Ann Thorac Surg 1997;64:1702-1706
© 1997 The Society of Thoracic Surgeons
Original Articles: Cardiovascular
Management of Infected Implantable Cardiac Defibrillators
Louis E. Samuels, MD,
Fania L. Samuels, MD,
Marla S. Kaufman, BA,
Rohinton J. Morris, MD,
Stanley K. Brockman, MD
Departments of Cardiothoracic Surgery and Medicine, Allegheny University Hospitals, Hahnemann Division, Philadelphia, Pennsylvania
Accepted for publication June 5, 1997.
 |
Abstract
|
|---|
Background. The implantable cardiac defibrillator (ICD) was introduced clinically in 1980 for the management of ventricular arrhythmias.
Methods. From January 31, 1989, through May 29, 1996, 329 ICD devices were implanted at Allegheny University Hospital, Hahnemann Division, Philadelphia, Pennsylvania. All device-related infections were examined.
Results. Fifteen patients (5%) experienced infection of the generator component of the ICD. There were 14 male and 1 female patients with a mean age of 62 years (range, 38 to 79 years). All infections involved the generator with or without other component involvement. Complete removal of the system was performed in 7 patients, partial removal in 5, and the entire system was left intact in 3. In 4 patients (27%), further procedures were performed to remove additional infection. Three patients (20%) died during the hospital stay.
Conclusions. Infection of ICD devices is a devastating event. We favor complete removal of the ICD generator and all the components when possible. Partial removal of the ICD unit (ie, generator only) is reserved for patients in whom the risk of complete removal is too high and infection is confined to the generator only.
|
Introduction
|
|---|
The clinical introduction of automatic implantable cardiac defibrillators (ICD) in 1980 [1] heralded the beginning of a technology that has revolutionized and supplemented the management of ventricular arrhythmias. The number of ICD devices inserted during the past decade and a half has continued to increase. The technologic advancements in lead design and generator function have made a significant impact on insertion techniques. Early generation systems required sternotomy or thoracotomy for direct access to the heart to place defibrillating patches and screw-in sensing leads. The cables from these components were attached to a defibrillator unit positioned in the subxiphoid or subcostal region. With the development of endovascular leads, a transmediastinal or transthoracic approach was replaced by a transvenous one. Finally, advances in microcircuitry and battery design permitted the development of a generator that was small enough to be placed in the subpectoral position. Most recently, infraclavicular subcutaneous implantation has been possible.
Despite the improvements in ICD design and insertion, complications related to the system still occur. Infection of ICD components is no exception. In an effort to more fully understand ICD infection, we reviewed the medical records of all patients in whom this event occurred. We recommend a strategy for management based on this experience.
|
Material and Methods
|
|---|
From January 31, 1989, through May 29, 1996, 329 ICD devices were implanted at Allegheny University Hospital, Hahnemann Division, Philadelphia, Pennsylvania. Fifteen patients experienced infections of the ICD generator. The indications for ICD insertion were ventricular arrhythmia refractory to medical therapy. The techniques of insertion included transmediastinal or transthoracic placement of defibrillating patches and sensing leads with subcostal location of the generator, or transvenous placement of defibrillating and sensing leads with either subcostal or subpectoral generator insertion. Management of the infected device involved either complete removal of the unit, partial removal of various components of the unit, relocation of the same unit, or drainage alone. In 1 case, cardiopulmonary bypass was required to remove the posterior patch and sensing leads. The majority of procedures were performed in the operating room under general anesthesia for initial implants and implants performed with coronary artery bypass grafting. Generator changes were performed in the electrophysiology laboratory with local anesthesia and intravenous sedation. The surgical procedures to treat infected ICD devices were performed in the operating room. A detailed analysis of the medical records and operative reports comprising all admissions of the involved patients was performed. Outpatient or telephone interviews were conducted for follow-up.
