Routine β-blockade in vascular surgery

doi:10.1016/S0967-2109(03)00127-3

Cardiovascular Surgery, Vol. 11, No. 6, pp. 459–463, 2003  2003 The International Society for Cardiovascular Surgery Published by Elsevier Ltd. All rights reserved. 0967-2109 $30.00

www.elsevier.com/locate/cardiosur

Routine β-blockade in vascular surgery Francesco Torella, Linda de Cossart, Sameh K. Dimitri and Paul R. Edwards Department of Surgery, Countess of Chester Hospital, Liverpool Road, Chester CH2 1UL, UK We have evaluated the safety and efficacy of routine β-blockade for the prevention of cardiac complications in a comprehensive series of patients undergoing major vascular surgery and amputation for atherosclerotic arterial disease. From 1 December 2001 to 31 May 2002, patients received perioperative β-blockade by atenolol. Outcomes in this period were compared to the immediately antecedent 6 months. The main outcome measure was the occurrence of cardiac complications. Fifty-three patients underwent surgery in the first period and 54 in the second. After introduction of routine β-blockade, only one patient suffered cardiac complications compared to 10 in the first period (P = 0.01). There were eight deaths in the first and two in the second period (P = 0.052). On multivariate analysis, treatment with β-blockers was the only variable significantly associated with a decrease in cardiac morbidity (OR = 0.12; 95% CI = 0.002–0.66; P = 0.014). Two patients suffered bronchospasm leading to discontinuation of atenolol. Routine perioperative β-blockade was safe and reduced the occurrence of cardiac complications after vascular surgery.  2003 The International Society for Cardiovascular Surgery. Published by Elsevier Ltd. All rights reserved. Keywords: β-blockers, vascular surgery, myocardial infarction, cardiac failure, arrhythmia

Introduction Mortality after vascular surgery has been reported at 6–14% in some studies [1–6]. The high prevalence of ischaemic heart disease underlies this, resulting in myocardial infarction, cardiac failure or arrhythmia [6–9]. Significant reduction of such complications has been reported by the use of perioperative βblockers in selected patients [10–13]. On 1 December 2001, we introduced routine β-blockade for all major arterial reconstructions and major amputations for atherosclerotic disease (occlusive or aneurysmal). To evaluate the efficacy and safety of this policy, we prospectively audited delivery of care and outcomes for a period of 6 months (up to 31 May 2002) and compared these with the results dur-

Correspondence to: F. Torella, 5 Hall Lane, Kelsall, Nr Tarporley, Cheshire CW6 0QY, UK. Tel.: +44-1829-751856; fax: +44-1244365139; e-mail: [email protected]

CARDIOVASCULAR SURGERY

DECEMBER 2003 VOL 11 NO 6

ing the immediately antecedent 6-month period (from 1 June 2001 to 30 November 2001).

Patients and methods Patients included in this audit were those undergoing scheduled or urgent arterial reconstructions, or major amputation for atherosclerotic arterial disease. Those undergoing emergency surgery (within 12 h of admission to hospital), vascular access for dialysis, minor amputations, or surgery for non-atherosclerotic disease were excluded. Beta-blockade with atenolol 50 mg once daily was initiated before commencement of surgery, but anaesthetists and surgeons were free to alter dosage according to clinical needs. Patients already on β-blockers were free to alter dosage according to clinical needs. Patients already on β-blockers were allowed to continue their own medications. Those in whom the enteral route could not be utilized received intravenous atenolol 5 mg once daily or according to clinical needs, up to a maximum dose of 10 mg twice daily. No target 459

Routine β-blockade in vascular surgery: F. Torella et al.

