Myocardial perfusion scintigraphy as a gatekeeper for invasive procedures

Among patients with stable angina pectoris, it is assumed that only those with reversible (stress-induced) myocardial ischemia and a minority of patients (1 in 5) with irreversible ischemia will benefit from coronary revascularization. However, patients with normal perfusion and the majority of patients with irreversible ischemia (4 in 5) will not benefit from this procedure.

To estimate how many coronary angiographies and revascularization procedures might be considered unnecessary if myocardial perfusion scintigraphy (MPS) was used early in the patient’s course of treatment, we performed a prospective study of 507 patients referred for coronary angiography. Previous studies on this issue are biased in that the result of MPS was known to the referring physician, which had subsequent effects on patient treatment and resource utilization. Because MPS has never been part of the routine work-up in Danish patients, it was ethically acceptable to conduct a study in which the result of primary MPS was kept secret and, thus, would not influence referral for coronary angiography and revascularization procedures.

Patients and methods

A consecutive series of 972 patients referred for coronary angiography because of known or suspected stable angina pectoris were eligible to participate in the study. Because of a limited examination capacity, only 507 patients were evaluated by MPS and, of these, 476 underwent coronary angiography about 2 months later, on average. The main reasons for exclusion were diabetes, age older than 75 years, and ongoing treatment in the cardiology clinic for known coronary artery disease (CAD). Among the 205 female and 302 male patients, the mean age was 57.6 ± 9.2 years; 67% of patients had arterial hypertension, 18% had previous myocardial infarction, and 14% had previous revascularization. Forty-eight percent of patients were taking β blockers, 27% were taking calcium antagonists, and 16% were taking long-acting nitrates.

Prior to MPS, all patients were examined by the same physician, whose skills with regard to classifying these patients clinically had been found to be very similar to those of a professor of cardiology.¹ This physician categorized each patient’s chest pain (typical, atypical, noncardiac, or no pain within the last 14 days) and the degree of pain according to the Canadian Cardiovascular Society.² MPS was performed using a 1-day dual-isotope protocol employing thallium-201 for rest and technetium-99m sestamibi for stress imaging, and electrocardiogram-gated single-photon emission computed gamma camera tomography to acquire images.

The scintigrams were read without the knowledge of other patient data and were categorized into 1 of 4 perfusion patterns: normal, reversible defects (new defects appearing during stress), mixed defects (worsening defects appearing during stress), and fixed defects (at-rest defects that remained unchanged during stress). Our use of this method has been reported elsewhere.^3,4

Coronary angiography was carried out using standard techniques, with ≥50% luminal diameter narrowing of an epicardial artery being considered significant CAD. The determination of treatment procedure was based entirely on history, physical examination, and angiographic results because the MPS findings were not revealed to the referring physician. Standard statistical analyses were employed.

Results

Of the 507 patients, 51% had normal MPS results, 40% had reversible (including mixed) defects, and 9% had fixed perfusion defects. Similarly, 53% of the 476 catheterized patients had insignificant CAD, whereas 17% had 1-vessel disease, 14% had 2-vessel disease, and 16% had 3-vessel disease. Coronary revascularization was performed in 168 patients (percutaneous coronary intervention [PCI] in 81 patients and coronary artery bypass graft [CABG] surgery in 87 patients). However, 16% of these procedures were carried out in 10% of the patients with normal MPS findings and 8% were carried out in 27% of the patients with fixed defects, whereas only about two thirds of patients with reversible ischemia were offered this treatment (

).

Figure

Assuming that only patients with reversible (stress-induced) myocardial ischemia and a minority of patients with “fixed” perfusion defects (present both at rest and during stress) would benefit from coronary revascularization, we calculated that, in the study population of 507 patients, 57% of catheterizations (all in patients with normal MPS results and four fifths in patients with fixed defects) were avoidable. Similarly, 22% of revascularizations were redundant. Translated to the entire consecutive series of 972 patients (41% of whom [ie, 361 of 883 catheterized patients] had insignificant angiographic disease or normal coronary angiography) and assuming that the rate of normal MPS was correspondingly lower in the whole series (whereas the ratio of reversible to fixed perfusion abnormalities was unchanged), 48% of catheterizations and 19% of revascularizations were avoidable.

