Evaluation of cardiovascular magnetic resonance imaging derived right atrial pressure estimation for risk classification in pulmonary hypertension

Background

Right atrial pressure (RAP) is used for risk assessment in pulmonary hypertension (PH) according to guideline recommendations. Non-invasive surrogate parameters such as the diameter of the inferior vena cava emerge inaccurate. Recently, models for RAP calculation based on RA and right ventricular (RV) volumes as well as atrial strain have been published. Consequently, we sought to evaluate these models for RAP calculation in PH.

Methods

The study population consisted of patients referred to right heart catheterisation (RHC) and cardiovascular magnetic resonance (CMR) imaging enrolled in a prospective monocentric registry. Patients were classified by RHC according to current guideline recommendations including low (RAP <8mmHg), intermediate (RAP 8-14mmHg) and high risk (RAP >14mmHg) subclassification. CMR analyses included RA/RV end-diastolic/systolic volumes (EDV/ESV) and deformation imaging (total (Es), passive conduit (Ee) and active (Ea) strain).

Results

The study population consisted of n=209 patients (n=55 without PH, n=72 pre-capillary, n=27 combined post- and pre-capillary pulmonary hypertension (CpcPH), n=15 isolated post-capillary pulmonary hypertension (IpcPH), n=34 exercise and n=6 unclassified PH). Correlation of invasive RAP and calculated RAP using recently proposed models based on 1. RA ESV (r=0.38, p<0.001), 2. RA ESV and RA Es (r=0.24, p<0.001) and 3. RA EDV/ESV and RV ESV (r=0.39, p<0.001) was overall moderate. All models overestimated invasive RAP (p<0.001). In contrast, RAP and afterload as mean pulmonary artery pressure (mPAP) showed higher correlation (r=0.52, p<0.001).

All atrial strains (p<0.001-0.022), atrial volumes (p<0.001-0.004) as well as model 1 and 3 (p=0.004 and 0.010) differed between low and intermediate risk groups. Only Ee differed significantly between intermediate and high-risk groups (p=0.038) and accurately identified high-risk patients (AUC 0.72). In the present population, a composite of RA Ee, RV EF and mPAP best identified high-risk patients (AUC 0.85 vs model 1 (AUC 0.64) and 3 (AUC 0.66), p<0.001).

Conclusions

Right atrial pressures are the composite result of pre-, afterload and cardiac function. Models based on right cardiac volumes and function alone emerge insufficient for RAP estimation and risk stratification in PH, underlining the persistent value of invasive haemodynamic assessments.