• Christine L Jellis, Teerapat Yingchoncharoen, Neville Gai, Kenya Kusunose, Zoran B Popović, Scott Flamm, and Deborah Kwon.
• Department of Cardiovascular Medicine, Desk J1-5, Heart and Vascular Institute, Cleveland Clinic, 9500 Euclid Avenue, Cleveland, OH, 44195, USA. jellisc@ccf.org.
• Int J Cardiovasc Imaging. 2018 Jan 1; 34 (1): 55-65.

AbstractRight ventricular (RV) fibrosis is increasingly recognized as the underlying pathological substrate in a variety of clinical conditions. We sought to employ cardiac magnetic resonance (CMR) techniques of strain imaging and longitudinal relaxation time (T1) mapping to better examine the relationship between RV function and structure. Our aim was to initially evaluate the feasibility of these techniques to evaluate the right ventricle. We then sought to explore the relationship between RV function and underlying fibrosis, along with examining the evolution of RV remodeling according to the amount of baseline fibrosis. Echocardiography was performed in 102 subjects with non-ischemic cardiomyopathy. Right ventricular parameters were assessed including: fractional area change (FAC) and longitudinal strain. The same cohort underwent CMR. Post-contrast T1 mapping was performed as a marker of fibrosis with a Look-Locker technique using inversion recovery imaging. Mid-ventricular post-contrast T1 values of the RV free wall, RV septum and lateral LV were calculated using prototype analysis software. Biventricular volumetric data including ejection fraction was measured by CMR using a cine short axis stack. CMR strain analysis was also performed to assess 2D RV longitudinal and radial strain. Simultaneous biochemical and anthropometric data were recorded. Subjects were followed over a median time of 29 months (IQR 20-37 months) with echocardiography to evaluate temporal change in RV FAC according to baseline post-contrast T1 values. Longitudinal data analysis was performed to adjust for patient loss during follow-up. Subjects (62% men, 51 ± 15 years) had mild to moderately impaired global RV systolic function (RVEF = 39 ± 15%; RVEDV = 187 ± 69 ml; RVESV = 119 ± 68 ml) and moderate left ventricular dysfunction at baseline (LVEF 30 ± 17%). Good correlation was observed between mean LV and RV post-contrast T1 values (r = 0.652, p < 0.001), with similar post-contrast T1 values maintained in both the RV free wall and septum (r = 0.761, p < 0.001). CMR RVEF demonstrated a proportional correlation with echocardiographic measures of RV longitudinal function and CMR RV strain (longitudinal r = -0.449, p = 0.001; radial r = -0.549, p < 0.001). RVEF was related to RV post-contrast T1 values, particularly in those with RV dysfunction (free wall T1 r = 0.259 p = 0.027; septal T1 r = 0.421 p < 0.001). RV strain was also related to RV post-contrast T1 values (r = -0.417, p = 0.002). Linear regression analysis demonstrated strain and post-contrast T1 values to be independently associated with RVEF. Subjects with severe RV dysfunction (CMR RVEF <25%) demonstrated lower RV CMR strain (longitudinal p = 0.018; radial p < 0.001), RV T1 values (free wall p = 0.013; septum <0.001) and RV longitudinal echocardiography parameters despite no difference in afterload. During follow-up, those with RV free wall post-contrast T1 values ≥ 350 ms demonstrated ongoing improvement in FAC (Δ6%), whilst values <350 ms were associated with deterioration in RV function (ΔFAC = -5%) (p = 0.026). CMR provides a comprehensive method by which to evaluate right ventricular function. Post-contrast T1 mapping and CMR strain imaging are technically feasible and provide incremental information regarding global RV function and structure. The proportional relationship between RV function and post-contrast T1 values supports that myocardial fibrosis is a causative factor of RV dysfunction in NICM, irrespective of RV afterload. This same structural milieu also appears integral to the propensity for both positive and negative RV remodeling long-term, suggestive that this is also determined by the degree of underlying RV fibrosis.

43 keywords...

hide…