Cardiac fibrosis is one of the hallmarks of pathological left ventricular (LV) remodeling,¹ which plays a significant role in the myocardial response to injury. Excessive cardiac fibrosis leads to progression of heart failure.² The objective of this study is to develop and evaluate a non-contrast cardiac magnetic resonance (CMR) approach for sensitive and quantitative assessment of excessive myocardial fibrosis.

Theory

According to two-site chemical exchange model, the spin-locking relaxation time T1ρ (ω₁) can be expressed for an on-resonance spin-locking module at a spin-locking frequency of ω₁:³

$$$R1\rho({\omega_{1}})=R_{2,0}+func({\omega_{1}},[collagen])$$$ (1)

Where: R_2,0 is an intrinsic transverse relaxation rate of water spins without chemical exchange contributions, which is assumed to be a constant for all ω₁ ; func is a function of site populations (free water and water bound to collagen), mean exchange rate at the two sites; and chemical shifts between two sites.

$$$R1\rho(0)-R1\rho(\omega_{1})=\triangle func(\omega_{1},[collagen])$$$ (2)

Myocardial fibrosis index (mFI) can be defined as

$$$mFI(\omega_{1})=T1(\omega_{1})-T1(0)=\frac{\triangle func \times T1\rho(0)}{1-\triangle func\times T1\rho(0)}\times T1\rho(0)$$$ (3)

Where $$$\triangle func \times T1\rho(0)$$$ is a function of myocardial fibrosis content and can be defined as pseudo ECV or pECV. Its value is between 0 and 1.

CMR in Patients

Ten patients with a variety of cardiomyopathy without obstructive coronary artery stenosis were prospectively scanned with and without administration of contrast media. CMR included non-contrast spin-locking T1ρ based imaging at two different spin-locking frequencies, native and post-contrast T1 mapping, as well as late gadolinium enhancement (LGE) imaging. A single dose of gadolinium contrast agent (0.1 mmol/kg, Magnevist, Schering) was administrated for contrast enhancement study. Three slices were obtained for each measurement. All Cardiac MR studies were performed on a Siemens 3T Trio System (Siemens Medical Solution, Malvern, PA) using an 8-element phased-array coil as the receiver.

Data Analysis

The mFI maps, native and post-contrast T1 maps, and ECV maps were all created using custom-made software. A standardized AHA 16-segment model was applied to for image analysis. The mFI maps was derived from the T1ρ dispersion data. The pECV values were calculated from T1ρ (0) and mFI using Eq. (3) on the myocardial segment basis. This index was compared with extracellular volume (ECV), calculated from native and post-contrast T1, for diffuse fibrosis detection. All correlations were determined by Pearson’s correlation coefficient.

For all myocardial segments free of artifacts, the correlation plots between ECV and 4 image contrasts, i.e., mFI, pECV, native T1, and post-contrast T1, are shown in Figure 1. A strong correlation between mFI and ECV was observed at r = 0.73, followed by moderate correlation for pECV and native T1, and a weak correlation for post-contrast T1. One a patient basis, excellent correlations were observed between ECV and mFI or pECV (Figure 2). Figure 3 demonstrates image examples obtained from the patient with alcoholic cardiomyopathy. One interesting observation was that the edema areas shown in the ECV map in the anterior and inferolateral regions were not seen in the corresponding mFI map. With increased diffuse fibrosis content, the change in the mFI exhibited up to 31 folds higher than changes in other non-contrast imaging parameters (native T1 and T1ρ).

This study introduced a new myocardial fibrosis index or mFI without the use of an exogenous contrast agent. The examinations in patients with non-ischemia cardiomyopathy revealed a strong correlation between the mFI and ECV, an established imaging marker for diffuse myocardial fibrosis. In comparison with other fibrosis imaging indices, such as native and post-contrast T1, the mFI demonstrated superior sensitivity.

The method for the improved sensitivity is based on T1ρ dispersion contrast, rather than absolute T1ρ contrast. For normal myocardial tissue, the T1ρ dispersion contrast between two SLFs should be relatively small.⁴ For myocardial tissue with excessive fibrosis, T1ρ dispersion contrast is expected to increase with fibrosis content and reach much higher intensity in the area with densely packed collagen, i..e, scar tissue. In comparison with native or post-contrast T1, this relatively higher sensitivity in mFI may be attributed by removing the nearly constant R_2,0, a parameter approximately 10-15 folds higher than mFI. This results in substantial higher sensitivity to detect chemical-exchange dominant T1ρ dispersion signals.⁵

This study is for the first time to introduce a non-contrast fibrosis index for the detection of myocardial fibrosis. This may have important implications for diagnosis and management of cardiac patients with cardiomyopathy and heart failure, particularly in those who have impaired renal function and/or need frequent surveillance for medical treatments.