Portal de Periódicos da CAPES

179 Deriving mean right atrial pressure from CMR: development of a model from the aspire registry

2024; Akadémiai Kiadó; Linguagem: Inglês

10.1136/heartjnl-2024-bcs.176

2732-0960

Tom Newman, Gareth Matthews, Hosamadin Assadi, Rui Li, Ciaran Grafton‐Clarke, Zia Mehmood, Bahman Kasmai, Chris Sawh, Liang Zhong, João L. Cavalcante, Ross J Thomson, Nay Aung, Rob J. van der Geest, Andrew J. Swift, Pankaj Garg,

Cardiovascular Health and Disease Prevention

Background Right atrial pressure (RAP) corresponds to fluid status and preload and is also important in prognostication for patients with heart failure and pulmonary hypertension. RAP can be measured invasively or non-invasively, but cannot currently be estimated by cardiac MRI (CMR). This study used paired CMR and invasive right heart catheter (RHC) measurements to develop a model to predict RAP from CMR. Methods The ASPIRE registry consists of patients referred for assessment of dyspnoea to Sheffield Teaching Hospitals between 2012 and 2020. Inclusion criteria were age >18 years, signs and symptoms of heart failure and adequate CMR image quality. Patients diagnosed with pulmonary arterial hypertension were excluded. RHC and CMR were performed in the same 24 hour period. CMR was performed with a 1.5 T GE HDx scanner. Chamber dimensions, strain, ejection fraction and stroke volume were acquired from 2- and 4- chamber cine-images using operator reviewed AI contours in MASS research software (Fig 1). Associations between invasive mean RAP (mRAP) and CMR metrics were assessed with Pearson's product-moment correlation coefficient. Stepwise multiple linear regression was used to develop models to predict mRAP. These were compared with receiver-operator curve analysis and DeLong's test. Results The cohort was made up of 672 patients, divided based upon invasive mRAP ≤ 8 mmHg (44%) and mRAP > 8 mmHg (56%) (table 1). Those with higher mRAP tended to be older, more likely to be male and have a higher diastolic blood pressure but there was no difference in rates of different types of heart failure (table 1). Metrics with the strongest correlation to invasive mRAP were right atrial (RA) dimensions, strain and ejection fraction with moderate correlation to right ventricular (RV) measurements (table 2). Four multivariable models were developed. Model 1 contained RA dimensions. RA strain was added to this to create Model 2. Model 3 contained RA and RV dimensions whilst model 4 incorporated RA end systolic volume (RAESV), the single strongest CMR predictor, and adjusted it for body mass index and sex. All four models had similar predictive capability (Fig 2 Panel A). Model 1 identified RAESV as the only variable required for mRAP prediction (coefficient = 0.06, p < 0.001) and was therefore preferred for simplicity. Using a threshold of mRAP >8 mmHg ROC analysis demonstrated an area under the curve of 0.78 (95% Confidence interval 0.75 to 0.81) (Fig 2 Panel B). Conclusions mRAP can be estimated with moderate confidence from CMR RAESV. Further studies are required to validate this model externally, including larger cohorts with reduced ejection fraction, and determine its clinical significance. Alternative clinical and MRI data could be explored to further enhance the models predictive capability. Application of this technique could enhance the role of CMR in non-invasive haemodynamic assessment and a wide range of cardiovascular pathophysiology. Conflict of Interest None