Winter 2005 Articles:
Perioperative Beta-Blockade: Protocols Before Proof?Author: Dr Karen Stuart-Smith, Consultant Anaesthetist, Glan Clwyd Hospital, Denbighshire, Wales.
Introduction
The efficacy of beta-blockers in limiting the complications of acute myocardial infarction (AMI) is well-documented.1 The efficacy of these drugs in the prevention of cardiac complications in patients undergoing non-cardiac surgery is far more controversial.2
The fundamental problem is that although it may be intellectually pleasing to consider that beta-blockade would reduce cardiovascular incidents associated with anaesthesia and surgery, the truth is that none of the available studies or meta-analyses have been able to show that this really happens in practice.3
A prospective, randomised, placebo-controlled trial, the PeriOperative ISchemic Evaluation (POISE) trial is currently recruiting 10,000 patients across 18 countries, to determine whether one beta-blocker, metoprolol CR, can produce a significant reduction in perioperative AMI and cardiac death. The patient population being recruited already fits at least one criterion for increased cardiovascular risk at entry to the trial, thus maximising the possibility of finding a positive treatment benefit (ClinicalTrials.gov, identifier NCT00182039).
The study is still in progress and the trial organisers have advised that practitioners await the outcome of the trial before making the prescription of perioperative beta-blockers routine practice.4 Nevertheless, a recent British government health initiative, the Safer Patient Initiative (SPI), advocated the use of perioperative beta-blockade in ‘high risk’ cardiac patients undergoing non-cardiac surgery. This recommendation has been withdrawn temporarily as several of the NHS Trusts involved in SPI have indicated that they would prefer to await the outcome of the POISE trial.
This article explores the theory behind the use of beta-blockers as perioperative cardioprotective agents, and asks whether the theory justifies the hype. I also discuss whether the POISE trial, in spite of being carefully designed and properly powered, will really be able to highlight a role for beta-blockers in the perioperative setting. Other methods of reducing perioperative cardiac mortality, which are already available to the anaesthetist, are discussed. Finally, it is asked whether SPI will achieve its goal of reducing perioperative cardiac mortality by introducing as yet unproved methods, or whether modification and improvement of current techniques might be more effective.
Assessing perioperative cardiac risk is fraught with problems. Two excellent reviews by Devereaux et al highlight the problem,5,6 and these are discussed briefly here. The first problem is that no recent prospective study (i.e. in the 21st Century) has been conducted to estimate the scale of the problem.
Older studies conducted some years ago suffer from the following problems when assessing present day risk:
- the average age of the current population is rising rapidly,
- most at-risk patients now arrive at hospital on some form of cardiac medication
- more complex and ambitious surgery is now undertaken in elderly patients than formerly
- HDU/ITU facilities are more readily available than in, say, the mid- 1980s when many of the older studies were conducted.
Each of these factors could affect the risk of perioperative AMI in any direction. For example, an older patient with ischaemic heart disease might undergo an extensive operation for bowel cancer but because of intraoperative invasive monitoring and a planned post-operative admission to ITU could go home without any adverse event. Conversely an apparently fit 60-year old could undergo a routine orthopaedic procedure and have an AMI three days later, almost certainly due to occult heart disease unrecognised even by the patient. Which one of these should have had the pre-operative beta-blocker?
A second problem highlighted by Devereaux and colleagues is that there are no standard diagnostic criteria for perioperative AMI. This is complicated by the fact that many post-operative MIs are probably silent. Diagnostic tests are usually only performed on those with obvious symptoms. These authors propose a set of diagnostic criteria to define the occurrence of a perioperative MI.
These criteria include ischaemic symptoms, a typical pattern of troponin elevation, ECG changes, and new wall-motion abnormalities on echocardiography. As can be surmised, performance of these tests on a large group of patients, in the hope of detecting both overt and occult AMIs, would be a substantial drain on the resources of the average NHS Trust. As a part of a nationwide initiative properly funded by central government, such a study would have huge epidemiological value. By contrast, asking individual Trusts to monitor their own perioperative AMI rate, as has recently been suggested by the SPI, is probably expensive, liable to produce inaccurate results, and administratively unworkable.
