ENDOVASCULAR graft repair of aortic aneurysms is currently being tested as an alternative to traditional open aortic reconstruction. During endovascular repair, a metallic intravascular stent is combined with a prosthetic graft and is positioned in the aortic segment to bridge the aneurysm excluding it from the circulation. A balloon is inflated to deploy the stent portion of the device. Because balloon inflation and device deployment results in approximately 60 s of aortic occlusion, temporary asystole is helpful to prevent severe hypertension or distal device migration, which may result in aortic damage by device embolization. High doses of adenosine have been used to induce temporary asystole or high degree AV nodal blockade during the device deployment. Unfortunately, its duration of action is unpredictable, and return of normal sinus rhythm may occur before complete device deployment.* In this case series, we report the use of induced controlled ventricular fibrillation during endovascular thoracic aortic repair.
The patient was a 45-yr-old woman with a medical history significant for hypertension and Takayasu's arteritis. Twenty-five years before admission, the patient underwent an aortodistal subclavian bypass. Seven years before admission, a pseudoaneurysm of the descending thoracic aorta was diagnosed originating from the aortosubclavian graft. The patient underwent a second thoracotomy, and her pseudoaneurysm was closed. Four months before the current admission, she developed a recurrent pseudoaneurysm and was considered for endoluminal repair because of the anticipated complexity of another reoperation on the thoracic aorta.
On the day of surgery, the patient was premedicated with oral lorazepam. After entering the operating suite, Zoll multifunction external electrodes (Zoll, Burlington, MA) were applied under the left breast and posteriorly to the left of midline at the level of the scapula. Radial artery catheterization was performed, and general anesthesia was induced with propofol, fentanyl, and pancuronium. An Edwards Swan-Ganz thermodilution Paceport catheter (Irvine, CA) was placed via the right jugular vein, and a ventricular lead (Edwards Flex-Tip Transluminal V-pacing probe) was properly positioned. Fluoroscopy confirmed that external defibrillator pad placement did not interfere with aortic arch imaging. The ventricular leads were connected via standard temporary transvenous pacemaker cables to an alternating current transformer (Medtronics, Minneapolis, MN). A 10-s trial of ventricular fibrillation was induced by activation of the alternating current transformer, and the patient was successfully defibrillated by discontinuation of the alternating current and administration of 1 200-J shock via the external pads. Baseline hemodynamics rapidly returned to baseline levels. After successful induction of ventricular fibrillation and defibrillation, the patient was actively cooled using a cooling blanket to 34 [degree sign] Celsius during surgical preparation.
An endovascular stent graft was fabricated from a large Palmaz balloon expandable stent and a woven Dacron vascular graft (Figure 1). The endograft was premounted on a 35-mm angioplasty balloon and inserted over a guidewire, via an open right transfemoral route. Fluoroscopy and transesophageal echocardiography (TEE) guidance allowed precise positioning of the device across the orifice of the pseudoaneurysm.
After confirmation of precise stent placement, the position of the ventricular lead was reconfirmed by the presence of 1:1 capture during ventricular pacing. The ventricular lead was reattached to the alternating current source, and the generator was activated. Ventricular fibrillation was instituted and was confirmed by a flat intraarterial pressure tracing (Figure 2). Over the next 90 s, the angioplasty balloon was inflated, the stent was deployed and preliminarily confirmed to cross and exclude the pseudoaneurysm by fluoroscopy and TEE. After deflation of the angioplasty balloon, the alternating current was discontinued, and the patient was defibrillated using 200 J of current. Because of hypertension (MAP 120) after defibrillation, sodium nitroprusside, 40 micro gram, was administered as a single bolus, and the patient's hemodynamic returned to baseline values. After exclusion of flow into the pseudoaneurysm by the stent was confirmed by TEE and by angiography, the right femoral artery access cut down was repaired. The patient was extubated 3 h after the conclusion of surgery and was discharged home 3 days after surgery.
The patient was a 52-yr-old man with a medical history significant for hypertension and a previous repair of a penetrating injury to the aortic arch presenting for an endovascular repair of an aortic arch pseudoaneurysm. On the day of surgery, a radial artery and a Paceport pulmonary artery catheter were inserted, and anesthesia was induced with fentanyl, thiopental, and pancuronium. After positioning the patient, a 10-s trial of ventricular fibrillation was induced, and the patient was successfully defibrillated using 200 J.
