Heart failure and atrial fibrillation are two of the most common heart diseases globally. Nearly half of all patients with heart failure also have atrial fibrillation. When heart failure and atrial fibrillation occur together, the risk of hospitalization and premature death increases significantly. However, there is a lack of reliable tools to assess how severely the heart is affected in these patients. This makes it difficult both to establish the correct diagnosis, tailor treatment, and predict who is at greatest risk of hospital admission or death from the disease.

One of the most important targets in heart failure is the filling pressure in the left ventricle. When this pressure is high, it means that the heart has difficulty receiving blood, leading to shortness of breath and fluid retention in the body. Today, filling pressure is usually estimated using ultrasound (echocardiography), but the available methods are primarily developed for patients without atrial fibrillation. In patients with both heart failure and atrial fibrillation, the measurements are so uncertain that they cannot be used as a reliable basis for clinical decision-making.

In this study, entitled Heart Failure combined with Atrial Fibrillation (HFcAF), the investigators will test new ultrasound methods that combine novel measures of cardiac chamber function with established techniques. Artificial intelligence will be used to identify the most useful combinations of parameters, select cardiac cycles that are best suited for analysis in atrial fibrillation, and automate and optimize the measurements. This approach may provide both more accurate and faster assessments, while also making the methods easier to implement in clinical practice. The aim is to improve the estimation of filling pressure so that it becomes more precise also in patients with atrial fibrillation. The investigators will then examine whether these improved methods can be used to predict which patients are at highest risk of hospitalization or death due to heart failure.

The study is designed as a prospective multicenter study, in which patients are recruited from several hospitals in different countries. This will make the results robust and generalizable to a wide range of patient populations. The investigators anticipate that the project will pave the way for better diagnostics and risk stratification in heart failure combined with atrial fibrillation and, in the longer term, contribute to improved guidelines and treatment for a large number of patients. If successful, the project will provide a new tool that can contribute to earlier and more targeted treatment, thereby improving quality of life and prognosis for a large group of patients.

Study Design:

This is a prospective, multicenter study of atrial fibrillation (AF) patients referred for right or left heart catheterization (RHC/LHC). The primary objective will be to develop and validate a clinically applicable algorithm combining echocardiographic and clinical parameters to differentiate normal from elevated left ventricular (LV) filling pressure (LVFP) in AF with ≥80% diagnostic accuracy. Secondary objectives will be to identify imaging predictors of abnormal pulmonary capillary wedge pressure (PCWP) elevation during exercise and to identify prognostic imaging markers associated with mortality and heart-failure outcomes over a 3-year follow-up period.

LVFP will be measured as PCWP during RHC and as LV end-diastolic pressure (LVEDP) during LHC. In cases studied during left atrial (LA) interventions, LA mean pressure will be used as a measure of LVFP. Echocardiography will be performed during, immediately before, or immediately after catheterization (≤8 hours separation, without cardiovascular medication changes). A dedicated exercise substudy will investigate imaging predictors of abnormal PCWP rise during exertion.

Recruitment of at least 400 patients is considered feasible based on experience from a previous international multicenter AF study. Patients will be recruited from 15 centers in the USA, Europe, Asia, and New Zealand over a period of 1.5 to 2 years.

Data collection:

A 12-lead electrocardiogram (ECG) and standard clinical data, including cuff blood pressure and standard blood tests with NT-proBNP, will be recorded.

Echocardiographic Imaging:

Echocardiographic measurements will be performed according to most recent American Society of Echocardiography / European Association of Cardiovascular Imaging guidelines. In addition, LA and right atrial (RA) strains will be measured. Echocardiographic recordings will be obtained by experienced investigators without knowledge of the invasive data. A core lab for echo analysis will be established. A minimum of 10 consecutive heartbeats will be recorded. Echo will be performed either simultaneously or within 8 hours of hemodynamic assessment. Equipment from different vendors will be used. Three-dimensional echocardiography will be used in selected patients.

Key Echocardiographic Views and Measurements

• Apical 4-chamber LV-focused: mitral inflow, tissue Doppler early diastolic velocity (e') (septal, lateral, average), systolic velocity (s'), isovolumetric relaxation time, LV global longitudinal strain (GLS), LV volumes and ejection fraction, LV mass
• Apical 4-chamber LA-focused: pulmonary vein Doppler (S and D velocities and their velocity-time integrals), LA strain, LA volumes
• Apical 5-chamber: aortic valve continuous-wave (CW) Doppler
• Apical 2- and 3-chamber: LV GLS, focused LV/LA imaging
• Right ventricular / pulmonary artery evaluation: tricuspid regurgitation velocity, right ventricular outflow tract (RVOT) pulsed-wave Doppler, RVOT acceleration time
• Subcostal inferior vena cava (IVC) view: IVC diameter and collapsibility.
• RA-focused view: RA reservoir strain

Cardiac Catheterization:

LVFP >15 mmHg will be considered elevated. LVFP will be averaged over 10 beats during end-expiration, using an index beat approach for selection of heart cycles.

Bicycle ergometer in supine position with simultaneous RHC will be performed in patients scheduled for this procedure as part of a diagnostic work-up and will be performed according to current clinical routine. This includes pressure measurements and focused echocardiographic study at rest and at peak exercise. The rate of pedaling will be 60 rotations per minute, and the workload will increase gradually to a moderate level.

Outcome Data:

All-cause mortality and heart failure hospitalizations will be the primary clinical outcomes. Outcome analysis will be extended to 3 years of follow-up.

Data Management and Analysis:

All imaging and invasive pressure measurements will be analyzed with investigators blinded to the corresponding data. Echocardiography will be reviewed centrally at the Echo Core Laboratory. Statistical analyses will include descriptive statistics, logistic regression, receiver operating characteristic curve analysis, and supervised machine-learning methods to derive a diagnostic algorithm. A two-sided P value <0.05 will be considered statistically significant.

Study Flowchart:

1. Screening → 2. Consent → 3. Clinical assessment & ECG → 4. Echocardiography → 5. RHC/LHC (± exercise) → 6. Data transfer to core labs → 7. Follow-up → 8. Analysis

Inclusion Criteria:

• Atrial fibrillation (paroxysmal, persistent, or permanent)
• Scheduled for RHC or LHC for clinical reasons
• Able to undergo echocardiography and invasive pressure measurement within 8 hours
• No cardiovascular medication changes between echo and catheterization
• Written informed consent provided

Exclusion Criteria:

• Mitral stenosis or mitral annular calcification causing severe functional stenosis
• Severe mitral or severe tricuspid regurgitation
• Prosthetic mitral valve
• Atrial fibrillation with rapid ventricular response >120 bpm
• Suboptimal echocardiographic imaging
• Conditions rendering PCWP or LVEDP unreliable
• Pregnant women
• Complex congenital heart disease
• LV assist device and patients with severe non-cardiac disease with poor prognosis
• Cardiac transplant patients
• End-stage liver disease

Patients with atrial fibrillation undergoing clinically indicated right or left heart catheterization.

De-identified individual participant data (IPD) underlying the results reported in this study, including the data dictionary, will be made available to qualified researchers upon reasonable request. Supporting documents, including the study protocol, statistical analysis plan, and analytic code, will also be available.

Data will become available from February 2026 and will remain available for 3 years.

Local Primary Investigator, administrative leader and the research director at each center will access the IPD and supporting information. Data will be provided in a de-identified format in accordance with General Data Protection Regulation and institutional regulations.