Impaired left ventricular performance as a prognostic marker in patients with precapillary pulmonary hypertension: cardiac magnetic resonance feature tracking study

Abstract

Introduction Pulmonary hypertension (PH) causes right ventricle (RV) dysfunction. Cardiac magnetic resonance (CMR) derived RV ejection fraction (EF) has prognostic value in this condition. However, left ventricle (LV) function and mechanics could be affected in patients with RV failure as expression of ventricular interdependence. Despite the fact that the prognostic merit of specific imaging findings is still far from robust, conventional and the novel CMR technology – feature tracking (FT) LV parameters - could be an important additional prognostic value for survival outcomes. Aims and Objectives The aim of this study was to evaluate conventional CMR and FT parameters of LV mechanics in patients with precapillary PH (pPH) and to define the prognostic markers for adverse clinical outcomes. Materials and methods CMR cine imaging data of 43 patients with confirmed pPH diagnosis on right heart catheterization (mean pulmonary artery pressure (mPAP 55.91±15.87mmHg, pulmonary capillary wedge pressure <15 mmHg) were retrospectively analyzed. Patients were divided in two groups according to survival outcomes (survival and nonsurvival) during the follow-up and treatment period. Patients with documented coronary and valvular heart disease also with atrial fibrillation were excluded from the study. LV and RV EF, indices of RV end diastolic and end systolic volumes (EDVi and ESVi) were evaluated using conventional CMR software (syngo.via; Siemens Healthcare). LV global longitudinal and circumferential strains (LV-LS, LV-CS) were assessed in long (two, three, four chambers) and short axes (basal, mid and apical) views using CMR FT software package (Medis Suite QStrain 2.0; Medis Medical Imaging Systems bv). The average of global LS and CS were calculated. Independent Samples Mann-Whitney U test was used to compare the distribution across categories and Spearman’s correlation was performed. Results The distribution of [...].