sST2 and CA125 identify distinct biological phenotypes and predict mortality in very old patients hospitalised for acute HFpEF

Heart failure with preserved ejection fraction (HFpEF) is the most prevalent form in the elderly and is associated with poor outcomes. Its management is challenging due to comorbidities, frailty and polypharmacy. Ageing-related mechanisms play a key prognostic role, often not captured by conventional assessment, supporting biomarker-based phenotypic classification and risk stratification.

To evaluate whether Soluble Tumorigenicity Suppressor 2 (sST2) and carbohydrate antigen 125 (CA125) based phenotyping reflects distinct biological profiles and predicts one-year mortality in elderly patients with acute HFpEF.

Prospective observational study with patients (≥80 years) admitted for acute HFpEF in an internal medicine department. Clinical characteristics and biochemical parameters were collected. Patients were classified into four phenotypes according to biomarker elevation: both low, sST2 high, CA125 high, and both high. One-year all-cause mortality was recorded.

Variables were summarised using appropriate descriptive statistics and compared across phenotypes. Survival was analysed using Kaplan-Meier. Independent predictors of mortality were evaluated using Cox regression models adjusted for clinically relevant covariates. Results reported as hazard ratios (HR), 95% confidence intervals, p<0.05 considered significant.

Eighty patients were included (mean age 88.3±4.0 years;75% women). As shown in Table 1, baseline characteristics were comparable across phenotypes (p>0.05). In contrast, marked biological differences emerged. Patients with both biomarkers elevated showed lower body mass index (BMI), higher brain natriuretic peptide (BNP) levels, greater frailty, and increased inflammatory burden. C-reactive protein (CRP) levels differed significantly across phenotypes, being highest in patients with isolated sST2 elevation (median 83.0 mg/L [19.3–128.2];p=0.038). Elevated sST2 was associated with lower estimated glomerular filtration rate (eGFR), consistent with evidence linking sST2 to renal dysfunction, myocardial fibrosis and systemic inflammation.

During one-year follow-up, all-cause mortality differed across phenotypes, being lowest in patients with both biomarkers low and highest in those with concomitant elevation (Figure 1;p=0.099).

In multivariable Cox analysis adjusted for age, sex, BMI, eGFR, CRP, BNP and frailty, the both-high phenotype was independently associated with increased mortality (HR 4.28 [1.25–14.64];p=0.020). Reduced eGFR (HR: 1.02 [1.00–1.04];p=0.034) and higher frailty score (HR: 1.91 [1.14–3.19];p=0.014) were also independent predictors of death. The multivariable model was significant (p=0.008).

sST2–CA125-based phenotyping identifies distinct biological profiles and a high-risk phenotype defined by concomitant biomarker elevation independently associated with one-year mortality in elderly patients with acute HFpEF, supporting biomarker-guided risk stratification and follow-up.

For image description, please refer to the figure legend and surrounding text.For image description, please refer to the figure legend and surrounding text.