EDITOR'S NOTE

The new millennium presents medicine with the challenge of providing the best possible patient outcomes at the least possible cost. PET MPI has emerged from its academic roots as a research tool to a new position as the standard for management of coronary artery disease. PET MPI can be used alone to manage most patients with coronary disease because it is sufficiently quantitative to guide clinical management decisions without the need for an invasive coronary angiogram.

This first in a series of newsletters explores how the nuclear physician can take a leadership role in bringing the power of PET MPI into widespread use for patients with coronary heart disease.--Michael E. Merhige, MD

Coronary artery disease (CAD) is the major cause of death in the United States. Up to 60% of all cases of acute myocardial infarction and sudden death occur unexpectedly, in previously apparently "healthy" people with no antecedent symptoms. This public health problem is driving the emergence of positron emission tomography (PET) from academic research centers to its new role as the workhorse for routine management of patients with known or suspected coronary heart disease. This first issue of Applications in Nuclear Cardiology addresses the shift to PET for routine clinical assessment of myocardial perfusion.

It is clear that many more invasive cardiology procedures are performed in the U.S., when compared with other less wealthy countries; yet these procedures do not result in improved clinical outcomes for patients. In patients who have had their first acute MI, coronary arteriography, coronary artery bypass surgery (CABG), and percutaneous transluminal coronary intervention (PTCI) are used 50% to 150% more frequently in the U.S. than Canada, yet there is no improvement in the incidence of coronary death/death plus recurrent MI at 1 year. ¹

Many well-conducted lipid lowering trials, which used coronary arteriography, have demonstrated that aggressive management of CAD with a very low-fat diet and lipid-lowering drugs can reduce acute MI and coronary death by 85% or more, despite only minimal angiographic improvement in coronary lumen diameter. ² In contrast, the more expensive and invasive approaches of PTCI and CABG do not reduce the incidence of subsequent coronary death or recurrent infarction. This is because it is the young, mild, plaque that is structurally vulnerable and, prone to sudden rupture, leading to death or an acute coronary syndrome. The older, more severe stenosis is actually structurally stable.

With the need for healthcare cost containment, the management of CAD in the new millennium is undergoing critical reappraisal. The therapeutic paradigm is now shifting away from expensive, anatomically driven treatment (e.g., coronary angiography followed by revascularization), toward identification of the vulnerable plaque, with subsequent biological stabilization of the atherosclerotic process, using aggressive lipid-lowering. Which patients will require invasive versus aggressive noninvasive management? How hard should we push lipid reduction to achieve coronary disease reversal in an individual patient?

Nuclear medicine is now in a unique position to provide answers to these critical questions for referring cardiologists, because of the power of positron emission tomography.

PET versus SPECT for myocardial perfusion imaging

PET myocardial perfusion imaging (MPI) permits cardiologists to manage CAD patients directly and noninvasively, usually without the need for coronary arteriography, because of physical advantages over conventional nuclear imaging using SPECT.

The chief physical advantages of PET MPI over SPECT imaging are:

1. Attenuation correction;

2. Depth-independent doubling of image resolution from 20 mm. Full-width half-maximum (FWHM) to 10 mm FWHM; and

3. Improved image contrast due to 30-fold higher count fluxes.

Figure 1 illustrates exercise sestamibi SPECT and dipyridamole PET images with Rb-82 performed within 2 weeks in the same (clinically stable) patient. The presence of a clinically occult posterobasal infarct, responsible for impaired left ventricular ejection fraction, is detected clearly with PET MPI, yet missed by SPECT.

PET MPI offers 95% sensitivity and specificity for CAD detection, ³ and uses pharmacologic vasodilation rather than exercise to assess coronary flow reserve. Figure 2 demonstrates PET MPI images in the same patient studied initially during exercise-induced ischemia, with 2-mm ST depression and then immediately thereafter with pharmacologic vasodilation. The images done with IV dipyridamole clearly define the extent and severity of this patient's mid-LAD disease better than imaging following exercise. This is because peak exercise increases coronary flows only two-fold, compared with resting levels. In contrast, pharmacologic vasodilation with IV adenosine or dipyridamole, raises flow in the normal coronary artery 4 to 5 times resting levels, permitting greater contrast between perfusion in normal versus diseased coronary territories. ⁴

Rubidium-82: Cost-effective PET tracer for MPI

The development of a generator-produced PET tracer for measurement of myocardial flow was a pivotal event in bringing PET imaging into clinical practice, since the need for a cyclotron for tracer production was eliminated. Rb-82 is a potassium analogue that is taken up by myocardial cells in proportion to blood flow, with a mechanism similar to thallium uptake. Rb-82 has an ultrashort radioactive half-life of 75 seconds, which allows for the delivery of relatively high doses (50 to 60 mCi) per injection, with resulting high quality images, with lower radioactivity exposure compared with the tracers used in SPECT MPI. The parent isotope, strontium-82 provides a fresh dose of Rb-82 every 10 minutes. The generator (Figure 3) must be replaced monthly, and costs approximately $25,000. This represents a fixed cost (with savings of 38%) compared with that of Tc99m-sestamibi.

Clinical performance of PET MPI

The study protocol for PET MPI is considerably shorter than even 1-day SPECT MPI. The complete study is accomplished within 1 hour.

