| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 1995;59:1092-1098
© 1995 The Society of Thoracic Surgeons
Divisions of Cardiothoracic Surgery and Cardiology, Medical College of Virginia and McGuire Veterans Administration Medical Center, Richmond, Virginia
 |
Abstract |
|---|
 TopFootnotes Abstract IntroductionMaterial and MethodsResultsCommentReferences |
|---|
|
Introduction |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentReferences |
|---|
Since their clinical introduction in 1980, implantable cardioverter defibrillators (ICDs) have revolutionized the treatment of patients with malignant ventricular tachyarrhythmias and sudden cardiac death. The clinical implantation of these devices has increased dramatically over the last 10 years, and it is estimated that 20,000 defibrillators are implanted annually in the United States at a cost of more than one billion dollars [1]. Traditionally, it was necessary to place the defibrillation patches directly on the epicardium to achieve adequate defibrillation within the energy constraints of these devices. To accomplish this, various surgical techniques have been employed including median sternotomy, left thoracotomy, and subcostal and subxiphoid approaches [2–4]. Using these ``thoracotomy'' approaches, surgical mortality rates at different centers have varied between 1% and 8%. There also has been significant perioperative morbidity involving between 15% and 42% of all patients [5–8].
From the time of the inception of ICDs, there was a uniform consensus that the ultimate goal in the technological evolution of ICDs would be the development of a transvenous defibrillation lead that could be implanted using a ``nonthoracotomy'' approach. The expectations were that these new devices would simplify implantation procedures, increase patient acceptance, expand clinical indications, and lower surgical morbidity and mortality. After more than a decade of intense basic and clinical research, technological advances led to the introduction of workable transvenous ICDs in the late 1980s [9–11]. As the transvenous defibrillation leads have improved and our basic understanding of defibrillation has increased, the use of nonthoracotomy systems has become widespread. They are presently the devices of choice in most patients undergoing ICD implantation. Initial reports have indicated that operative mortality is indeed lower with transvenous lead systems [1, 12].
The purpose of this report was to examine 100 consecutive nonthoracotomy implantations at our institution. The implantation risks and clinical outcome of these patients were compared with those of 102 consecutive thoracotomy implantations performed by the same surgeon.
|
Material and Methods |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentReferences |
|---|
The indications for ICDs in these patients were either (1) documented syncope, cardiac arrest, or ventricular fibrillation unrelated to transient or reversible clinical events with clinical or electrophysiologic evaluation that suggested the patient to be at risk for recurrent cardiac arrest and (2) sustained drug-refractory ventricular tachyarrhythmias.
Surgical Approach
Several different operative approaches were used in patients undergoing a thoracotomy implantation. All procedures were performed in the operating room, under general anesthesia, using invasive monitoring (arterial line, central venous line). The majority of implantations (n = 86) were performed using a subcostal approach that we have described previously [2, 13]. Five patients underwent implantation via a left thoracotomy, and a median sternotomy was used in 11 patients. Epicardial defibrillation patches were implanted in either anteroposterior or medial-lateral orientations and sutured to the pericardium. Patches were placed both in intrapericardial and extrapericardial positions. The size of the implanted patch was dictated by patient anatomy. The following epicardial patches were used: CPI model L67, surface area = 27.9 cm2; CPI model L66, surface area = 13.9 cm2; Telectronics model 106, surface area = 28 cm2; and Telectronics model 107, surface area = 39 cm2.
The pace/sense leads were inserted transvenously via the left or right subclavian vein in 50 patients. In-line bipolar leads were used in these patients with either passive or active fixation mechanisms. Epicardial screw-in leads were used in the other 52 patients. R-wave amplitudes were greater than 4 mV in all patients. In patients receiving second- and third-generation ICDs, pacing thresholds were measured and were adequate (<2.0 V; pulse width, 0.5 ms) in each case.
The pulse generators were implanted below the rectus muscle through a left subcostal incision. The various pulse generators used included the following: CPI 1550 (n = 31), 1555 (n = 1), 1600 (n = 9), 1700 (n = 14), and 1715 (n = 5) (CPI, Minneapolis, MN); Ventritex V100C (n = 2) (Ventritex Inc, Sunnyvale, CA); and Telectronics 4203 (n = 3) and 4210 (n = 29) (Telectronics Pacing Systems, Englewood, CA).