The next issue is the likely pathophysiology of perioperative cardiac events, and whether beta-blockers are really the best preventative agents in this setting. The presumed protective effect of these drugs is inhibition of the effects of catecholamines released in response to surgical stress.
Beta-blockers reduce heart rate and cardiac contractility, decreasing myocardial oxygen demand. Prolonged diastolic filling aids coronary artery perfusion. Beta-blockers are also anti-arrythmogenic. These actions probably limit myocardial ischaemia1 although direct evidence that this happens in the surgical patient is not available.
Indeed, there is no direct evidence to suggest that raised catecholamine levels are directly linked to perioperative cardiac complications. The evidence is as fuzzy as the long-running debate over the role of steroid therapy in sepsis.
Further, these drugs do cause hypotension and heart failure in some patients, although the incidence of this in the surgical population is not yet clear. The POISE trial will be very helpful in this area. Until this information is available, the best guide seems to be that in perioperative situations where myocardial ischaemia is already present, beta-blocker therapy is very likely to be beneficial, in the same way that these agents are already used in acute myocardial infarction treated in the coronary care unit.
Looking more closely at the pathophysiology of perioperative cardiac ischaemia, the principal causes of cardiac stress are:
- the inevitable and ubiquitous response to anaesthesia and surgery, i.e. the release of inflammatory mediators and the induction of the hypercoagulable state, and
- the more preventable ‘side effects’ such as pain, hypothermia, bleeding, hypoglycaemia, hypoxia and sepsis (the reader is referred to Devereaux et al 2005 for further expansion of these issues).
Most readers would class all these phenomena together as the ‘stress response’ to surgery, and there are many ways in which its impact can already be limited, using current technology, some of it very simple. Routine prescription of subcutaneous unfractionated heparin counteracts the hypercoagulable state. Keeping the patient warm before, during and after surgery, and making sure the oxygen mask stays on the postoperative patient’s face are obvious ways of preventing hypothermia and hypoxia respectively. Good insulin control in both insulin dependent and non-insulin dependent diabetics should be a matter of routine.
Control of bleeding is largely up to the surgeon, but accurate measurement and prompt replacement of blood loss are the anaethetists’ role. The source of pyrexias should be sought and treated. Many of these recommendations are already being pursued by the Safer Patient Initiative, and rightly so.
One aspect of immediate importance to the patient is pain control, which can be indifferent in many parts of the country. Nationwide, we do not train our junior doctors and nurses well enough in the management of post-operative pain and nausea, yet these are the things that matter most to the patient and cause the greatest stress. Before routine beta-blockers are introduced, this area of patient management needs urgent overhaul.
Conclusion
In summary, although perioperative cardiac ischeamia is a significant problem, many questions surround its detection, its prevalence, its aetiology and its management. In addition to its main purpose, the POISE trial should go some way to answering these questions. With regard to beta-blockers as cardioprotective agents in the surgical setting, it is less clear what the outcome of the trial will be. It is likely that certain subgroups of patients with coronary artery disease, such as those undergoing peripheral vascular surgery, may well benefit from perioperative beta-blockade.
The advantages of preemptive beta-blockers in other groups may be less certain, and it is vitally important that we await the conclusions of POISE on this topic. Administering medications on a theoretical basis without awaiting the outcome of a full trial can occasionally have disastrous consequences (see article on selective COX-2 inhibitors, page 30).
In the meantime, there are many areas of patient management, some already recommended by SPI, which will reduce the incidence of perioperative acute myocardial ischaemia.
References
- Lopez-Sendon J et al. Eur Heart J 2004; 25:1341-1362
- Bolsin S and Colson M BMJ 2005; 331:919-920
- Devereaux PJ et al BMJ 2005a; 331:313-321
- Devereaux PJ et al CMAJ 2004; 171:245-247
- Deveraux PJ et al CMAJ 2005b; 173:627-634
- Devereaux PJ et al CMAJ 2005c; 173:779-788
Anesthesia as Women’s Work: The Historical Role of the Female AnesthetistAuthor: Robert D. Watson, Jr., M.D. and A.J. Wright, M.L.S., Department of Anesthesiology, University of Alabama School of Medicine.
