Detection of Patients at Risk of Developing Heart Failure Responsive Tomineralocorticoid Receptor Antagonists (MRAs)

New Insights and Opportunities

Bertram Pitt; James Brian Byrd


Eur Heart J. 2021;42(6):697-699. 

Mineralocorticoid receptor antagonists (MRAs), in particular spironolactone, have been used alone and in combination with a thiazide diuretic for the treatment of primary hypertension for >70 years.[1] Their use was originally based upon the effect of spironolactone on blocking the effects of aldosterone on the renal absorption of sodium and excretion of potassium. Subsequently, information regarding vascular inflammation and an increase in oxidative stress and vascular remodelling has reinforced the use of MRAs in patients with primary hypertension (PH). MRAs have also been used for the treatment of primary aldosteronism (PA) after the seminal work of Conn at the University of Michigan.[2] The incidence of PA has been estimated to be ~10% in patients with undifferentiated hypertension. Spironolactone is also the drug of choice in patients with resistant hypertension (RH),[3] >20% of whom have PA.[4,5] The use of MRAs, however, is relatively uncommon in patients with hypertension, even in those with RH.

The use of MRAs in patients with heart failure (HF) was driven by: (i) the knowledge that while angiotensin-converting enzyme inhibitors (ACEis) and angiotensin receptor blockers (ARBs) suppressed aldosterone in patients with PH over the short term, over the long term (~6 months or more) many patients had an increase in aldosterone levels back to or above baseline levels;[4] (ii) the finding that aldosterone induced myocardial fibrosis[5] that could be blocked by spironolactone;[6] and (iii) the finding that a dose of spironolactone of 12.5–50 mg/day was effective in reducing naturetic peptide levels in patients with HF with reduced ejection fraction (HFrEF) with only a moderate increase in serum potassium.[7] These factors resulted in the design of the RALES trial which showed that spironolactone was effective in reducing cardiovascular mortality and hospitalizations for HF in patients with NYHA class III–IV HF.[8] A substudy from RALES suggested that the major benefit of spironolactone in patients with HFrEF was in those patients with evidence of ongoing collagen formation.[9] The importance of myocardial fibrosis in the pathophysiology of HFrEF and atrial fibrillation as well as the effectiveness of MRAs in preventing myocardial fibrosis have since been confirmed in numerous pre-clinical and clinical studies. MRAs have also been shown to be effective in preventing vascular and renal fibrosis. The most effective way to detect patients who might best respond to an MRA remains, however, uncertain. Biomarkers of ongoing collagen formation and degradation such as procollagen type I and type III, markers of collagen cross-linking, galectin 3, echocardiographic strain rate imaging, cardiac magnetic resonance imaging (MRI) measurement of urinary or serum renin/aldosterone alone or with sodium loading, or administration of captopril have all been suggested to be of value.

Further insight into the selection of patients at risk for HF who might be responsive to an MRA comes from the report of the HOMAGE trial by Cleland et al.[10] in this issue of the European Heart Journal. The HOMAGE investigators set out to determine the effect of the MRA spironolactone 50 mg/day on markers of collagen formation and degradation as well as cardiac function over a 9-month period in patients at an increased risk of developing HF who also had an increase in plasma brain naturetic peptide (BNP). They excluded patients with evidence of HF or atrial fibrillation and those taking loop diuretics at baseline. The primary endpoint was the interaction between serum galectin 3 and the change in serum procollagen type-III N-terminal propeptide (PIIIP). They also evaluated other markers of collagen formation and degradation, including procollagen type-I C-terminal propeptide (PICP) and collagen type-1 C-terminal telopeptide (CITP). They found that galectin 3 and PIIIP did not help to identify patients who had a reduction in other biomarkers responsive to spironolactone. They did, however, find that patients randomized to spironolactone had a reduction in PICP and PICP/CITP, along with a reduction in systolic blood pressure, left atrial volume, and NT-proBNP levels. Of note was their finding that spironolactone reduced PICP and CITP at 1 month after randomization compared with the final visit, suggesting an early increase in the rate of collagen turnover that does not persist at that level.

Of importance for the future use of MRAs was the fact that patients entered into HOMAGE had a normal left venticular end-diastolic volume (LVEDV) but an increase in left atrial (LA) volume and LV mass at baseline, compatible with the history of hypertension in most of the patients entered into HOMAGE, all of which were reduced by spironolactone. The authors suggested that the favourable effects of spironolactone on cardiac structure and function could delay or prevent the clinical onset of HF. They also suggested that many of their patients probably had evidence of HF with preserved ejection fraction (HFpEF) and that those randomized to spironolactone achieved favourable effects on cardiac structure and function, but unfortunately the trial was not powered to demonstrate an improvement in clinical symptoms or exercise performance. The findings in HOMAGE are nevertheless of importance, especially in light of other recent findings. (i) The incidence of PA (or non-suppressible aldosterone production) in patients who ostensibly have PH—thought to be ~10%—is actually >20% using newer diagnostic criteria in patients with stage 2 hypertension, as well as in patients with RH.[11] Using these new diagnostic criteria, PA has also been detected in some patients without hypertension.[11] While spironolactone is the drug of choice for patients with RH, this new finding suggests that an MRA should be considered as first- or second-line therapy for a far larger number of patients currently diagnosed as having PH. (ii) A recent report from the Multi Ethnic study of Atherosclerosis (MESA) found that an increase in serum aldosterone was associated with coronary artery calcification and subclinical coronary artery disease (CAD) as well as an increase in all-cause mortality in those with a suppressed renin level, suggesting the presence of PA.[12] Thus, it is likely that many of the patients entered into HOMAGE, almost all of whom had either CAD or hypertension, had undiagnosed PA and therefore are candidates for treatment with an MRA. (iii) Recent MRI findings by Redheuil et al.[13] suggest that the pattern of myocardial fibrosis associated with PA is different from that in patients with PH. Patients with PA have an increase in extracellular myocardial fibrosis which was not seen in patients with PH.

Thus, we now know that far more patients with hypertension have non-suppressible aldosterone production or PA, and that the findings on cardiac MRI of extracellular myocardial fibrosis or, as suggested by HOMAGE, an increase in the PICP/CITP ratio could identify a subset of patients with hypertension, as well as some without hypertension, who are at risk of developing HF and who would be likely to respond to an MRA and thus prevent vascular stiffening, the progression from hypertension to HF as well as to chronic kidney disease, dementia, and Alzheimer's disease—all which have been associated with an increase in fibrosis and vascular stiffness (Figure 1). The quest for other indicators of autonomous aldosterone production, markers of ongoing collagen formation and degradation, and inappropriate activation of the mineralocorticoid receptor assumes greater importance with the knowledge that a relatively large percentage of patients with what we now classify as having PH actually have PA and thus should be receiving therapy to prevent the deleterious effects of aldosterone on the vasculature. One of the authors, J.B.B., is working to develop a novel biomarker of mineralocorticoid receptor activation using urinary extracellular vesicles,[14] and other approaches are also being explored.

Figure 1.

Evolving paradigms for the treatment of inappropriate or excess mineralocorticoid receptor activation with mineralocorticoid receptor antagonists.

In conclusion, while the major use of MRAs currently is in patients with HFrEF and to a lesser degree HFpEF, PA, and RH, it is reasonable to predict that on the basis of HOMAGE and the other recent findings outlined above, in the near future the major use of MRAs will be in patients with hypertension more generally, using new means to identify inappropriate mineralocorticoid receptor activation to prevent rather than to treat HF, RH, and chronic kidney disease.