Non-invasive pulse-wave-analysis for screening of HFpEF in at-risk patients: findings from the ARTS-PREDICTION study

Background

Heart failure with preserved ejection fraction (HFpEF) is an increasing public health burden linked to cardiovascular comorbidities. Simple, reliable screening tools are needed to identify high-risk individuals for early detection of cardiac dysfunction and timely treatment. Arterial stiffness commonly coexists with HFpEF and may precede clinical manifestation or reflect early pathophysiology, suggesting potential value for early detection of unrecognized HFpEF.

Purpose

To evaluate whether non-invasive pulse wave analysis (PWA)-derived markers identify prevalent or emerging HFpEF phenotypes in at-risk individuals.

Methods

In the ongoing prospective ARTerial Stiffness as a PREdictor of DIastolic DysfunCTION (ARTS-PREDICTION) study, patients referred to outpatient clinics for non-heart failure-related consultations underwent PWA (arteriograph) and longitudinal follow-up with clinical, laboratory and echocardiographic assessment (Fig. 1A). Eligible patients had ≥1 HFpEF risk factor (age >60 years, arterial hypertension, diabetes, chronic kidney disease, atrial fibrillation or obesity). Exclusion criteria included known HFpEF, LVEF <50%, cardiomyopathies, severe valvular heart disease, or pulmonary disease. PWA comprised aortic augmentation index (AIX) and pulse wave velocity (PWV). HFpEF phenotype was defined as new diagnosis of HFA-PEFF Score ≥5 at index visit or within one-year follow-up. HFpEF-ABA score served as reference. Associations were assessed using age- and sex-adjusted logistic regression. Diagnostic performance was evaluated by area under the receiver operating characteristic curve (AUC) analysis and net reclassification improvement (NRI).

Results

From 02/24-02/26, 130 patients with HFpEF risk factors (54±17y, 31% female) completed the study protocol. A HFpEF phenotype was newly diagnosed in 27 patients (21%); only 10 (37%) had overt symptoms (NYHA II: 5; III: 5). Most diagnoses occurred at index visit (78%), with 22% detected during follow-up.

Baseline characteristics are shown in Table 1. Patients with a HFpEF phenotype were older and had lower eGFR, higher NT-proBNP, higher AIX and worse diastolic function; PWV was similar. AIX was independently associated with a HFpEF phenotype (OR per SD 2.29, 95%CI 1.26-4.16, p=0.007). For detecting unrecognized HFpEF (Fig. 1B), AIX showed an AUC of 0.76 (95%CI 0.66-0.85), increasing to 0.84 (95%CI 0.77-0.92) after age adjustment and outperforming the HFpEF-ABA score (AUC 0.74, 95%CI 0.63-0.85). Combining AIX with HFpEF-ABA yielded an AUC of 0.83 (95%CI 0.74-0.91). Age-adjusted AIX improved risk classification over HFpEF-ABA (continuous NRI 0.58, 95%CI 0.17-0.99).

Conclusions

In this prospective study, PWA-derived AIX improved discrimination of patients with an unrecognized HFpEF phenotype, outperforming established clinical-only scores. PWA-derived measurements may represent a simple and efficient tool for screening patients at risk for unrecognized HFpEF.Figure 1Table 1