Mineralocorticoid Receptor Antagonism Prevents Aortic Plaque Progression and Reduces Left Ventricular Mass and Fibrosis in Patients With Type 2 Diabetes and Chronic Kidney Disease: The MAGMA Trial

Abstract

Background: Persistent mineralocorticoid receptor activation is a pathologic response in type 2 diabetes and chronic kidney disease. Whereas mineralocorticoid receptor antagonists are beneficial in reducing cardiovascular complications, direct mechanistic pathways for these effects in humans are lacking.

Methods: The MAGMA trial (Mineralocorticoid Receptor Antagonism Clinical Evaluation in Atherosclerosis) was a randomized, double-blind, placebo-controlled trial in patients with high-risk type 2 diabetes with chronic kidney disease (not receiving dialysis) on maximum tolerated renin-angiotensin system blockade. The primary end point was change in thoracic aortic wall volume, expressed as absolute or percent value (ΔTWV or ΔPWV), using 3T magnetic resonance imaging at 12 months. Secondary end points were changes in left ventricle (LV) mass; LV fibrosis, measured as a change in myocardial native T1; and 24-hour ambulatory and central aortic blood pressures. Tertiary end points included plasma proteomic changes in 7596 plasma proteins using an aptamer-based assay.

Results: A total of 79 patients were randomized to placebo (n=42) or 25 mg of spironolactone daily (n=37). After a modified intent-to-treat, including available baseline data of study end points, patients who completed the trial protocol were included in the final analyses. At the 12-month follow-up, the average change in PWV was 7.1±10.7% in the placebo group and 0.87±10.0% in the spironolactone group (P=0.028), and ΔTWV was 1.2±1.7 cm³ in the placebo group and 0.037±1.9 cm³ in the spironolactone group (P=0.022). Change in LV mass was 3.1±8.4 g in the placebo group and -5.8±8.4 g in the spironolactone group (P=0.001). Changes in LV T1 values were significantly different between the placebo and spironolactone groups (26.0±41.9 ms in the placebo group versus a decrease of -10.1±36.3 ms in the spironolactone group; P=6.33×10^-4). Mediation analysis revealed that the spironolactone effect on thoracic aortic wall volume and myocardial mass remained significant after adjustment for ambulatory and central blood pressures. Proteomic analysis revealed a dominant effect of spironolactone on pathways involving oxidative stress, inflammation, and leukocyte activation.

Conclusions: Among patients with diabetes with moderate to severe chronic kidney disease at elevated cardiovascular risk, treatment with spironolactone prevented progression of aortic wall volume and resulted in regression of LV mass and favorable alterations in native T1, suggesting amelioration of left-ventricular fibrosis.

Registration: URL: https://www.clinicaltrials.gov; Unique identifier: NCT02169089.

Keywords: atherosclerosis; diabetes mellitus, type 2; mineralocorticoid receptor antagonists; renal insufficiency, chronic.

FIGURE 1:

Consort diagram for MAGMA trial.

FIGURE 2.

Estimation of the primary outcome measure in the thoracic aorta wall volume as measured by PWV and TWV. Proximal landmark established distal to the left subclavian ostium; Distal landmark is 100 mm distal to proximal landmark.

FIGURE 3.

Comparisons of changes in primary endpoint in both randomized treatment groups (placebo and spironolactone). Change in percent thoracic aorta wall volume PWV. B. Change in delta thoracic aorta wall volume (ΔTWV) between treatment groups. Delta change was defined as the value at follow-up visit (#8, 12m) minus the value at baseline visit (#2, 0m). Group means ± SD are shown for each group, and median values are shown at horizontal line of whisker boxes. Progression status (0,1) is overlaid on the plots. Progressors (1) were defined as those with either PWV > 0 or ΔTWV > 0, while non-progressors (0) were defined as those with PWV ≤ 0 or ΔTWV ≤ 0. Significance codes: (.) p < 0.10, (*) p < 0.05, (**) p < 0.01, (***) p < 0.001, (****) p < 0.0001. C. Change in TWV between visits within treatment group. D. Change in TWV between treatment groups within visit. E. Interaction plot of TWV changes between visits within treatment group. Group ranges (color ribbon) and group mean ± SD (annotated dotted lines) are shown for each group. F. Lines plot of TWV changes between visits within treatment group by subject.

FIGURE 4.

Categorical data analysis results of spironolactone treatment on atheroma volume progression. Progressors (1) were defined as those with either PWV > 0 or ΔTWV > 0, while non-progressors (0) were defined as those with PWV ≤ 0 or ΔTWV ≤ 0. A. Description summary of the analysis. B. Alluvial Sankey plot of atheroma volume progression by treatment. Alluvia are annotated by the progression status variable PROG. C. Balloon plot of contingency table and chi-squared test of independence. Chi-squared test of drug treatment and atheroma volume progression status (PROG), considering the observed data, i.e. the first two follow-up visits (Visits #2, #8). D. Binomial GLMEM response model fitting. Results show treatment effect estimate on atheroma volume progression after adjusting for other main effects, their interaction, and baseline Demographics, Laboratory and Clinical covariates. CI: 95% Confidence Interval of the Odds Ratio (intercept estimates not shown for conciseness).

FIGURE 5.

Comparisons of changes in secondary endpoint left ventricular mass (LVM) in both randomized treatment groups (placebo and spironolactone). A. Change in delta left ventricular mass (ΔLVM) between treatment groups. Delta change was defined as the value at follow-up visit (#8, 12m) minus the value at baseline visit (#2, 0m). Group means ± SD are shown for each group, and median values are shown at horizontal line of whisker boxes. Progression status (0,1) is overlaid on the plots. Progressors (1) were defined as those with either PWV > 0 or ΔTWV > 0, while non-progressors (0) were defined as those with PWV ≤ 0 or ΔTWV ≤ 0. Significance codes: (.) p < 0.10, (*) p < 0.05, (**) p < 0.01, (***) p < 0.001, (****) p < 0.0001. B. Change in LVM between visits within treatment group. C. Change in LVM between treatment groups within visit. D. Interaction plot of LVM changes between visits within treatment group. Group ranges (color ribbon) and group mean ± SD (annotated dotted lines) are shown for each group. E. Lines plot of LVM changes between visits within treatment group by subject.

FIGURE 6.

Differential expression analysis results of proteomic markers between experimental conditions. A. Multi-dimensional scaling (MDS) scatterplot of patients. Lines highlighted by experimental groups: placebo (blue) vs. spironolactone (green), baseline (0m, light) vs. follow-up (3m, dark) visits. Axes captions denote the percentage explained variance (PEV) captured by each MDS component. B Volcano plot showing differentially expressed proteins (215), associated either with an increase (88, red) or a decrease (127, green) of the interaction effect (differential effect of spironolactone vs. placebo over time, i.e. in baseline vs. follow-up visits - see methods). The volcano plot is a scatterplot of statistical significance vs. magnitude of change, where every point represents a single protein. The y-axis represents the absolute value of the moderated test statistic (mod. t- test) and the x-axis represents the moderated log-fold change (mod. Log2(FC)) of differential expression. C-D. Interaction plots of top 10 differentially expressed proteins, associated either with an increase (C, top) or a decrease (D, bottom) of the interaction effect. Dotted lines represent the approximated 95% confidence interval bands of median expression values.