Cardiovascular magnetic resonance in an adult human population: serial observations from the multi-ethnic study of atherosclerosis

Abstract

The Multi-Ethnic Study of Atherosclerosis (MESA) is the first large-scale multi-ethnic population study in the U.S. to use advanced cardiovascular magnetic resonance (CMR) imaging. MESA participants were free of cardiovascular disease at baseline between 2000 and 2002, and were followed up between 2009 and 2011 with repeated CMR examinations as part of MESA. CMR allows the clinician to visualize and accurately quantify volume and dimensions of all four cardiac chambers; measure systolic and diastolic ventricular function; assess myocardial fibrosis; assess vessel lumen size, vessel wall morphology, and vessel stiffness. CMR has a number of advantages over other imaging modalities such as echocardiography, computed tomography, and invasive angiography, and has been proposed as a diagnostic strategy for high-risk populations. MESA has been extensively evaluating CMR imaging biomarkers, as markers of subclinical disease, in the last 15 years for low-risk populations. On a more practical level, some of the imaging biomarkers developed and studied are translatable to at-risk populations. In this review, we discuss the progression of subclinical cardiovascular disease and the mechanisms responsible for the transition to symptomatic clinical outcomes based on our findings from MESA.

Keywords: Ageing; Cardiovascular disease; Fibrosis; Heart failure; Torsion.

Fig. 1

The natural history of myocardial function in an adult human population. LVEDV statistically significantly decreased over 10 years for each age category a and LV mass increased in Men b and M/V ratio increased c despite LV mass did not progressively increase b and d. Although SV progressively falls, LVEF maintains due to progressive decline in LV volumes e. Aging is associated with the development of a concentric remodeling pattern secondary to a progressive decline in LV volume f. LV = left ventricular; LVEF = left ventricular ejection fraction; LVEDV = Left ventricular end diastolic volume; LVESV = Left ventricular end systolic volume; M/V = mass-to-volume ratio Figures prepared based on data from Eng et al. [8], and Cheng et al. [5]

Fig. 2

Left ventricular remodeling and incident cardiac events in MESA. The risk of incident cardiovascular disease is greater for those individuals who develop the concentric remodeling at a younger compared with older age (black arrows, a). Higher LV mass (hypertrophy) is associated with incident heart failure b. Additional LV mass and log NT-pro BNP produces the largest increased in c-statistic and improves the NRI beyond traditional risk factors c. Figures prepared based on data from Cheng et al. [5], Bluemke et al. [21], and Chahal et al. [22]. Abbreviations as in Fig. 1

Fig. 3

Fibrosis and left ventricular function in MESA. The subject without a LGE-defined scar had a relatively lower post contrast T1 time at 25 min with altered regional myocardial shortening and LVEF compared to the subject without an LGE-defined myocardial scar and no evidence of extra-cellular expansion. The subject with a LGE-defined replacement fibrosis had a progressively reduced systolic function. Figures prepared based on data from Donekal et al. [35]. LGE = late gadolinium-enhanced image; LVEF = left ventricular ejection fraction

Fig. 4

Cardiac deformation imaging using tagged CMR in asymptomatic individuals. Impaired Ecc is associated with lower myocardial perfusion and blood flow a Impaired LV circumferential shortening (Ecc) provides predictive value for incident heart failure during 5.5 years follow-up b LV torsion, however, is greater with older age despite lower stroke volume and myocardial shortening c. LV wringing motion might represents a compensatory mechanism for systolic dysfunction to maintain LVEF. Figures prepared based on date from Rosen et al. [38], Choi et al. [45], and Yoneyama et al. [6]. LAD = left anterior descending artery; LCx = left circumflex artery; RCA = right coronary artery. Abbreviations as in Fig. 1

Fig. 5

Potential pathways in the progression from asymptomatic to symptomatic cardiovascular events from MESA CMR remodeling and fibrosis data. Aging leads to concentric remodeling (higher M/V ratio and maintained LV mass) without no focal scar (pathway 1) [5, 6, 69]. Typically hypertension progresses to concentric hypertrophy (higher M/V ratio and LV mass) without focal scar (pathway 2) [6, 31]. The pathway from concentric remodeling and hypertrophy to replacement fibrosis without MI (pathway 3) [31]. Individuals with concentric remodeling can develop vascular events with preserved LVEF (pathway 4) [5, 21]. Individuals with concentric remodeling with MI or severe coronary calcification can contribute to replacement fibrosis (pathway 5) [29]. Individuals with concentric hypertrophy can develop symptomatic vascular events either with replacement scar (pathway 6) [28] or without (pathway 7) [21, 22]. Black arrows depict MESA CMR results, and thicker dot arrow, a known pathway. MI = myocardial infarction. Abbreviations as in Fig. 1

Fig. 6

Left atrial measurements by tissue-tracking CMR in asymptomatic individuals. LA volume, LA strain, LA strain rate, and LV volume in a asymptomatic participant free of cardiovascular daises as a control (block line) and a MESA participant who developed CHF (orange line). A case of CHF had a relatively greater LA volumes with impaired LA strains (S max and S preA) and LA strain rates (SRs, SRe, and SRa) than a control. Volumes are index to body surface area. LA function was analyzed by using a tissue-tracking method with semiauto mated software (multimodality tissue tracking [MTT] version 5.0; Toshiba, Tochigi, Japan). CHF; congestive heart failure; LA = left atrial; LV = left ventricular; S max = maximum strain; S preA = pre atrial contraction strain; SRa = strain rate at atrial contraction; SRe = strain rate at LV early diastole; SRs = maximum strain rate, V max = maximum indexed volume; V min = minimum indexed volume; V pre A = pre-atrial contraction indexed volume

Fig. 7

Aortic imaging using CMR in in asymptomatic individuals. Decreased proximal aorta distensibility predicted incident cardiovascular disease a MRA without contrast shows no significant stenosis in the proximal RCA, and black blood images shows normal wall and eccentric plaque b Black blood images obtained the carotid artery, and identify lipid core c. Figures prepared based on date from Redheuil et al. [62], Miao et al. [66] and Wasserman et al. [67]