Automatic Implantable Cardioverter-Defibrillator: Techniques of Implantation and Results Richard J. Thurer, M.D., Richard M. Luceri, M.D., and Hooshang Bolooki, M.D. ABSTRACT An initial experience with use of the automatic implantable cardioverter-defibrillator (AICD) is described. Twelve patients received the device. One death has occurred during a mean follow-up of 15 months, and it was due to causes other than arrhythmias. Appropriate device discharge terminating a malignant arrhythmia occurred in 9 patients (75%).The observed survival (92%)far exceeds that to be expected in survivors of sudden death treated by conventional means. There have been no operative deaths. Morbidity has been minimal, although three reoperations were required in 2 patients because of lead dislodgment. The AICD has been demonstrated to be effective in treating patients at risk for sudden arrhythmic death. It can be employed safely with minimum morbidity using a variety of implantation techniques. The automatic implantable cardioverter-defibrillator (AICD) has been developed for the treatment of patients who survive an episode of sudden cardiac death and fail drug therapy [l].The device currently undergoing clinical evaluation is designed to sense episodes of ventricular tachycardia or ventricular fibrillation and deliver a shock of 25 to 30 J across the ventricular muscle mass. Early studies  indicated that the device reliably senses malignant tachyarrhythmias and is effective in causing their termination. This has resulted in a marked reduction in deaths due to arrhythmias in patients in whom the device has been implanted compared with historical controls treated by conventional means using drugs . In addition, overall survival in this group of patients is improved. Stringent criteria for patient selection have been followed. Initially, patients were considered for implantation of the device only after two episodes of sudden cardiac death not associated with acute myocardial infarction and expected drug failure on the basis of electrophysiological studies. More recently, indications have been liberalized and the device is now considered for use in patients who have survived one episode of sudden cardiac death and who either fail electrophysiologically guided antiarrhythmic drug therapy or whose arFrom the Department of Surgery, Division of Thoracic and Cardiovascular Surgery, and the Department of Medicine, Division of Cardiology, University of Miami School of Medicine, Miami, FL. Presented at the Thirty-second Annual Meeting of the Southern Thoracic Surgical Association, Boca Raton, FL, Nov 7-9, 1985. Address reprint requests to Dr.Thurer, Division of Thoracic and Cardiovascular Surgery (R-114), University of Miami School of Medicine, PO Box 016960, Miami, FL 33101.
143 Ann Thorac Surg 42:143-147, Aug 1986
rhythmia does not allow satisfactory evaluation of drug therapy. Patients in this latter group generally have an arrhythmia that is not inducible in the electrophysiology laboratory. Many methods have been used in an attempt to prevent recurrent cardiac death. They include drug therapy, pacemaker devices, direct surgical procedures such as endocardial resection and encircling endocardial ventriculotomy with and without coronary artery bypass grafting and aneurysmectomy , and the AICD. We believe that the AICD is an effective method in properly selected patients, and report here our initial experience with 12 patients undergoing implantation of this device without other cardiac surgical procedures. Despite this relatively limited experience, we have come to prefer certain implantation techniques in particular clinical situations. Intraoperative testing, follow-up, and pacemaker interactions with the defibrillator will be discussed.