During surgical preparation and preliminary stent-graft placement, the patient was actively cooled to 34 [degree sign] Celsius. During the surgical procedure, the patient required five periods of induced ventricular fibrillation lasting 55–90 s with 8–17 min between periods of induced fibrillation. The patient was successfully defibrillated with 200 J after each period of ventricular fibrillation. The patient was consistently hypertensive after each defibrillation and required sodium nitroprusside, 20–40 micro gram, boluses. The patient was extubated 2 h after the surgical procedure and was discharged home 3 days after surgery.
The patient was a 75-yr-old woman with a medical history significant for hypertension and end-stage renal disease, who presented for endovascular stent-graft repair of a ruptured descending thoracic aneurysm. After placement of a radial artery and a Paceport pulmonary artery catheter, anesthesia was induced with fentanyl, etomidate, and pancuronium. Before incision, a 10-s trial of ventricular fibrillation was induced, and the patient was successfully defibrillated using 200 J.
During the surgical procedure, the patient required one period of induced ventricular fibrillation lasting 172 s. The patient was successfully defibrillated with 200 J. Immediately after defibrillation, the patient was hypotensive, but responded immediately to epinephrine, 16 micro gram. Technical difficulties with the stent-graft device necessitated surgical conversion to an open thoracoabdominal aortic reconstruction, which was successfully performed with a 26-mm Dacron graft.
Endovascular stent graft repair of abdominal aortic aneurysms was first suggested by Dotter in 1969 and reached clinical application with the work of Parodi et al. in 1991. Endovascular grafts have the distinct advantage of being a less invasive technique compared with conventional arterial reconstructions, resulting from the unique ability to insert these grafts from remote access sites.
When balloon expandable stents are used to anchor the endovascular conduit to the underlying arterial wall, distal migration of the device may occur during deployment as a result of the propulsive pressure of systole. This problem may be managed by the administration of adenosine to induced asystole, although the results of adenosine often are unpredictable. Early return of normal sinus rhythm after adenosine administration may result in inadequate time for device deployment and balloon deflation. Concern is greatest with intrathoracic deployment when temporary complete aortic occlusion during device deployment can exert the greatest propulsive force on the incompletely deployed aortic endograft. Although a large dose of adenosine would be safe in the presence of a ventricular pacing lead, this approach was not chosen because we believed that it was unlikely to provide the necessary period of asystole. Furthermore, if the period of adenosine-induced asystole was too short, damage may result before other actions may be taken.
An alternative technique to induce temporary asystole for endograft deployment uses ventricular fibrillation. Ventricular fibrillation may be induced by either rapid ramp pacing or by applying alternating current to the endocardial surfacing using an AC fibrillator via transvenous pacing electrodes. In the present case, an alternating current source was chosen because of the ease of its use and the ability to control the duration of ventricular fibrillation.
Previous studies have examined the safety of induced ventricular fibrillation in the perioperative period during implantable cardiac defibrillator (ICD) implantation. Repeated induction of ventricular fibrillation and defibrillation during ICD implantation resulted in no detectable differences in either cardiac index or stroke volume. In echocardiographic studies, no differences in ejection fraction or left ventricular end diastolic volume were detected after successful repeated induction and reversal of ventricular fibrillation. 
Ventricular fibrillation may result in cerebral ischemia. During ICD testing, ischemia related changes in electroencephalography (EEG) occurred approximately 8 s after induction of ventricular fibrillation and lasted 64 +/- 49 s. Adams et al. confirmed the appearance of these ischemia-related EEG changes, although no patients exhibited a new neurologic deficit, exacerbation of previous deficits, or changes in neuropsychometric performance. 
Because the period required for complete device deployment was unknown, intentional mild hypothermia was induced to provide cerebral and myocardial protection. Mild hypothermia has conferred significant cerebral protection in many experimental preparations. Hypothermia may also provide some degree of myocardial protection during ventricular fibrillation. In the present case reports, these beneficial effects of hypothermia were believed to outweigh the deleterious effects, which include the possible increased incidence of perioperative morbid cardiac events or wound infections. [10,11]
In this case series, we describe the use of induced ventricular fibrillation for temporary cardiac arrest during endovascular repair of thoracic aortic aneurysms. This technique has the advantage of predictability of cardiac arrest time over high-dose adenosine necessary for endovascular repair and the advantage of ease of induction of ventricular fibrillation over electrophysiologic stimulation. Further work is necessary to delineate the use of induced ventricular fibrillation during endovascular repair of the aorta.
*Our previous unpublished experience with abdominal aortic endograft-stent repairs.