Author + information
- †National Institute of Cardiology, Rio de Janeiro, Brazil
- ‡Clínica de Diagnostico por Imagem (CDPI), Rio de Janeiro, Brazil
- ↵∗Reprint requests and correspondence:
Dr. Ilan Gottlieb, Department of Radiology, National Institute of Cardiology, Rua das Laranjeiras, 374-Laranjeiras, Rio de Janeiro 22240-006, Brazil.
“It's hard to make predictions—especially about the future.”
—Robert Storm Petersen (1)
Predictions are even harder in medicine, when so much individual variability and system complexity occurs. Nonetheless, we should not shun this important task. Few imaging modalities offer a more comprehensive phenotyping of the heart than cardiac magnetic resonance (CMR). It is currently the noninvasive gold standard for quantification of cardiac volumes and function (2), provides excellent anatomic depiction of cardiac structures and masses—benefiting from its unique ability to differentiate and characterize tissues—and has been shown to be at least as accurate as single-photon emission tomography for myocardial ischemia detection (3). CMR has also been shown to detect myocardial fibrosis and necrosis (4), microvascular obstruction and hemorrhage (5,6), and myocardial edema (7). Most of these diagnostic capabilities have been shown to contain prognostic information that allows for better risk assessment and ensuing patient care. In this issue of the Journal, El Aidi et al. (8) have published a systematic review of the literature examining the prognostic ability of different information provided by CMR in acute, chronic, and suspected coronary artery disease (CAD). For their well-performed work, and especially for the elegant compilation of so much information in comprehensive tables, the authors are to be congratulated.
Left ventricular ejection fraction (LVEF) is the most widely used cardiac systolic performance measure. Despite having some dependence on physiological conditions (i.e., pre-load and afterload), its reduction below normal thresholds expresses the failure of many compensatory mechanisms, and is related to advanced stages of most cardiac pathologies. Deprived of its reserve, the failing heart is both unable to respond to physiological necessities and is more susceptible to fatal arrhythmias (9). For at least 40 years (10), LVEF, as assessed by different invasive and noninvasive methods, has successfully been tested for prognostication in a variety of clinical scenarios. Accordingly, El Aidi et al. (8) have shown that when the LVEF falls by 10%, the risk of events increases 13% to 15% in acute coronary syndromes and increases 18% to 28% in chronic or suspected CAD. CMR is the gold standard noninvasive imaging technique to assess LVEF because of its high accuracy in chamber volumetric measurements associated with very low measurement variability (2).
CMR can detect obstructive CAD by either inducing wall motion abnormalities (WMA) after an adrenergic stressor (usually dobutamine) (11) or by inducing myocardial perfusion gradients after a vasodilatory challenge (usually with adenosine or dipyridamole) (3). Although one could potentially combine WMA and perfusion assessments, this does not enhance overall accuracy in detecting obstructive CAD (12), and in practice, it is not usually performed. It should be noted that vasodilators do not necessarily induce ischemia, but rather generate perfusion imbalances between myocardial territories as seen by signal intensity gradients during first-pass dynamic imaging. Because blood flow distal to an obstructive lesion at peak vasodilatory challenge may be similar to rest, using vasodilators for inducing WMAs have been shown to have low sensitivity for obstructive CAD detection (13), and are rarely used. Conversely, when true myocardial ischemia is generated by dobutamine stress, the patient's heart rate is frequently higher than 150 beats/min, which, when added to the discomfort generated by the drug, makes stress contractility plus perfusion imaging impractical. Curiously, both WMA and perfusion assessments appear to have similar diagnostic accuracies, but as El Aidi et al. (8) have shown, perfusion abnormalities appear to have greater prognostic impact (adjusted hazard ratio [HR]: 3.02 to 7.77) than WMA (adjusted HR: 1.87 to 2.99). The reason for this is not completely clear, and is worthy of further scientific investigation, because this is one of the most important messages from their work. But regardless of the methodology used for obstructive CAD detection, one has to bear in mind that the causa mortis is usually plaque instability leading to myocardial infarction (MI), for which higher obstruction grades are a major risk factor (14). Hence, prognostic information is mostly driven by the extension of myocardial ischemia (directly related to the number and location of obstructive lesions) and to ischemia intensity (directly related to the degree of stenosis).
Perhaps the most sought for information from CMR is myocardial fibrosis, as seen by late gadolinium enhancement (LGE). The myocardium is a densely packed cellular structure, with aligned myocytes lying side by side and with relatively little extracellular matrix (only approximately 25% of the tissue). Being an extracellular agent, gadolinium has little distribution in normal myocardium, but its concentration increases greatly in necrotic or fibrotic tissue. LGE has been shown to have great diagnostic importance in both ischemic and nonischemic cardiomyopathies (15,16). Because LGE can be seen with as little as 5 g of myocardial fibrosis, it is the most sensitive noninvasive method for fibrosis or necrosis detection. The presence of myocardial LGE has been extensively shown to be associated with worse prognosis, and not surprisingly, El Aidi et al. (8) have confirmed previous results from single studies, finding an adjusted HR between 2.82 and 9.43 across different studies of chronic or suspected CAD.
Regarding acute coronary syndromes, the authors present some confusing results. Theoretically, infarct size (as seen by CMR LGE) should be more closely related to prognosis than acute LVEF is, because myocardial stunning may obscure the long-term LVEF. But El Aidi et al. (8) have found that although LVEF was an independent predictor of major adverse cardiac events, infarct size was a predictor only on univariate analysis. In fact, they report that CMR contained no prognostic information at all regarding hard events in acute MI, which is very unlikely, both because of physiological reasons and because there is ample evidence that LVEF measured by echocardiography is related to hard events in the acute coronary syndrome setting (17). The reason for the authors' findings is the somewhat arbitrary (although the line has to be drawn somewhere) determination that any variable that has <1,000 patients studied did not present enough evidence for conclusions. Regardless of the validity of this 1,000-patient threshold, this brings into light the urgent need of CMR studies in the acute MI setting. It would also be interesting to understand how all CMR variables combined interact with the prognostic models, especially using individual patient data meta-analyses.
↵∗ Editorials published in the Journal of the American College of Cardiology reflect the views of the authors and do not necessarily represent the views of JACC or the American College of Cardiology.
Both authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- American College of Cardiology Foundation
- ↵Robert Storm Petersen. Available at: http://quoteinvestigator.com/2013/10/20/no-predict/. Accessed December 12, 2013.
- Lima J.A.,
- Judd R.M.,
- Bazille A.,
- Schulman S.P.,
- Atalar E.,
- Zerhouni E.A.
- El Aidi H.,
- Adams A.,
- Moons K.G.M.,
- et al.
- Klem I.,
- Heitner J.F.,
- Shah D.J.,
- et al.