Although the first modern anesthetics were developed and administered by men6, surgeons soon began recruiting nurses, who were almost exclusively female, to provide anesthesia care for their patients1. A combination of social and economic factors contributed to the selection of women as anesthesia providers.
Foremost, the lack of competitive compensation for anesthesia care precluded men from being anesthetists exclusively2, since men needed a well-compensated job in order to fulfill their societal obligation of being the family’s financial breadwinner3. Women were able to take lower-paying jobs, one of which was anesthesia. In addition, the anesthetist was considered secondary and subservient to the surgeon – the “Captain of the Ship.”1
One fascinating concept that figured prominently into the early choice of women as anesthetists was that women have a natural ability that makes them superior anesthetists to men. Surgeons of the early twentieth century “recognized the importance of feminine qualities in the administration of anesthetics.”4
Nurse anesthesia historian Virginia Thatcher wrote in 1953 that the woman anesthetist “brought to her work a natural aptitude that made her superior to the man.”5
Dr. George Crile, early head and neck surgeon at the Cleveland Clinic, declared: “I do not think a man can ever have the finesse in the administration of an anesthetic that a woman has.”7
Other surgeons of the early 1900’s echoed similar sentiments, with one stating that the female anesthetist “has certain qualities that a man does not possess,”8 such as a “musical feminine voice.”9
Arguments for the anesthetic superiority of women to men focused on psychological and social skills rather than the more technical skills of anesthesia. Alice Magaw, the “Mother of Anesthesia,” was “deeply concerned with the psychological preparation of patients for surgery and used ‘suggestion’ to soothe her patients prior to surgery.”
She refined a technique of psychological preparation designed to increase the safety of the anesthetic by requiring less intraoperative anesthesia2. Motherly qualities, such as a soft voice and gentle touch, were considered among the most valuable in a female anesthetist’s arsenal.10
The dramatic change in American society’s perception of gender roles that occurred in the twentieth century were echoed in the changing perceptions of women’s roles in anesthesia. The social skills that help put patients at ease preoperatively are no longer considered uniquely feminine. Further, technological and pharmacologic advances have rendered “psychological preparation” less important to the everyday practice of anesthesia. The difficulty with which these dramatic changes in perception came about is evidenced by the conflict within the American Association of Nurse Anesthetists in the late 1940’s on the issue of whether to accept male nurse anesthetists for membership.11
In addition to women’s natural anesthetic talents, the notion that women were the ideal anesthetists was also bolstered by the convergence of two complementary concepts – the prevailing social belief that women should be subservient to men and the “Captain of the Ship” concept, which was the status quo in the operating room. The male surgeon could be clearly in charge of all aspects of perioperative care as the female anesthetist carried out his wishes. “Certainly the perception was that the surgeon could more easily control the ‘nurse (woman) anesthetist’ than the ‘physician (male) anesthetist’”.12
In 1896, British anesthetist Dr. Frederic Hewitt wrote, “Anesthesia was born a slave; and she has ever remained the faithful handmaid of her master Surgery.”13 His use of the feminine “she” and the word handmaid to refer to anesthesia reflects the perception that a subservient woman would make the best anesthetist.
While the social changes of the twentieth century have also abolished the notion that women are inferior or subservient to men, anesthesia’s role as Surgery’s “handmaid” continues to be a source of professional conflict today. Ira Gunn describes what happened when a USA Today reporter asked a male surgeon who was in charge in the operating room. “The surgeon stated that surgeons were in charge. When he asked a male anesthesiologist the same question, the anesthesiologist stated that anesthesiologists were in charge.”12
In the modern practice of anesthesia, gender is no longer seen as a factor in the skill of an anesthetist. However, the obvious and important role of women in early anesthesia cannot be overstated. The excellent care that women anesthetists provided to their patients paved the way for further advancements in the sciences of surgery and anesthesia13. Understanding the role of early female anesthetists also reveals the source of the professional conflict that exists among modern surgeons and anesthesiologists.