Material and Methods This study was reviewed by the Human Studies Committee of the University of Miami School of Medicine and approved on July 11, 1983. The Device The AICD was developed by Mirowski and originally manufactured by INTEC Systems, recently acquired by Cardiac Pacemakers, Inc., St. Paul, MN . The unit weighs 290 gm and contains lithium batteries capable of charging two capacitors to approximately 700 V in 7 to 9 seconds. The detection system requires that two specific criteria be satisfied. One is ventricular rate, and the other is termed the probability density function, which refers to the time the ECG signal recorded is away from the baseline. In ventricular fibrillation, the signal spends little time at baseline. The unit, therefore, recognizes the rhythm to be one that triggers the defibrillation sequence. One version of the device (not used in this study) employs only rate to trigger it. Two bipolar electrode systems are used. One serves both for defibrillation and sensing the electrical waveform, and consists of a patch over the left ventricle and either a spring electrode placed at the junction between the superior vena cava (SVC) and right atrium or a patch over the right atrium or right ventricle. The second electrode pair senses rate and serves for R wave synchronization during device discharge. This electrode may be a bipolar transvenous right ventricular (RV) lead or two epicardial electrodes placed 1 to 2 cm apart. The output sequence requires 5 to 15 seconds to diagnose the arrhythmia and an additional 5 to 15 seconds to
144 The Annals of Thoracic Surgery
Vol 42 No 2 August 1986
charge the output capacitors. A shock of approximately 25 J is delivered between 15 and 30 seconds following onset of the arrhythmia. Should the initial shock not convert the arrhythmia, a second shock of 30 J will be given after another 10 to 30 seconds. A third and fourth shock of 30 J will be delivered if the arrhythmia persists. The unit will give no further shocks unless it recognizes a rhythm other than ventricular tachycardia or fibrillation for 30 seconds. The system can be monitored following implantation. Using a magnet and detector (AID Check-B, INTEC Systems), the number of delivered pulses and the charging time of the capacitors can be measured. The charging time is a reflection of battery life. The device can be activated or inactivated by magnet application. An audible signal indicates the status of the device. Expected life of the device is three years or a total of 100 discharges 161.
Patient Population Twelve patients underwent implantation of the AICD from August 25, 1983, to March 6, 1985. There were 8 men and 4 women. Age at implantation ranged from 45 to 72 years (mean, 63 years). In 10 patients the underlying cause was atherosclerotic cardiovascular disease. Patients with coronary artery disease were not considered candidates for revascularization by virtue of controlled angina, noncritical lesions, or exceedingly poor ventricular function. Two patients were thought to have congestive cardiomyopathy. Both had normal coronary arteries on angiography . Patient selection has been rigidly controlled. All patients experienced at least one cardiac arrest due to ventricular tachyarrhythmia or had recurrent life-threatening ventricular tachycardia. Patients with inducible ventricular arrhythmias at baseline electrophysiological study all failed serial studies of drug efficacy. All patients had failed at least three drugs alone or in combination. In 3 patients, all women, cardiac arrest had occurred while they were on a regimen of amiodarone therapy. All patients underwent complete cardiac catheterization. One patient had undergone a coronary bypass procedure ten years prior to AICD implantation. Restudy indicated patent grafts. All other patients had either normal coronary arteries or anatomy and left ventricular function, and were not thought suitable for coronary bypass grafting, aneurysm resection, or other direct surgical approaches. As a group, left ventricular function was poor with a mean ejection fraction of 32% (range, 15 to 45%). In the first 5 patients the initially described method was used, that is, transvenous placement of the SVC spring electrode and RV sensing electrode along with epicardial placement of the left ventricular (LV) patch electrode through a left thoracotomy [7, 81. In 5 other patients, two patches were used. The approach was through a left thoracotomy with intrapericardial placement of the two patches and a transvenous RV sensing
lead. Median sternotomy and an entirely epicardial approach were used in 2 patients . Pacemakers also were required in 4 patients. Two patients had received pacemakers prior to defibrillator implantation; one was a transvenous VVI unit that was not functioning properly because of lead placement in the coronary sinus and one, a unipolar DDD unit. Both patients had sick sinus syndrome. One patient required a pacemaker postoperatively for sick sinus syndrome. A unipolar W I unit was placed elsewhere 4 months following AICD implantation. One patient required pacemaker implantation at the time of defibrillator implantation. Following discharge of the defibrillator, an adequate escape rhythm was not present. A transvenous bipolar VVI unit was implanted. A tined lead (Cordis 327-563) was positioned in a more superior RV position than the sensing lead of the AICD and the pulse generator (Cordis Multicor gamma 336A) placed in the right infraclavicular region. In 1 patient, two additional epicardial pacing leads were placed on the diaphragmatic surface of the heart to establish bipolar VVI pacing. All epicardial leads were the same type (Medtronic 4951).
