Ventricular fibrillation (VF) and sudden death are recognized possible sequelae of myocardial infarction, with or without the presence of cardiac failure (1). When arrhythmia persists despite optimal pharmacological treatment, the insertion of an implantable cardioverter defibrillator (ICD) is considered. The value of an ICD for such patients has been reported in a randomized, controlled study of 704 patients (2), which concluded that this approach reduced the risk of sudden death in high-risk patients with coronary disease by approximately 27% and that the overall mortality was reduced by approximately 17%.
VF induction is part of the process of ICD implantation, as is the determination of the minimally effective defibrillation energy threshold (DFT) of the device (3). ICDs are currently implanted subcutaneously, and the leads are inserted via the subclavian vein. This procedure may be performed under local or regional analgesia and sedation (4,5) or under general anesthesia (GA) by using a variety of volatile or IV drugs (5,6). The latter may impair ICD efficiency at a given energy output used to defibrillate the heart (6). If this happens, the programmed voltage or energy output could fail when spontaneous fibrillation occurs. Also, the induction of defibrillation at unnecessarily high intensity direct current stimulation has a risk for metabolic and functional deterioration of the heart (7).
The two major volatile anesthetics that have been used either alone or in multianesthetic drug regimens during the last 20 yr, halothane and isoflurane, affect the electrophysiological properties of the heart (6,8). It has been suggested that these anesthetics might themselves cause arrhythmia and interfere with the induction of experimental ventricular tachycardia (VT) and defibrillation (5). However, animal studies comparing the use of halothane and barbiturates found no abnormal effect of halothane on the DFT (9). Jarvis and Lahtinen (10) and Gill et al. (9) studied the effects of isoflurane, halothane, and barbiturates in animal models, and they, too, found no difference among these drugs on the DFT. This study is the first to investigate the influence of halothane, isoflurane, and IV fentanyl (fentanyl is the most frequent intraoperative opioid used for cardiac patients) (11) on the DFT during ICD implantation in humans. We hypothesized that volatile anesthetics would affect DFT requirement during ICD implantation, as opposed to fentanyl or a local anesthetic infiltration combined with an intermittent propofol-awake sedation regimen.
The study was approved by the institutional investigational review board. Eighty consecutive consenting patients were enrolled in this randomized, blinded, and crossover study in which the Ventak® Mini™ IV (Guidant, St. Paul, MN) ICD device was implanted. The patients who received ventilatory support, those taking inotropic or vasoactive drugs, those with a preexisting pacemaker, and those who refused to sign an informed consent form were excluded from the study. None of the patients scheduled to receive halothane had been exposed to it before. There was no use of placebo, for ethical reasons.
All the patients were premedicated with hydroxyzine 25 mg the evening before the procedure. Individual daily chronic medications were continued up to the time of preparation for the implantation. Blood electrolyte profiles and drug plasma levels (when applicable) were checked, and the results were available to the anesthesiologist: abnormal results prompted exclusion of the patient from the study.
A 16-gauge peripheral venous catheter and a radial artery catheter were inserted into all patients before the procedure. Patients were randomly divided into one of three GA groups, in which anesthesia was maintained by adding halothane, isoflurane, or IV fentanyl, or into a fourth group, which received local (lidocaine) anesthesia in combination with a sedative dose of IV propofol (Fig. 1). The team that implanted the device was blinded to the patient’s GA protocol.
The induction of anesthesia in the three GA groups was induced with 4 mg/kg of sodium thiopental administered over 1 min, and endotracheal tube insertion was facilitated by 1 mg/kg of succinylcholine. GA was then maintained throughout the procedure with 2/1 L/min of N2O/oxygen fresh gas flow, and a single dose of 0.1 mg/kg of pancuronium was administered over 2 min. GA was initially supplemented by a 0.7% end-tidal concentration of halothane, 1% isoflurane, or 1.5 μg/kg of IV fentanyl. The ICD’s leads were inserted via the subclavian vein and positioned on the endocardial surface aided by chest radiograph fluoroscopy. When this was accomplished and the initial volatile anesthetics’ concentrations had been constantly maintained for ≥5 min, the first set of measurements was performed: externally induced VF was invoked, and defibrillation was tested in decreas-ing intervals of 5 J, as described by Fitzpatrick et al. (3). The minimally effective energy output that was achieved determined the DFT.