In today’s fast paced environment of rapid turnover and operating room efficiency, anesthesia care providers would serve their patients well to remember that a soft touch and “sweet words of comfort”9 were effective anxiolytics long before effective pharmacologic premedication was available.
References
- Bankert M. Watchful Care: A History of America’s Nurse Anesthetists. New York: Continuum, 1989, p. 50
- Wauguman WR, Foster SD. CRNAs: an enviable legacy of patient safety. Adv Prac Nurs Q 1(1): 21-28, 1995, [p. 22]
- Bankert M. History of nurse anesthesia. In: Waugaman WR, et al, eds. Principles and Practice of Nurse Anesthesia. Norwalk, Conn.: Appleton & Lange, 1992, p. 9
- Hunziker-Dean J. Voice and touch: Florence Henderson on the skills of an ether specialist. AANA J 67(3): 263-269, June 1999 [p. 263]
- Thatcher V. History of Anesthesia with Emphasis on the Nurse Specialist. Philadelphia: Lippincott, 1953, p. 88
- Thatcher V. History of Anesthesia with Emphasis on the Nurse Specialist. Philadelphia: Lippincott, 1953, pp. 3-13
- Crile GW. Greetings. Bull Natl Assoc Nurse Anesth 4(4): 182- 184, November 1936
- Philemon E. Truesdale quoted in Thatcher V. History of Anesthesia with Emphasis on the Nurse Specialist. Philadelphia: Lippincott, 1953, p. 111
- Tendler MJ. Spinal anesthesia and the nurse anesthetist. . Bull Natl Assoc Nurse Anesth 7(3): 141-146, August 1939 [pp. 142- 143]
- Bankert M. Watchful Care: A History of America’s Nurse Anesthetists. New York: Continuum, 1989, pp. 50-1
- Bankert M. History of nurse anesthesia. In: Waugaman WR, et al, eds. Principles and Practice of Nurse Anesthesia. Norwalk, Conn.: Appleton & Lange, 1992, p. 8
- Gunn IP Professional identity and historical roots. CRNA 10(1):41-47, February 1999 [p. 42]
- Hewitt FW. The past, present, and future of anaesthesia. Practitioner 57: 347-356, 1896
AAGBI winter scientificA look at the forthcoming Congress in London
Click here to view a PDF file of the Congress floor plan and stands.
COX-2 Inhibitors for Arthritic Pain - A good idea gone bad?Author: Dr Karen Stuart-Smith
As the Western population ages and becomes more obese, the burden of osteoarthritis is increasing rapidly.1 Quite apart from the ever-lengthening queue for major joint surgery, pain management in osteoarthritis places heavy demands on primary care teams. Non-steroidal anti-inflammatory agents (NSAIDs) such as diclofenac and naproxen were developed to provide relief from the pain and inflammation of arthritic joints. The mechanism of action of these drugs is straightforward. Cyclooxygenase (COX) is an enzyme which catalyses the break down of cell-membrane-derived arachidonic acid to form several vasoactive products, including thromboxane A2 and various prostanoids.
Prostanoids mediate the pain and inflammation associated with arthritic joints. NSAIDs act by inhibiting cyclooxygenase and so reducing prostanoid formation. As a class they are very effective at relieving joint pain and inflammation, and have been widely prescribed. However prostanoids are ubiquitous in the body and have multiple beneficial effects, not least of which is protection of the gastric mucosa from the effects of gastric acid. As NSAIDs became more widely prescribed, there was a rise in reported gastrointestinal complications, mainly gastric erosions, ulceration and even perforation leading to death.
