The predictive capabilities of a novel cardiovascular magnetic resonance derived marker of cardiopulmonary reserve on established prognostic surrogate markers in patients with pulmonary vascular disease: results of a longitudinal pilot study

Abstract

Background: No unified method exists to effectively predict and monitor progression of pulmonary arterial hypertension (PAH). We assessed the longitudinal relationship between a novel marker of cardiopulmonary reserve and established prognostic surrogate markers in patients with pulmonary vascular disease.

Methods and results: Twenty participants with confirmed (n = 14) or at high risk (n = 6) for PAH underwent cardiovascular magnetic resonance (CMR) at baseline and after ~6 months of guideline-appropriate management. Ten PAH participants underwent RHC within 48 h of each CMR. RHC (mean pulmonary arterial pressure, mPAP; pulmonary vascular resistance index, PVRI; cardiac index, CI) and phase-contrast CMR (mean pulmonary arterial blood flow velocity, meanPAvel) measurements were taken at rest and during continuous adenosine infusion (70/140/210 mcg/kg/min). Initial meanPAvel's (rest and hyperemic) were correlated with validated surrogate prognostic parameters (CMR: RV ejection fraction, RVEF; RV end systolic volume indexed, RVESVI; RHC: PVRI, CI; biomarker: NT-pro brain natriuretic peptide, NTpBNP; clinical: 6-min walk distance, 6MWD), a measure of pulmonary arterial stiffness (elastic modulus) and volumetric estimation of RV ventriculoarterial (VA) coupling. Changes in meanPAvel's were correlated with changes in comparator parameters over time. At initial assessment, meanPAvel at rest correlated significantly with PVRI (inversely), CI (positively) and elastic modulus (inversely) (R ² > 0.37,P < 0.05 for all), whereas meanPAvel at peak hyperemia correlated significantly with PVRI, RVEF, RVESVI, 6MWD, elastic modulus and VA coupling (R ² > 0.30,P < 0.05 for all). Neither resting or hyperemia-derived meanPAvel correlated with NTpBNP levels. Initial meanPAvel at rest correlated significantly with RVEF, RVESVI, CI and VA coupling at follow up assessment (R ² > 0.2,P < 0.05 for all) and initial meanPAvel at peak hyperemia correlated with RVEF, RVESVI, PVRI and VA coupling (R ² > 0.37,P < 0.05 for all). Change in meanPAvel at rest over time did not show statistically significant correlation with change in prognostic parameters, while change in meanPAvel at peak hyperemia did show a significant relationship with ΔRVEF, ΔRVESVI, ΔNTpBNP and ΔCI (R ² > 0.24,P < 0.05 for all).

Conclusion: MeanPAvel during peak hyperemia correlated with invasive, non-invasive and clinical prognostic parameters at different time points. Further studies with predefined clinical endpoints are required to evaluated if this novel tool is a marker of disease progression in patients with pulmonary vascular disease.

Fig. 1

Reference sequences for main pulmonary arterial (MPA) phase contrast imaging were two double-oblique orthogonal views along the main axis of the pulmonary trunk (a). The endocardial border of the MPA was manually outlined at all 20 reconstructed cardiac phases (b and c) permitting flow velocity profiles to be generated at rest and hyperemia (d and e respectively: flow velocity in ml/s on y axis and time in milliseconds on x axis). MeanPAvel was calculated as the average blood flow velocity across all cardiac phases. Ao = aorta, RV = right ventricle, RVOT = right ventricular outflow tract

Fig. 2

Changes in RHC-derived hemodynamics (panel 1) and CMR-derived meanPAvel (initial assessment, panel 2a; follow-up assessment, panel 2b) in response to adenosine in PAH and High Risk participants. At RHC, intravenous adenosine produced a dose dependent reduction in PVRI driven predominantly by a higher CI rather than a lower mPAP or transpulmonary gradient (not illustrated). At CMR, these changes were reflected by a dose-dependent rise in meanPAvel. Changes were more pronounced in High Risk participants relative to those with PAH

Fig. 3

Scatterplots illustrating the relationship between RVEF, meanPAvel at rest and meanPAvel at hyperemia at different time points R = Pearson correlation coefficient.