heart rate was predefined, but β-blockade was omitted if the systolic arterial pressure fell below 100 mmHg or the heart rate fell below 45 bpm. Atenolol was not given to patients with known contraindications, including asthma, uncontrolled cardiac failure, II or III degree AV block, sick sinus syndrome, Prinzmetal’s angina, myasthenia gravis, sinus bradycardia (⬍50 bpm), concurrent treatment with verapamil and hypersensitivity. Beta-blockade was discontinued on the seventh post-operative day or on discharge (whichever came first) unless otherwise indicated, or if significant side effects occurred. Correct administration of β-blockers was monitored by reviewing drug administration charts in all cases. The main outcome measure was the occurrence of cardiac complications defined as follows: (I) myocardial infarction, suspected on clinical grounds and/or electrocardiographic ischaemic changes and confirmed by Troponin I assay (⬎0.1 µg/l) or postmortem findings. (II) Left ventricular failure, with clinical and radiological signs of pulmonary oedema requiring long-term (⬎4 days) modification in treatment with cardiac medications or diuretics, and/or confirmed by a deterioration in left ventricular function on echocardiography. (III) New arrhythmia, including atrial fibrillation/flutter, II–III degree atrioventricular block, supraventricular/ventricular tachycardia or ventricular fibrillation requiring pharmacological treatment for at least 12 h, defibrillation, cardioversion or pacing. Secondary outcome measures were in hospital or 30-day mortality (whichever came latest), non-cardiac complications and side effects of β-blockade. Statistical analysis Data were analysed on an intention to treat basis. Proportions were presented with absolute numbers and percentages, and compared with the chi square test with a continuity correction for 2 × 2 tables, or the Fisher’s exact test when cell counts were low. Continuous variables were presented with medians and inter-quartile ranges (IQR) and compared with the Mann–Whitney U-test. Two analyses were performed: initially, the two 6-month periods were compared; subsequently, factors associated with cardiac complications were identified by univariate analyses. Factors of potential relevance (Pⱕ0.2) were then entered in a multivariate logistic regression model to confirm independent association with cardiac morbidity.

Results Fifty-three patients were audited in the first period and 54 in the second. The two groups were comparable for age, sex and perioperative cardiac risk factors, however, more patients in the second period 460

were on a statin (Table 1). Three patients in the second period did not receive perioperative β-blockade, two because of severe asthma and one in violation of the protocol. Beta-blockers were correctly administered in all the others. Only one patient suffered cardiac complications in the second 6-month period (fatal myocardial infarction and left ventricular failure), compared to 10 in the first period (P = 0.01). These included six myocardial infarctions (two fatal, after a carotid endarterectomy and a major amputation), five episodes of left ventricular failure and five episodes of arrhythmia (one in an amputee who subsequently died of a calculous cholecystitis). One non-fatal myocardial infarction in the first period occurred after aortic surgery in a woman on long-term atenolol. Preoperative factors with a possible association with cardiac morbidity were hypertension (P = 0.19) and previous or current smoking (P = 0.2) (Table 2). Other potential associations were also found with use of nitrates (P = 0.18) and use of digoxin (P = 0.15) (Table 3). When all these variables were entered in a multivariate model, only perioperative β-blockade was found to Table 1 Preoperative risk factors and use of cardiovascular medications before and after introduction of routine β-blockade

Median age (IQR) Men/women Operation CEA Aortic surgery Infra-inguinala Amputation Elective/urgent surgery ASA grade II III IV Comorbidity Ischaemic heart disease Respiratory disease Diabetes Hypertension Cardiovascular drugs β-Blockers Nitrates Ca++ channel blockers ACE inhibitorsb Antiplatelet agents Statins Diuretics Digoxin Smoking habit Never Ex Current

Before βblockers

After βblockers

P

74 (66–79) 39/14

72 (65–78) 32/22

0.27 0.17 0.34

10 8 22 13 37/16

12 15 17 10 32/22

12 34 7

6 39 9

24 8 19 25

27 10 19 19

0.77 0.83 1 0.29

8 9 17 12 34 15 17 3

51 6 12 14 43 30 22 4

⬍0.001 0.55 0.35 0.87 0.12 0.008 0.47 1 0.85

10 15 28

10 18 26

0.35 0.28

CEA = carotid endarterectomy. a Includes procedures from iliac inflow. b Includes three patients on angiotensin II receptor blockers.