After this first DFT measurement, the skin was excised (either under GA or lidocaine infiltration) for the ICD implantation, and the inhaled anesthetics were turned off. At least 5 min after the end-expired concentrations of the inhaled anesthetics were recorded as being 0, the patients of the three GA groups received fentanyl 1 μg/kg. One minute later, a second set of DFT measurements was performed, taking place 25–30 min after the first one. The third set was taken at the end of the surgical procedure, with the GA groups being maintained solely under the effect of N2O/oxygen and when at least 30 min had passed after the second measurement.
The protocol in the lidocaine plus propofol group consisted of 5 mL of 1.5% lidocaine injected subcutaneously at the beginning of the procedure to enable the insertion of the electrodes. An additional 13 mL of 1.5% lidocaine was injected 30 min later, between the first and the second set of measurements, for the formation of the subcutaneous pocket where the ICD was later inserted (see above). One minute before each of the three fibrillation/defibrillation tests that were performed at the same time intervals as in the GA groups (see above), 0.35 mg/kg of IV propofol was slowly injected to induce awake sedation (12). This dose has been used satisfactorily in our institution during defibrillation procedures in patients with atrial fibrillation.
Throughout the study period and while they were in the postanesthesia care unit (PACU), the patients were continuously monitored for invasive systolic and diastolic blood pressures, heart rate, respiratory rate, pulse oximetry (Spo2), and end-tidal CO2 (ETco2). During the procedure itself, all the GA patients had been mechanically ventilated, aiming at an ETco2 of 32–36 mm Hg. The end-tidal alveolar concentrations of the exhaled volatile anesthetics were continuously recorded by a Cardiocap Ultima® monitor (Datex, Helsinki, Finland).
Before being discharged, all the study patients were asked whether they were satisfied with the anesthetic technique. The lidocaine plus propofol patients were also asked whether they recalled any of the DF events.
The analyses were performed by using SPSS for Windows, Version 9 (SPSS Inc., Chicago, IL). A prestudy power table, in which δ = 3.5 J (mean difference in DFT values recorded in a pilot study), α = 0.05, and power = 0.9, indicated the need for 16 patients in each group. Demographic data (age and weight) and background characteristics (baseline heart and respiratory rates, systolic and diastolic blood pressures, and intraoperative respiratory and hemodynamic values), as well as the ASA and the New York Heart Association (NYHA) classes and the ejection fraction (EF) of the four study groups, were compared by using one-way analysis of variance (after anesthesia effect). Patients’ medical history and chronic medications were analyzed with the Pearson χ2 test. The effects of the various types of anesthesia on the corresponding and subsequent DFTs were analyzed with two-way analysis of variance with repeated measures (after anesthesia or time effects). All values are expressed as mean ± sd, with significance defined as P ≤ 0.05.
The study patients’ demographic and medical data are reported in Table 1; all patients completed all phases of the study. The preoperative general health status (ASA and NYHA classes) was similar for all the patients. All the studied patients had normal blood count and biochemistry analysis. None required any electrolyte supplement before, during, or after the procedure. Previous relevant medical history and medications (Table 1) showed no differences among the groups, except for the left ventricular EF, which was significantly lower in the isoflurane-treated patients compared with the other groups. There was no correlation between EF and the ASA or NYHA data in the various groups. Finally, the perioperative hemodynamic values of the four groups were similar (selected data are reported in Table 2), as were the respiratory rate, ETco2, and Spo2 (data not shown).
The DFT values obtained during the three sets of measurements are reported in Table 3. The DFTs were similar in the three GA groups in the first set (N2O/oxygen-based GA added to halothane, isoflurane, or fentanyl), but they were significantly higher than in the lidocaine plus propofol sedation group (P = 0.01; main anesthesia effect). In the second set (∼30 min later, at an end-tidal = 0% of the volatile anesthetics for ∼5 min), the DFTs in the two volatile groups were significantly lower than their values in the first set (P = 0.01 for halothane and P = 0.02 for isoflurane; main anesthesia effect). In the fentanyl group, the second set’s DFT was similar to those in the two volatile groups but lower than its own DFT in the first set (although the values did not reach statistical significance). The DFT value for the lidocaine plus propofol sedation group was slightly lower than those of the three GA groups but was similar to the first set’s value.
