Thoracic Epidural Anesthesia for Coronary Bypass Surgery Affects Autonomic Neural Function and Arrhythmias

Yashiki, Noriyoshi MD; Watanabe, Go MD, PhD; Tomita, Shigeyuki MD, PhD; Nishida, Satoru MD, PhD; Yasuda, Tamotsu MD, PhD; Arai, Sadahiko MD, PhD

Innovations: Technology & Techniques in Cardiothoracic & Vascular Surgery:
doi: 10.1097/01.gim.0000198517.54063.74
Original Articles

Background: In recent years, the invasiveness of coronary reconstruction has been markedly reduced. Awake off-pump coronary artery bypass (AOCAB), coronary bypass surgery with thoracic epidural anesthesia (TEA) without general anesthesia and cardiopulmonary bypass), has been reported in the literature. Because the details of this technique are still unclear, we evaluated its usefulness by examining the autonomic neural state and the incidence of arrhythmia.

Methods: Fifty-five patients who underwent elective coronary artery bypass grafting (CABG) between April and December 2003 were included in the study. Patients who underwent CABG under high TEA alone comprised group A, those who underwent CABG under general anesthesia combined with TEA comprised group B, and those who underwent CABG under general anesthesia alone comprised group C. Holter electrocardiography was performed before and after surgery, and perioperative electrocardiograms were recorded (before surgery and during surgery, postoperative days 0-3, and postoperative day 7). On obtained electrocardiograms, the autonomic neural state was evaluated by analysis of heart rate variability, and the incidence of atrial fibrillation.

Results: Concerning the autonomic neural state, sympathetic inhibition was observed during TEA in both groups A and B. After discontinuation of TEA, sympathetic activity recovered. Vagal activity was not inhibited in group A, but decreased during surgery and gradually recovered after surgery in group B. Evaluation of the balance between sympathetic and vagal activities showed that sympathetic activity became predominant rapidly on postoperative day 2 in group B but gradually after surgery in group A. The incidence of postoperative atrial fibrillation was the highest in group B.

Conclusions: In AOCAB, because there is no vagal inhibition, vagal dominance can be maintained after surgery. This may be associated with the lower incidence of postoperative atrial fibrillation in group A compared with group B. Further studies are necessary to evaluate the details of AOCAB.

Author Information

From the Department of General and Cardiothoracic Surgery, Kanazawa University School of Medicine, Kanazawa, Japan.

Address correspondence and reprint requests to Noriyoshi Yashiki, MD, Department of General and Cardiothoracic Surgery, Kanazawa University School of Medicine, Kanazawa, Japan; e-mail:

Article Outline

IIn recent years, the invasiveness of coronary reconstruction has been markedly reduced by doing coronary bypass surgery without cardiopulmonary bypass, called off-pump coronary artery bypass (OPCAB).2–4 Karagoz et al. reported Awake OPCAB (AOCAB) performed with thoracic epidural anesthesia and without general anesthesia.1 Our institution introduced this technique in 2003. We have evaluated the autonomic neural state and arrhythmias before, during, and after AOCAB.

Back to Top | Article Outline


The subjects were 55 patients who underwent elective coronary artery bypass grafting (CABG) between April and December 2003. Patients not receiving anticoagulation therapy immediately before surgery and those who desired thoracic epidural anesthesia (TEA) alone and in whom the operation time was expected to be short were managed by TEA alone (group A). Patients not receiving anticoagulant therapy and not wanting to be awake during operation had CABG under general anesthesia combined with high TEA (group B). Patients who received anticoagulation therapy until immediately before surgery were managed by general anesthesia alone because of the risk of paraplegia due to epidural hematoma (group C). Informed consent was obtained. Patients with acute myocardial infarction or preoperative atrial fibrillation (Af), receiving antiarrhythmic drugs other than β-blockers or having emergency operation or minimally invasive direct coronary artery bypass surgery were excluded. Oral administration of aspirin was discontinued 1 week before surgery. Oral β-blockers were discontinued the day before surgery.