CARDIOVASCULAR SURGERY

DECEMBER 2003 VOL 11 NO 6

Routine β-blockade in vascular surgery: F. Torella et al. Table 2 Univariate associations between preoperative risk factors and perioperative cardiac complications Cardiac complications

Age ⬍65 ⱖ65 Sex Men Women Operation CEA Aortic surgery Infra-inguinala Amputation Timing of surgery Elective Urgent ASA grade II III IV Ischaemic heart disease No Yes Respiratory disease No Yes Diabetes No Yes Hypertension No Yes Smoking habit Never Ex Current

No

Yes

18 78

3 8

63 33

8 3

20 21 37 18

2 2 2 5

62 34

7 4

16 65 15

2 8 1

49 47

7 4

79 17

10 1

62 34

7 4

59 37

4 7

P

0.45 0.75 0.21

1 0.85

0.53 0.69

Cardiac complications

β-Blockers No Yes Nitrates No Yes Ca++ channel blockers no yes ACE inhibitorsa No Yes Antiplatelet agents No Yes Statins No Yes Diuretics No Yes Digoxin No Yes

No

Yes

39 57

9 2

84 12

8 3

71 25

7 4

72 24

9 2

25 71

5 6

54 42

8 3

61 35

7 4

91 5

9 2

P

0.01 0.18 0.49 1 0.29 0.35 1 0.15

1 a

Includes three patients on angiotensin II receptor blockers.

0.19 0.2 20 28 48

0 5 6

CEA = carotid endarterectomy. a Includes procedures from iliac inflow.

be independently associated with a decrease in cardiac complications, with an odds ratio (95% CI) of 0.12 (0.02–0.66) (P = 0.014). In this study, perioperative β-blockade resulted in a 16% absolute reduction in cardiac complications (from 19% to 3%); if confirmed, this would be equivalent to a number needed to treat of 6.25. Four patients on βblockers had significant intra-operative bradycardia requiring a single dose of intravenous glycopyrrholate or atropine, but continued to receive post-operative β-blockade without further ill effects. Two other patients suffered significant bronchospasm 2 and 5 days after an aortic aneurysm repair and an antegrade aorto-mesenteric bypass, respectively, leading to discontinuation of atenolol but no other consequence. The overall complication rate, including patients who subsequently died, was 43% (23/53) in the first and 35% (19/54) in the second period (P = 0.5). There were eight deaths in the first and two CARDIOVASCULAR SURGERY

Table 3 Univariate associations between use of cardiovascular medications and perioperative cardiac complications

DECEMBER 2003 VOL 11 NO 6

in the second period (P = 0.052), but most deaths were not attributable to cardiac causes. Additional causes of deaths were respiratory failure (3), cerebral haemorrhage after carotid endarterectomy (1), intraoperative coagulopathy in a patient with liver cirrhosis (1), pulmonary embolism (1) and gastrointestinal haemorrhage in a patient with advanced gastric cancer who underwent a femoro-popliteal bypass for severe ischaemia (1). Five deaths occurred after major amputations, two after infra-inguinal and carotid surgery, respectively, and one after aortic surgery. Mortality was 6% (4/69) after elective and 19% (6/39) after urgent surgery.

