- Email: [email protected]
Descending Aortic Stent Graft Collapse During Frozen Elephant Trunk Repair: Detection Using Invasive Blood Pressure Monitoring and Intraoperative Transesophageal Echocardiography
Descending Aortic Stent Graft Collapse During Frozen Elephant Trunk Repair: Detection Using Invasive Blood Pressure Monitoring and Intraoperative Transesophageal Echocardiography Devon Cole, MD, Aaron Seller, DO, and Yong G. Peng, MD, PhD, FASE
T
HE DIFFICULTY OF repairing continuous aortic disease from the ascending aorta and aortic arch through the descending thoracic aorta was addressed by Borst in 1983, when he described an “elephant-trunk” technique that joined ascending aorta and aortic arch replacement, followed by placement of a free-floating length of tube graft extended into the descending thoracic aorta.1 This technique provided the advantage of having graft already in the descending thoracic aorta for correction of more distal disease at a later stage, thereby avoiding the hazards of a surgical approach to arch structures when completing distal thoracic aorta repair. In the 1990s, surgeons began combining ascending aorta and aortic arch repair with deployment of an endovascular stent graft into the descending thoracic aorta, called the frozen elephant-trunk (FET) procedure.2-4 The benefit of an FET stent graft is a tight seal in the descending aorta at an area that is not accessed easily through an anterior approach (Fig 1C, green arrow). It also turns a complex, 2-stage procedure into 1 surgery. FET has been used for chronic aneurysms of the aorta, proximal and distal to the left subclavian artery,5 and acute and chronic dissections of both Stanford types A and B, for which it primarily is used in patients with complicated type-B dissections.6 The authors report a case in which a stent collapse was detected after an FET procedure. This devastating complication can lead to distal organ hypoperfusion syndrome. The case highlights the fact that using proximal and distal arterial pressure reading discrepancies and intraoperative transesophageal echocardiography (TEE) can ensure timely diagnosis and guide effective surgical correction. FET is performed via median sternotomy with tube graft replacement of the ascending aorta and aortic arch with the patient under deep hypothermic circulatory arrest. Arch reconstruction may involve supra-aortic debranching of the great vessels with tube graft prostheses anastomosis versus hemiarch technique anastomosis. For an FET procedure, the distal arch tube graft is joined to a descending endovascular stent graft deployed anterograde into the exposed descending thoracic aorta (Fig 2). Exclusion of the left subclavian artery by the endovascular stent is possible, which then would require bypass. However, using landing zone 3 of the descending aorta obviates the necessity for bypass (Fig 1A, yellow arrow). This case report describes the events that led to the detection of stent graft collapse using intraoperative TEE during an FET procedure. The authors became suspicious after separating the patient from cardiopulmonary bypass when discrepancies between left radial and left femoral invasive pressure monitoring persisted stent graft collapse at the proximal descending aorta, showing a loss of more than 75% of internal stent diameter, with limited high-velocity, color-flow, and highpressure gradients. These TEE views also were used to assess the effectiveness of balloon graft expansion and the efficacy of an additional endostent at the proximal descending aorta to reexpand FET. The University of Florida Institutional Review Board granted permission to publish this report.
CASE PRESENTATION
A 77-year-old, 70-kg, white male with known Stanford type-B dissection returned with intolerable chest discomfort radiating to his back. Continued medical management of blood pressure followed by a series of repeat chest computed tomography (CT) angiography showed little change in his descending aorta dissection. However, in comparison with a prior CT scan, an increased change in false lumen flow in the descending aorta and a secondary aneurysmal dilation of the ascending aorta to 5.5 cm and arch to 4.1 cm, respectively, were noted. Comorbidities included a known 4.0-cm fusiform, infrarenal abdominal aortic aneurysm with a 0.5-cm mural thrombus followed up without change for the past 5 years, a 40-plus pack-year smoking history, hypertension, ischemic colitis, and coronary artery disease status postdrug-eluting stent to the proximal right coronary artery. The patient’s baseline creatinine was 1.2 mg/dL, and electrocardiography demonstrated sinus rhythm with multiform premature ventricular complexes. Left heart catheterization showed mildly reduced left ventricular function, estimated ejection fraction of 50%, and mild global hypokinesis. Using CT angiography, the patient was classified as experiencing chronic Stanford type-B dissection originating at the aortic wall distal to the left subclavian takeoff (Fig 1B, blue arrow), extending down to the level of the celiac plexus. The aneurysmal ascending aorta had a cross-sectional diameter of 5.5 cm, with preservation of root structures and an arch diameter of 4.1 cm. Because the patient had little relief from chest and back pain after medical management and he experienced an interval increase in ascending aorta dilation, the decision for surgery was made. Because of his preserved cardiovascular function and the overall health of his organ systems, it was decided that the patient was a good surgical candidate. There was no significant carotid artery stenosis, history of cerebral vascular accident, or transient ischemic attack, and surgical mortality and stroke risk were believed to be reduced by the hybrid endovascular approach. The surgical plan for aorta reconstruction included an ascending graft from the sinotubular junction through the arch with graft side arms for arch vessel anastomosis, with distal graft connection to the descending aortic
From the Department of Anesthesiology, University of Florida College of Medicine, Gainesville, FL. Address reprint requests to Yong G. Peng, MD, PhD, FASE, Department of Anesthesiology, University of Florida College of Medicine, 1600 SW Archer Road, PO Box 100254, Gainesville, FL 32610. E-mail: [email protected]fl.edu © 2016 Elsevier Inc. All rights reserved. 1053-0770/2602-0033$36.00/0 http://dx.doi.org/10.1053/j.jvca.2015.12.016 Key words: thoracic aneurysm, frozen elephant trunk, arterial blood pressure monitoring, transesophageal echocardiography, stent collapse, organ hypoperfusion
Journal of Cardiothoracic and Vascular Anesthesia, Vol ], No ] (Month), 2016: pp ]]]–]]]
1
2
COLE ET AL
Fig 1. (A) Ishimaru classification of thoracic aortic landing zones for repair with endovascular stent graft; zone 3, distal to left subclavian artery (yellow arrow). (B) Sagittal view of chronic Stanford type-B dissection with aneurysmal ascending aorta; descending dissection flap originates distal to the left subclavian artery takeoff (blue arrow). (C) Schematic of tube graft reconstruction of ascending aorta and aortic arch joined to the endovascular stent graft with frozen elephant trunk landing zone distal to the left subclavian takeoff and tight seal in descending thoracic aorta (green arrow).