Initially, the patient is positioned in the PET camera, with low-intensity lasers mounted in the camera gantry. A 50 to 60 mCi rest injection is performed and counts are acquired over 6 minutes, following a 70-second wait to allow blood pool and lung clearance. This is followed by the transmission scan, using a built-in rod of Germanium 68 (a positron emitter) which circles the patient for approximately 20 minutes, providing attenuation correction. This is followed by IV infusion of dipyridamole or adenosine as used in SPECT MPI and repeat Rb-82 injection during the peak coronary hyperemic effect. Images are reconstructed immediately and are reviewed with patients before they leave the nuclear medicine laboratory.

Introduction of PET MPI with Rb-82 into clinical practice

Barriers to the widespread implementation of PET MPI have been largely resolved. In 1991, the FDA approved the use of Rb-82 in patients for the detection of CAD. In 1995, the American College of Cardiology, American Heart Association, and the American Society of Nuclear Cardiology redefined PET MPI as a "Category 1" procedure, the same category as SPECT MPI. The critical event that brought PET MPI to patients, however, was the 1995 approval by the Healthcare Finance Administration of reimbursement for PET MPI done with Rb-82 in Medicare and Medicaid patients.

The position of the professional societies, and most third-party payers, is that PET MPI should be considered in the routine management of CAD patients only if the cost of PET MPI is similar to SPECT in the same community. Obviously, the PET MPI study fee cannot be made the same as the fee for SPECT MPI primarily due to the difference in camera price. A SPECT system costs $250,000 to $350,000, while a PET system costs $1.8 to $2.4 million.

Experience with PET MPI in private clinical practice

After the approval of reimbursement for PET MPI in 1995, our practice contemplated purchasing a PET camera for routine management of patients with CAD. We reasoned that the increased accuracy of PET MPI would offer a competitive advantage over other area practices by providing noninvasive diagnostic accuracy, essentially equivalent to direct coronary arteriography. We hypothesized that the overall costs of CAD management would be reduced by using PET MPI rather than SPECT imaging, despite the increase in the fee for PET, because the improved accuracy of PET would eliminate unnecessary test layering, invasive arteriograms, PTCI, and CABG surgery.

We began to study patients in our office November 1, 1995, using a Positron HZL-R PET camera (Positron Corp., Houston, TX) and Rb-82. Previous econometric modeling had predicted that PET MPI would be the most cost-effective diagnostic test for CAD management if the pretest likelihood of CAD was <70%, when compared with nonnuclear exercise treadmill testing, exercise SPECT, and direct coronary angiography. ⁵ Patients referred to our Nuclear Cardiology department underwent rigorous evaluation of pretest disease likelihood using the "CADENZA"software program (Advanced Heuristics, Bainbridge Island, WA). This program uses published epidemiological data on the prevalence of CAD. Patient demographics, coronary risk factors, type of chest pain, and previous test data are entered, and the pretest likelihood of CAD (or active ischemia in patients with known CAD) is determined.

As shown in Table 1, patients with intermediate pretest disease likelihood 3/470% were sent directly to PET MPI. We followed 1490 sequential patients after PET MPI for subsequent invasive procedure utilization and clinical outcomes, for an average of 12 months. We also estimated CAD management costs using the procedure fees shown in Table 2. The data for these PET patients were compared with retrospective data on a matched sequential group of CAD patients imaged with SPECT prior to the installation of the PET camera.

The mean pretest likelihood of CAD in the first 102 patients routed to PET MPI was 37%, which was matched to 102 patients studied with SPECT. As experience with PET increased, physicians began to rely on the technology without using a subsequent angiogram as frequently as they did with SPECT imaging. In addition, "sicker" patients were referred to PET. The pretest likelihood of CAD increased in the total PET group (n = 1490) to 43%.

Figure 4 shows utilization of invasive procedures in patients studied with PET versus SPECT MPI. The likelihood of subsequent coronary arteriography in patients studied with SPECT was 31%, which is appropriate for a population with a 37% pretest likelihood of CAD. The rate of false-positive SPECT studies was reasonable for clinical practice at 15%, substantially and significantly higher than the 5% rate for PET MPI, which reflects a specificity of 95%, consistent with the published literature. With PET MPI, subsequent coronary arteriography was reduced by more than half, compared with SPECT.

There was no difference in the costs of CAD diagnosis: $2500 per patient whether SPECT or PET MPI was used. The higher PET study fee was balanced by the excess use of angiography in the SPECT group (Figure 5). Overall, there was a 25% cost savings in patients managed routinely with PET MPI because of the reduction in CABG.

Figure 6 shows 1-year outcomes in patients studied with SPECT versus PET MPI. There was no difference in total mortality, but a strong trend toward reduced coronary mortality and acute myocardial infarction in the patients who underwent management with PET. Review of cases demonstrated that the difference appeared to be related, in part, to the avoidance of perioperative complications following CABG.

From this experience, we concluded that routine use of PET MPI in carefully preselected patients with an intermediate pretest likelihood of CAD, results in: a 50% reduction in utilization of coronary angiography and bypass surgery, excellent clinical outcomes, and a 25% reduction in CAD management costs. ⁶

Conclusion

This article illustrates the practicality of adding PET capability for myocardial perfusion imaging to a Nuclear Medicine department using SPECT imaging, or even replacing SPECT MPI altogether. Nuclear medicine physicians using PET MPI can serve as consultants to the managing cardiologist, providing expert imaging of myocardial perfusion that is more accurate and clinically reliable than competitors using SPECT. Third-party payers in our area are beginning to consider not reimbursing for SPECT studies when PET MPI is available (e.g., for the morbidly obese patient).