Nonthoracotomy devices also were implanted in the operating room under general anesthesia. The transvenous defibrillation electrode (CPI Endotak, 99 patients; Medtronic Transvene [Medtronic, Inc, Minneapolis, MN], 1 patient) was inserted under fluoroscopic guidance. All patients underwent successful subclavian venipuncture using the Seldinger technique. The left subclavian vein was used preferentially for lead insertion (92 patients), whereas the right subclavian vein was used in 8 patients. Satisfactory R waves (greater than 4 mV), pacing thresholds (less than 1.5 V), and lead resistances were obtained in all patients.
Forty-two patients required a subcutaneous patch (CPI model 0063, electrode surface area = 280 mm2). This patch was inserted through a small left submammary incision and placed along the left lateral chest wall. The transvenous defibrillation and subcutaneous patch leads were tunneled percutaneously to a left subcostal incision. The pulse generators were implanted below the rectus muscle in a carefully formed pocket. The pulse generators used included the following: CPI 1555 (n = 2), 1600 (n = 25), 1705 (n = 25), and 1715 (n = 3) and Ventritex V100C (n = 45).
Defibrillation Threshold Testing
All patients underwent device testing intraoperatively after the induction of ventricular fibrillation. Three successful shocks with a minimal 10 J safety margin between the defibrillation threshold and the maximal stored energy of the pulse generator were required for implantation. The defibrillation threshold was defined as the lowest successful energy required for ventricular fibrillation termination. Failure at low energy levels was not always required to define the precise defibrillation threshold.
Follow-up postoperative electrophysiologic testing was performed in all patients on the discharge drug regimen. Patients were seen in follow-up 2 weeks after implantation. Outpatient follow-up in the Electrophysiology Clinic was scheduled every 2 to 4 months depending on the particular requirements of the implanted device.
Analysis of Results and Definitions
The study end point was either death due to any cause or completion of follow-up. The follow-up period ended on October 1, 1994. Perioperative mortality was defined as all-cause mortality either within the 30-day interval after implantation or any in-hospital death occurring during the admission for defibrillator implantation. Sudden cardiac death was defined as a sudden, unexpected death occurring without symptoms or within 1 hour of onset of symptoms and included unwitnessed death. Sudden cardiac death and total survival curves were calculated using a life-table method for actuarial analysis. Statistical significance was examined with the Wilcoxon rank sum test. This analysis was limited to the first 18 postoperative months because of the short period of follow-up in the nonthoracotomy group. Perioperative mortality was included in the total survival data. Complications were defined as symptomatic or asymptomatic clinical conditions that had the potential for adverse effects on clinical outcome.
Data are presented as mean ± standard deviation. Proportions were analyzed by the 
2 test, except when the sample size was small, in which case a Fisher exact test was employed. Comparisons of group data were performed by an analysis of variance. A p value of 0.05 or less was considered statistically significant.
|
Results |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentReferences |
|---|
Fifty-one patients had a history of either New York Heart Association class III or class IV heart failure. Mean left ventricular ejection fraction was 0.31 ± 0.12. The clinical characteristics of both the nonthoracotomy and thoracotomy groups are summarized in Table 1
. The two groups were comparable for age, sex distribution, comorbidity, cardiac disease status, functional class, and ejection fraction.
|
). In the thoracotomy group, the number of antiarrhythmic drugs tested was slightly higher than in the nonthoracotomy group. The results of the preoperative electrophysiologic study were similar in the two groups.
|
One hundred six patients underwent attempted implantation of a nonthoracotomy or transvenous system. A nonthoracotomy system was implanted successfully in 100 patients (94%). Of the 6 patients who had high defibrillation thresholds (greater than 20 J) with the nonthoracotomy or transvenous approach, 5 patients underwent successful implantation of a thoracotomy system. However, in 2 of the patients, the defibrillation threshold remained high at 25 J. For the sake of analysis, these 5 patients were considered to have thoracotomy implants. There was no operative mortality in this group of patients. In 1 patient, adequate defibrillation thresholds were not obtained with either a nonthoracotomy or thoracotomy approach. No ICD was implanted. This patient was excluded from analysis.