|
Results
|
|---|
From January 31, 1989, through May 29, 1996, 329 ICD devices were implanted at Allegheny University Hospital, Hahnemann Division, Philadelphia, PA. Fifteen patients (5%) experienced infection of the device (generator, lead, or patch). There were 14 men and 1 woman with a mean age of 62 years (range, 38 to 79 years). Eleven patients presented with infection within 1 year of implantation (mean, 32 days; range, 8 days to 9 months) and 4 patients presented after 1 year (mean, 18 months; range, 13 months to 2 years). Ten devices were first-time implants. Five devices were generator changes. Eight devices were placed without concomitant procedures. Seven devices were placed at the time of open heart operations (eg, coronary artery bypass grafting). The approach to insertion was transmediastinal in 7 units, transvenous in 4, transthoracic in 3, and subxiphoid in 1. There were 5 Staphylococcus epidermidis infections, 3 Staphylococcus aureus, 3 combined (Escherichia coli/S aureus, Serratia marcesans/S aureus, S epidermidis/S aureus), 1 ß-hemolytic streptococcus, and 1 Klebsiella pneumoniae. Cultures from 2 patients failed to grow organisms (Table 1
). Complete removal of the system (generator, leads, patches) was performed in 7 patients, partial removal (ie, generator only) in 5, and the entire system was left intact in 3. In 4 (27%) patients, further procedures were performed to remove additional infected components of the unit. After management of the infection, new devices were reimplanted in 5 patients. Seven patients did not have new devices reimplanted. In 3 patients, the original device was left intact. Three patients (20%) died during the hospital stay. The causes of death were sepsis, ventricular arrhythmia, and heart failure (Table 2). No further infectious complications occurred in the remaining 12 patients who were discharged. Follow-up was complete. Nine of the patients were alive at a mean of 21 months (range, 4 to 66 months). There were three deaths (one lymphoma, two congestive heart failure) at a mean of 23 months (range, 1 to 58 months).
|
Comment
|
|---|
The first clinical internal defibrillation was performed by Beck and coworkers in 1947 [2]. The first transthoracic direct current defibrillation was by Zoll and associates in 1956 [3]. Experimental work by Mirowski and colleagues [4] and Shuder and associates [5] on termination of ventricular fibrillation with an automatic and implantable defibrillator were conducted in the 1970s. An implantable cardiac defibrillator for clinical use was first performed by Mirowski and coworkers in 1980 [1]. For the past 15 years, an escalating number of ICD devices have been implanted across the country. Techniques of insertion have varied according to the era in which the device was implanted, the technology available at the time, and the preference of the surgeon and electrophysiologist. Despite improvements in technologic design and simplicity of insertion, complications continue to occur. Device-related infection is no exception. Depending on the system implanted, the degree of infection, and the approach of the surgeon, various management styles have been described. In general, management has been tailored to the individual case. The main issues are whether total or partial removal of the ICD system is necessary and whether the risk of total removal in certain circumstances is too great.
In 1981, Choo and associates [6] described the superiority of total removal of infected epicardial pacemakers compared with nonsurgical or partial removal. In 1984, Mirowski and others [7] reported primary infectious complications associated with 6% of implanted ICD devices. They advocated total removal. Marchlinski and colleagues [8] in 1986 found a similar incidence and also advised total removal. Although total removal represented the most definitive method by which to eradicate all potentially infected foreign material, the challenge of removing all hardware from the heart was not without risk. Dissection of the epicardial patches, particularly the posterior patch, and sensing leads is difficult because of the extensive adhesions between the patch and the epicardium. In some situations, cardiopulmonary bypass is necessary, thereby increasing the complexity and risk of the procedure. Furthermore, in a number of cases, the epicardial components of the system were not contaminated. As such, several investigators [9–11] attempted to determine preoperatively whether material beyond the generator was infected.
In 1988, Kelly and colleagues [9] reported the value of gallium scanning in identifying infected ICD units. In 3 patients, they [9] showed that gallium uptake at the generator and along the path of the electrodes up to the heart represented spread of infection in this territory. Thus, total removal was necessary. In the same year, Almassi and coworkers [10] described the computed tomographic (CT) scan findings of infected ICD devices and proposed a uniform method of management consisting of complete removal with povidone-iodine irrigation and intravenous antibiotics. They argued that a fluid density between the heart and an ICD patch, particularly in the presence of a previously normal CT appearance, indicated that a local infection was likely. The radiographic and CT scan findings of infected ICD devices were reinforced by Goodman and associates [11] in 1989. Distortion or crumpling of the patches on CT scan beyond the fourth postoperative week were confirmatory radiographic findings in patients with clinically suspected infected ICD devices in their study [11]. In our opinion, these studies [9–11] are valuable when positive findings are present. However, the absence of gallium uptake along the electrodes or around the heart, absence of fluid around the heart on CT scan, or absence of crumpling of patches on plain roentgenography does not eliminate the possibility of underlying infection at these locations. The converse is also true. The presence of these findings may represent postoperative or inflammatory changes that are noninfectious. Therefore, conclusions from radiographic findings must be interpreted within the clinical context of the case.