In groups A and B, an anesthesiologist inserted an epidural catheter into T1-2 or T2-3 using a Tuohy needle (B. Braun Aesculap Japan. Co., Ltd.) in the operation room on the day before surgery. On the day of surgery, no atropine sulfate or related drugs were given. After patients entered the operation room, continuous administration of a mixture of 2% lidocaine (40 mL) and fentanyl hydrochloride (250 μg) was initiated via the epidural catheter. In group A, when the radial artery or great saphenous vein were was used, axillary nerve block or femoral nerve block, respectively, was added. After confirmation of the anesthesia range, surgery was initiated. In group B, anesthesia was routinely performed. In group C, no pretreatment with drugs such as atropine sulfate was given and anesthesia was routinely performed.

After routine median sternotomy, the internal thoracic artery was prepared. In group A (spontaneous respiration), attention was paid to the possible development of pneumothorax. A stabilizer was used for the visual field as in routine OPCAB. For coronary perfusion during anastomosis, a coronary active perfusion system with blood collected from the femoral artery was used.5

Two-channel 24-hour Holter electrocardiograms were recorded on a memory card using a digital Holter recorder FM-100 (Fukuda Denshi, Japan), before, during, and after surgery over 4 consecutive days and on postoperative day 7. All recordings were stored using the Ambulatory ECG analysis system SCM-5000 (Fukuda Denshi), after which arrhythmia and the activity of the autonomic nervous system were analyzed.

Autonomic nervous system activity was evaluated by examining the heart rate variability (HRV) taken from the Holter recordings using a commercially available software algorithm (HPS-RRA(Win), Fukuda Denshi). During the analysis, only normal beats were measured, and artifacts or extra systolic beats were eliminated. RR interval sequences of 512 seconds were sampled, and a fast Fourier transform was used to estimate the power spectral densities of the RR interval variability. Frequency domain measures of RR interval variability were computed by integrating the power spectrum over the frequency intervals. The power spectra (milliseconds squared) were quantified by measuring the areas across the following frequency bands: low frequency (LF: range 0.04–0.15 Hz), reflecting modulation from the sympathetic and parasympathetic nervous system; high frequency (HF: range 0.15–0.4 Hz), reflecting modulation by the parasympathetic nervous system; and the LF/HF ratio, interpreted as a marker of sympathovagal balance. The analysis was performed for a specified period, which was measured when each patient was sleeping at night before and after surgery (postoperative days 0, 1, 2, 3, and 7). During surgery, it was measured over 4 periods; entry into the operating room, during the beginning of surgery, during anastomosis of the coronary artery, and while closing the wound. If the heart rhythm was Af while HRV analysis was being performed, the analysis was canceled. This only occurred in 1 case.

Atrial fibrillation was defined as an irregular narrow complex rhythm in the absence of bundle branch block without discrete P waves if it persisted for more than ten beats. The incidence of Af, which was revealed by analyzing Holter ECG, was compared among the groups before surgery, and for postoperative days 0, 1, 2, 3, and 7. If Af occurred, β-blocker and digitalis were given for its treatment.

For the examination of HRV, each value was compared with the preoperative value for the same group and the value of AOCAB for each respective time. The incidence of Af was then compared with the value of AOCAB for each respective time. All data for the 3 groups are expressed as means and standard deviations. Discrete variables were analyzed by chi-square analyses, and findings among groups were compared by ANOVA. Statistical analyses were conducted using SPSS software version 10.1.3 J (SPSS Inc.), for which a value of P less than 0.05 was considered to indicate significance.

Back to Top | Article Outline


Data on 55 patients were obtained (AOCAB, group A, n = 17; OPCAB with TEA, group B, n = 21; OPCAB without TEA, group C, n = 17). In 5 patients, surgery was initiated under TEA alone, but intubation or a laryngeal mask became necessary due to development of pneumothorax. Of the 5 patients, 3 were included in group A because surgery could be completed with maintenance of spontaneous respiration. The other 2 patients were included in group B because of induction of general anesthesia.