With the nonthoracotomy systems, 58 patients had successful implantation with the transvenous defibrillation lead alone. The probability of a successful lead-alone implantation was related directly to the type of shock waveform. With devices capable of delivering only monophasic shock waveforms (n = 52), only 36% of implantations were successful with the defibrillation lead alone. The remainder of patients required a subcutaneous patch. In generators capable of delivering a biphasic shock (n = 48), 81% of implantations were successful with the defibrillation lead alone. This difference was statistically significant (p < 0.001). In lead-alone implantations, the distal right ventricular coil was the cathode and the proximal superior vena cava coil the anode in all except 1 patient. In the patients requiring a subcutaneous patch, most devices (56/58) were implanted with the distal right ventricular coil as the cathode and the subcutaneous patch and the proximal coil as a combined anode.
Implantation electrophysiologic data were examined for both the thoracotomy and nonthoracotomy systems. Although there was no significant difference in R-wave amplitude of the pace/sense lead between the nonthoracotomy and thoracotomy systems (10.8 ± 5.5 versus 11.6 ± 5.8 mV; respectively, p = not significant), there was a slightly and significantly lower pacing threshold in the non-thoracotomy group (0.8 ± 0.7 versus 1.0 ± 0.8 V; p = 0.03). The mean defibrillation threshold was not statistically different between the nonthoracotomy and thoracotomy ICDs (17.5 ± 5.2 versus 16.6 ± 5.8 J; p = not significant).
Clinical Outcome
The mean number of days in the intensive care unit was significantly greater in patients undergoing a thoracotomy (3.3 ± 3.9 versus 1.7 ± 1.7 days; p < 0.0005) (Fig 1). With the nonthoracotomy approach, the postoperative stay also was shortened significantly by almost 50% compared with our thoracotomy series.
|
). Pulmonary complications were particularly more common after a thoracotomy approach. Moreover, the pulmonary complications after nonthoracotomy implantations were minor and usually related to subclavian entry (one small hemothorax not requiring a chest tube, one pneumothorax). In contrast, the pulmonary complications in the thoracotomy group were more serious and usually resulted in prolonged hospital courses. In this group, 5 patients had a culture-positive pneumonia, there were five episodes of adult respiratory distress syndrome, and there was one pulmonary embolus. The incidence of the other specific complications was too small to allow for a statistical comparison; however, the total incidence of complications was significantly higher in patients receiving a thoracotomy system.
|
). In the nonthoracotomy group, the one death was due to an acute myocardial infarction. In the thoracotomy group, 1 patient died of an acute myocardial infarction and 1 of incessant ventricular tachycardia, and one death was due to severe adult respiratory distress syndrome in a patient receiving amiodarone. There appeared to be a correlation between left ventricular ejection fraction and death in the thoracotomy group. All three deaths occurred in patients with an ejection fraction of 0.30 or less. In this subgroup, perioperative mortality with the thoracotomy system was 6%. In the 56 consecutive patients with ejection fractions of 0.30 or less in the nonthoracotomy group, there was no perioperative mortality. This difference approached statistical significance (p = 0.06).
Follow-up
The patients undergoing thoracotomy implantation were followed up for a mean of 26 ± 16 months. Thirty-one of the 102 patients were either lost to follow-up or followed up at other institutions. The nonthoracotomy patients were followed up for a mean of 9 ± 8 months. Nineteen patients were either lost to follow-up or followed up elsewhere. Because of the varying lengths of follow-up, comparison between the two groups as to their overall survival and event-free survival are difficult to interpret. There have been eight late deaths in the nonthoracotomy group and 24 in the thoracotomy group. The overall actuarial survival at 1 year was 87% ± 3% and did not differ between the two groups (Fig 2A). Sudden death survival was 100% and 97% ± 2%, respectively, at 1 year in the nonthoracotomy and thoracotomy groups (Fig 2B).
|
). The majority of the lead-related complications were identified by undersensing or oversensing problems and usually were due to lead fracture. There were three cases of lead dislodgements, two involving transvenous pace/sense leads in the thoracotomy systems and one dislodgement of a Medtronic Transvene defibrillation lead. There was only one late infection in the 202 patients. This occurred 30 days postoperatively and required device explantation. The late pulmonary complications consisted of one case of persistent pleural effusion requiring repeated thoracenteses and one case of amiodarone lung toxicity.
|
|
Comment |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentReferences |
|---|
The present study reports our single-institution experience comparing perioperative morbidity and mortality in consecutive patients undergoing implantation of thoracotomy and nonthoracotomy ICD systems. Similar to other reports, the majority of the patients were male and suffered from ischemic heart disease [1, 12, 20]. The electrophysiologic presentation and the results of the preoperative electrophysiologic study also were not significantly different between the groups. In the thoracotomy group, the number of antiarrhythmic drugs tested was slightly higher than in the nonthoracotomy group (2.0 ± 1.5 versus 1.6 ± 0.09; p = 0.054). This likely reflects the tendency of our electrophysiologists to refer patients for nonthoracotomy ICDs earlier in their course because of the less invasive nature of the procedure.