Discussion The compelling evidence that β-blockers improve cardiac function and, in some circumstances, reduce post-operative cardiac events, caused us to consider how to use β-blockade in practice. On a pragmatic basis and in absence of randomised controlled trials on unselected patients, we chose to initiate routine β-blockade in a consecutive series of patient in our unit. Careful audit was maintained. Following the introduction of routine β-blockade, we noted a reduced incidence of cardiac complications. Perhaps more importantly, atenolol was relatively safe, with six patients suffering temporary side effects, leading 461

Routine β-blockade in vascular surgery: F. Torella et al.

to discontinuation of treatment in only two. Most of our patients requiring vascular surgery did not have contraindications to β-blockade. Our results are in agreement with the literature, although most published studies did not focus particularly on vascular surgery. Already in 1987 however, a case-control trial by Pasternack et al. [15] showed a significant reduction in the incidence of myocardial infarction and arrhythmiae in patients treated with perioperative metoprolol undergoing aortic aneurysm repair. More recently, Raby et al. randomised 26 high risk vascular surgical patients, identified by 24-h Holter monitoring, to perioperative esmolol or placebo [13]. Although the trial was too small for hard clinical endpoints, it demonstrated a marked reduction in the proportion of patients experiencing perioperative myocardial ischaemia. Interestingly, there have been similar findings in patients undergoing percutaneous coronary artery intervention, a group with universal ischaemic heart disease by definition. A recent large observational study on 1675 consecutive patients demonstrated a significantly lower incidence of perioperative rise in creatine kinase-MB (13.2% versus 22.1%) and mortality (0.78% versus 1.96%) with β-blockade [16]. Two further randomised trials support the use of perioperative β-blockers in vascular surgery. Mangano et al. showed that perioperative atenolol dramatically reduced mortality and cardiac events for up to 8 months after major surgery (mainly vascular) in patients with either proven cardiac disease or at least two cardiac risk factors [11]. Poldermans et al. demonstrated a reduction in cardiac complications from 35% to 3.5% by perioperative bisoprolol in patients with reversible perfusion defects on dobutamine stress echocardiography undergoing vascular surgery [12]. Thus, both trials, in contrast to our series, involved a degree of patients selection. Despite recent advances, there is persistent controversy about the preoperative management of cardiac risk. Current thinking seems to support the use of clinical assessment based on scoring systems followed by selective non-invasive cardiac investigations [10]. The most recently proposed scoring systems include the Revised Cardiac Risk Index [17] and the criteria used by Mangano et al. in their trial [11]. These systems may have significant limitations. The Revised Cardiac Risk Index was developed in a single American institution and has not been widely validated in other settings. Furthermore, it performs poorly in aortic reconstructions [17], which carry the highest cardiac risk in vascular surgery [1–3, 6, 9]. Mangano’s criteria were developed on a relatively small number of patients (n = 200) and, again, require wider validation. As they are less strict, their specificity would be expected to be lower than that of the Revised Cardiac Risk Index. Interestingly, 93% of our patients would have been classed at high risk 462

according to Mangano’s criteria, demonstrating the high prevalence of cardiac disease and risk factors among those undergoing vascular surgery. Ideally, patients identified at risk should undergo non-invasive cardiac investigations. However, in the UK, practice varies widely [18]. This reflects not only lack of agreement on the best test but also variable access to diagnostic facilities. Dobutamine stress echocardiography and dipyridamole-thallium cardiac scintigraphy are probably the most sensitive cardiac investigations for ischaemic complications, but specificity is low [19, 20]. They are not universally available in UK hospitals (and possibly in many other countries) and require skilled interpretation hence published results are not automatically reproducible. The waiting time for both tests in our institution is prohibitive and, as over one third of our cases require urgent surgery, their routine use is not an option. The role of coronary angiography and coronary revascularization is uncertain, hence cardiac catheterisation is not currently recommended [21]. To date, the use of routine perioperative βblockade may offer the best possible outcome, although a proportion of patients will be treated unnecessarily. Several β-blockers have been used in perioperative myocardial protection, including metoprolol [12, 15], atenolol [11, 22], esmolol [13] and bisoprolol [12], but there is little evidence that one agent is superior to another. We chose atenolol as it is cheap, available both for oral and intravenous administration, and widely investigated previously [11, 22]. A post-operative treatment of 7 days was deemed appropriate as most cardiac events are expected to occur during the first post-operative week; furthermore, the same treatment duration has been used with success by the Study of Perioperative Ischaemia Research Group [11, 22]. Where indicated, we continued treatment beyond 7 days, and all our patients with ischaemic heart disease who were not on a βblocker pre-operatively were discharged on atenolol. We attempted to initiate β-blockade as early as possible but, in urgent cases, atenolol was usually started within 24 h of surgery. Introducing β-blockade well in advance offers the theoretical advantage of detecting side effects before the operation. Interestingly however, our two patients who discontinued atenolol for bronchospasm suffered this complication post-operatively, 4 and 18 days after commencing treatment, respectively. Other investigators have introduced β-blockade on induction of anaesthesia, with little evidence of major complications [11, 22]. In our pragmatic approach, we did not attempt to achieve a target heart rate, as this may not be a practical approach in routine clinical practice. There is also little evidence to suggest that a target heart rate is desirable. Recent studies indicate that each patient CARDIOVASCULAR SURGERY