In the third set (30 min after the second set; only N2O/oxygen was delivered to maintain GA in the three designated groups), the two volatile groups and the fentanyl groups had DFTs that were significantly lower (P < 0.01; main anesthesia effects) than they were in the first set. The third set’s DFT in the lidocaine plus propofol sedation group was almost identical to that of the two previous ones and, at the same time, similar to the corresponding third set’s values in the three GA groups. The three sets of measurements thus indicated an overall intergroup difference among the four groups during the first set of measurements and an intragroup variability in the three GA groups’ measurements for all three sets.
Neither dysrhythmia nor aberrant conduction nor prolongation of the QT interval was encountered in any of the patients. Before their discharge to the ward, none of the lidocaine plus propofol patients claimed to have any recall of the fibrillation/defibrillation procedure or to have felt pain, and all expressed their satisfaction and willingness to undergo a similar maneuver with the same technique. None of the patients had postanesthesia nausea or vomiting. All the study patients were discharged uneventfully to the intermediate cardiac care unit according to the PACU discharge protocol.
This study is the first to demonstrate that halothane, isoflurane, and fentanyl, when added to N2O/oxygen-based GA, affect the DFT during the process of ICD implantation in humans. Compared with local anesthesia combined with intermittent small-dose propofol, these three anesthetics increased the DFT. These results also indicate that neither repetitive fibrillation/defibrillation events over time nor N2O/oxygen itself was the cause for the negatively affected myocardial response to the defibrillator, but rather halothane, isoflurane, and fentanyl themselves. Subcutaneous lidocaine plus intermittent small-dose IV propofol minimized the DFT level while providing patient satisfaction.
The need for adjustment of defibrillation energy requirements of an ICD at the time of its implantation arises from unpredictable individual changes and variations of the myocardial electrical threshold. These include previous diseases, electrolyte changes, concurrent antiarrhythmic treatment, site of contact of the ICD electrode, tissue scarring, the method of measurement of the energy, and the technique of fibrillation/defibrillation (3). The ICD is programmed on the basis of the data retrieved during its implantation for it to subsequently work properly (6), including the occasionally unnecessarily high intensity of defibrillation shocks, which themselves could produce abnormal arrhythmogenic responses (7). Because all anesthetics may potentially alter myocardial electrophysiology (6,9,13), it is essential to minimize any external effect that could interfere with the myocardial electrical threshold response to defibrillation during ICD implantation.
This study protocol enabled us to exclude the eventuality of lidocaine having an effect on DFT values. The 5 mL of 1.5% lidocaine (approximately 1 mg/kg) that was injected subcutaneously before the first DFT measurement set and the additional 13 mL (approximately 2.5 mg/kg) that was added to enable the formation of the subcutaneous pocket were 50% or less the dose (7 mg/kg) that had been used in patients undergoing inguinal hernia repair (14). The latter single-shot infiltration (as opposed to our divided single dose given with 30-minute intervals) was associated with a peak plasma concentration of 0.54 mg/mL, which is substantially below the therapeutic window for lidocaine (1.5–5 μg/mL) (15).