stent deployed antegrade through the exposed aorta (Fig 1C). For consideration of hybrid debranching versus open arch repair, data on long-term durability and survival are lacking.7 The hybrid approach was selected as a less-invasive alternative, although 4 observational studies showed a nonsignificant trend toward increased neurologic events and late-term mortality in hybrid debranching groups.8 General anesthesia was induced with propofol, fentanyl, and vecuronium and was maintained with isoflurane in a 50% oxygen/air mixture. Oral intubation was uneventful, and hemodynamics were stable without the need for vasodilator or vasopressor support. Baseline lactate was 0.6 mmol/L. Invasive blood pressure monitoring was measured via left radial and left femoral arteries and central pressures by right internal jugular vein pulmonary artery catheter; right arm circulation was reserved for antegrade cerebral perfusion. A baseline TEE examination was performed with no signs of valvular regurgitation and normal biventricular function.
Fig 2. Anterograde deployment of frozen elephant trunk TX-2 stent graft over stiff guidewire in exposed descending thoracic aorta.
After sternotomy and systemic heparinization, a 2-stage venous cannula was placed in the right atrium and a retrograde cardioplegia catheter was placed the coronary sinus. The right axillary artery was cannulated with a catheter (EOPA,Medtronic, Minneapolis, MN) for antegrade cerebral perfusion during hypothermic circulatory arrest. At this point, the patient was placed on cardiopulmonary bypass and his temperature was cooled to 261C. Circulatory arrest commenced with the patient in the deep Trendelenburg position, and bilateral electroencephalography showed sustained isoelectric burst suppression. Dissection of the innominate, left carotid, and left subclavian arch arteries was completed for arch reconstruction with a 34-mm Gelweave prosthesis, 14 10 10 mm (VascutekTerumo, Renfrewshire, Scotland), with a side-arm branch for attachment to the cardiopulmonary bypass circuit. A wire then was passed from the left femoral artery into the thoracic aorta true lumen, confirmed using TEE with descending aorta short-
Fig 3. Intraoperative vital signs with invasive arterial blood pressures, showing a 15 mmHg mean arterial pressure gradient difference between left radial artery of 79 mmHg (red circle) and left femoral artery of 64 mmHg (white box).
DETECTION OF DESCENDING AORTIC STENT GRAFT COLLAPSE
axis and long-axis views. The wire was pushed up out of the exposed proximal descending thoracic aorta into the surgical field. An antegrade catheter was placed over the exposed femoral wire and was removed after being exchanged for a stiff antegrade wire used for anterograde deployment of a 36 202-mm Zenith TX 2 endograft (Cook Medical, Bloomington, IN) into the exposed descending thoracic aorta (see Fig 2) landing zone in the proximal end distal to the left subclavian takeoff. The stent device was deployed without event, and there were no open view signs of stent compression during initial deployment. The proximal end of the stent graft then was anastomosed to the 34-mm Gelweave graft just distal to the left subclavian artery takeoff (Fig 1C, schematic). Cardiopulmonary bypass was reinstituted via the side-arm branch on the 34-mm graft to reperfuse the body and brain via right axillary access. Total circulatory arrest time was 29 minutes with bilateral cerebral oximetry (INVOS, Somanetics, Troy, MI). Although intraoperative perfusion performance was optimized via antegrade cerebral flow throughout the entire circulatory arrest phase, there was a symmetric 33% reduction in saturations measured. Hemodilution of hematocrit from 38% to 29% after bypass pump prime was considered the major contributor to this finding. Antegrade cerebral perfusion via the right axillary artery was maintained at 500 mL/min. Although near-infrared spectroscopy is good, it is not perfect and has not been shown to reduce postoperative stroke risk.9 Red cell transfusion was used to increase hematocrit 430% with a concomitant increase in cerebral saturations.
3
Anastomosis of the arch vessels was performed in a sequential manner with continued antegrade perfusion via the right axillary artery until completion of anastomosis of the left subclavian, left carotid, and right innominate arteries and the sinotubular junction was completed (Fig 1C, schematic). After this, the patient’s temperature was rewarmed to a core of 36.51C, and the authors were able to wean the patient from cardiopulmonary bypass. Total cardiopulmonary bypass time was 3 hours, 46 minutes. At separation from cardiopulmonary bypass, the patient’s mean arterial blood pressure differences fluctuated between 15 and 20 mmHg from the left radial (Fig 3, red circle) to left femoral (Fig 3, white box) arteries. Epinephrine infusion at 0.04 mg/kg/min was started, with the following hemodynamic parameters: central venous pressure 6 mmHg, pulmonary artery pressure 26/13, cardiac output 6.1 L/min, cardiac index 3 L/min/m2, and mixed venous oxygen saturation 76%. When 20 mmHg pressure differences persisted between the left radial and left femoral arteries, the patency and calibration of arterial pressure monitoring systems were checked. Suspicion of a stent graft malfunction arose when no other plausible explanation was identified. Suspicion of a stent graft collapse then was assessed using TEE. A 2-dimensional cross-sectional view of the descending aorta short axis at the proximal descending aorta revealed stent graft collapse, with a 475% loss of diameter (Fig 4A: true lumen, red circle; false lumen, blue circle), just distal to the arch anastomosis. When compared with more a distal descending stent graft, full
Fig 4. Two-dimensional transesophageal echocardiography cross-sectional view of descending aorta short axis at proximal descending aorta showing stent graft collapse, 475% loss of diameter (A) (true lumen: red circle: false lumen: blue circle) just distal to arch anastomosis, compared with more distal descending stent graft with full expansion (B) (true lumen: red circle; false lumen: blue circle). Descending aorta long-axis view at proximal descending aorta shows stent graft severity of lumen compression with limited, high-velocity color-flow Doppler (C) (yellow) compared with more distal descending stent graft with laminar color-flow Doppler (D) (red).
