Prognostic value of the 2025 ASE/EACVI diastolic function algorithm in patients with heart failure
I Miyamoto, T Nagai, M Nakao, S Ishizaka, M Kambayashi, Y Kato, S Imagawa, Y Tokuda, M Takahashi, J Matsumoto, K Motoi, Y Wakamatsu, M Machida, T Saito, T AnzaiAbstract
Background
Elevated left ventricular filling pressure is a key pathophysiological feature of heart failure (HF) and is closely associated with adverse clinical outcomes. The 2016 ASE/EACVI framework improved standardization of diastolic assessment but frequently resulted in indeterminate classifications, limiting its clinical interpretability. The 2025 update introduced a structured stepwise algorithm for left atrial pressure evaluation designed to eliminate indeterminate categories. However, its prognostic significance in large contemporary HF populations remains uncertain.
Purpose
To evaluate classification patterns and prognostic value of the 2025 diastolic dysfunction grading in large contemporary HF cohort.
Methods
We examined 1,435 patients with chronic HF enrolled in a prospective multicentre registry (29 sites) between January 2020 and June 2025. Patients incompatible with the 2025 algorithm were excluded according to the guideline. Diastolic function was graded using the 2016 and 2025 algorithms.1,2 The primary outcome was a composite of all-cause death and hospitalisation for worsening HF.
Results
According to the 2025 algorithm, 122 patients (8.5%) were classified as normal, 555 (38.7%) as grade I, 641 (44.7%) as grade II, and 117 (8.2%) as grade III, with no indeterminate cases. Reclassification from the 2016 framework occurred across multiple categories, and all previously indeterminate patients were reassigned to definitive grades (Figure 1). During a median follow-up of 848 days (IQR 539–1347), the primary outcome occurred in 273 patients. The cumulative incidence increased progressively across diastolic grades (Figure 2). After adjustment for age, sex, systolic blood pressure, NYHA class, diabetes mellitus, haemoglobin, sodium, eGFR, NT-proBNP, and LVEF, grade I (HR 1.97; 95% CI 0.94–4.12), grade II (HR 2.08; 95% CI 1.01–4.29), and grade III (HR 3.14; 95% CI 1.44–6.84) were associated with increasing risk compared with normal diastolic function. After adjustment for the MAGGIC risk score and NT-proBNP, grade II (HR 2.22; 95% CI 1.09–4.56) and grade III (HR 2.99; 95% CI 1.39–6.47) remained significantly associated with the primary outcome. Addition of the 2025 algorithm to the MAGGIC risk score increased the c-index from 0.690 to 0.708 (P = 0.007). There was no significant interaction between diastolic grade and HF phenotype defined by LVEF (P = 0.61). Analyses using the 2016 framework showed similar prognostic associations. (Figure 1).
Conclusions
In a large contemporary HF cohort, the 2025 ASE/EACVI diastolic function algorithm was independently associated with adverse clinical outcomes and provided incremental prognostic value beyond established risk scores. Although its overall prognostic discrimination was comparable to the 2016 framework, the elimination of indeterminate classifications enables complete categorisation across the diastolic spectrum and facilitates structured risk stratification.