A pharmacological effect on DFT of the small dose of propofol (administered in the lidocaine group) could also be reliably excluded. Our results showed no changes in DFT measurements, blood pressure, or heart rate throughout the serial measurement. This was unlike the case in which 320 mg of IV propofol that had been injected as a single dose caused the DFT to increase to >31 joules (16). We also reasoned that there was no accumulative effect of propofol in our dose regimen based on kinetic simulation, where the injection of 2 mg/kg of IV propofol resulted in a 3.5 μg/mL plasma concentration after one minute, which decreased to 320 ng/mL 25 minutes later, far below the 50% inhibitory concentration of the drug (17). Our group (18) had warned against a possible cardiovascular depression effect of the drug. Nevertheless, propofol has gained worldwide popularity as a useful drug for both deep and awake sedation, and it has proven effective without major deleterious hemodynamic effects even in patients recovering from cardiac surgery (19). Interestingly, although propofol may cause the prolongation of the QT interval during the implantation of an ICD (20) —an event that did not occur in our study—this has not kept the drug from gaining the reputation as being safe for this procedure (4,21), even in patients with severe left ventricular dysfunction (21). Finally, although it is possible to implant an ICD under local anesthesia alone (4), we opted for the use of sedation during the induction of fibrillation, for ethical reasons. Thus, the dose and administration technique of propofol in association with subcutaneous lidocaine infiltration that we now describe seems not to have interfered with DFT and proved to be effective in inducing awake sedation and full patient satisfaction.
By virtue of its being fentanyl free, the last set of measurements in the three GA groups permitted us to infer a lack of an effect on DFT by N2O itself. Twenty-five to 30 minutes after an IV bolus of fentanyl 2 μg/kg, both the plasma and the effect-site concentrations were found to be far below therapeutic levels in various studies that used pharmacokinetic simulation (22,23). The estimated peak effect-site concentration after this IV bolus was approximately 3 ng/mL at 3–5 minutes and 1.4 ng/mL 20 minutes later (23). Also, 24 minutes after the injection of a bolus of 3 μg/kg (two to three times the fentanyl dose in our study), the brain effect-site concentration was 40% of the peak concentration, which was also far below the 50% inhibitory concentration level (22). This would suggest that both the plasma and effect-site concentrations in our GA groups were far below the therapeutic levels, especially at 30 minutes after the second bolus of fentanyl. It is noteworthy that the concentrations of fentanyl in highly perfused (vessel-rich) tissues, such as the heart, lungs, and brain, were found to pharmacokinetically parallel the concentration of fentanyl in the plasma and thus were designated as part of the central compartment (24). Furthermore, the third set’s values in the three GA groups (anesthesia maintained by only N2O) were not different from the DFTs measured for the same third set in the lidocaine plus propofol group. This could indicate that N2O did not have any net effect on the DFT or that it was negligible at most and that the increased DFTs values recorded in the three GA groups should instead be attributed to the two volatile anesthetics and to fentanyl. Finally, given the similarity of all the DFT values in the lidocaine plus propofol group, it can also be concluded that the repetitive fibrillation/defibrillation over time had no effect on the DFT values in this group or, presumably, in any of the GA groups.
Numerous studies during the last two decades have investigated the effects of different anesthetics on cardiac electrophysiology and on the DFT. Most of these studies were conducted on animals, and their results were conflicting. Both isoflurane and halothane were proven to minimally affect the atrial/sinoatrial nodal automaticity, the ventricular conductance, the rate of ventricular escape beats, and the idioventricular rhythm in the dog (8,25). Hunt and Ross (6), however, demonstrated that halothane prolonged the PR interval, extended the ventricular refractory period, and interfered with the ability to induce VT in dogs. Moreover, in one study (13), both isoflurane and halothane were arrhythmogenic. We believe that our current findings add new important data for establishing the possible detrimental effects of volatile anesthetics and of fentanyl on the electrophysiological properties of the human heart. Namely, we found that halothane, isoflurane, and fentanyl, administered as part of an N2O/oxygen-based anesthesia, caused the heart to become more resistant to defibrillation during ICD implantation compared with solely N2O/oxygen-based anesthesia or with local infiltration of lidocaine plus small-dose propofol anesthesia. Although new volatile anesthetics have been introduced clinically in the last decade and halothane has seldom been used in adults in many countries, its characteristic arrhythmogenic effects and its being the most frequently investigated volatile anesthetic in this milieu were the main reasons that we chose it as a model to compare with other anesthetics/techniques in this arrhythmia-prone patient population.
Because it was previously suggested that fentanyl 1.5 μg/kg had no effect on the electrophysiological performance of the heart (26), it was surprising to note an altered degree of resistance to defibrillation in the fentanyl patients. We have discussed and concluded above that N2O per se did not cause an increase in the DFT. However, the results of the first set of measurements could indicate that N2O plus fentanyl might induce resistance to defibrillation. This possible additive effect would not exclude the above-mentioned lack of an effect of fentanyl (26), and this is supported by a previously reported study (27).
