovascular stenting was required in 13 (39%) using either self-expanding or balloon-expandable covered or bare-metal stents, as appropriate (Fig 4). Among the subset of 11 patients who presented with malperfusion, 73% required branch vessel stenting. By comparison, only 27% of 15 patients who underwent repair for rupture required adjunctive stenting (P = .04). The locations of adjunctive stenting included the iliac artery in five cases, the mesenteric arteries in five, and the renal arteries in four.
Fig 4.
Thick maximum intensity projection of adjunctive aortic branch vessel stenting that was required to treat malperfusion due to residual static flaps. Note the self-expanding nitinol stent (Protégé, ev3, Plymouth, Minn) in the superior mesenteric artery and a balloon-expandable covered stent (iCAST, Atrium Medical, Hudson, NH) in the left renal artery.
Additional surgical or endovascular procedures were required in 16 patients (48%) after their endograft repair, including 6 aortic bandings (4 distally, 2 proximally and distally), 3 aortic replacements, 2 exploratory laparotomies with small bowel resections, 1 bilateral iliofemoral bypass for disrupted iliac vessels, and 2 tube thoracostomies. More specifically, aortic banding was performed to treat either a persistent false lumen flow or type I endoleak. Briefly, a 26- or 30-mm Dacron tube graft (InterVascular, Datascope, Montvale, NJ) was bi-valved, resulting in an 8- or 9.5-cm-wide patch that was cut to approximate length (π × desired diameter). The aorta was circumferentially exposed either at the proximal or distal landing zones of the endograft through a thoracotomy with division of one or more intercostal arteries, as necessary. The graft was wrapped around the aorta, and the ends were stapled together. The seam was gradually tightened until an epiaortic ultrasound probe detected cessation of flow.
Three patients underwent direct surgical replacements of the descending thoracic aorta to treat persistent false lumen flow. One patient had a proximal endoleak associated with an ascending aortic aneurysm, and the ascending aorta and arch were replaced. In another patient, after a failed proximal banding for a persistent endoleak, the arch and proximal descending aorta were replaced with distal anastomosis performed directly to the remaining endograft. In the third patient, a distal aortic banding was complicated by a pseudoaneurysm and the aorta distal to the endograft was replaced. Exploratory laparotomies and segmental resection of small-bowel infarctions were performed in the two patients with malperfusion immediately after their endovascular repair. These were likely secondary to their initial visceral ischemia (vs intraoperative emboli).
The 30-day in-hospital mortality rate was 21% (7 of 33), with a mean time to death of 19 days (median, 12; range, 1-69 days). All of the deaths occurred in those who presented initially with either malperfusion (n = 2, 18%) or rupture (n = 5, 33%). The causes of death were cardiac arrest in 3 (43%), progressive multisystem organ failure in 2 (29%), rupture in 1 (14%), and unknown in 1 (14%). Of the three patients who died of cardiac arrest, one was hemodynamically unstable, required escalating pharmacologic supportive care, and the family decided to withdraw support; one had persistent hypotension of unclear etiology, and one had an acute type A dissection with pericardial tamponade at autopsy.13, 14
Overall, 25 of 33 patients (76%) experienced some type of major complication. These included respiratory failure in 11 (33%), renal failure in 4 (12%), permanent SCI in 5 (15%), and stroke in 4 (12%). SCI and stroke deserve additional clarification. Six patients had preoperative and 11 had postoperative spinal drainage. No SCI developed in those who had preoperative drainage. Symptoms of lower extremity weakness developed in eight patients (24%) who did not have preoperative drainage. This was completely reversed in three patients with prompt placement of a spinal catheter, but five were left with some degree of permanent deficit. By indication, in the 11 patients with malperfusion, six never had a spinal drain (SCI developed in one, which never resolved, but a drain could not be placed due to hemodynamic instability) and five were placed postoperatively (3 for SCI: 2 reversed, 1 permanent). In the 15 patients treated for rupture, seven did not have spinal drainage, two were placed preoperatively, and six postoperatively (4 for SCI: 1 reversed, 3 permanent). All three patients with intractable pain had preoperative spinal drainage. The remaining four patients did not receive spinal drainage. No complications occurred related to spinal drainage. With regards to strokes, two patients had posterior circulation strokes, one had an anterior circulation stroke, and one had both anterior and posterior circulation strokes. In all of the patients, the repair had extended to the left common carotid artery. However, the left carotid artery flow was not compromised, and all had either a dominant right or codominant vertebral arteries and did not require preemptive subclavian artery revascularizations.
