Plasma procollagen I and III N-terminal propeptides as adjunctive indicators of structural remodelling in acute heart failure

Myocardial structural remodelling driven by extracellular matrix (ECM) dysregulation contributes to disease progression and adverse outcomes in patients with acute heart failure (HF). Although imaging modalities can characterize remodelling, their availability remains limited in resource-constrained settings. Circulating collagen turnover biomarkers, including procollagen type I and III N-terminal propeptides (PINP and PIIINP), may reflect ECM remodelling, but their diagnostic utility across HF phenotypes remains uncertain. In this cross-sectional study, 200 adults were enrolled, including 150 patients hospitalized with acute HF (50 HF with reduced ejection fraction [HFrEF], 50 HF with mildly reduced ejection fraction [HFmrEF], and 50 HF with preserved ejection fraction [HFpEF]) and 50 healthy controls. Plasma PINP and PIIINP concentrations were measured by enzyme-linked immunosorbent assay. Structural remodelling was assessed using left atrial volume index (LAVI), with remodelling defined as LAVI ≥ 34 mL/m2. Principal component analysis (PCA) was exploratorily performed using standardized ECM biomarkers (PINP, PIIINP, soluble suppression of tumorigenicity [sST2], and matrix metalloproteinases 9 [MMP-9]) to derive a composite biomarker score (PC1). Multivariable logistic regression identified independent predictors of LAVI-defined remodelling. Discriminatory performance was evaluated using receiver operating characteristic (ROC) curves and area under the curve (AUC) analysis. Patients with HF were older than controls (55.6 vs. 39.3 years; p < 0.001) and had significantly higher plasma PIIINP concentrations (33.3 vs. 25.7 ng/mL; p < 0.001). LAVI-defined structural remodelling was present in 74% (95% confidence interval [CI] 66–81%) of patients with HF and was most prevalent in HFrEF (84%; p = 0.040). PINP correlated positively with PIIINP (r = 0.44; p = 0.001). In multivariable analysis, higher left ventricular mass index (LVMI) was independently associated with remodelling (adjusted odds ratio [OR] 1.02 per 1 g/m2; p = 0.003), whereas higher PIIINP levels (adjusted OR 0.95 per 1 ng/mL; p = 0.009) and higher Kansas City Cardiomyopathy Questionnaire (KCCQ) score (adjusted OR 0.95 per 1-point; p = 0.021) were inversely associated. Both PINP and PIIINP demonstrated limited discrimination for LAVI-defined remodelling (AUC 0.57 [0.46–0.67], AUC 0.58 [0.46–0.69] respectively), and PC1 did not improve performance (AUC 0.58 [0.50–0.66]. In exploratory phenotype-stratified analyses, PIIINP demonstrated the highest discriminatory performance in HFmrEF (AUC 0.80 [0.66–0.92]), but should be interpreted cautiously given the small subgroup size. In acute HF, circulating procollagen biomarkers reflect ECM turnover but do not reliably discriminate structural remodelling. The inverse association between PIIINP and LAVI suggests a dissociation between dynamic collagen turnover and structural burden in acute setting. These findings support prioritizing imaging-based assessment, with biomarkers interpreted cautiously as indicators of biological activity rather than structural remodelling.