The intergroup difference in the EF had no effect on the measured DFTs, a finding that is in agreement with other studies. Steinbeck et al. (28) concluded that the incidence of VT and the DFT values were not different in patients with an EF <30% compared with those with an EF >30%, even among the former, who could have experienced the most severe left ventricular deterioration during ICD implantation.
There was no difference in the hemodynamic profile among the four study groups. The patients in the lidocaine plus propofol group did not respond to skin incision by movement, tachycardia, hypertension, or other symptoms that may be considered sympathetically mediated; the third set of DFT measurements in the GA groups were similarly free of such reactions. These observations also suggest that the sympathetic tone did not affect the DFT during any set or for any group. Furthermore, an increase in the sympathetic tone is expected to be arrhythmogenic, and this would cause an increase rather than a decrease in the minimal requirement of energy.
Two limitations of this preliminary study should be mentioned. The first is the lack of dose-response comparisons of the tested anesthetics: this was not possible because of ethical considerations and the need for a much larger number of patients to perform it. In addition, the study was specifically designed to examine whether the various anesthetics at the clinically acceptable doses or concentrations affected DFT. The second limitation is the lack of a proper placebo-control group of patients, again because of ethical restrictions.
This study investigated the role of volatile, IV, and local anesthesia techniques for ICD implantation. The disparate results support the validity of the combination of lidocaine with intermittent small-dose propofol as an equally safe and clinically applicable mode of anesthesia in these high-risk patients during ICD implantation. Whereas N2O/oxygen-based GA with halothane, isoflurane, or fentanyl at the tested clinical concentrations or doses increased the minimal DFT during cardioverter defibrillator implantation, the subcutaneous lidocaine plus IV propofol technique minimized it while providing equal patient satisfaction.
We thank Esther Eshkol for editorial assistance.
1. Bigger JT. Prophylactic use of implanted cardiac defibrillators in patients at high risk for ventricular arrhythmias after coronary artery bypass graft surgery: Coronary Artery Bypass Graft (CABG) Patch Trial Investigators. N Engl J Med 1997; 337: 1569–75.
2. Buxton AE, Lee KL, Fisher JD, et al. A randomized study of the prevention of sudden death in patients with coronary artery disease: Multicenter Unsustained Tachycardia Trial Investigators. N Engl J Med 1999; 341: 1882–90.
3. Fitzpatrick AP, Lesh MD, Epstein LM, et al. Electrophysiological laboratory, electrophysiologist-implanted, non-thoracotomy-implantable cardioverter/defibrillators. Circulation 1994; 89: 2503–8.
4. Lipscomb KJ, Linker NJ, Fitzpatrick AP. Subpectoral implantation of a cardioverter defibrillator under local anesthesia. Heart 1998; 79: 253–5.
5. Pinosky ML, Reeves ST, Fishman RL, et al. Intravenous sedation for placement of automatic implantable cardioverter-defibrillators. J Cardiothorac Vasc Anesth 1996; 10: 764–6.
6. Hunt GB, Ross DL. Comparison of the effect of three anesthetic agents on induction of ventricular tachycardia in a canine model of myocardial infarction. Circulation 1988; 78: 221–6.
7. Kodama I, Shibata N, Sakuma I, et al. Aftereffects of high-intensity DC stimulation on the electromechanical performance of ventricular muscle. Am J Physiol 1994; 267: H248–58.
8. Atlee JL, Brownlee SW, Burstrom RE. Conscious state comparisons of the volatile anesthetics on specialized atrioventricular conduction times in dogs. Anesthesiology 1986; 64: 703–10.
9. Gill RM, Sweeney RJ, Reid PR. The defibrillation threshold: a comparison of anesthetic and measurement methods. Pacing Clin Electrophysiol 1993; 16: 708–14.
10. Jarvis AS, Lahtinen SP. A pilot study: defibrillation thresholds in dogs are similar with isoflurane, halothane and pentobarbital. Pacing Clin Electrophysiol 1994; 17: 280–5.