In this study, nonthoracotomy systems were implanted successfully in the great majority of patients. One hundred of 106 patients (94%) who were brought to the operating room for a nonthoracotomy system were able to achieve adequate defibrillation thresholds and avoid a thoracotomy. This finding is consistent with other reports showing implantation success rates between 72% and 95% [1, 12, 19]. There was a learning curve with these devices, with five of our six failures occurring during the first year of our experience. It is important to point out that since July 1992, all patients evaluated for defibrillator implantation at our institution have undergone an attempted nonthoracotomy or transvenous approach. Thus, nonthoracotomy devices are widely applicable to the entire spectrum of patients with malignant ventricular tachyarrhythmias. Moreover, defibrillation thresholds for our nonthoracotomy group were not significantly different from those of our thoracotomy series. This demonstrates a comparable efficacy with both systems.
The efficacy of the nonthoracotomy systems was influenced strongly by the morphology of the defibrillation shock waveform. Devices capable of delivering biphasic shock waveforms were more likely to have successful nonthoracotomy implantation. Five of the six failures occurred with monophasic devices. This reflected a failure rate of 10% with monophasic devices and only 2% with biphasic devices. Moreover, ICD systems with biphasic shocks were significantly more likely to be successful with the transvenous defibrillation lead alone than monophasic shock devices (81% versus 36%; p < 0.001). Thus, our study, in agreement with previous reports [21, 22], documents the improved efficacy of biphasic shock waveforms for transvenous defibrillation.
Nonthoracotomy ICDs also had significantly lower pacing thresholds than thoracotomy devices. This finding is in agreement with other studies [12]. The improved pacing capability, although small, may be important in third-generation devices with antitachycardia pacing.
The principal advantage of nonthoracotomy systems was the smoother and more benign postoperative course seen in these patients. There was a significant decrease in both intensive care unit stay and postoperative complications when compared with thoracotomy systems. The most striking decrease was seen in pulmonary complications, from 18% to 3%. Moreover, the severe pulmonary complications such as pneumonia and adult respiratory distress syndrome seen with thoracotomy devices were not present in those patients undergoing a nonthoracotomy approach. The difference is likely to be attributable to the absence of a thoracotomy and the significantly decreased amount of postoperative pain. Overall, there were almost three times as many complications with the thoracotomy systems. As opposed to other reports [12, 20], lead dislodgement was not a significant problem with nonthoracotomy systems, occurring in only 1 of our 100 patients.
There was a low incidence of infection (1 in 202 patients, 0.5%) and no erosions in this series. Previous reports have indicated infection rates up to 7% with an incidence of erosion of approximately 1% [5, 23]. At our institution, the following measures are taken to prevent infection: use of prophylactic antibiotics, limitation of operating room time, placing all leads in the operating room under sterile conditions, performing procedures first on the schedule, avoiding tissue trauma, achieving meticulous hemostasis, irrigating the wounds with antibiotic solution, and placing the generators below the rectus muscle. It is our belief that submuscular placement of the generator is important in both minimizing infection and preventing erosion.
By decreasing the postoperative complications, the nonthoracotomy approach significantly shortened postoperative stay. The significant decrease in both intensive care unit and postoperative hospital stay would be expected to decrease the substantial implantation costs associated with these devices.
Perioperative mortality was lower with nonthoracotomy devices. However, this did not reach statistical significance, probably due to the relatively small size of our two patient groups. A larger multicenter trial involving more than 1,200 patients revealed similar mortality rates of 1% and 4% with nonthoracotomy and thoracotomy devices, respectively, and this difference did reach statistical significance [1]. The present study suggested a trend toward lower mortality with nonthoracotomy systems, particularly in high-risk patients (those with left ventricular ejection fractions of 0.30 or less). There were no deaths in this group undergoing a nonthoracotomy implantation, whereas the mortality rate with a thoracotomy approach was 6% (p = 0.06). A history of class III or class IV congestive heart failure has been suggested to be related to mortality in other series [1, 24]. However, only 1 of our 4 patients who died had a history of congestive heart failure.
