Quantitative Stress Perfusion Cardiac Magnetic Resonance Improves Prognostication

Crossroads for Increased Use of This Powerful Technology

Corresponding Author

• Download figure
• Open in new tab
• Download powerpoint

• perfusion CMR
• prognosis
• quantitative

Cardiac magnetic resonance (CMR) stress perfusion imaging is commonly viewed as expensive and complicated and is not used as commonly in the United States as single-photon emission computed tomography myocardial perfusion imaging (MPI). Continued developments in CMR and extensive validations challenge these practice patterns. In this issue of iJACC, Sammut et al. (1) take these validations another important step forward by using quantitative stress CMR perfusion to predict major adverse cardiac events (MACE).

Although strict meta-analysis representing thousands of patients have also shown the prognostic value of stress CMR perfusion (2), one of the unique aspects of the study by Sammut et al. (1) is the comparisons of prognosis determined from the extent of visually apparent perfusion defects (qualitative analysis) versus the extent of perfusion defects determined from quantitative CMR perfusion. Myocardial perfusion was quantified with Fermi function deconvolution of first-pass adenosine stress perfusion with a dual bolus contrast injection scheme. The visual extent of perfusion defects was the number of abnormal sectors found by the consensus of 2 experts. For quantitative analysis, the number of sectors with myocardial perfusion reserve (MPR) <1.5 defined the extent of the perfusion defect. They found that the quantitative results, summarized as MPR, did a better job of discriminating low and high risk than the expert clinical perfusion interpretation (visual analysis). Net reclassification improvement and integrated discrimination improvement analyses support the incremental value of stress-induced perfusion defects compared with age, sex, and presence of LGE. It is interesting that the MPR threshold of 1.5 that was adopted from the positron emission tomography (PET) literature also worked for CMR prognosis. That a measurable quantity has better prognostic discrimination than the consensus of 2 experts is a clue that more information can be squeezed out of these CMR images. This is an important step forward in understanding the value of quantitative CMR perfusion imaging because there have been relatively few clinical studies that have shown clear benefits over qualitative interpretation (3,4).

Although it is hoped that quantitative metrics will help less experienced readers achieve similar results, all new methods need to be viewed critically. Although quantitative MPR detected more patients (∼22) with MACE than expert visual analysis in the study by Sammut et al. (1), there were patients who were missed by both methods (∼9) and some patients who were missed by MPR but detected by visual analysis (∼13). It is critical to recognize that the quantitative CMR analysis of MPR was not immediately available and processing time was time consuming. The investigators were careful to exclude revascularization in the first 90 days after the test to avoid bias related to the clinically available CMR results. Excluding early revascularization is important for minimizing the influence of test results on the prognosis study endpoints because late revascularization represents a significant fraction of the outcomes in studies of this nature. Future studies will need to define how to resolve discordance between the visual and quantitative results.

In general, stress perfusion CMR has matured substantially over the past 2 decades. In a meta-analysis comparing the various stress myocardial perfusion methods, CMR, PET, and current procedural terminology had higher diagnostic accuracy than the more commonly used SPECT and stress echo (5). In a meta-analysis that compared anatomic and physiological evaluations of coronary stenosis with invasive fractional flow reserve (FFR) as the reference standard, both CT and invasive coronary angiographic assessment of stenosis had poor specificity for physiologically significant coronary stenosis (39% and 67%, respectively) (6). Although computational fluid dynamic estimation of FFR from the CT coronary angiography (FFR_CT) improved the specificity of CT without losing significant sensitivity, CMR outperformed all other modalities, including FFR_CT, stress echocardiography, SPECT MPI, and invasive coronary stenosis assessment without FFR (6). The role of stress CMR is now well accepted for multiple indications in both European (7,8) and U.S. guidelines (9).

Despite these major advances, use of stress CMR in the United States remains low. As summarized in Table 1, Medicare reimbursement for CMR is within 3% of SPECT reimbursement for 2016. Thus, the low use of CMR cannot be explained by lack of validation studies, prognosis studies, or Medicare costs of examinations. As the evidence mounts, we will have to see how practice patterns and guidelines adapt to new technology. Objectively and quantitatively, stress perfusion CMR appears to be in a strong position for assuming a larger role in the management of stable coronary disease.