11. Wray DL, Fine RH, Huges CW, Thomas SJ. Anesthesia for cardiac surgery. In: Barash PG, Cullen BF, Stoelting RK, eds. Clinical anesthesia. 2nd ed. Philadelphia: JB Lippincott Co, 1992: 1021–50.
12. Pacifico A, Wheelan KR, Nasir N Jr, et al. Long-term follow-up of cardioverter-defibrillator implanted under conscious sedation in prepectoral subfascial position. Circulation 1997; 95: 946–50.
13. Stadnicka A, Bosnjak ZJ, Kampine JP, Kwok W. Modulation of cardiac inward rectifier K+
current by halothane and isoflurane. Anesth Analg 2000; 90: 824–33.
14. Karatassas A, Morris RG, Walsh D, et al. Evaluation of the safety of inguinal hernia repair in the elderly using lidocaine infiltration anaesthesia. Aust N Z J Surg 1993; 63: 266–9.
15. Roden DM. Antiarrhythmic drugs. In: Hardman JG, Limbrid LE, eds. Goodman & Gilman’s The pharmacological basis of therapeutics. 9th ed. New York: McGraw-Hill, 1996: 839–74.
16. Cohen TJ, Chengot T, Quan C, Peller AP. Elevation of defibrillation thresholds with propofol during implantable cardioverter-defibrillator testing. J Invasive Cardiol 2000; 12: 121–3.
17. Doenicke AW, Roizen MF, Rau J, et al. Pharmacokinetics and pharmacodynamics of propofol in a new solvent. Anesth Analg 1997; 85: 1399–403.
18. Weinbroum AA, Halpern P, Rudick V, et al. Midazolam versus propofol for long-term sedation in the ICU: a randomized prospective comparison. Intensive Care Med 1997; 23: 1258–63.
19. Wahr JA, Plunkett JJ, Ramsay JG, et al. Cardiovascular responses during sedation after coronary revascularization: incidence of myocardial ischemia and hemodynamic episodes with propofol versus midazolam—Institutions of the McSPI Research Group. Anesthesiology 1996; 84: 1350–60.
20. Michaloudis D, Fraidakis O, Kanoupakis E, et al. Idiopathic prolonged QT interval and QT dispersion: the effects of propofol during implantation of cardioverter-defibrillator. Eur J Anaesthesiol 1999; 16: 842–7.
21. Lehmann A, Boldt J, Rompert R, et al. Target-controlled infusion or manually controlled infusion of propofol in high-risk patients with severely reduced left ventricular function. J Cardiothorac Vasc Anesth 2001; 15: 445–50.
22. Youngs EJ, Shafer SL. Pharmacokinetic parameters relevant to recovery from opioids. Anesthesiology 1994; 81: 833–42.
23. Ebling WF, Lee EN, Stansky DR. Understanding pharmacokinetics and pharmacodynamics through computer simulation: the comparative clinical profile of fentanyl and alfentanil. Anesthesiology 1990; 72: 650–8.
24. Hug CC Jr, Murphy MR. Tissue redistribution of fentanyl and termination of its effects in rats. Anesthesiology 1981; 55: 369–75.
25. Laszlo A, Polic S, Atlee JL, et al. Anesthetics and automaticity in latent pacemaker fibers. I. Effects of halothane, enflurane and isoflurane on automaticity from overdrive suppression in Purkinje fibers derived from canine hearts. Anesthesiology 1991; 75: 98–105.
26. Canessa R, Lema G, Urzua J, et al. Anesthesia for elective cardioversion: a comparison of four anesthetic agents. J Cardiothorac Vasc Anesth 1991; 5: 566–8.
27. Moffitt EA, Scovil JE, Barker RA, et al. The effect of nitrous oxide on myocardial metabolism and hemodynamics during fentanyl or enflurane anesthesia in patients with coronary disease. Anesth Analg 1984; 63: 1071–5.
© 2002 International Anesthesia Research Society
28. Steinbeck G, Dorwarth U, Mattke S, et al. Hemodynamic deterioration during ICD implant: predictors of high-risk patients. Am Heart J 1994; 127: 1064–7.