4
expansion was seen (Fig 4B: true lumen, red circle; false lumen, blue circle). The descending aorta long-axis view at the proximal descending aorta showed stent graft severity of lumen compression with limited, high-velocity, color-flow Doppler (Fig 4C, yellow) stent graft with laminar color-flow Doppler (Fig 4D, red). Continuous-wave Doppler showed a peak gradient across the compressed area to be approximately 50 to 80 mmHg, except the angle of incidence was 420 degrees. To more reliably assess the severity of stent graft collapse, intravascular ultrasound (IVUS) and hemodynamic pressure gradient via pull back from the femoral artery catheter were considered, but patient stability and signs of adequate perfusion evidenced with relative normal arterial gas values (pH 7.36, HCO3– 22.8 mmol/L, base excess –1.9 mmol/L, lactate 2.4 mmol/L) influenced the decision to observe and follow up with contrast CT angiography. Follow-up CT imaging showed a compressed stent graft at the genu of the thoracic descending aorta with 490% loss of diameter (Fig 5, red arrow). Serial arterVial blood gas measures showed a pH 7.34 to 7.45; lactic acid of 1.9, initially trending down to 0.9; and a serum creatinine rise from 1.28 (baseline) to 1.91, with urine output maintained at 50 to 100 mL/hour with nesiritide and bumetanide infusions. Serial venous blood gas saturations were 65% to 77%, with the patient weaned from epinephrine infusion in the first 12 hours postoperatively, and the Swan cardiac index remained at 2.6 to 3 L/min/m2. The patient required mechanical ventilation for more than 24 hours after the first procedure due to signs of volume overload after multiple blood products were administered, with chest x-ray showing edematous vasculature and positive end-expiratory pressure of 10 mmHg required to maintain partial oxygen pressure 470 mmHg. The patient remained intubated and returned to the operating room 2 days later for diminishing distal pedal pulses. Thoracic endovascular aortic repair was performed for expansion of the stent graft collapse. At the beginning of the second procedure, TEE continuous-wave Doppler showed a gradient across the
Fig 5. Postoperative chest computed tomography angiography with sagittal view showing endovascular stent graft collapse (red arrow) at genu of descending aorta.
COLE ET AL
compressed area of approximately 80 to 100 mmHg. IVUS of the thoracic and abdominal aorta showed 90% compression of the stent graft diameter at the genu of the descending thoracic aorta compared with 2.8 cm at the distal descending thoracic aorta. After IVUS confirmation of stent graft collapse, a Zenith 38 152-mm nontapered stent (Cook Medical) was placed after the collapsed stent graft zone was dilated with a Coda balloon (Cook Medical). The in-stent device was deployed without event in zones 3 and 4 of the aorta, and the balloon was dilated in both proximal and distal aspects and the genu where the graft appeared to be collapsed. Follow-up IVUS and angiography demonstrated continued collapse but much improved flow. In addition, the patient’s femoral sheath waveform and his lower extremity Doppler signals were much improved after deployment of the in-stent device. The patient’s mean arterial pressure gradient remained 10 mmHg greater between radial and femoral artery lines; however, after deployment of the additional stent graft, a strong triphasic Doppler signal was demonstrated in both feet, and the patient had good urine output. The patient’s postoperative course worsened, with development of acute kidney injury, sepsis, atrial fibrillation with rapid ventricular rate, and respiratory insufficiency with acute respiratory distress syndrome. His respiratory status declined despite advanced mechanical ventilation, sedation, and a paralytic agent. A family decision for withdrawal of care was made 18 days after the initial surgery. DISCUSSION
This case report outlines a serious complication of endovascular stent graft collapse during an FET procedure. The stent graft collapse was suspected because of an invasive blood pressure discrepancy and was confirmed by TEE, which was instrumental in identifying the area of collapse and evaluating the flow using color-flow Doppler. Because of the angle of incidence, pressure gradients demonstrated on continuous-wave Doppler may not be as reliable. The relative risk of stent collapse during FET has yet to be defined. Several possible causes of collapse were proposed: the stent graft collapse may have resulted from a large mural thrombus at the proximal descending aorta, chronic nature of the dissection, and/or the acute angle of the arch to aortic descent. The phenomenon of endovascular stent graft collapse also has been described as a failure of stent conformation to the aorta wall (apposition), which can lead to endoleak and graft collapse.10 Another possible mechanism of collapse may involve thrombus or compression from surrounding false lumen. Results from the STARZ-TX2 (Study of Thoracic Aortic Aneurysm Repair with the Zenith TX2 TAA Endovascular Graft) clinical trial showed that kinking was noted in 1.6% (2 of 123) of patients at 12 months and compression was noted in 0.9% (1 of 108), but none of these 3 patients experienced adverse clinical sequelae or required a secondary intervention.11 Experience from using Gore’s TAG endovascular stent (Gore, Flagstaff, AZ) has shown that small-diameter aortas, difficult aortic arches, and poor sealing zones increased the risk
5
DETECTION OF DESCENDING AORTIC STENT GRAFT COLLAPSE
for stent graft complications.12 Canaud et al reported 4 cases of stent graft collapse among 285 patients treated with thoracic endovascular aortic repair, in which all 4 patients were treated with the TAG stent graft. Endovascular management was performed in 3 of the 4 patients; the other patient underwent stent graft explant and open repair. None of these patients died. A lack of device wall apposition and acute aortic arch angle (range 921-1181) were observed in all 4 patients.13 Although FET’s 1-stage method has lowered perioperative mortality for patients with extensive aortic disease, morbidity with the incidence of stroke remains a concern. Circulatory arrest time is regarded as an independent risk factor for stroke, despite the routine use of selective antegrade cerebral perfusion.14 Fortunately, the patient in this case report did not exhibit any physical signs of a cerebral vascular event, and intraoperative perfusion performance was optimized during circulatory arrest by continuous antegrade cerebral perfusion, cerebral oximetry, and bispectral index monitoring. The authors recognized that using bilateral radial artery pressures is beneficial for