The mean hospital length of stay was 17.2 ± 16.5 days (median, 11; range, 1-69 days). At discharge, only 14 (42%) of the original 33 patients were able to go home, 12 (36%) were transferred to another inpatient facility, and one (3%) left against medical advice.
Postoperative imaging obtained ≤1 month of the TEVAR (mean, 13 days) was available in 20 patients (61%). Of these, there were seven endoleaks (35%), consisting of two type Ia and five type II-5, and 15 (75%) had complete false lumen thrombosis of the thoracic aorta. At a mean follow-up of 5.5 months, only 20 patients (61%) were alive and only 15 (45%) were living at home.
Discussion
The early mortality rate after endovascular stent graft repair of acute, complicated aortic dissections may be up to 25%,15, 16, 17, 18, 19, 20 with a significant number of complications involving bowel infarction, extremity gangrene, pneumonia, and renal failure.16, 17 Although false lumen patency has been shown to be an independent predictor of dissection-related death,21 endovascular therapy for aortic dissection has been associated with only a 75% rate of false lumen thrombosis.19 Alternative techniques to treat malperfusion syndromes have included primary percutaneous fenestrations with or without adjunctive branch vessel stenting. In a review of 165 cases of renal malperfusion treated with this technique, Barnes et al22 reported a mortality rate of 21%. Using a variety of endovascular stenting or fenestration, or both, Slonim et al23 reported a mortality of 25% despite a successful revascularization rate of 93%.
Endovascular stent graft repair of complicated acute dissections may offer several theoretic advantages over other catheter-directed therapies such as primary percutaneous fenestrations. Endograft repair aims to restore the native aortic anatomy by closure of the primary tear, redirection of blood flow through the true lumen, and depressurization of the false lumen. Percutaneous fenestration, in contrast, aims to increase true lumen perfusion by equalizing the pressures in both lumens by creating additional tears in the flap, but does not address the underlying anatomic abnormality of the dissection itself. In the era of endograft therapy, we believe that the role of percutaneous fenestrations should be limited to those patients who lack a suitable proximal landing zone or complex multilumen dissections that are not corrected by closure of the primary tear. Similarly, extra-anatomic surgical revascularizations only address the ischemic complication without treatment of the dissection itself. In both of these treatment modalities, the patient still faces an ongoing risk of rupture or extension of the dissection, which could result in recurrent malperfusion and late aneurysmal degeneration of the false lumen.
Despite the minimally invasive nature of endovascular repair, the early mortality (21%) and morbidity (76%) associated with this therapy was not insignificant. The favorable early clinical results of endovascular repair seen for elective repairs of atherosclerotic, degenerative thoracic aortic aneurysms have not been mirrored in the current series. Clearly, this is related to the compromised pathophysiology of these patients at the time of their initial presentation and the unavailability of a suitable endograft system that is able to accommodate some of the unique anatomic issues that involve aortic dissections.
In a single-center review of acute, complicated aortic dissections managed by a variety of techniques, the mortality rate of those who presented with malperfusion (28%) was much higher than those who did not (23% vs 16%, P = .26). One-third of the patients who presented with malperfusion in this series were managed with peripheral intervention alone (including percutaneous fenestration, branch vessel stenting, or surgical bypass), whereas the remainder were treated medically or had direct aortic replacements.
The mortality rates for patients treated with peripheral intervention were lower than those who underwent aortic therapy (12% vs 28%, P = .19).24 Duebener et al18 reviewed a series of patients who underwent endografting for acute type B dissections and found an overall mortality rate of 20%, although there was no mortality among the two patients presenting with rupture.18 In a study by Verhoye et al20 involving 16 patients with complicated type B aortic dissections who had endovascular repairs, early mortality was 25%, and fully half of the deaths were attributable to ruptures.