DECEMBER 2003 VOL 11 NO 6

Routine β-blockade in vascular surgery: F. Torella et al.

has an individual ischaemic threshold at a particular heart rate [13], but we did not routinely perform specific investigations to identify such threshold. Our study has important limitations. The most obvious relates to its design, which used historical controls for comparison and a heterogeneous group of procedures. Also, despite statistical significance, the relatively small sample size could not accurately measure the effect of the treatment on cardiac complications. We feel that the most significant finding was of the safety and applicability of perioperative βblockade, with little evidence of serious adverse events. Thus, awaiting further evidence, it continues to be part of our preoperative preparation. The reduction in cardiac complications we observed could be due to factors other than β-blockers. In particular, there was a significantly higher use of statins in patients receiving routine β-blockade, following the findings of the Heart Protection Study [14]. These agents are known to stabilise atherosclerotic plaques and could contribute to perioperative myocardial protection. Our analysis however failed to show such protective effect (Table 3). Furthermore, perioperative β-blockade was the only factor associated with a reduction in cardiac complications on multivariate analysis. Further research is clearly required. If a protective effect of routine β-blockade is confirmed, it may be possible to reduce the need for cardiac investigations before vascular surgery, with profound implication on patient management, costs and the organization of vascular services. If no significant effect from βblockade is shown, the argument for mandatory preoperative evaluation will be strengthened, enforcing a minimum requirement in terms of diagnostic facilities for established and prospective vascular units. Several unanswered questions remain, including issues on patient selection, choice of drug, duration of treatment, and whether a target heart rate is necessary or not. Meanwhile, our practical experience would suggest that routine β-blockade is relatively safe and should be offered to this complex and high risk group of patients sooner rather than later.

5. 6. 7. 8.

9.

10. 11.

12.

13. 14. 15. 16.

17.

18. 19.

References 1. Lawrence, P. F., Gazak, C., Bhirangi, L. et al. The epidemiology of surgically repaired aneurysms in the United States. J Vasc Surg, 1999, 30, 632–640. 2. Mortality results for randomised controlled trial of early elective surgery or ultrasonographic surveillance for small abdominal aortic aneurysms. The UK Small Aneurysm Trial Participants. Lancet 1998, 353, 1649–55. 3. Bayly, P. J., Matthews, J. N., Dobson, P. M. et al. In-hospital mortality from abdominal aortic surgery in Great Britain and Ireland: Vascular Anaesthesia Society audit. Br J Surg, 2001, 88, 687–692. 4. Seeger, J. M., Pretus, H. A., Carlton, L. C. et al. Potential pre-

CARDIOVASCULAR SURGERY

DECEMBER 2003 VOL 11 NO 6

20. 21.

