Association of right ventricular structure and function with HFpEF and HFrEF: the Multi-Ethnic Study of Atherosclerosis (MESA)

Presenting Author(s)

• 

  Bharath Ambale-Venkatesh

  Physicist
  The John Hopkins Hospital
  Baltimore, Maryland, United States

Primary Author(s)

• 

  Bharath Ambale-Venkatesh

  Physicist
  The John Hopkins Hospital
  Baltimore, Maryland, United States

Co-Author(s)

• MM

  Monica Mukherjee, MD, MPH

  Assistant Professor of Medicine
  Johns Hopkins University School of Medicine
  Baltimore, Maryland, United States
• 

  Rob J. Van Der Geest, PhD

  Associate Professor in Medical Imaging
  Leiden University Medical Center
  Leiden, Zuid-Holland, Netherlands
• AA

  Ashkan Abdollahi

  Post-doctoral fellow
  The Johns Hopkins University, United Arab Emirates
• OC

  Omar Chehab, MD, MSc

  Cardiac Imaging Fellow / MD, MSc
  The John Hopkins Hospital, United States
• RZ

  Ralph Zeitoun, MD

  MD
  The Johns Hopkins Hospital, United States
• VV

  Vinithra Varadarajan, MD

  Cardiologist
  Division of Cardiology, Johns Hopkins University, Baltimore, Maryland 21287-0409, USA., United States
• GB

  Graham Barr, MD, PhD

  professor
  Columbia University, New York, United States
• CW

  Colin Wu

  Biostatician
  NIH
  Bethesda, Maryland, United States
• NA

  Norrina Allen

  Professor
  Northwestern University, United States
• WP

  Wendy Post, MD, MSc

  Cardiology / Professor
  The John Hopkins Hospital, United States
• SK

  Steven Kawut

  Professor
  University of Pennsylvania, United States
• 

  David A. Bluemke, MD

  MD
  University of Wisconsin-Madison
  Madison, Wisconsin, United States
• JL

  Joao A. C Lima, MD

  Professor
  The Johns Hopkins University
  Baltimore, Maryland, United States

Background: The association of right sided heart function with heart failure (HF) has been less studied. Most prior work has focused on the coupling of pulmonary hypertension and right ventricle (RV) function. [1-3] In this study, we aim to study the association of RV structure and function with incident HF with preserved and reduced ejection fraction (HFpEF and HFrEF)).

Methods: MESA has been described in the literature.[4] In the baseline exam, 5004 participants underwent CMR. RV structure and function were quantified in 4204 participants using cine fast gradient echo short-axis images using QMass (Medis).[5] All volumes (end-diastolic – RVEDV, end-systolic – RVESV, and stroke volume -RVSV) and mass were indexed to body surface area. RV strain was calculated as the ratio of the change in mid-ventricular RV free wall contour length to the original length. HF events were adjudicated as previously described.[6] Cox proportional hazards models with stepwise backwards variable selection assessed the power of RV parameters to predict HFpEF and HFrEF over a 19-year follow-up period. Models adjusted for HF risk factors, left ventricular (LV) ejection fraction and mass, and intermediate atherosclerotic cardiovascular, HFrEF/HFpEF, or atrial fibrillation events.

Results: The mean age of the participants was 61±10 years, 52.5% female, 43% hypertensive, 36% former smokers, 13% current smokers and 11.6% diabetic. Over 19 years of follow-up, 110 HFpEF and 124 HFrEF events were observed. 17 participants with HF for whom EF information at the time of the event was not available were excluded from all analyses. RVEF was further categorized based on reference ranges into 3-categories: < 42%,42-74%, and >74%.[7]

After adjustment for LVEF, none of the RV parameters significantly predicted incident HFrEF. Higher RVEF (both continuous and categorical >74%) and reduced RV strain (lower absolute strain values) were associated with incident HFpEF after adjustments. Higher RVEF was significantly associated even after adjustment for RV strain in the full model. Participants in the high RVEF group were predominantly women, older, had a lower BMI, higher HDL cholesterol, fewer smokers, and more likely hypertensive and had higher NT pro-BNP (see Table 2) as compared to those in the normal range. Participants in the high RVEF group also had lower RV mass and volumes, but increased RV mass-to-volume ratio and increased RV strain (greater absolute values).