Discussion

The frequency of about 50% of patients in our study population and about 40% of the entire consecutive series with normal MPS results (and normal coronary angiography) may appear high. However, this corresponds to a frequency of 44% reported in an American study on the cost-effectiveness of MPS, in which more than 5000 stable angina pectoris patients underwent coronary angiography without undergoing preceding MPS.⁵ This may reflect not only that the availability of the invasive procedures may increase their use,^6,7 but also that the clinical selection of these patients is perhaps not as reliable as anticipated. Thus, 42% of our patients with typical angina had normal perfusion, and, conversely, 33% of patients with no chest pain within the last 14 days before MPS and 28% of those with noncardiac pain had, in fact, reversible perfusion defects (

).

Figure

This may also explain why 10% of patients with normal perfusion and 27% of patients with fixed defects underwent PCI or CABG surgery. Thus, it appears that the indications for these procedures are symptom driven or coronary angiography driven, rather than being based on the demonstration of underlying myocardial ischemia. In addition, this may, at least in part, explain why 36% of patients with reversible perfusion abnormalities were not offered invasive therapy (

).

Figure

One limitation of our study was that MPS was not performed in the entire consecutive series of patients and that several higher-than-average-risk patients, that is, patients with diabetes, older age, and ongoing management for stable angina, were not included. On the other hand, it is often stated that MPS is most useful in patients with an intermediate likelihood of disease.^8-10 Nonetheless, assuming that the ratio of normal MPS findings to normal coronary angiography results were similar in the entire sample, we could estimate the frequency of normal perfusion and the number of avoidable coronary angiography and revascularization procedures in the whole series of patients.

If one accepts that myocardial ischemia is the substrate for true angina and that probably only a minority of patients with fixed perfusion defects will benefit from revascularization, the use of MPS as a gatekeeper would mean that about half of catheterizations and one fifth of revascularizations would become avoidable. Thus, patients with normal MPS results and most of the patients with fixed abnormalities are not candidates for PCI or CABG surgery and, therefore, should not undergo coronary angiography either. Instead, many of the patients with reversible abnormalities who did not undergo revascularization (

) would have been offered this procedure if they had undergone primary MPS.

Figure

Such a new strategy would be ethically sound because the outcome in patients treated with knowledge of MPS results is not inferior to the outcome in patients managed without this information.^5,11 It is noteworthy that the prognosis of patients with normal MPS results is comparable with that of the background population,^8,10 such that any invasive procedure would mean an increase in risk, which should be outweighed by the chance to relieve the patient’s angina by means of invasive management.

However, for obvious reasons, this chance is low in patients with normal perfusion or a fixed perfusion defect. All this speaks in favor of a noninvasive treatment strategy in patients with normal myocardial perfusion, even if they have angina pectoris. In such patients, other causes of chest pain should be sought. Patients with fixed perfusion abnormalities constitute but a small part of the stable angina pectoris population, which, in our study, was less than 10% of the sample. In accordance with the general belief, we assumed that only a few of these patients might benefit from coronary revascularization. In another study using myocardial perfusion imaging to predict the effectiveness of coronary revascularization in patients with stable angina pectoris, we showed that none of the patients benefited from coronary revascularization, which was not superior to medical or no treatment.¹²

Some fear that 3-vessel or left main CAD may not show up on MPS if perfusion is equally reduced in most parts of the myocardium. However, in our study, only 5 of the 476 patients (1.1%) who underwent coronary angiography had normal MPS results despite serious CAD and, thus, were at risk for being “overlooked.” In such patients, continued severe chest pain would still be an indication for coronary angiography, despite normal MPS results. Moreover, it is often stated that angina and not hypoperfusion is the main indication for coronary revascularization.⁹ Current guidelines recommend that patients with typical angina or severe chest pain, or both, should be referred directly for coronary angiography and possible revascularization, especially if the patient has important risk factors. As many as 42% of our 290 patients with typical angina had normal perfusion, and because less than 10% of all our patients had severe chest pain, insisting on referral of these patients directly to coronary angiography would have little influence on the consequences of using MPS as a gatekeeper.

Conclusion

We evaluated the potential benefit of using MPS as a gatekeeper for patients with stable angina pectoris prior to performing coronary angiography and revascularization procedures. Results showed that a significant number of coronary angiography and revascularization procedures might be avoided in patients with known or suspected stable angina pectoris if MPS is used prior to these invasive procedures.