Our practice has been to routinely cover most of the descending thoracic aorta and an average of 87% of the aorta was covered in this series. The issue of how much of the thoracic aorta should be covered (proximal tear alone vs entire dissected segment) remains controversial. The disadvantages of extended aortic coverage include increased cost of additional devices, risk of SCI, procedural time, and contrast volume. On the other hand, our rationale was as follows:
In the current study, SCI developed in nearly a quarter of those who did not receive prophylactic spinal drainage, two-thirds of whom were left with permanent deficits even after prompt placement of a spinal drain. Based on this study and observations made of our entire TEVAR practice, we now place prophylactic drains in those patients (1) requiring >15 cm thoracic aortic coverage, (2) coverage of the last 5 cm of the distal thoracic aorta, (3) who do not have coagulopathy, and (4) who are hemodynamically stable. Postoperatively, although theoretically more physiologic, we have not routinely monitored intrathecal pressures for regulation of spinal drainage but have only used volumetric measures. To date, significant complications have not occurred from this approach.
Conclusions
Emergency endovascular repair of acute, complicated type B aortic dissections may be feasible, with outcomes that seem comparable with other therapies; however, these repairs are technically challenging and are still associated with significant mortality and morbidity, which is likely reflective of the pathophysiology of the underlying disease. The overall role of this therapy relative to other established surgical or interventional options remains to be better defined before its generalized adoption in the treatment of acute aortic dissections.
Author contributions
References
- Epidemiology and clinicopathology of aortic dissection. Chest. 2000;117:1271–1278
- Long-term survival in patients presenting with type B acute aortic dissection: insights from the International Registry of Acute Aortic Dissection. Circulation. 2006;114:2226–2231
- . Aortic dissection: perspectives in the era of stent-graft repair. J Vasc Surg. 2006;43(suppl A):30A–43A
- Endovascular repair of acute type B aortic dissection: early and mid-term results. J Vasc Surg. 2006;43:1090–1095
- . Utility of the aortic fenestration technique in the management of acute aortic dissections. Ann Thorac Cardiovasc Surg. 2007;13:296–300
- Expert consensus document on the treatment of descending thoracic aortic disease using endovascular stent-grafts. Ann Thorac Surg. 2008;85:S1–S41
- . Acute aortic dissection. Lancet. 2008;372:55–66
- . Transluminal placement of endovascular stent-grafts for the treatment of type A aortic dissection with an entry tear in the descending thoracic aorta. J Vasc Surg. 2001;34:1023–1028
- . Pathologic variants of thoracic aortic dissections (Penetrating atherosclerotic ulcers and intramural hematomas). Cardiol Clin. 1999;17:637-7
- Acute intramural hematoma of the aorta: a mystery in evolution. Circulation. 2005;111:1063–1070
- Intramural hemorrhage of the thoracic aorta (Diagnostic and therapeutic implications). Circulation. 1995;92:1465–1472
- Natural history and serial morphology of aortic intramural hematoma: a novel variant of aortic dissection. Am Heart J. 1997;134:495–507
- Retrograde type A dissection after endovascular stenting of the descending thoracic aorta (Is the risk real?). Eur J Cardiothorac Surg. 2008;33:1014–1018
- Type A dissection following endovascular thoracic aortic stent-graft repair. J Endovasc Ther. 2005;12:74–81
- . Epidemiology and long-term prognostic factors in acute type B aortic dissection. Ann Vasc Surg. 2007;21:415–422
- Complications after endovascular repair of acute symptomatic and chronic expanding Stanford type B aortic dissections. J Thorac Cardiovasc Surg. 2006;132:361–368
- Endovascular stent-graft placement for the treatment of acute aortic dissection. N Engl J Med. 1999;340:1546–1552
- Emergency endovascular stent-grafting for life-threatening acute type B aortic dissections. Ann Thorac Surg. 2004;78:1261–1266
- Endovascular stent-graft placement in aortic dissection: a meta-analysis. Eur Heart J. 2006;27:489–498
- . Complicated acute type B aortic dissection: midterm results of emergency endovascular stent-grafting. J Thorac Cardiovasc Surg. 2008;136:424–430
- Effects of the patent false lumen on the long-term outcome of type B acute aortic dissection. Eur J Cardiothorac Surg. 2004;26:359–366
- A single-center experience treating renal malperfusion after aortic dissection with central aortic fenestration and renal artery stenting. J Vasc Surg. 2008;47:903–910
- . Percutaneous balloon fenestration and stenting for life-threatening ischemic complications in patients with acute aortic dissection. J Thorac Cardiovasc Surg. 1999;117:1118–1126
- Contemporary management of aortic branch compromise resulting from acute aortic dissection. J Vasc Surg. 2001;33:1185–1192
Competition of interest: none.
PII: S0741-5214(08)01797-7
doi:10.1016/j.jvs.2008.09.071
© 2009 Published by Elsevier Inc.
