Positron emission tomography for the assessment of myocardial viability: an evidence-based analysis

The objective was to update the 2001 systematic review conducted by the Institute For Clinical Evaluative Sciences (ICES) on the use of positron emission tomography (PET) in assessing myocardial viability. The update consisted of a review and analysis of the research evidence published since the 2001 ICES review to determine the effectiveness and cost-effectiveness of PET in detecting left ventricular (LV) viability and predicting patient outcomes after revascularization in comparison with other noninvasive techniques.

Left Ventricular Viability Heart failure is a complex syndrome that impairs the contractile ability of the heart to maintain adequate blood circulation, resulting in poor functional capacity and increased risk of morbidity and mortality. It is the leading cause of hospitalization in elderly Canadians. In more than two-thirds of cases, heart failure is secondary to coronary heart disease. It has been shown that dysfunctional myocardium resulting from coronary heart disease (CAD) may recover contractile function (i.e. considered viable). Dysfunctional but viable myocardium may have been stunned by a brief episode of ischemia, followed by restoration of perfusion, and may regain function spontaneously. It is believed that repetitive stunning results in hibernating myocardium that will only regain contractile function upon revascularization. For people with CAD and severe LV dysfunction (left ventricular ejection fraction [LVEF] <35%) refractory to medical therapy, coronary artery bypass and heart transplantation are the only treatment options. The opportunity for a heart transplant is limited by scarcityof donor hearts. Coronary artery bypass in these patients is associated with high perioperative complications; however, there is evidence that revascularization in the presence of dysfunctional but viable myocardium is associated with survival benefits and lower rates of cardiac events. The assessment of left ventricular (LV) viability is, therefore, critical in deciding whether a patient with coronary artery disease and severe LV dysfunction should undergo revascularization, receive a heart transplant, or remain on medical therapy. Assessment of Left Ventricular Viability Techniques for assessing myocardial viability depend on the measurement of a specific characteristic of viable myocytes such as cell membrane integrity, preserved metabolism, mitochondria integrity, and preserved contractile reserve. In Ontario, single photon emission computed tomography (SPECT) using radioactive (201)thallium is the most commonly used technique followed by dobutamine echocardiography. Newer techniques include SPECT using technetium tracers, cardiac magnetic resonance imaging, and PET, the subject of this review. POSITRON EMISSION TOMOGRAPHY: PET is a nuclear imaging technique based on the metabolism of radioactive analogs of normal substrates such as glucose and water. The radiopharmaceutical used most frequently in myocardial viability assessment is F18 fluorodeoxyglucose (FDG), a glucose analog. The procedure involves the intravenous administration of FDG under controlled glycemic conditions, and imaging with a PET scanner. The images are reconstructed using computer software and analyzed visually or semi-quantitatively, often in conjunction with perfusion images. Dysfunctional but stunned myocardium is characterized by normal perfusion and normal FDG uptake; hibernating myocardium exhibits reduced perfusion and normal/enhanced FDG uptake (perfusion/metabolism mismatch), whereas scar tissue is characterized by reduction in both perfusion and FDG uptake (perfusion/metabolism match).

The Medical Advisory Secretariat used a search strategy similar to that used in the 2001 ICES review to identify English language reports of health technology assessments and primary studies in selected databases, published from January 1, 2001 to April 20, 2005. Patients of interest were those with CAD and severe ventricular dysfunction being considered for revascularization that had undergone viability assessment using either PET and/or other noninvasive techniques. The outcomes of interest were diagnostic and predictive accuracy with respect to recovery of regional or global LV function, long-term survival and cardiac events, and quality of life. Other outcomes of interest were impact on treatment decision, adverse events, and cost-effectiveness ratios. Of 456 citations, 8 systematic reviews/meta-analyses and 37 reports on primary studies met the selection criteria. The reports were categorized using the Medical Advisory Secretariat levels of evidence system, and the quality of the reports was assessed using the criteria of the Quality Assessment of Diagnostic Accuracy Studies (QUADAS) developed by the Centre for Dissemination of Research (National Health Service, United Kingdom). Analysis of sensitivity, specificity, predictive values and likelihood ratios were conducted for all data as well as stratified by mean left ventricular ejection fraction (LVEF). There were no randomized controlled trials. The included studies compared PET with one or more other noninvasive viability tests on the same group of patients or examined the long-term outcomes of PET viability assessments. The quality assessment showed that about 50% or more of the studies had selection bias, interpreted tests without blinding, excluded uninterpretable segments in the analysis, or did not have clearly stated selection criteria. Data from the above studies were integrated with data from the 2001 ICES review for analysis and interpretation.

The evidence was derived from populations with moderate to severe ischemic LV dysfunction with an overall quality that ranges from moderate to low.PET appears to be a safe technique for assessing myocardial viability.CAD patients with moderate to severe ischemic LV dysfunction and residual viable myocardium had significantly lower 2-year mortality rate (3.2%) and higher event-free survival rates (92% at 3 years) when treated with revascularization than those who were not revascularized but were treated medically (16% mortality at 2-years and 48% 3-year event-free survival).A large meta-analysis and moderate quality studies of diagnostic accuracy consistently showed that compared to other noninvasive diagnostic tests such as thallium SPECT and echocardiography, FDG PET has:Higher sensitivity (median 90%, range 71%-100%) and better negative likelihood ratio (median 0.16, range 0-0.38; ideal <0.1) for predicting regional myocardial function recovery after revascularization.Specificity (median 73%, range 33%-91%) that is similar to other radionuclide imaging but lower than that of dobutamine echocardiographyLess useful positive likelihood ratio (median 3.1, range 1.4 -9.2; ideal>10) for predicting segmental function recovery.Taking positive and negative likelihood ratios together suggests that FDG PET and dobutamine echocardiography may produce small but sometimes important changes in the probability of recovering regional wall motion after revascularization.Given its higher sensitivity, PET is less likely to produce false positive results in myocardial viability. PET, therefore, has the potential to identify some patients who might benefit from revascularization, but who would not have been identified as suitable candidates for revascularization using thallium SPECT or dobutamine echocardiography.PET appears to be superior to other nuclear imaging techniques including SPECT with (201)thallium or technetium labelled tracers, although recent studies suggest that FDG SPECT may have comparable diagnostic accuracy as FDG PET for predicting regional and global LV function recovery.No firm conclusion can be reached about the incremental value of PET over other noninvasive techniques for predicting global function improvement or long-term outcomes in the most important target population (patients with severe ischemic LV dysfunction) due to lack of direct comparison.An Ontario-based economic analysis showed that in people with CAD and severe LV dysfunction and who were found to have no viable myocardium or indeterminate results by thallium SPECT, the use of PET as a follow-up assessment would likely result in lower cost and better 5-year survival compared to the use of thallium SPECT alone. The projected annual budget impact of adding PET under the above scenario was estimated to range from $1.5 million to $2.3 million.

In patients with severe LV dysfunction, that are deemed to have no viable myocardium or indeterminate results in assessments using other noninvasive tests, PET may have a role in further identifying patients who may benefit from revascularization. No firm conclusion can be drawn on the impact of PET viability assessment on long-term clinical outcomes in the most important target population (i.e. patients with severe LV dysfunction).