Biomarker-based assessment of collagen cross-linking identifies patients at risk of heart failure more likely to benefit from spironolactone effects on left atrial remodelling. Insights from the HOMAGE clinical trial

S. Ravassa, B. Lopez, J.P. Ferreira, N. Girerd, E. Bozec, P. Pellicori, B. Mariottoni, F. Cosmi, M. Hazebroek, J.A.J. Verdonschot, J. Cuthbert, J. Petutschnigg, M.U. Moreno, S. Heymans, J.A. Staessen, B. Pieske, F. Edelmann, A.L. Clark, J.G.F. Cleland, F. ZannadJ. Diez*, A. Gonzalez*, HOMAGE Trial Committees and Investigators

*Corresponding author for this work

Research output: Contribution to journalArticleAcademicpeer-review

9 Citations (Web of Science)


Aims The HOMAGE randomized trial found that spironolactone reduced left atrial volume index (LAVI), E:A ratio, and a marker of collagen type I synthesis (procollagen type I C-terminal propeptide) in patients at risk of heart failure (HF). Previous trials showed that patients with HF, preserved ejection fraction and low serum collagen type I C-terminal telopeptide to matrix metalloproteinase-1 ratio (CITP:MMP-1), associated with high collagen cross-linking, had less improvement in diastolic function with spironolactone. We evaluated the interaction between serum CITP:MMP-1 and spironolactone on cardiac function in the HOMAGE trial.

Methods and results Patients at risk of HF were randomized to spironolactone (n = 260) or not (n = 255). Blood sampling and echocardiography were done at baseline, one and nine months. CITP:MMP-1 was used as an indirect measure of collagen cross-linking. Higher baseline CITP:MMP-1 (i.e. lower collagen cross-linking) was associated with greater reductions in LAVI with spironolactone at both one (p = 0.003) and nine (p = 0.01) months, but no interaction was observed for E:A ratio. Spironolactone reduced LAVI after one and nine months only for those patients in the third tertile of CITP:MMP-1 (estimated lowest collagen cross-linking) [mean difference(sspiro/control): -1.77 (95% confidence interval, CI -2.94 to -0.59) and -2.52 (95% CI -4.46 to -0.58) mL/m(2); interaction p(across-tertiles) = 0.005; interaction p(third tertile) = 0.008] with a similar trend for N-terminal pro-B-type natriuretic peptide which was consistently reduced by spironolactone only in the lowest collagen cross-linking tertile [mean differences(spiro/control): -0.47 (95% CI -0.66 to -0.28) and -0.31 (95% CI -0.59 to -0.04) ng/L; interaction p(across-tertiles) = 0.09; interaction p(third tertile) < 0.001].

Conclusions These findings suggest that, for patients at risk of HF, the effects of spironolactone on left atrial remodelling may be more prominent in patients with less collagen cross-linking (indirectly assessed by serum CITP:MMP-1).


Patients at risk of heart failure from the HOMAGE clinical trial were classified according to the baseline degree of myocardial collagen cross-linking, non-invasively assessed by the serum collagen type I C-terminal telopeptide (CITP) to matrix metalloproteinase-1 (MMP-1) ratio (CITP:MMP-1). As highly cross-linked collagen fibres are more resistant to degradation and CITP is a cross-linked peptide, for a given MMP-1 quantity less CITP will be released and, subsequently, serum CITP:MMP-1 will be lower. Whereas patients with low collagen cross-linking (high CITP:MMP-1) benefit from the cardioprotective effects of treatment with spironolactone on left atrial remodelling [i.e. a decrease in left atrial volume index (LAVI)] and on N-terminal pro-B-type natriuretic peptide (NT-proBNP) levels, these beneficial effects are not found in patients with higher collagen cross-linking (low CITP:MMP-1).

Original languageEnglish
Pages (from-to)321-331
Number of pages11
JournalEuropean journal of heart failure
Issue number2
Early online date9 Dec 2021
Publication statusPublished - Feb 2022


  • Heart failure
  • Spironolactone
  • Atrial remodelling
  • Collagen cross-linking

Cite this