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
Susana Ravassa # 1 2 , Begoña López # 1 2 , João Pedro Ferreira 3 , Nicolas Girerd 3 , Erwan Bozec 3 , Pierpaolo Pellicori 4 , Beatrice Mariottoni 5 , Franco Cosmi 5 , Mark Hazebroek 6 , Job A J Verdonschot 6 , Joe Cuthbert 7 , Johannes Petutschnigg 8 , María U Moreno 1 2 , Stephane Heymans 6 , Jan A Staessen 9 10 , Burkert Pieske 8 11 , Frank Edelmann 8 , Andrew L Clark 7 , John G F Cleland 4 , Faiez Zannad 3 , Javier Díez 1 2 12 , Arantxa González 1 2 , HOMAGE Trial Committees and Investigators
Aims: The HOMAGE randomised trial found that spironolactone reduced left atrial volume index (LAVI), E:A ratio, and a marker of collagen type I synthesis (PICP) 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 (CITP:MMP-1) ratio, 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 ratio (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-differencesspiro/control : -1.77 [95%CI:-2.94 to -0.59] and -2.52 [95%CI:-4.46 to -0.58] mL/m2 ; interaction-Pacross-tertiles =0.005; interaction-P3rd tertile =0.008) with a similar trend for NT-proBNP which was consistently reduced by spironolactone only in the lowest collagen cross-linking tertile (mean-differencesspiro/control :-0.47[95%CI:-0.66 to -0.28] and -0.31[95% CI:-0.59 to -0.04] ng/L; interaction-Pacross-tertiles =0.09; interaction-P3rd tertile <0.0001).
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).