The idea that the infection may not involve the entire ICD unit led some investigators [12–14] to perform partial removal. Although we have been taught that all foreign material must be removed in the face of infection, the risk of complete removal, especially the posterior patches, was so great that partial removal was performed. In 1990, Taylor and colleagues [12] reported a patient in whom the ICD generator was salvaged by continuous irrigation of the infected pocket with antibiotics. Wunderly and others [13] described removal of the infected generator, resterilization of the unit with ethylene oxide, and reinsertion after 2 weeks of intravenous antibiotics followed by 6 weeks of additional antibiotics. In 1993, Gupta and associates [14] described a unique application of the rectus abdominis muscle flap for generator pocket infections. Provided that the leads were not involved with infection, salvage of the ICD generator was made possible by wrapping the unit in the muscle flap. Contrary to these studies, Siclari and coworkers [15] argued that complete removal was mandatory. In 8 patients with infected ICD devices, local treatment alone was unsuccessful in all but one. Local irrigation of the ICD unit and electrodes with antibiotic solution before implant eliminated infection in their last 50 devices [15]. In our experience, partial removal of the ICD was possible and successful. Although no uniform approach was applied, complete eradication of infection was achieved, provided certain principles were adhered to. Namely, removal of all infected material was mandatory. If the patches and sensing leads were not infected, then removal of them was not performed. In one case (no. 6), in which complete removal of patches, sensing leads, and generator was performed, the patient died intraoperatively. Because of extensive adhesions between the posterior patch and the epicardium, complete removal required cardiopulmonary bypass. After removal of all foreign material, the patient could not be weaned from cardiopulmonary bypass and died of heart failure. The patches and sensing leads were uninvolved with infection and were found to be culture negative. For that reason, we have become selective in recommending complete removal of infected ICD systems. A similar case was described by Kassanoff and others [16] in which there was extreme difficulty in explanting the patches because of firm adhesions formed onto the epicardium. They stressed early diagnosis of infection and early explantation to avoid difficulties in removal. Since infection most commonly begins in the generator pocket, it is not unreasonable to remove the generator, divide the leads into a separate sterile area, and leave the remainder of the system (eg, patches and sensing leads) alone. This approach is particularly applicable in the patient in whom the risk of patch and sensing lead removal is formidable. When we choose this partial removal approach, we investigate for patch or sensing lead involvement with CT scan and tagged white blood cell scan. Although imperfect, positive findings indicate a more aggressive and complete approach. Negative findings permit a sense of cautious optimism that the material left behind is sterile. Nevertheless, vigilance for residual infection is necessary, because in 2 patients (nos. 10 and 13) complete removal was required after partial removal proved unsatisfactory. The issue of partial versus complete removal, in our opinion, is more of a problem with the older and larger subcostal generator and epicardial patch and sensing lead systems. Although many of these systems are still in place, the current subpectoral-transvenous approach is likely to eliminate the problem.
At present, the majority of ICD devices implanted use a transvenous approach. There are several advantages of this new technology. First, insertion is quicker and easier. There is less morbidity postoperatively. Management of complications is simpler. In 1992, Saggau and associates [17] reported the superiority of endocardial versus epicardial ICD implantation. In addition to ease of insertion, they reported less infection with the endocardial approach. In 1994, Hammel and coworkers [18] described the relative ease of managing infected ICD devices of nonthoracotomy devices. Complete removal is easier and more readily achieved. Pfeiffer and others [19] noted less infection with nonthoracotomy systems. As such, this new technology translates into simpler and easier management of infected ICD devices. Complete removal, therefore, is more realistic in every case.
In addition to the treatment of infected ICD devices, preventive measures are important to avoid this outcome. As outlined by Spratt and others [20], infection was related to placement during other procedures (eg, coronary artery bypass grafting), reoperation, and concomitant infection. We found this to be the case in our series as well. Therefore, careful consideration of the timing of device implantation is paramount. Another area of importance is the environment in which the device is implanted. Shahian and associates [21] described detailed guidelines for prevention of infection in the perioperative period. Meticulous preoperative skin inspection and preparation, operative safeguards including double gloves and new gloves when handling the device, and postoperative wound inspection with early drainage of fluid collections were advised.
In conclusion, infection of ICD devices is a devastating event. Complete removal of the infected ICD units was successful in eradicating infection in all cases. Partial removal required further revision in a quarter of the patients. We favor complete removal of the ICD generator and all the components when possible. Partial removal of the ICD unit (ie, generator only) is reserved for patients in whom the risk of complete removal is too high and infection is confined to the generator only.
|
Footnotes
|
|---|
Address reprint requests to Dr Samuels, Dept of Cardiothoracic Surgery, Hahnemann Division, Allegheny University Hospitals, Broad and Vine Sts, MS 111, Philadelphia, PA 19102-1192.
|
References
|
|---|
- 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 1980;303:322–4.[Medline]
- Beck CS, Pritchard WH, Feil HS. Ventricular fibrillation of long duration abolished by electric shock. JAMA 1947;135:985–6.