In traoperative Testing Thorough intraoperative testing is necessary to be certain that any potential adverse interactions will not occur. Following placement of required leads and patches, ventricular tachycardia or fibrillation or both is induced, usually with ramp pacing of the right ventricle at 20 mA. Defibrillation thresholds are measured using the INTEC external cardioverter-defibrillator, which is adjustable for energy output from 1 to 40 J. A final test is done with the implantable device after induction of ventricular fibrillation to ensure proper function postoperatively. From 3 to 16 (mean, 7) arrhythmia inductions were performed in each patient. Occasionally movement of the patch or SVC spring electrodes was necessary to achieve acceptable defibrillation thresholds. On two occasions the SVC electrode was replaced with a patch over the right atrium because of high defibrillation thresholds with the SVC spring electrode-LV patch configuration. Clinical intraoperative and postoperative data are summarized in the Table.
Pacemaker-Defibrillator Interactions Concomitant pacing was necessary in 4 patients. Certain problems can occur as a result of interactions between the pacemaker and the defibrillator unit. Interactions can take a variety of forms. With a tachyarrhythmia, for instance, the pacer could fail to sense, and continue firing in the asynchronous mode. The AICD could ignore the tachyarrhythmia in favor of the regular pacing impulse and fail to shock the patient. Double counting for both pacing impulse and the evoked ventricular depolarization can result in defibrillator discharge during paced rhythm, since the rate requirement of the AICD is exceeded. The use of dual-chambered devices can easily
145 Thurer, Luceri, Bolooki: Automatic Implantable Cardioverter-Defibrillator
Summary of Patient Data ~
Patient No., Age (yr), Sex
Associated Rhythm Disturbances
VT or VF
Induction in OR
Follow-up First Shock (mo)
Spring and patch, transvenous sensing Spring and patch, transvenous sensing Spring and patch, transvenous sensing
1. 45, F
Transvenous, L thoracotomy
2. 64, M
Transvenous, L thoracotomy
3. 50, M
Sick sinus syndrome
Transvenous, L thoracotomy
4. 70, F
Transvenous, S thoracotomy
Spring and patch, transvenous sensing
5. 70, M
Asystole after shock
Transvenous, L thoracotomy
6. 58, F
7. 72, M
Sick sinus syndrome, previous VVI lead in coronary sinus ...
8. 63, M
Transvenous, L thoracotomy
9. 69, F
Transvenous, S thoracotomy
10. 49, M
Transvenous, L thoracotomy
11. 65, M
Transvenous, L thoracotomy
12. 70, M
Spring and patch, transvenous sensing, transvenous pacing (pacemaker) Double patch, epicardial sensing, epicardial pacing Double patch,b transvenous sensing Double patch,b transvenous sensing Double patch, transvenous sensing Double patch, transvenous sensing Double patch, transvenous sensing Double patch, epicardial sensing
Transvenous, L thoracotomy
Sick sinus syndrome, previous DDD unipolar unit
Unipolar VVI pacer 4 mo after implantation; died 5 mo after implantation of acute MI Reoperation x 2 sensing lead and SVC spring repositioned, pulse generator relocated Reoperation: SVC spring repositioned
'Spring and patch means a spring electrode at the junction between the superior vena cava and right atrium, and a left ventricular patch; double patch means either a right atrial or right ventricular patch and a left ventricular patch. bSpring and patch was attempted initially. EF = ejection fraction; VT = ventricular tachycardia; VF cardiovascular disease; MI = myocardial infarction.
ventricular fibrillation; OR
operating room; L
146 The Annals of Thoracic Surgery Vol 42 No 2 August 1986
D 0 C UME NT E D A P P R O P R I A T E OR HIGHLY PROBABLE (58%)
P R O B A B L E OR U N D O C U M E N T E D
Fig 1. Results after implantation of the automatic cardioverterdefibrillator. Shocks, both appropriate and inappropriate, are shown. (LV = left ventricular.)
result in exceeding the rate requirement of the AICD during dual-chambered pacing at relatively low rates. These problems are potentially more common with unipolar pacing because of the magnitude of the pacemaker stimulus artifact. Accordingly, multiprogrammable bipolar systems are preferred. Sensing leads of the defibrillator should be placed at some distance from pacing leads. When epicardial leads are used, the defibrillator sensing leads should be placed close together (1 to 2 cm apart) and at a distance from the pacing electrodes. The pulse generator itself must be placed at a distance from the AICD. A magnet or magnetic programming wand must not be applied near the defibrillator during pacemaker testing, since the defibrillator may inadvertently be deactivated.