The pharmaceutical companies took a closer look at the cyclooxygenase enzyme. There are at least two isoforms, COX-1 and COX-2. COX-1 is described as ‘constitutive’, i.e. present in various cells continuously producing a steady supply of prostanoids. It is this isoform which normally provides gastric protection. COX-2, by contrast, is ‘inducible’, i.e. the enzyme is expressed in the presence of stimuli such as inflammation, and, simply put, was thought to be the enzyme catalysing the formation of prostanoids mediating the symptoms of arthritis. Inhibition of COX-2 without inhibition of COX-1 should, in theory, reduce joint pain and inflammation without having a catastrophic effect on the gastric mucosa. Unfortunately this line of reasoning is flawed because, as we shall see, the drug developers failed to take account of several factors.
Firstly, global inhibition of a ubiquitous enzyme is likely to have unexpected side effects. Secondly, the potential effect of selective COX-2 inhibition on the balance between antithrombotic prostanoids such as prostacyclin, and prothrombotic COX products such as thromboxane, appears to have been overlooked. Thirdly, COX-2 appears to be constitutive as well as inducible, and in its constitutive form has several important roles in the maintenance of renal tubular function and endothelium-dependent arterial relaxation. All of these issues have a bearing on the story of the development of the COX-2 inhibitors (the ‘coxibs’) and the subsequent complete withdrawal of some of the drugs in this class.
The first selective COX-2 inhibitors to undergo extensive clinical trials were rofecoxib (Vioxx) and celecoxib (Celebrex). The Vioxx and Gastrointestinal Outcomes Research (VIGOR) trial, which randomised 8,076 patients to receive either high dose rofecoxib or a standard dose of naproxen, showed that all gastric ulcer events and ulcer complications were significantly lower in the rofecoxib group.2 The number needed to treat (NNT) to obtain this benefit was encouragingly low (NNT 42 for reduction in proven symptomatic ulcer, NNT 125 for reduction in ulcer complications). However, there appeared to be a higher incidence of acute myocardial infarction and other life-threatening vascular events in the rofecoxib group. Meta-analysis has since suggested a number needed to harm of 78,3 a fine balance. At the time of the VIGOR trial, the apparent increased risk of acute myocardial infarction in the rofecoxib group did not reach significance. The time course of the study was only 9 months, which, as we shall see, was probably not long enough for the adverse effects of rofecoxib to be clearly demonstrated. At the time, the company successfully argued that the difference between the two groups was due to a cardioprotective (aspirin-like) effect of naproxen, rather than an adverse effect of rofecoxib.
The Celecoxib Long-term Arthritis Safety Study (CLASS) was more controversial. This study, which sought to compare celecoxib with both ibuprofen and diclofenac, had several design flaws.2 The most important of these was that 6 months worth of data were extrapolated to an ‘expected’ 1 year outcome, which predicted a lower rate of gastric ulceration and ulcer complications in the control group. In fact, at 1 year, there was no significant difference in gastrointestinal events between celecoxib and the two non-selective NSAIDs, somewhat weakening the main reason for prescribing this drug. In contrast to rofecoxib, there appeared to be no increase in cardiovascular events. However, there was a fundamental and crucial difference between the VIGOR and CLASS trials, which was that patients in the CLASS trial were allowed to take low-dose aspirin, whereas those in the VIGOR trial were not. Subsequent studies absolving celecoxib have also allowed the use of low-dose aspirin. The significance of this will be discussed below.
More concrete evidence for an increased cardiovascular risk associated with selective COX-2 inhibitors came from randomised controlled trials of these drugs in a completely separate group of patients. Long-standing anecdotal evidence had suggested that patients taking aspirin or other NSAIDs demonstrated a reduction in mortality from colorectal cancer. This was confirmed when clinical studies showed that NSAIDs caused regression of pre-existing adenomas in individuals with Familial Adenomatous Polyposis. The mechanism is unclear, but probably involves inhibition of COX-2-derived prostaglandins that promote tumourigenesis in colonic epithelium.