22.

dictors of outcome in patients with tissue loss who undergo infrainguinal vein bypass grafting. J Vasc Surg, 1999, 30, 427– 435. Harrington, M. E., Harrington, E. B., Haimov, M. et al. Axillofemoral bypass: compromised bypass for compromised patients. J Vasc Surg, 1994, 20, 195–201. Wong, J. C., Torella, F., Haynes, S. L. et al. Autologous versus allogeneic transfusion in aortic surgery: a multicenter randomized clinical trial. Ann Surg, 2002, 235, 145–151. Pasternack, P. F., Imparato, A. M., Baumann, F. G. et al. The hemodynamics of beta-blockade in patients undergoing abdominal aortic aneurysm repair. Circulation, 1987, 76(III), 1–7. Paciaroni, M., Eliasziw, M., Kappelle, L. J. et al. Medical complications associated with carotid endarterectomy. North American symptomatic carotid endarterectomy trial (NASCET). Stroke, 1999, 30, 1759–1763. Haggart, P. C., Adam, D. J., Ludman, P. F. et al. Comparison of cardiac troponin I and creatine kinase ratios in the detection of myocardial injury after aortic surgery. Br J Surg, 2001, 88, 1196–1200. Auerbach, A. D. and Goldman, L. Beta-blockers and reduction of cardiac events in noncardiac surgery: scientific review. JAMA, 2002, 287, 1435–1444. Mangano, D. T., Layug, E. L., Wallace, A. et al. Effect of atenolol on mortality and cardiovascular morbidity after noncardiac surgery. Multicenter Study of Perioperative Ischemia Research Group. N Engl J Med, 1996, 335, 1713–1720. Poldermans, D., Boersma, E., Bax, J. J. et al. The effect of bisoprolol on perioperative mortality and myocardial infarction in high-risk patients undergoing vascular surgery. Dutch Echocardiographic Cardiac Risk Evaluation Applying Stress Echocardiography Study Group. N Engl J Med, 1999, 341, 1789–1794. Raby, K. E., Brull, S. J., Timimi, F. et al. The effect of heart rate control on myocardial ischemia among high-risk patients after vascular surgery. Anesth Analg, 1999, 88, 477–482. MRC/BHF Heart Protection Study of cholesterol lowering with simvastatin in 20,536 high-risk individuals: a randomised placebo-controlled trial. Lancet 2002, 360, 7–22. Pasternack, P. F., Imparato, A. M., Baumann, F. G. et al. The hemodynamics of β-blockade in patients undergoing aortic aneurysm repair. Circulation, 1987, 76(Suppl III), 1–7. Sharma, S. K., Kini, A., Marmur, J. D. et al. Cardioprotective effect of previous β-blocker therapy in reducing creatine kinaseMB elevation after coronary intervention. Circulation, 2000, 102, 166–172. Lee, T. H., Marcantonio, E. R., Mangione, C. M. et al. Derivation and prospective validation of a simple index for prediction of cardiac risk of major noncardiac surgery. Circulation, 1999, 100, 1043–1049. Michaels, J. A., Kayne, S. P. K. and Galland, R. B. A survey of methods used for cardiac risk assessment prior to major vascular surgery. Eur J Vasc Endovasc Surg, 1996, 11, 221–224. Das, M. K., Pellikka, P. A., Mahoney, D. W. et al. Assessment of cardiac risk before nonvascular surgery: dobutamine stress echocardiography in 530 patients. J Am Coll Cardiol, 2000, 35, 1647–1653. Boucher, C. A., Brewster, D. C., Darling, R. C. et al. Determination of cardiac risk by dipyridamole-thallium imaging before peripheral vascular surgery. N Engl J Med, 1985, 312, 389–394. Guidelines for assessing and managing the perioperative risk from coronary artery disease associated with major noncardiac surgery. American College of Physicians. Ann Intern Med 1997, 127, 309–312. Wallace, A., Layug, B., Tateo, I. et al. Prophylactic atenolol reduces postoperative myocardial ischemia. Anesthesiology, 1998, 88, 7–17.

Paper accepted 1 July 2003

463