Prognostic Value of CMR-Derived Left Ventricular Filling Pressure in Broad Referral Populations

Figures

• 

  Figure 1. Multivariable Association of Cardiac Magnetic Resonance-Derived Left Ventricular Filling Pressure (Continuous) With the Composite Outcome Forest plot showing the results of a multivariable Fine-Gray model with LVFPcmr entered as a continuous variable. Variables marked with an asterisk were modeled using restricted cubic splines, with HRsub reported for Q3 vs Q1, and P values derived from the joint Wald test of spline terms. CCS = Canadian Cardiovascular Society; CMR = cardiac magnetic resonance; EF = ejection fraction; HCM = hypertrophic cardiomyopathy; HRsub = subdistribution HR; EP = electrophysiology; LGE = late gadolinium enhancement; LV = left ventricle; LVFPcmr = cardiac magnetic resonance-derived left ventricle filling pressure; NICM = nonischemic cardiomyopathy; PVC = premature ventricular contractions; RV = right ventricle.

• 

  Figure 2. Association of continuous Cardiac Magnetic Resonance-Derived Left Ventricular Filling Pressure With the Outcome by Left Ventricular Ejection Fraction Partial effect plot showing adjusted relative HRsub for LVFPcmr modeled using restricted cubic splines vs the optimal cutpoint of 18 mm Hg, while adjusting for other variables shown in Figure 1 and including an interaction term with LVEF categories. Abbreviations as in Figure 1.

• 

  Figure 3. Multivariable Association of Dichotomized Cardiac Magnetic Resonance-Derived Left Ventricular Filling Pressure With the Composite Outcome Forest plot showing the results of a multivariable Fine-Gray model with LVFPcmr dichotomized at the calculated optimal cutpoint of 18 mm Hg. Variables marked with an asterisk were modeled using restricted cubic splines, with HRsub reported for Q3 vs Q1, and P values derived from the joint Wald test of spline terms. Abbreviations as in Figure 1.

• 

  Figure 4. Cumulative incidence of outcome for Cardiac Magnetic Resonance-Derived Left Ventricular Filling Pressure with Left Ventricular Ejection Fraction >50% Unadjusted cumulative incidence plots of the composite heart failure outcome for LVFPcmr dichotomized at the optimal cutpoint of 18 mm Hg, in patients with LVEF >50%. (A to C) Overall cohort and patients with and without known CAD. 466, 94, and 372 patients were omitted from each cohort, respectively, due to missing LVFPcmr (ie, values were not imputed for these plots). P value is reported from Gray test. CAD = coronary artery disease; LVFP = left ventricle filling pressure; other abbreviations as in Figure 1.

• 

  Central Illustration. Overview of the Main Findings of the Study Sex-specific CMR-derived LVFP was independently associated with the future heart failure composite outcome in the overall cohort, as well as subgroups with and without coronary artery disease, with the strongest effect magnitude in the subgroup having preserved LV ejection fraction (>50%). aHR = adjusted HR; CAD = coronary artery disease; CMR = cardiac magnetic resonance; EF = ejection fraction; LV = left ventricle; LVFP = left ventricle filling pressure; LVFPcmr = cardiac magnetic resonance-derived left ventricle filling pressure.