Late follow-up revealed no significant differences in overall survival or sudden death survival between our nonthoracotomy and thoracotomy series. Comparisons were carried out only for the first 18 months in each group because of the significantly shorter follow-up in the nonthoracotomy group. Nonthoracotomy devices were very effective in preventing sudden cardiac death. At 18 months of follow-up, sudden cardiac death survival was 100%. The low incidence and the varying length of follow-up made a statistical comparison of late complications between thoracotomy and nonthoracotomy patients impossible. Lead-related problems were the most significant source of late morbidity in both groups. This is in agreement with our previous report of the international experience during a third-generation defibrillator multicenter trial, in which late lead complications occurred in 13% of thoracotomy systems [25]. The majority of these problems were due to lead fractures. The lead adapters were a particular weak link in these systems, with complications with some models exceeding 20% [25]. It was and remains our recommendation to avoid all types of lead adapters in both thoracotomy and nonthoracotomy systems.
In the nonthoracotomy systems, our lead-related complications involved one late migration of a Medtronic Transvene lead, and one lead that required revision in the subclavicular pocket because of patient discomfort. The other three lead problems involved either undersensing or oversensing with Endotak leads. None of these patients showed a change in their defibrillation threshold at the time of revision. Radiographic examination of these leads at the time of the revision revealed ``crimping'' of the defibrillation lead under the clavicle. This was a particular problem in earlier models of the Endotak lead [26]. Because of this problem, it is now our policy to use a modified subclavian entry with a more lateral approach to avoid subclavian crush [27]. Since changing our technique in early 1994, we have seen no further episodes of this problem. In our opinion, the transvenous defibrillation lead remains the Achilles heel of nonthoracotomy systems. Hopefully, as has happened with dual-chamber pacemakers, future developments in lead technology will minimize these problems.
The principal limitation of this study is that it is a nonrandomized, retrospective comparison. However, with the documented efficacy of the less invasive nonthoracotomy systems, a randomized, prospective trial is unlikely to be performed in this country. The shortcomings of this report are partially overcome by several facts. It does represent a large single-institution study comparing perioperative morbidity and mortality between nonthoracotomy and thoracotomy devices. These systems all were implanted by the same surgeon over a relatively short period of time. The patient populations also were comparable in every respect.
Another limitation of this report is the short period of follow-up for the nonthoracotomy group. This made comparisons of long-term survival difficult to interpret. Further follow-up will be needed to establish the long-term efficacy of nonthoracotomy systems. The early results are encouraging. However, late lead complications remain a concern, and close surveillance will be mandatory.
This study shows nonthoracotomy ICDs to be a safe and efficacious treatment for all patients with malignant ventricular tachyarrhythmias. These devices resulted in significantly less perioperative morbidity and shorter intensive care unit and hospital stays when compared with our previous experience with thoracotomy implantation. Nonthoracotomy systems should lead to a significant decrease in the substantial health care costs associated with ICD implantation. It is our opinion that they are the devices of choice in all patients, unless a concomitant cardiac operation needs to be performed or there is a direct contraindication to the positioning of a transvenous lead. Because of the improved efficacy of biphasic shocks, all nonthoracotomy ICDs should incorporate this feature.
Future biomedical engineering advances are likely to decrease the size of the devices, increase their defibrillation efficacy, and improve transvenous lead design. Thus, it is likely that the clinical indications for these devices will continue to expand. However, the lack of prospective, randomized trials in the literature comparing defibrillators with other types of medical management remains a significant drawback to the development of a rational clinical approach to patients with ventricular tachyarrhythmias. The results of several large clinical trials currently in progress soon should delineate which patients will achieve long-term benefit from ICD implantation.
|
Footnotes |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentReferences |
|---|
Address reprint requests to Dr Damiano, Division of Cardiothoracic Surgery, Medical College of Virginia, PO Box 980068, Richmond, VA 23298-0068.
|
References |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  A. Oto, A. Yildirir, G. Kabakci, and E. Atalar Inappropriate Shocks Diagnosed by Stored Electrograms of Implantable Cardioverter Defibrillators: Two Case Reports Angiology, May 1, 2000; 51(5): 425 - 430. [Abstract] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