Conclusion: In this large multi-ethnic cohort study high RVEF and low absolute RV strain were predictive of HFpEF. The higher RVEF group consisted predominantly of women with a seemingly healthier risk profile (including better RV strain) yet had higher levels of NT-proBNP at baseline indicating higher cardiac stress. In conclusion, we have identified a subset of individuals with a unique phenotype where the RV is in a hyperdynamic state – and is associated with higher risk of HFpEF, independent of their RV contractility.

Table 1. The table shows hazard ratios (confidence intervals) from Cox proportional hazards models used to study the association of RV parameters with incident heart failure with reduced ( HFrEF) and preserved (HFpEF) ejection fraction. The multivariable models were adjusted for demographics (age, gender, race), cardiovascular risk factors (hypertension medication use, systolic blood pressure, total and HDL cholesterol levels, lipid medication use, diabetes status, body mass index and cigarette smoking status), and interim events (atrial fibrillation and coronary heart disease). *p < 0.05. RVEDV: RV end-diastolic volume, RVESV: RV end-systolic volume, RVEF: RV ejection fraction, RVSV: RV stroke volume, RVEDM: RV end-diastolic mass.

	HFrEF (n=124)		HFpEF (n=110)
	Univariable	Multivariable	Univariable	Multivariable
RVEDV indexed, ml/m2	1.02 [1.001-1.03]*	-	0.99 [0.97-1.00]	-
RVESV indexed, ml/m2	1.04 [1.02-1.07]*	-	0.96 [0.93-0.99]*	-
RVEF, %	0.97 [0.94-0.99]*	-	1.04 [1.01-1.08]*	1.03 [1.01-1.07]*
RVSV indexed, ml/m2	1.01 [0.99-1.03]	-	1.00 [0.97-1.02]	-
RVEDM, g/m2	1.12 [1.02-1.22]*	-	0.92 [0.82-1.02]	-
RV tangential strain, %	0.99 [0.98-1.01]	-	0.97 [0.96 – 0.99]*	0.98 [0.97-0.99]*

Table 2. Characteristics of participants based on their RV ejection fraction. *p < 0.05 in comparison with the normal group (RVEF: 52-80%). RVEDV: RV end-diastolic volume, RVESV: RV end-systolic volume, RVEF: RV ejection fraction, RVSV: RV stroke volume, RVEDM: RV end-diastolic mass.

	RVEF: 42-74% (n=2931)	RVEF >74% (n=1273)
Age, yrs	60.8±10.0	63.1±10.2*
%Female	44.3	71.6*
% Hypertensive	39.9	50.0*
% Diabetic	11.9	10.9
% Former/Current smoker	50.2	44.0*
Body Mass Index, kg/m2	28.0±4.8	27.6±5.3*
Race/Ethnicity,% Caucasian Chinese-American African-American Hispanic	39.1 11.4 27.5 22.0	39.1 14.7* 24.0* 22.1
Total Cholesterol, mg/dl	193±35	197±35*
HDL cholesterol, mg/dl	49.6±14.2	54.6±16.2*
RVEDV indexed, ml/m2	68.4±12.8	62.5±11.4*
RVESV indexed, ml/m2	22.5±5.8	14.0±3.3*
RVSV indexed, ml/m2	46.0±8.8	48.5±8.8*
RVEDM, g/m2	11.6±1.9	10.9±1.8*
RVMVR, g/ml	0.17±0.02	0.18±0.02*
NT pro-BNP levels, pg/ml	45 [19-91]	69[36-137]*
RV tangential strain, %	-18.1±9.7	-23.0±11.7

Figure 1. Kaplan-Meier curve showing the survival rate by RV ejection fraction (RVEF) categories for the endpoint of heart failure with preserved ejection fraction (HFpEF).