validating systemic pressures across the aortic arch; however, right arm circulation was reserved for antegrade cerebral perfusion, and the single left radial arterial tracing confirmed the absence of left subclavian artery stenosis after stent deployment. The large pressure discrepancy that persisted between left radial and left femoral artery readings prompted the authors to seek the possible causes. To assess distal central perfusion pressure, the authors relied on left femoral artery pressure. Other complications with FET include endoleak, cerebral spinal cord dysfunction, spinal cord infarction, and aortic rupture. During FET, thoracic endovascular stent graft collapse is an exceedingly rare event. One possible option would have been to extend the bypass phase and reopen the distal anastomosis to resolve the stent collapse under direct visualization. Open repair for chronic type-B dissection can be performed safely but is a relatively morbid surgery.15 Although results are encouraging from endovascular aortic repair of acute type-B dissection, chronic type-B dissections pose unique challenges.6 A multicenter review of collapsed thoracic endografts from 2005 to 2009 suggested that endograft collapse could be managed successfully using endovascular techniques in most
cases.16 Tadros et al described successful management of delayed (430 days) endograft-related complications using open conversion with device explantation. Another case report of delayed endovascular treatment of descending aortic stent graft collapse highlighted potential complications of visceral and lower limb ischemia but showed that management using additional endoprostheses had both good clinical and technical results.17 This case highlighted the valuable role of TEE in visualizing stent deployment and assessing overall stent graft function in real time after an FET procedure. TEE views of the proximal descending aorta and short and long axes are obtained easily. These views generate clear images of descending aortic stents and/or grafts and functional evaluation using color-flow Doppler. The risks for this patient’s stent graft collapse were a large mural thrombus at the proximal descending aorta, the chronic nature of dissection, and the acute angle of the arch to aortic descent. In this case, intraoperative decisions to examine clinically based signs of adequate perfusion and successful separation from cardiopulmonary bypass were made in recognition of considerable risk to the patient to undergo stent graft explantation and open repair. CONCLUSION
The FET procedure is a single-stage, hybrid surgical and endovascular procedure to repair thoracic aortic disease of the ascending aorta, aortic arch, and proximal descending aorta. FET has the benefit of being a single surgery but also has its risks. This is a case report of intraoperative endovascular stent graft collapse by compression at the proximal descending thoracic aorta. This first was suspected when vital sign monitoring showed a major difference between arterial blood pressures proximal and distal to the aortic arch after the patient was separated from cardiopulmonary bypass. Stent-graft compression was confirmed using TEE views of the proximal descending aorta that showed a loss of more than 75% of internal stent diameter, limited high-velocity color flow, and pressure gradient across the collapsed zone. TEE views were used to assess the effectiveness of balloon graft expansion with an additional endostent.
REFERENCES 1. Borst HG, Walterbusch G, Schaps D: Extensive aortic replacement using “elephant trunk” prosthesis. Thorac Cardiovasc Surg 31: 37-40, 1983 2. Dake MD, Miller DC, Semba CP, et al: Transluminal placement of endovascular stent-grafts for the treatment of descending thoracic aortic aneurysms. N Engl J Med 331:1729-1734, 1994 3. Kato M, Ohnishi K, Kaneko M, et al: New graft-implanting method for thoracic aortic aneurysm or dissection with a stented graft. Circulation 94:188-193, 1996 4. Buth J, Penn O, Tielbeek A, et al: Combined approach to stent-graft treatment of an aortic arch aneurysm. J Endovasc Surg 5:329-332, 1998 5. Karck M, Chavan A, Khaladj N, et al: The frozen elephant trunk technique for the treatment of extensive thoracic aortic aneurysms: Operative results and follow-up. Eur J Cardiothorac Surg 28:286-290, 2005 6. Pacini D, Tsagakis K, Jakob H, et al: The frozen elephant trunk for the treatment of chronic dissection of the thoracic aorta: A multicenter experience. Ann Thorac Surg 92:1663-1670, 2011
7. Ouzounian M, LeMaire S, Coselli J: Open aortic arch repair: State-of-the-art and future perspectives. Sem Thorac Cardiovasc Surg 25:107-115, 2013 8. Benedetto U, Melina G, Angeloni E, et al: Current results of open total arch replacement versus hybrid thoracic endovascular aortic repair for aortic arch aneurysm: A meta-analysis of comparative studies. J Thorac Cardiovasc Surg 145:305-306, 2013 9. Olsson C, Thelin S: Regional cerebral saturation monitoring with near-infrared spectroscopy during selective antegrade cerebral perfusion: Diagnostic performance and relationship to postoperative stroke. J Thorac Cardiovasc Surg 131:371-379, 2006 10. Lee WA, Martin TD, Hess PJ Jr, et al: First United States experience of the TX2 Pro-Form thoracic delivery system. J Vasc Surg 52:1459-1463, 2010 11. Matsumura JS, Cambria RP, Dake MD, et al: International controlled clinical trial of thoracic endovascular aneurysm repair with
6
the Zenith TX2 endovascular graft: 1-year results. J Vasc Surg 47: 247-257, 2008 12. Muhs BE, Balm R, White GH, et al: Anatomic factors associated with acute endograft collapse after Gore TAG treatment of thoracic aortic dissection or traumatic rupture. J Vasc Surg 45:655-661, 2007 13. Canaud L, Alric P, Desgranges P, et al: Factors favoring stentgraft collapse after thoracic endovascular aortic repair. J Thorac Cardiovasc Surg 139:1153-1157, 2010 14. Bachet J, Guilmet D, Goudot B, et al: Antegrade cerebral perfusion with cold blood: A 13-year experience. Ann Thorac Surg 67: 1874-1878, 1999
COLE ET AL
15. Roselli E: Thoracic endovascular aortic repair versus open surgery for type-B chronic dissection. J Thorac Cardiovasc Surg 149: S163-S167, 2015 16. Tadros R, Lipsitz E, Chaer R, et al: A multicenter experience of the management of collapsed thoracic endografts. J Vasc Surg 53: 1217-1222, 2011 17. Nano G, Mazzaccaro D, Malacrida G, et al: Delayed endovascular treatment of descending aorta stent graft collapse in a patient treated for post-traumatic aortic rupture: A case report. J Cardiothorac Surg 6:76-82, 2011
T