The Adenomatous Polyp Prevention on Vioxx (APPROVe) trial was designed as a three year, randomised control trial to determine whether rofecoxib could reduce the risk of development of neoplastic polys in patients with a history of colorectal adenomas.4 The trial had to be terminated early, when analysis of cardiovascular events at 18 months showed a significantly increased risk of thrombotic events, principally myocardial infarction and stroke, in the rofecoxib group as opposed to the placebo group. Rofecoxib was voluntarily withdrawn from the market as a direct result of the findings of the APPROVe trial.
 |
| Figure 1. taken from ‘Cardiovascular Risk Associated with Celecoxib in a Clinical Trial for Colorectal Adenoma Prevention’,
by Scott D. Solomon, M.D. et al.
March 17, 2005, Vol. 352, no. 11, pg 1077.
Copyright © 2005 Massachusetts Medical Society. |
A similar randomised trial looking at the effect of celecoxib on adenoma formation (the Adenoma Prevention with Celecoxib (APC) trial) was also terminated early as a result of a significantly greater number of cardiovascular events in the celecoxib group as opposed to placebo (Figure).5 The effect was dose-related, and only became apparent from 12 months onwards. This is significant, as the original VIGOR and CLASS studies were terminated after 9 months and 1 year respectively, which was too early for adverse cardiovascular events to be revealed. The subsequent international effects of these findings on the COX-2 inhibitors and the companies that marketed them is well documented in the media.
What went so wrong with an apparently simple concept? The answer of course, is that the theory was far too simple and neglected to take account of the ubiquitous presence and multiple actions of COX products in the body. COX-1 and COX-2 enzymes produce various substances that have opposite effects. Blockade of COX-2 products means that COX-1 products are either produced in greater quantity or act unopposed.
The accepted explanation for the increased number of thrombotic events in patients taking selective COX-2 inhibitors is that inhibition of COX-2 results in the loss of platelet and endothelial prostacylin production, and a rise in thromboxane production. The result is an increased tendency to thrombus formation and vasoconstriction - the perfect setting for a myocardial or cerebral thrombotic event. This risk is increased in an older population where atherosclerosis is the norm even if there is no overt vascular disease. The apparent delay of 12 to 18 months before a significant increase in cardiovascular events probably reflects the fact that the longer a patient is on the drug, the greater the chance that a clot will develop, and, as in any population study, it may take some time for this risk to reach statistical significance. This is not quite the same thing as saying that having an individual patient on a selective COX-2 inhibitor for a short period of time is less risky. The irony is that the most effective way to prevent thrombotic events when prescribing COX-2 inhibitors is to place the patient on aspirin-a non-selective NSAID. This may contribute to the apparent difference in cardiovascular safety between rofecoxib and celecoxib in the early trials, since aspirin could be prescribed in the CLASS trial but not the VIGOR trial. An equally important issue which appears to have been overlooked is that prostaglandins are essential for normal renal tubular endothelial function. In addition, prostaglandin produced constitutively from renal arterial endothelium maintains flow to the glomeruli. Inhibition of these prostaglandins leads to reduced glomerular flow and fluid retention. The damaging effect of non-selective COX inhibitors on renal function is well-documented, but few authors seem to have considered the effect of the COX-2 inhibitors on the kidney.
The fact that COX-2 inhibitors cause a rise in systolic pressure and predispose to heart failure6 suggests the potential for significant renal damage. Studies are slowly emerging on this topic, but at the time of writing a comprehensive prospective study of the collective long-term cardiovascular and renal effects of COX-2 inhibitors is lacking.
The lessons of the COX-2 story are 1) know the basic science 2) ask yourself if more than one thing could happen if you give this drug and 3) wait for the results of the entire study before placing the drug on the market. These rules will not ensure a safe drug launch, as even the best drugs may have unpredictable side effects. However, they may spare you some embarrassment in the courtroom.
References
- Wolf AD and Pfleger B. Bull WHO 2003; 81:646-656
- Drug and Therapeutics Bulletin. 2005; 43:1-8
- Wright JM. CMAJ 2002; 167:1131-1137
- Bresalier RS et al NEJM 2005; 352:1092-1102
- Solomon SD et al NEJM 2005; 352: 1071-1080
- Hudson M et al BMJ 2005; 330:1370-1375
Issue: June 2008