DEATU ABORTED DEATH
E X P E C T E D SURVIVAL (25%)
T= MONTHS (I%= 1 6 )
Fig 2. Comparison between actual and expected survival based on postoperative shocks for malignant arrhythmia. The points on the expected survival curve indicate the first appropriate shock after implantation in response to a potentially lethal arrhythmia. This would be the survival curve if the automatic implantable cardioverterdefibrillator had not converted the arrhythmia and the patient had not survived the event.
Results There were no operative deaths. One late death occurred at 5 months. The patient sustained an acute myocardial infarction and died in electromechanical dissociation. He was monitored at the time of death in an emergency room. The AICD did not discharge and functioned normally. No discharges had occurred from the time of implantation to the time of death. Follow-up was documented in all patients, and the data reflect follow-up to October 1, 1985. Follow-up ranged from 5 months (the 1 death) to 25 months with a mean follow-up of more than 15 months. Patient survival was 92% with a mean follow-up of 15 months (Fig 1). Expected survival in this group of patients would have been 25% if we assume that death would have occurred at the time of the first appropriate device discharge for ventricular tachycardia or fibrillation (Fig 2). Appropriate discharges occurred in 9 of the 12 patients, and were documented as appropriate or highly probable in 7 and probable in 2. One documented inappropriate shock occurred as a result of dislodgment of the sensing lead. Reoperation was required on three occasions in 2 patients for sensing lead dislodgment once and for SVC spring electrode dislodgment twice (see Table).
Bipolar sensing lead dislodgment because of “twiddler’s syndrome” [lo] was noted on one occasion. The patient was seen with multiple inappropriate shocks presumably caused by myopotential sensing of pectoral muscle activity when the lead tip was in the subclavian vein. A nontined lead was replaced with a tined lead, which was well fixed to the pectoral fascia. The patient later was noted to have the SVC spring electrode in the subclavian vein, and it could not be advanced simply. A new lead was placed through the right subclavian vein. The pulse generator was repositioned at this time because of imminent skin erosion in this thin woman. Another patient was noted on follow-up to have the SVC spring electrode in the subclavian vein. In this patient, simple advancement of the lead into the SVC was possible. There were no wound infections. No incidence of skin erosion was noted despite the large size of the device and the small size of most of the female patients. In 1 patient, device repositioning was performed in conjunction with lead replacement because of thinning of the skin. Acceptance of the device was generally good. Most patients felt reassured by its presence even if it had not
147 Thurer, Luceri, Bolooki: Automatic Implantable Cardioverter-Defibrillator
yet discharged. Device discharge is clearly unpleasant but not particularly painful. Patients who have experienced external cardioversion state that AICD discharge is considerably less uncomfortable. One patient likened the device discharge to being kicked in the chest “but from the inside.” Most patients in our series have not lost consciousness prior to being shocked and therefore brace themselves for the sensation.
