Urinary Proteomic Signature of Mineralocorticoid Receptor Antagonism by Spironolactone: Evidence from the Randomized-Controlled HOMAGE and PRIORITY Trials

Abstract

BACKGROUND: Mineralocorticoid receptor (MR) activation induces fibrosis. Urinary proteomic profiling (UPP) detects thousands of sequenced peptides, mainly derived from collagen. No previous study applied UPP to generate insights in the antifibrotic actions of MR antagonism. METHODS: Based on urine sample availability, subsets of the open HOMAGE trial (n=290; 23.8% women; median age: 73 years) and the double-blind PRIORITY trial (n=110; 21.8% women; 64 years) were analyzed as discovery and replication data sources. In the open HOMAGE trial, patients at risk of heart failure were randomized to usual therapy or usual therapy combined with spironolactone 25-50 mg/d. In the double-blind PRIORITY trial, type 2 diabetic patients with normal renal function were randomized to placebo or spironolactone 25 mg/d, both given on top of usual therapy. UPP relied on capillary electrophoresis coupled with mass spectrometry. In HOMAGE, the PICP/CITP ratio was calculated from serum PICP and serum CITP, which are markers of type-1 collagen synthesis and degradation, respectively. After rank-normalization of the biomarker distributions, between-group differences in the biomarker changes were analyzed by multivariable models. Correlations between the changes in urinary peptides and in serum CITP, derived from mature type-1 collagen, were compared between groups, using Fisher Z transform. RESULTS: In the HOMAGE and PRIORITY analytical subsets, patients had detectable signals of 1498 urinary peptides. Follow-up totaled 9 months in HOMAGE and was 30 months (median) in PRIORITY. All changes in urinary peptides that remained significantly different (P<0.05) between randomization groups after accounting for baseline levels, covariables and multiple testing were collagen fragments. In HOMAGE and PRIORITY spironolactone reduced 16/27 and 10/13 collagen-derived urinary peptides. In HOMAGE, from baseline to 9 months, serum PICP and PICP/CITP decreased from 79.0 to 75.4 µg/L and from 21.3 to 18.3, respectively (P≤0.0129). Correlations between changes from baseline to follow-up in urinary type-1 collagen fragments and CITP were positive often reaching significance if fragments increased during follow-up, but were nonsignificant if fragments decreased during follow-up. There were no between-group differences in these correlations. CONCLUSIONS: MR antagonism predominantly reduces collagen-derived urinary peptides. Inhibition of collagen synthesis by lowering the amount available for breakdown may be a contributing mechanism.