- Zoll PM, Linethal AJ, Gibson W, et al. Termination of ventricular fibrillation in man by externally applied electric shock. N Engl J Med 1956;254:727–32.[Medline]
- Mirowski M, Mower MM, Staewen WS, Tabatznik B, Mendeloff AI. Standby automatic defibrillator: an approach to prevention of sudden coronary death. Arch Intern Med 1970;126:158–61.[Medline]
- Shuder JC, Stoeckle H, Gold JH, West JA, Keskar PY. Experimental ventricular defibrillation with an automatic and completely implanted system. Trans Am Soc Artif Int Organs 1970;16:207–12.[Medline]
- Choo MH, Holmes DR Jr, Gersh BJ, et al. Infected epicardial pacemaker systems. J Thorac Cardiovasc Surg 1981;82:794–6.[Abstract]
- Mirowski M, Reid PR, Mower M, et al. Clinical performance of the implantable cardioverter-defibrillator. PACE 1984;7:1345–50.[Medline]
- Marchlinski FE, Flores BT, Buxton AE, et al. The automatic implantable cardioverter defibrillator: efficacy, complications, and device failures. Ann Intern Med 1986;104:481–8.[Medline]
- Kelly PA, Wallace S, Tucker B, et al. Postoperative infection with the automatic implantable cardioverter defibrillator: clinical presentation and use of the gallium scan in diagnosis. PACE 1988;11:1220–5.[Medline]
- Almassi GH, Olinger GN, Troup PJ, Chapman PD, Goodman LR. Delayed infection of the automatic implantable cardioverter-defibrillator. J Thorac Cardiovasc Surg 1988;95:908–11.[Abstract]
- Goodman LR, Almassi GH, Troup PJ, et al. Complications of automatic implantable cardioverter defibrillators; radiographic, CT, and echocardiographic evaluation. Radiology 1989;170:447–52.[Abstract/Free Full Text]
- Taylor RL, Cohen DJ, Widman LE, Chilton RJ, O'Rourke RA. Infection of an implantable cardioverter defibrillator: management without removal of the device in selected cases. PACE 1990;13:1352–5.[Medline]
- Wunderly D, Maloney J, Edel T, et al. Infections in implantable cardioverter defibrillator patients. PACE 1990;13:1360–4.[Medline]
- Gupta S, Prevel C, Shaheen K, et al. Wound complications and treatment of the infected implantable cardioverter defibrillator generator. J Card Surg 1993;8:671–7.[Medline]
- Siclari F, Klein H, Troster HJ, Borst HG. Infectious complications after AICD implantation [Abstract]. PACE 1990;13:547.
- Kassanoff AH, Levin CB, Wyndham CRC, Mills LJ. Implantable cardioverter defibrillator infection causing constrictive pericarditis. Chest 1992;102:960–3.[Abstract/Free Full Text]
- Saggau W, Sack F-U, Lange R, et al. Superiority of endocardial versus epicardial implantation of the implantable cardioverter defibrillator (ICD). Eur J Cardiothorac Surg 1992;6:195–200.[Abstract]
- Hammel D, Scheld HH, Block M, Breithardt G. Nonthoracotomy defibrillator implantation: a single-center experience with 200 patients. Ann Thorac Surg 1994;58:321–7.[Abstract]
- Pfeiffer D, Jung W, Fehske W, et al. Complications of pacemaker-defibrillator devices: diagnosis and management. Am Heart J 1994;127:1073–80.[Medline]
- Spratt KA, Blumberg EA, Wood CA, Kutalek SP, Reboli AC. Infections of implantable cardioverter defibrillators: approach to management. Clin Infect Dis 1993;17:679–85.[Medline]
- Shahian DM, Williamson WA, Martin D, Venditti FJ Jr. Infection of implantable cardioverter defibrillator systems: a preventable complication? PACE 1993;16:1956–60.[Medline]
This article has been cited by other articles:
|
|
|
|
 
M. R. Moon, C. J. Camillo, and M. J. Gleva
Laser-assist during extraction of chronically implanted pacemaker and defibrillator leads
Ann. Thorac. Surg.,
June 1, 2002;
73(6):
1893 - 1896.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
J. D. Chua, B. L. Wilkoff, I. Lee, N. Juratli, D. L. Longworth, and S. M. Gordon
Diagnosis and Management of Infections Involving Implantable Electrophysiologic Cardiac Devices
Ann Intern Med,
October 17, 2000;
133(8):
604 - 608.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