HE DIFFICULTY OF repairing continuous aortic disease from the ascending aorta and aortic arch through the descending thoracic aorta was addressed by Borst in 1983, when he described an “elephant-trunk” technique that joined ascending aorta and aortic arch replacement, followed by placement of a free-floating length of tube graft extended into the descending thoracic aorta.1 This technique provided the advantage of having graft already in the descending thoracic aorta for correction of more distal disease at a later stage, thereby avoiding the hazards of a surgical approach to arch structures when completing distal thoracic aorta repair. In the 1990s, surgeons began combining ascending aorta and aortic arch repair with deployment of an endovascular stent graft into the descending thoracic aorta, called the frozen elephant-trunk (FET) procedure.2-4 The benefit of an FET stent graft is a tight seal in the descending aorta at an area that is not accessed easily through an anterior approach (Fig 1C, green arrow). It also turns a complex, 2-stage procedure into 1 surgery. FET has been used for chronic aneurysms of the aorta, proximal and distal to the left subclavian artery,5 and acute and chronic dissections of both Stanford types A and B, for which it primarily is used in patients with complicated type-B dissections.6 The authors report a case in which a stent collapse was detected after an FET procedure. This devastating complication can lead to distal organ hypoperfusion syndrome. The case highlights the fact that using proximal and distal arterial pressure reading discrepancies and intraoperative transesophageal echocardiography (TEE) can ensure timely diagnosis and guide effective surgical correction. FET is performed via median sternotomy with tube graft replacement of the ascending aorta and aortic arch with the patient under deep hypothermic circulatory arrest. Arch reconstruction may involve supra-aortic debranching of the great vessels with tube graft prostheses anastomosis versus hemiarch technique anastomosis. For an FET procedure, the distal arch tube graft is joined to a descending endovascular stent graft deployed anterograde into the exposed descending thoracic aorta (Fig 2). Exclusion of the left subclavian artery by the endovascular stent is possible, which then would require bypass. However, using landing zone 3 of the descending aorta obviates the necessity for bypass (Fig 1A, yellow arrow). This case report describes the events that led to the detection of stent graft collapse using intraoperative TEE during an FET procedure. The authors became suspicious after separating the patient from cardiopulmonary bypass when discrepancies between left radial and left femoral invasive pressure monitoring persisted stent graft collapse at the proximal descending aorta, showing a loss of more than 75% of internal stent diameter, with limited high-velocity, color-flow, and highpressure gradients. These TEE views also were used to assess the effectiveness of balloon graft expansion and the efficacy of an additional endostent at the proximal descending aorta to reexpand FET. The University of Florida Institutional Review Board granted permission to publish this report.
CASE PRESENTATION
A 77-year-old, 70-kg, white male with known Stanford type-B dissection returned with intolerable chest discomfort radiating to his back. Continued medical management of blood pressure followed by a series of repeat chest computed tomography (CT) angiography showed little change in his descending aorta dissection. However, in comparison with a prior CT scan, an increased change in false lumen flow in the descending aorta and a secondary aneurysmal dilation of the ascending aorta to 5.5 cm and arch to 4.1 cm, respectively, were noted. Comorbidities included a known 4.0-cm fusiform, infrarenal abdominal aortic aneurysm with a 0.5-cm mural thrombus followed up without change for the past 5 years, a 40-plus pack-year smoking history, hypertension, ischemic colitis, and coronary artery disease status postdrug-eluting stent to the proximal right coronary artery. The patient’s baseline creatinine was 1.2 mg/dL, and electrocardiography demonstrated sinus rhythm with multiform premature ventricular complexes. Left heart catheterization showed mildly reduced left ventricular function, estimated ejection fraction of 50%, and mild global hypokinesis. Using CT angiography, the patient was classified as experiencing chronic Stanford type-B dissection originating at the aortic wall distal to the left subclavian takeoff (Fig 1B, blue arrow), extending down to the level of the celiac plexus. The aneurysmal ascending aorta had a cross-sectional diameter of 5.5 cm, with preservation of root structures and an arch diameter of 4.1 cm. Because the patient had little relief from chest and back pain after medical management and he experienced an interval increase in ascending aorta dilation, the decision for surgery was made. Because of his preserved cardiovascular function and the overall health of his organ systems, it was decided that the patient was a good surgical candidate. There was no significant carotid artery stenosis, history of cerebral vascular accident, or transient ischemic attack, and surgical mortality and stroke risk were believed to be reduced by the hybrid endovascular approach. The surgical plan for aorta reconstruction included an ascending graft from the sinotubular junction through the arch with graft side arms for arch vessel anastomosis, with distal graft connection to the descending aortic
From the Department of Anesthesiology, University of Florida College of Medicine, Gainesville, FL. Address reprint requests to Yong G. Peng, MD, PhD, FASE, Department of Anesthesiology, University of Florida College of Medicine, 1600 SW Archer Road, PO Box 100254, Gainesville, FL 32610. E-mail: [email protected]fl.edu © 2016 Elsevier Inc. All rights reserved. 1053-0770/2602-0033$36.00/0 http://dx.doi.org/10.1053/j.jvca.2015.12.016 Key words: thoracic aneurysm, frozen elephant trunk, arterial blood pressure monitoring, transesophageal echocardiography, stent collapse, organ hypoperfusion
Journal of Cardiothoracic and Vascular Anesthesia, Vol ], No ] (Month), 2016: pp ]]]–]]]
1
2
COLE ET AL
Fig 1. (A) Ishimaru classification of thoracic aortic landing zones for repair with endovascular stent graft; zone 3, distal to left subclavian artery (yellow arrow). (B) Sagittal view of chronic Stanford type-B dissection with aneurysmal ascending aorta; descending dissection flap originates distal to the left subclavian artery takeoff (blue arrow). (C) Schematic of tube graft reconstruction of ascending aorta and aortic arch joined to the endovascular stent graft with frozen elephant trunk landing zone distal to the left subclavian takeoff and tight seal in descending thoracic aorta (green arrow).