Comment The improved survival in this group of selected patients receiving the AICD device compared with any similar group not receiving the device is dramatic. The mortality in comparable patients with an incomplete response to amiodarone may reach 30 to 40% at one year. The single death in the follow-up period was not due to arrhythmia but did reflect the severe underlying atherosclerotic disease in that patient. Indeed, if deaths due to arrhythmia are removed from the group of patients with previous cardiac arrest and atherosclerotic disease or cardiomyopathy, the pattern of end-stage disease is likely to be markedly changed. Despite the severe nature of the underlying cardiac disease in these patients, which is indicated by impaired ejection fraction and electrical instability, implantation of the device can be safely accomplished. Although thoracotomy or sternotomy was performed and multiple inductions of lethal arrhythmias were required, no operative deaths or major morbidity were encountered. Despite the large size of the device, the multiple incisions required, and the small size of a number of the elderly female patients, no instance of infection or erosion was noted. We believe that this device, although only in its initial stages of technical development, is highly effective in treating recurrent ventricular tachycardia and ventricular fibrillation. On the basis of this experience, we prefer median sternotomy with epicardial placement of the required leads and patches. We believe the two-patch configuration is superior to the SVC spring-LV patch configuration. In 2 patients, the SVC spring-LV patch configuration was abandoned intraoperatively in favor of the RA patch-LV patch scheme because of unacceptably high defibrillation thresholds. The two-patch configuration can easily be accomplished through the limited left anterior thoracotomy approach when the pericardium is relatively free from adhesions. The median sternotomy approach with epicardial placement was initially used in a patient with a poorly functioning pacing lead that could not be removed safely. With the initially described approach, this patient would have had three intracardiac leads: the abandoned pacing lead, the new pacing lead, and the SVC spring electrode. This patient and subsequent patients tolerated median sternotomy quite well. The epicardial approach avoids the need for fluoros-
copy and additional incisions and tunnels. Also, the potential for lead dislodgment, seen on three occasions, is avoided. Therefore, at present, we use the median sternotomy approach with epicardial placement in patients who have not undergone previous cardiac operation. In patients operated on previously, we use the originally described method with limited left thoracotomy, placement of the LV patch outside the pericardium, and placement of the SVC spring and RV sensing leads by the transvenous approach. We have not used the subxiphoid approach [ll, 121 largely because of the requirement to move the patches in a number of patients to achieve low defibrillation thresholds. We believe our ability to do this would be limited with this approach. The AICD has been demonstrated to be effective in treating selected patients at risk for sudden arrhythmic death. It can be employed safely with minimum morbidity using a variety of implantation techniques.
References 1. Mirowski M, Mower MM, Staewen WS, et al: Standby automatic defibrillator: an approach to prevention of sudden coronary death. Arch Intern Med 126:158, 1970 2. Mirowski M, Reid PR, Mower MM, et al: Termination of malignant ventricular arrhythmias with an implanted automatic defibrillator in human beings. N Engl J Med 303:322, 1980 3. Reid PR, Mirowski M, Mower MM, et al: Clinical evaluation of the internal automatic cardioverter-defibrillator in survivors of sudden cardiac death. Am J Cardiol51:1608, 1983 4. Battaglini J, Wendt D, LaPenna W, et al: Surgical prevention of sudden death. Clin Prog Pacing Electrophysiol 3:105, 1985 5. Mower MM, Reid PR, Watkins L Jr, et al: Automatic implantable cardioverter-defibrillator: structural characteristics. PACE 7:Part 2:1331, 1984 6. Reid PR: The automatic implantable defibrillator: current status and future development. In Barold SS (ed): Modern Cardiac Pacing. Mt. Kisco, NY, Futura, 1985, pp 745-757 7. Watkins L Jr, Mirowski M, Mower MM, et al: Automatic defibrillation in man: the initial surgical experience. J Thorac Cardiovasc Surg 82:492, 1981 8. Watkins L Jr, Mirowski M, Reid PR, et al: Surgical techniques for implanting the automatic implantable defibrillator. PACE 7:Part 2:1357, 1984 9. Brodman R, Fisher JD, Furman S, et al: Implantation of automatic cardioverter-defibrillators via median sternotomy. PACE 7:Part 2:1363, 1984 10. Bayliss CE, Beanlands DS, Baird RJ: The pacemaker twiddler’s syndrome: a new complication of implantable transvenous pacemakers. Can Med Assoc J 99:371, 1968 11. Watkins L Jr, Mirowski M, Mower MM, et al: Implantation of the automatic defibrillator: the subxiphoid approach. Ann Thorac Surg 34:515, 1982 12. Lawrie GM, Griffin JC, Wyndham CRC: Epicardial implantation of the automatic implantable defibrillator by left subcostal thoracotomy. PACE 7:Part 2:1370, 1984