Preoperative data on the patients are shown in Table 1. UAP was frequently observed in group C, but the other parameters did not differ among the groups.

Data during surgery are shown in Table 2. The number of anastomoses was significantly higher in group B than in group A. The operation time was the shortest in group A.

The state of the use of β-blockers is shown in Table 3. The percentage of patients postoperatively treated with β-blockers was higher in group B than in group A.

Table 4 shows data on TEA. The dose of lidocaine and the time of its use did not differ between groups A and B. TEA was discontinued on postoperative day 2.

Figure 1 shows the results on LF. LF decreased after induction of anesthesia in all the 3 groups and became high at the time of coronary anastomosis (Ope3) in group C. In groups A and B, LF was low on the day of surgery and postoperative day 1 but began to increase on postoperative day 2 when TEA was discontinued.

Figure 2 shows the results on HF. In group A, no changes were observed in HF. This may be because of no influences of TEA on the vagus nerve. In group C, HF remained low even on postoperative day 7, which may be the influence of general anesthesia. group B showed inhibition of HF but earlier recovery than group C. This may be because general anesthesia was shallow due to the combined use of TEA, and vagal inhibition was slight.

The results of LF/HF are shown in Figure 3. LF/HF was low after the initiation of surgery in groups A and B under TEA but began to increase on postoperative day 2 when TEA was discontinued. In particular, group B showed recovery on postoperative day 2 to a level that was similar to the preoperative level and higher than the level in group A. In group C, postoperative LF/HF remained at a level similar to the preoperative level.

The incidence of postoperative Af is shown in Figure 4. The incidence of Af was high on postoperative days 2 and 3, particularly in group B.

Back to Top | Article Outline


There have been various studies on the state of sympathetic activity and arrhythmia. It has been reported that ventricular arrhythmia, which develops during acute myocardial ischemia, tends to be induced by sympathetic excitation,6 and sympathetic stimulation and acute myocardial ischemia tend to induce ventricular fibrillation while vagal stimulation increases the threshold for ventricular fibrillation.7–9 These studies suggest that TEA inhibits only sympathetic activity without inhibiting vagal activity, which has antiarrhythmic effects.10

There is concern about a decrease in cardiac function due to TEA. However, a previous study showed no changes in coronary perfusion pressure and no influence on cardiac output, stroke volume, systemic vascular resistance, or pulmonary vascular resistance.11

With TEA, inhalation anesthesia during surgery can be reduced and early removal of the tube is possible, which reduces respiratory complications.11,12 However, because the details of AOCAB were unclear, we evaluated the autonomic neural state and the incidence of arrhythmia in this study.

Routine general anesthesia not only produces loss of pain and consciousness but also inhibits autonomic activity, including both sympathetic and vagal activities. Therefore, in group C, postoperative sympathetic and vagal activities were low.

When general anesthesia is performed in combination with TEA, it can be reduced because TEA considerably inhibits pain. In addition, because TEA inhibits sympathetic activity but not vagal activity, there is less inhibition of vagal activity with combined TEA and general anesthesia than with general anesthesia alone. Therefore, in group B, vagal activity remained dominant during TEA, but sympathetic activity began to recover on postoperative day 2 when TEA was discontinued, rapidly becoming dominant.

When CABG is performed under TEA alone, sympathetic activity alone is inhibited without inhibition of vagal activity. In group A, vagal activity was naturally dominant while TEA was used immediately after surgery. In this group, sympathetic activity recovered after discontinuation of TEA but did not rapidly become dominant unlike group B because of the absence of vagal inhibition.

Atrial fibrillation often develops after CABG, but its cause is unclear. Studies on the association between Af developing after various surgeries and autonomic activity have shown that Af develops due to vagal excitation in the absence of heart disease12 and due to sympathetic excitation in the presence of heart disease.13–15 This is definitely one of the causes, though the cause of postoperative Af can not be explained by autonomic balance alone.