stent deployed antegrade through the exposed aorta (Fig 1C). For consideration of hybrid debranching versus open arch repair, data on long-term durability and survival are lacking.7 The hybrid approach was selected as a less-invasive alternative, although 4 observational studies showed a nonsignificant trend toward increased neurologic events and late-term mortality in hybrid debranching groups.8 General anesthesia was induced with propofol, fentanyl, and vecuronium and was maintained with isoflurane in a 50% oxygen/air mixture. Oral intubation was uneventful, and hemodynamics were stable without the need for vasodilator or vasopressor support. Baseline lactate was 0.6 mmol/L. Invasive blood pressure monitoring was measured via left radial and left femoral arteries and central pressures by right internal jugular vein pulmonary artery catheter; right arm circulation was reserved for antegrade cerebral perfusion. A baseline TEE examination was performed with no signs of valvular regurgitation and normal biventricular function.
Fig 2. Anterograde deployment of frozen elephant trunk TX-2 stent graft over stiff guidewire in exposed descending thoracic aorta.
After sternotomy and systemic heparinization, a 2-stage venous cannula was placed in the right atrium and a retrograde cardioplegia catheter was placed the coronary sinus. The right axillary artery was cannulated with a catheter (EOPA,Medtronic, Minneapolis, MN) for antegrade cerebral perfusion during hypothermic circulatory arrest. At this point, the patient was placed on cardiopulmonary bypass and his temperature was cooled to 261C. Circulatory arrest commenced with the patient in the deep Trendelenburg position, and bilateral electroencephalography showed sustained isoelectric burst suppression. Dissection of the innominate, left carotid, and left subclavian arch arteries was completed for arch reconstruction with a 34-mm Gelweave prosthesis, 14 10 10 mm (VascutekTerumo, Renfrewshire, Scotland), with a side-arm branch for attachment to the cardiopulmonary bypass circuit. A wire then was passed from the left femoral artery into the thoracic aorta true lumen, confirmed using TEE with descending aorta short-
Fig 3. Intraoperative vital signs with invasive arterial blood pressures, showing a 15 mmHg mean arterial pressure gradient difference between left radial artery of 79 mmHg (red circle) and left femoral artery of 64 mmHg (white box).
DETECTION OF DESCENDING AORTIC STENT GRAFT COLLAPSE
axis and long-axis views. The wire was pushed up out of the exposed proximal descending thoracic aorta into the surgical field. An antegrade catheter was placed over the exposed femoral wire and was removed after being exchanged for a stiff antegrade wire used for anterograde deployment of a 36 202-mm Zenith TX 2 endograft (Cook Medical, Bloomington, IN) into the exposed descending thoracic aorta (see Fig 2) landing zone in the proximal end distal to the left subclavian takeoff. The stent device was deployed without event, and there were no open view signs of stent compression during initial deployment. The proximal end of the stent graft then was anastomosed to the 34-mm Gelweave graft just distal to the left subclavian artery takeoff (Fig 1C, schematic). Cardiopulmonary bypass was reinstituted via the side-arm branch on the 34-mm graft to reperfuse the body and brain via right axillary access. Total circulatory arrest time was 29 minutes with bilateral cerebral oximetry (INVOS, Somanetics, Troy, MI). Although intraoperative perfusion performance was optimized via antegrade cerebral flow throughout the entire circulatory arrest phase, there was a symmetric 33% reduction in saturations measured. Hemodilution of hematocrit from 38% to 29% after bypass pump prime was considered the major contributor to this finding. Antegrade cerebral perfusion via the right axillary artery was maintained at 500 mL/min. Although near-infrared spectroscopy is good, it is not perfect and has not been shown to reduce postoperative stroke risk.9 Red cell transfusion was used to increase hematocrit 430% with a concomitant increase in cerebral saturations.
3
Anastomosis of the arch vessels was performed in a sequential manner with continued antegrade perfusion via the right axillary artery until completion of anastomosis of the left subclavian, left carotid, and right innominate arteries and the sinotubular junction was completed (Fig 1C, schematic). After this, the patient’s temperature was rewarmed to a core of 36.51C, and the authors were able to wean the patient from cardiopulmonary bypass. Total cardiopulmonary bypass time was 3 hours, 46 minutes. At separation from cardiopulmonary bypass, the patient’s mean arterial blood pressure differences fluctuated between 15 and 20 mmHg from the left radial (Fig 3, red circle) to left femoral (Fig 3, white box) arteries. Epinephrine infusion at 0.04 mg/kg/min was started, with the following hemodynamic parameters: central venous pressure 6 mmHg, pulmonary artery pressure 26/13, cardiac output 6.1 L/min, cardiac index 3 L/min/m2, and mixed venous oxygen saturation 76%. When 20 mmHg pressure differences persisted between the left radial and left femoral arteries, the patency and calibration of arterial pressure monitoring systems were checked. Suspicion of a stent graft malfunction arose when no other plausible explanation was identified. Suspicion of a stent graft collapse then was assessed using TEE. A 2-dimensional cross-sectional view of the descending aorta short axis at the proximal descending aorta revealed stent graft collapse, with a 475% loss of diameter (Fig 4A: true lumen, red circle; false lumen, blue circle), just distal to the arch anastomosis. When compared with more a distal descending stent graft, full
Fig 4. Two-dimensional transesophageal echocardiography cross-sectional view of descending aorta short axis at proximal descending aorta showing stent graft collapse, 475% loss of diameter (A) (true lumen: red circle: false lumen: blue circle) just distal to arch anastomosis, compared with more distal descending stent graft with full expansion (B) (true lumen: red circle; false lumen: blue circle). Descending aorta long-axis view at proximal descending aorta shows stent graft severity of lumen compression with limited, high-velocity color-flow Doppler (C) (yellow) compared with more distal descending stent graft with laminar color-flow Doppler (D) (red).
