Preoperative screening for coronary disease: who needs it and how do you do it?

Abstract

See article on page 1591

Perioperative myocardial infarction is a major cause of complications and death among patients undergoing non‐cardiac surgery.¹ Patients undergoing vascular surgery are at particularly high risk owing to the very high prevalence of significant, yet often occult, coronary disease.² Preoperative assessment of these patients can be challenging. Exercise tolerance is frequently limited by claudication, and interpretation of treadmill exercise testing is hampered by abnormalities on the resting ECG. Pharmacological stress testing with dobutamine stress echocardiography (DSE) or myocardial perfusion imaging (MPI) is more applicable but also more time consuming and costly and routine coronary angiography is associated with an increased risk of complications in this population. Furthermore, there is no evidence to support “prophylactic” revascularisation to reduce perioperative risk. A systematic approach has been suggested³ with the aim of reducing perioperative and longer‐term risk based on an initial clinical assessment and allowing stratification of patients into specific risk categories. This can then guide the selective use of non‐invasive testing to refine the risk stratification and direct specific intervention to further assess or modify cardiac risk.

Assessing cardiac risk

The Revised Cardiac Risk Index (RCRI) provides a simple and accurate estimate of risk which has been extensively tested.⁴ This involves clinical assessment and ECG evaluation and identifies six key clinical variables: high‐risk surgical procedure (abdominal, thoracic or major vascular surgery), history of ischaemic heart disease (myocardial infarction (MI), angina, use of glyceryl trinitrate, positive stress test, Q waves, previous revascularisation), history of heart failure, history of stroke or transient ischaemic attack, preoperative insulin treatment and preoperative serum creatinine levels >177 μmol/l. Based on the presence of 0, 1, 2 or ⩾3 of these predictors, the rate of major cardiac complications (MI, pulmonary oedema, ventricular fibrillation or other primary cardiac arrest) was estimated to be 0.4%, 0.9%, 7% and 11%, respectively. This allows a group of patients at low risk to proceed to surgery without needing further risk stratification: those patients with none or only one of the key clinical risk factors might proceed to their non‐cardiac surgery without further cardiac investigation. Patients at higher risk should probably have further cardiac investigations, although it may be felt that those undergoing very low risk surgery or those who have an excellent functional capacity, or both, might proceed in some circumstances without further evaluation.

How patients who have had previous coronary intervention or who have had previous cardiac evaluation are handled is not clear, but current recommendations are that patients who have undergone coronary artery bypass grafting within 5 years without recurrence of symptoms or cardiac evaluation within 2 years without subsequent deterioration need no further cardiac evaluation.³ The assessment of those who have had previous coronary angioplasty depends on the time period since the angioplasty, the likelihood of restenosis and the symptomatic status of the patient.

Those patients who have at least an intermediate level of risk (RCRI 2+) should generally undergo further risk stratification with non‐invasive imaging. Exercise treadmill and pharmacological stress testing (DSE and MPI) have good negative predictive values. In patients with positive tests the likelihood of perioperative events varies depending on the patient population studied, but most studies have positive predictive values in the range of 10–30%. In general, a negative non‐invasive test, particularly DSE and MPI, indicates a low operative cardiac risk. Pharmacological stress has better discriminatory power than exercise testing and is more widely applicable as it can be used in patients who cannot exercise and those with abnormal baseline ECGs. DSE may be preferable to MPI because of a slightly higher specificity⁵ and the ability to assess in detail left ventricular function, valves and pulmonary artery pressure. In 300 consecutive patients presenting for major vascular surgery, DSE had positive and negative predictive values of 38% and 100%, respectively. In addition, the heart rate at which ischaemia occurred defined a group with low ischaemic threshold (<70% maximum predicted heart rate) who were at particularly high risk of morbidity and mortality (66% perioperative events, 43% death or MI).⁶

The recent advance of coronary imaging with mutidetector CT coronary angiography (CTCA) has brought a new technique to the clinical arena. Although the technique at present has many false positives, it has a negative predictive value of 95–100%⁷ and may provide a useful non‐invasive test for ruling out significant coronary stenoses in selected patients of low to intermediate cardiac risk. In this issue of Heart, Russo et al examine the value of 16‐slice CTCA for detection of coronary stenoses ⩾50% in patients without a history of ischaemic heart disease undergoing elective non‐coronary cardiac surgery (mitral and aortic valve surgery, aortic repair) (see article on page 1591).⁸ Of 116 patients evaluated, 72 (62%) had no stenoses ⩾50% detected by CTCA. Although confirmatory coronary angiography was not performed, this group underwent surgery without severe in‐hospital cardiac complications.