In group B, the incidence of Af was especially high on postoperative day 2, which may be because sympathetic activity rapidly became dominant on this day due to the discontinuation of TEA. In group A, the incidence of Af was relatively low, possibly because vagal activity was dominant even after recovery of sympathetic activity after the discontinuation of TEA.

In conclusion, the incidence of Af was low after AOCAB, which may be due to dominant vagal activity. AOCAB, which requires more advanced skills than OPCAB, may not be indicated in all patients. For patients who cannot undergo general anesthesia, AOCAB, which may be associated with a low incidence of postoperative Af, should be considered. Because this technique is still new, further studies are necessary to evaluate its effect on the incidence of postoperative Af, long-term graft patency, and long-term prognosis.

Back to Top | Article Outline


1. Karagoz HY, Sonmez B, Bakkaloglu B, et al. Coronary artery bypass grafting in the conscious patient without endotracheal general anesthesia. Ann Thorac Surg. 2000;70:91–6.
2. Kolessov VI. Mammary artery-coronary artery anastomosis as method of treatment for angina pectoris. J Thorac Cardiovasc Surg. 1967;54:535–44.
3. Calafiore AM, Giammarco GD, Teodori G, et al. Left anterior descending coronary artery grafting via left anterior small thoracotomy without cardiopulmonary bypass. Ann Thorac Surg. 1996;61:1658–63.
4. Watanabe G, Misaki T, Kotoh K, et al. Multiple minimally invasive direct coronary artery bypass grafting for the complete revascularization of the left ventricle. Ann Thorac Surg. 1999;68:131–6.
5. Kamiya H, Watanabe G, Doi T, et al. Coronary active perfusion system can maintain myocardial blood flow and tissue oxygenation. Eur J Cardiothorac Surg. 2002;22:410–4.
6. Corr PB, Gillis RA. Autonomic neural influences on the dyshythmias resulting from myocardial infarction. Circ Res. 1978;43:1–9.
7. Kent KM, Smith ER, Redwood DR, et al. Electrical stability of acutely ischemic myocardium. Influence of heart rate and vagal stimulation. Circulation. 1973;47:291–8.
8. Yoon MS, Han J, Tse WW, et al. Effects of vagal stimulation, atropine, and propranorol on fibrillation threshold of normal and ischemic ventricles. Am Heart J. 1977;93:60–5.
9. Zuanetti G, De Ferrari GM, Priori BG, et al. Prospective effect of vagal stimulation on reperfusion arrhythmias in cats. Circ Res. 1987;61:429–35.
10. Turfrey DJ, Ray DA, Sutcliffe NP, et al. Thoracic epidural anaesthesia for coronary artery bypass graft surgery. Anaesthesia. 1997;52:1090–1113.
11. Blomberg S, Emanuelsson H, Ricksten SE. Thoracic epidural anesthesia and central hemodynamics in patients with unstable angina pectoris. Anesth Analg. 1989;69:558–62.
12. Maarten P, van den Berg MD, Haaksma J, et al. Heart rate variability in patients with atrial fibrillation is related to vagal tone. Circulation. 1997;96:1209–16.
13. Herweg B, Dalal P, Nagy B, Schweitzer P. Power spectral analysis of heart period variability of preceding sinus rhythm before initiation of paroxysmal atrial fibrillation. Am J Cardiol. 1998;82:869–74.
14. Dimmer C, Tavernier R, Gjorgov N, et al. Variations of autonomic tone preceding onset of atrial fibrillation after coronary artery bypass grafting. Am J Cardiol. 1998;82:22–5.
15. KalmanJM, Munawar M. Atrial fibrillation after coronary artery bypass grafting is associated with sympathetic activation. Ann Thorac Surg. 1995;60:1709–15.

Awake off-pump coronary artery bypass; Thoracic epidural anesthesia; autonomic neural state; Heart rate variability; Atrial fibrillation

© 2005 Lippincott Williams & Wilkins, Inc.