4
expansion was seen (Fig 4B: true lumen, red circle; false lumen, blue circle). The descending aorta long-axis view at the proximal descending aorta showed stent graft severity of lumen compression with limited, high-velocity, color-flow Doppler (Fig 4C, yellow) stent graft with laminar color-flow Doppler (Fig 4D, red). Continuous-wave Doppler showed a peak gradient across the compressed area to be approximately 50 to 80 mmHg, except the angle of incidence was 420 degrees. To more reliably assess the severity of stent graft collapse, intravascular ultrasound (IVUS) and hemodynamic pressure gradient via pull back from the femoral artery catheter were considered, but patient stability and signs of adequate perfusion evidenced with relative normal arterial gas values (pH 7.36, HCO3– 22.8 mmol/L, base excess –1.9 mmol/L, lactate 2.4 mmol/L) influenced the decision to observe and follow up with contrast CT angiography. Follow-up CT imaging showed a compressed stent graft at the genu of the thoracic descending aorta with 490% loss of diameter (Fig 5, red arrow). Serial arterVial blood gas measures showed a pH 7.34 to 7.45; lactic acid of 1.9, initially trending down to 0.9; and a serum creatinine rise from 1.28 (baseline) to 1.91, with urine output maintained at 50 to 100 mL/hour with nesiritide and bumetanide infusions. Serial venous blood gas saturations were 65% to 77%, with the patient weaned from epinephrine infusion in the first 12 hours postoperatively, and the Swan cardiac index remained at 2.6 to 3 L/min/m2. The patient required mechanical ventilation for more than 24 hours after the first procedure due to signs of volume overload after multiple blood products were administered, with chest x-ray showing edematous vasculature and positive end-expiratory pressure of 10 mmHg required to maintain partial oxygen pressure 470 mmHg. The patient remained intubated and returned to the operating room 2 days later for diminishing distal pedal pulses. Thoracic endovascular aortic repair was performed for expansion of the stent graft collapse. At the beginning of the second procedure, TEE continuous-wave Doppler showed a gradient across the
Fig 5. Postoperative chest computed tomography angiography with sagittal view showing endovascular stent graft collapse (red arrow) at genu of descending aorta.
COLE ET AL
compressed area of approximately 80 to 100 mmHg. IVUS of the thoracic and abdominal aorta showed 90% compression of the stent graft diameter at the genu of the descending thoracic aorta compared with 2.8 cm at the distal descending thoracic aorta. After IVUS confirmation of stent graft collapse, a Zenith 38 152-mm nontapered stent (Cook Medical) was placed after the collapsed stent graft zone was dilated with a Coda balloon (Cook Medical). The in-stent device was deployed without event in zones 3 and 4 of the aorta, and the balloon was dilated in both proximal and distal aspects and the genu where the graft appeared to be collapsed. Follow-up IVUS and angiography demonstrated continued collapse but much improved flow. In addition, the patient’s femoral sheath waveform and his lower extremity Doppler signals were much improved after deployment of the in-stent device. The patient’s mean arterial pressure gradient remained 10 mmHg greater between radial and femoral artery lines; however, after deployment of the additional stent graft, a strong triphasic Doppler signal was demonstrated in both feet, and the patient had good urine output. The patient’s postoperative course worsened, with development of acute kidney injury, sepsis, atrial fibrillation with rapid ventricular rate, and respiratory insufficiency with acute respiratory distress syndrome. His respiratory status declined despite advanced mechanical ventilation, sedation, and a paralytic agent. A family decision for withdrawal of care was made 18 days after the initial surgery. DISCUSSION
This case report outlines a serious complication of endovascular stent graft collapse during an FET procedure. The stent graft collapse was suspected because of an invasive blood pressure discrepancy and was confirmed by TEE, which was instrumental in identifying the area of collapse and evaluating the flow using color-flow Doppler. Because of the angle of incidence, pressure gradients demonstrated on continuous-wave Doppler may not be as reliable. The relative risk of stent collapse during FET has yet to be defined. Several possible causes of collapse were proposed: the stent graft collapse may have resulted from a large mural thrombus at the proximal descending aorta, chronic nature of the dissection, and/or the acute angle of the arch to aortic descent. The phenomenon of endovascular stent graft collapse also has been described as a failure of stent conformation to the aorta wall (apposition), which can lead to endoleak and graft collapse.10 Another possible mechanism of collapse may involve thrombus or compression from surrounding false lumen. Results from the STARZ-TX2 (Study of Thoracic Aortic Aneurysm Repair with the Zenith TX2 TAA Endovascular Graft) clinical trial showed that kinking was noted in 1.6% (2 of 123) of patients at 12 months and compression was noted in 0.9% (1 of 108), but none of these 3 patients experienced adverse clinical sequelae or required a secondary intervention.11 Experience from using Gore’s TAG endovascular stent (Gore, Flagstaff, AZ) has shown that small-diameter aortas, difficult aortic arches, and poor sealing zones increased the risk
5
DETECTION OF DESCENDING AORTIC STENT GRAFT COLLAPSE
for stent graft complications.12 Canaud et al reported 4 cases of stent graft collapse among 285 patients treated with thoracic endovascular aortic repair, in which all 4 patients were treated with the TAG stent graft. Endovascular management was performed in 3 of the 4 patients; the other patient underwent stent graft explant and open repair. None of these patients died. A lack of device wall apposition and acute aortic arch angle (range 921-1181) were observed in all 4 patients.13 Although FET’s 1-stage method has lowered perioperative mortality for patients with extensive aortic disease, morbidity with the incidence of stroke remains a concern. Circulatory arrest time is regarded as an independent risk factor for stroke, despite the routine use of selective antegrade cerebral perfusion.14 Fortunately, the patient in this case report did not exhibit any physical signs of a cerebral vascular event, and intraoperative perfusion performance was optimized during circulatory arrest by continuous antegrade cerebral perfusion, cerebral oximetry, and bispectral index monitoring. The authors