These are promising results in this patient group. However, the technique is limited by a relatively low specificity and a high proportion of non‐evaluable segments owing to artefact, arrhythmia or coronary calcification, which are common in the high risk older population. In the present study, 27 of the initial 143 patients enrolled were not evaluated and CTCA incorrectly diagnosed coronary stenosis ⩾50% in 8/44 cases. The high ionising radiation and contrast dose must also be taken into account, although one of the competing technologies, MPI also has a significant radiation dose. In the wide setting of non‐cardiac surgery, CTCA might currently have a role in ruling out coronary disease in those patients at intermediate risk with equivocal or suboptimal functional tests. Before valve surgery, where coronary evaluation is routinely performed, preliminary CTCA has been shown to be an effective gatekeeper for invasive angiography.⁹ As the technique improves and radiation doses are reduced, CTCA is likely to become an important screening test for significant coronary disease in this and other clinical settings.

Modifying cardiac risk

The benefit of perioperative β blockers has been the subject of a number of reviews, with conflicting conclusions. Current evidence is weak and limited by small numbers of randomised patients¹⁰ and lack of data about the target population, ideal dose and route of administration. A recent observational trial using data from almost 700 000 patients¹¹ suggested that β blockers given within 2 days of any non‐cardiac surgery provided protective benefit only in patients at higher risk (RCRI ⩾2) In patients at low risk (RCRI of 0), β blockers were associated with increased complications. The ACC/AHA give class I or IIa recommendation for β blockers to be continued in patients already taking them and prescribed for patients identified as at high risk (defined by clinical risk factors or ischaemia on preoperative tests) who are undergoing intermediate or high‐risk procedures.¹² Current Canadian (POISE) and European (DECREASE IV) trials should study sample size limitations and provide further data on the value of β blockers in patients at intermediate risk.

Patients with high clinical risk (RCRI >3) including a background of coronary disease, particularly those with continuing symptoms of angina and/or high‐risk features on non‐invasive testing should be considered for coronary angiography.³,¹³ Although it might seem logical to deal with coronary arteries with tight stenoses before non‐cardiac surgery, no randomised trials exist to support “prophylactic” coronary revascularisation to reduce risk in the absence of conventional indications. Recently, McFalls et al randomised 510 patients undergoing vascular surgery to either revascularisation (coronary angioplasty or bypass surgery) or medical treatment (>80% in both groups receiving β blockers).¹⁴ There was no difference in long‐term outcome. The study included only patients with stable coronary disease and excluded those with left main stem stenosis, ejection fraction <20% and severe aortic stenosis. In this trial, non‐cardiac surgery was performed as early as 2 weeks after coronary angioplasty, which might be insufficient to allow adequate re‐endothelialisation of the stent to reduce the risks of stent thrombosis. The major limitations of this study relate to the highly selected nature of the population and relatively small numbers, with correspondingly low power to detect a difference between the two groups. No specific data are available for management of patients with drug‐eluting stents (DES) requiring urgent non‐cardiac surgery and the use of DES in patients before surgery should be avoided.

Many centres, including our own, would intervene on those patients with tight coronary stenoses supplying a significant territory. If angioplasty is contemplated the type of stent used needs to be carefully considered. In most instances this should be a bare metal stent as re‐endothelialisation generally occurs within 2–4 weeks and the non‐cardiac surgery can then be performed after the patient has stopped taking clopidogrel, which might otherwise cause bleeding problems for the surgeon, particularly in combination with aspirin. If a patient has a DES in situ and is taking long‐term aspirin and clopidogrel to prevent in‐stent thrombosis, very careful judgments need to be made about the timing of surgery and when/if the clopidogrel might be stopped. The complexity of the previous angioplasty, including length and width of the stent(s) inserted, the time period that has elapsed since the angioplasty, the urgency of the non‐cardiac surgery and the bleeding risk from the surgery should all be taken into account. These decisions are generally best made jointly by the cardiologist and surgeon.

Conclusions

Current guidelines provide a systematic evidence‐based approach involving clinical risk assessment to guide the selective use of diagnostic tests and appropriate preventive interventions. However, more data and better information are still needed. Preoperative screening should also be regarded as an opportunity to study longer‐term cardiac risk and to secure aggressive secondary prevention. In the future there will be advances in coronary imaging and revascularisation (minimally invasive bypass surgery, new stents) as well as anaesthetic and surgical techniques (endovascular aortic aneurysm repair) which will continue to improve operative outcomes in patients at higher risk.

Abbreviations

CTCA - CT coronary angiography

DES - drug‐eluting stent(s)

DSE - dobutamine stress echocardiography

MI - myocardial infarction

MPI - myocardial perfusion imaging

RCRI - Revised Cardiac Risk Index