Can 3D RVEF be Prognostic for the Non-Ischemic Cardiomyopathy Patient but not the Ischemic Cardiomyopathy Patient? A Cardiovascular MRI Study.

BACKGROUND: While left ventricular ejection fraction (LVEF) has been shown to have prognostic value in ischemic cardiomyopathy (ICMX) patients, right ventricular ejection fraction (RVEF) has not been systematically evaluated in either ICMX or non-ischemic cardiomyopathy (NICMX) patients. Moreover, an accurate estimation of RVEF is problematic due to the geometry of the right ventricle (RV). Over the years, there have been improvements in the resolution, image acquisition and post-processing software for cardiac magnetic resonance imaging (CMR), such that CMR has become the "gold standard" for measuring RV volumetrics and RVEF. We hypothesize that CMR defines RVEF more so than LVEF and might have prognostic capabilities in ischemic and non-ischemic cardiomyopathy patients (ICMX and NICMX). METHODS: Patients that underwent CMR at our institution between January 2005 and October 2012 were retrospectively selected if three-dimensional (3D) LVEF < 35%. Patients were further divided into ICMX and NICMX groups. The electronic medical record (EMR) database inquiry determined all-cause mortality and major adverse cardiovascular events (MACE). Additionally, a Social Security Death Index (SSI) database inquiry was performed to determine all-cause mortality in patients who were lost to follow-up. Patients were further sub-grouped on the basis of 3D RVEF ≥ 20%. Separately, patients were sub-grouped by LVEF ≥ 20% in both ICMX and NICMX cases. A cut-off of ≥20% was chosen for the RVEF based on the results of prior studies showing significance based on Kaplan⁻Meier (KM) survival curves. Cumulative event rates were estimated for each subgroup using the KM analysis and were compared using the log-rank test. The 3D RV/LVEFs were compared to all-cause mortality and MACE. ICMX patients were defined using the World Health Organization (WHO) criteria. RESULTS: From a 7000-patient CMR database, 753 heart failure patients were selected. Eighty-seven patients met WHO definition of ICMX and NICMX (43 ICMX and 44 NICMX). The study patients were followed for a median of 3 years (Interquartile range or IQR 1.5⁻6.5 years). The mean age of patients was 58 ± 13 years; 79% were male. In ICMX, mean 3D LVEF was 21% ± 6% and mean 3D RVEF was 38% ± 14%, while for NICMX, mean 3D LVEF was 16% ± 6% and mean 3D RVEF was 30% ± 14% (p < 0.005 for intra- and inter-group comparison). It should be noted that LVEF < RVEF in both groups and the ejection fraction (EF) in NICMX was less than the corresponding EF in ICMX. Overall mortality was higher in ICMX than NICMX (12/40, 30% vs. 7/43, 16%; p < 0.05). Patients were stratified based on both RVEF and LVEF with a threshold of EF ≥ 20% separately. RVEF but not LVEF was a significant predictor of death for NICMX (χ² = 8; p < 0.005), while LVEF did not predict death in ICMX (χ² = 2, p = not significant). Similarly, time to MACE was predicted by RVEF for NICMX (χ² = 9; p < 0.005) but not by LVEF in ICMX (χ² = 1; p = NS). Importantly, RVEF, while predictive of NICMX MACE, did not emerge as a predictor of survival or MACE in ICMX. CONCLUSIONS: Via 3D CMR in non-ischemic CMX patients, RVEF has important value in predicting death and time to first MACE while 3D LVEF is far less predictive.

Cardio-hepatic risk assessment by CMR imaging in liver transplant candidates.

BACKGROUND: The preoperative workup of orthotopic liver transplantation (OLT) patients is practically complex given the need for multiple imaging modalities. We recently demonstrated in our proof-of-concept study the value of a one-stop-shop approach using cardiovascular MRI (CMR) to address this complex problem. However, this approach requires further validation in a larger cohort, as detection of hepatocellular carcinoma (HCC) as well as cardiovascular risk assessment is critically important in these patients. We hypothesized that coronary risk assessment and HCC detectability is acceptable using the one-stop-shop CMR approach. METHODS: In this observational study, patients underwent CMRI evaluation including cardiac function, stress CMR, thoracoabdominal MRA, and abdominal MRI on a standard MRI scanner in one examination. RESULTS: Over 8 years, 252 OLT candidates underwent evaluation in the cardiac MRI suit. The completion rates for each segment of the CMR examination were 99% for function, 95% completed stress CMR, 93% completed LGE for viability, 85% for liver MRI, and 87% for MRA. A negative CMR stress examination had 100% CAD event-free survival at 12 months. A total of 63 (29%) patients proceeded to OLT. Explant pathology confirmed detection/exclusion of HCC. CONCLUSIONS: This study further defines the population suitable for the one-stop-shop CMR concept for preop evaluation of OLT candidates providing a road map for integrated testing in this complex patient population for evaluation of cardiac risk and detection of HCC lesions.

Use of bio-informatics assessment schema (BIAS) to improve diagnosis and prognosis of myocardial perfusion data: results from the NHLBI-sponsored women's ischemia syndrome evaluation (WISE).

BACKGROUND: We introduce an algorithmic approach to optimize diagnostic and prognostic value of gated cardiac single photon emission computed tomography (SPECT) and magnetic resonance (MR) myocardial perfusion imaging (MPI) modalities in women with suspected myocardial ischemia. The novel approach: bio-informatics assessment schema (BIAS) forms a mathematical model utilizing MPI data and cardiac metrics generated by one modality to predict the MPI status of another modality. The model identifies cardiac features that either enhance or mask the image-based evidence of ischemia. For each patient, the BIAS model value is used to set an appropriate threshold for the detection of ischemia. METHODS: Women (n=130), with symptoms and signs of suspected myocardial ischemia, underwent MPI assessment for regional perfusion defects using two different modalities: gated SPECT and MR. To determine perfusion status, MR data were evaluated qualitatively (MRIQL) and semi-quantitatively (MRISQ) while SPECT data were evaluated using conventional clinical criteria. Evaluators were masked to results of the alternate modality. These MPI status readings were designated "original". Two regression models designated "BIAS" models were generated to model MPI status obtained with one modality (e.g., MRI) compared with a second modality (e.g., SPECT), but importantly, the BIAS models did not include the primary Original MPI reading of the predicting modality. Instead, the BIAS models included auxiliary measurements like left ventricular chamber volumes and myocardial wall thickness. For each modality, the BIAS model was used to set a progressive threshold for interpretation of MPI status. Women were then followed for 38±14 months for the development of a first major adverse cardiovascular event [MACE: CV death, nonfatal myocardial infarction (MI) or hospitalization for heart failure]. Original and BIAS-augmented perfusion status were compared in their ability to detect coronary artery disease (CAD) and for prediction of MACE. RESULTS: Adverse events occurred in 14 (11%) women and CAD was present in 13 (10%). There was a positive correlation of maximum coronary artery stenosis and BIAS score for MRI and SPECT (P<0.001). Receiver operator characteristic (ROC) analysis was conducted and showed an increase in the area under the curve of the BIAS-augmented MPI interpretation of MACE vs. the original for MRISQ (0.78 vs. 0.54), MRIQL (0.78 vs. 0.64), SPECT (0.82 vs. 0.63) and the average of the three readings (0.80±0.02 vs. 0.60±0.05, P<0.05). CONCLUSIONS: Increasing values of the BIAS score generated by both MRI and SPECT corresponded to the increasing prevalence of CAD and MACE. The BIAS-augmented detection of ischemia better predicted MACE compared with the Original reading for the MPI data for both MRI and SPECT.