recognized that using bilateral radial artery pressures is beneficial for validating systemic pressures across the aortic arch; however, right arm circulation was reserved for antegrade cerebral perfusion, and the single left radial arterial tracing confirmed the absence of left subclavian artery stenosis after stent deployment. The large pressure discrepancy that persisted between left radial and left femoral artery readings prompted the authors to seek the possible causes. To assess distal central perfusion pressure, the authors relied on left femoral artery pressure. Other complications with FET include endoleak, cerebral spinal cord dysfunction, spinal cord infarction, and aortic rupture. During FET, thoracic endovascular stent graft collapse is an exceedingly rare event. One possible option would have been to extend the bypass phase and reopen the distal anastomosis to resolve the stent collapse under direct visualization. Open repair for chronic type-B dissection can be performed safely but is a relatively morbid surgery.15 Although results are encouraging from endovascular aortic repair of acute type-B dissection, chronic type-B dissections pose unique challenges.6 A multicenter review of collapsed thoracic endografts from 2005 to 2009 suggested that endograft collapse could be managed successfully using endovascular techniques in most
cases.16 Tadros et al described successful management of delayed (430 days) endograft-related complications using open conversion with device explantation. Another case report of delayed endovascular treatment of descending aortic stent graft collapse highlighted potential complications of visceral and lower limb ischemia but showed that management using additional endoprostheses had both good clinical and technical results.17 This case highlighted the valuable role of TEE in visualizing stent deployment and assessing overall stent graft function in real time after an FET procedure. TEE views of the proximal descending aorta and short and long axes are obtained easily. These views generate clear images of descending aortic stents and/or grafts and functional evaluation using color-flow Doppler. The risks for this patient’s stent graft collapse were a large mural thrombus at the proximal descending aorta, the chronic nature of dissection, and the acute angle of the arch to aortic descent. In this case, intraoperative decisions to examine clinically based signs of adequate perfusion and successful separation from cardiopulmonary bypass were made in recognition of considerable risk to the patient to undergo stent graft explantation and open repair. CONCLUSION
The FET procedure is a single-stage, hybrid surgical and endovascular procedure to repair thoracic aortic disease of the ascending aorta, aortic arch, and proximal descending aorta. FET has the benefit of being a single surgery but also has its risks. This is a case report of intraoperative endovascular stent graft collapse by compression at the proximal descending thoracic aorta. This first was suspected when vital sign monitoring showed a major difference between arterial blood pressures proximal and distal to the aortic arch after the patient was separated from cardiopulmonary bypass. Stent-graft compression was confirmed using TEE views of the proximal descending aorta that showed a loss of more than 75% of internal stent diameter, limited high-velocity color flow, and pressure gradient across the collapsed zone. TEE views were used to assess the effectiveness of balloon graft expansion with an additional endostent.
REFERENCES 1. Borst HG, Walterbusch G, Schaps D: Extensive aortic replacement using “elephant trunk” prosthesis. Thorac Cardiovasc Surg 31: 37-40, 1983 2. Dake MD, Miller DC, Semba CP, et al: Transluminal placement of endovascular stent-grafts for the treatment of descending thoracic aortic aneurysms. N Engl J Med 331:1729-1734, 1994 3. Kato M, Ohnishi K, Kaneko M, et al: New graft-implanting method for thoracic aortic aneurysm or dissection with a stented graft. Circulation 94:188-193, 1996 4. Buth J, Penn O, Tielbeek A, et al: Combined approach to stent-graft treatment of an aortic arch aneurysm. J Endovasc Surg 5:329-332, 1998 5. Karck M, Chavan A, Khaladj N, et al: The frozen elephant trunk technique for the treatment of extensive thoracic aortic aneurysms: Operative results and follow-up. Eur J Cardiothorac Surg 28:286-290, 2005 6. Pacini D, Tsagakis K, Jakob H, et al: The frozen elephant trunk for the treatment of chronic dissection of the thoracic aorta: A multicenter experience. Ann Thorac Surg 92:1663-1670, 2011
7. Ouzounian M, LeMaire S, Coselli J: Open aortic arch repair: State-of-the-art and future perspectives. Sem Thorac Cardiovasc Surg 25:107-115, 2013 8. Benedetto U, Melina G, Angeloni E, et al: Current results of open total arch replacement versus hybrid thoracic endovascular aortic repair for aortic arch aneurysm: A meta-analysis of comparative studies. J Thorac Cardiovasc Surg 145:305-306, 2013 9. Olsson C, Thelin S: Regional cerebral saturation monitoring with near-infrared spectroscopy during selective antegrade cerebral perfusion: Diagnostic performance and relationship to postoperative stroke. J Thorac Cardiovasc Surg 131:371-379, 2006 10. Lee WA, Martin TD, Hess PJ Jr, et al: First United States experience of the TX2 Pro-Form thoracic delivery system. J Vasc Surg 52:1459-1463, 2010 11. Matsumura JS, Cambria RP, Dake MD, et al: International controlled clinical trial of thoracic endovascular aneurysm repair with
6
the Zenith TX2 endovascular graft: 1-year results. J Vasc Surg 47: 247-257, 2008 12. Muhs BE, Balm R, White GH, et al: Anatomic factors associated with acute endograft collapse after Gore TAG treatment of thoracic aortic dissection or traumatic rupture. J Vasc Surg 45:655-661, 2007 13. Canaud L, Alric P, Desgranges P, et al: Factors favoring stentgraft collapse after thoracic endovascular aortic repair. J Thorac Cardiovasc Surg 139:1153-1157, 2010 14. Bachet J, Guilmet D, Goudot B, et al: Antegrade cerebral perfusion with cold blood: A 13-year experience. Ann Thorac Surg 67: 1874-1878, 1999
COLE ET AL
15. Roselli E: Thoracic endovascular aortic repair versus open surgery for type-B chronic dissection. J Thorac Cardiovasc Surg 149: S163-S167, 2015 16. Tadros R, Lipsitz E, Chaer R, et al: A multicenter experience of the management of collapsed thoracic endografts. J Vasc Surg 53: 1217-1222, 2011 17. Nano G, Mazzaccaro D, Malacrida G, et al: Delayed endovascular treatment of descending aorta stent graft collapse in a patient treated for post-traumatic aortic rupture: A case report. J Cardiothorac Surg 6:76-82, 2011