Steroidal or non-steroidal MRAs: should we still enable RAASi use through K binders?

ABSTRACT

Renin–angiotensin–aldosterone system inhibitors (RAASi) and mineralocorticoid receptor antagonists (MRAs) are important interventions to improve outcomes in patients with chronic kidney disease and heart failure, but their use is limited in some patients by the development of hyperkalemia. The risk of hyperkalemia may differ between agents, with one trial showing lower risk of hyperkalemia with the novel non-steroidal MRA finerenone compared with steroidal MRA spironolactone. Novel potassium binders, including patiromer and sodium zirconium cyclosilicate, are available interventions to manage hyperkalemia and enable continuation of RAASi and MRAs in patients who could benefit from these treatments. These agents bind free potassium ions in the lumen of the gastrointestinal tract to prevent the absorption of dietary potassium and increase potassium secretion. Several studies showed that potassium binders are effective compared with placebo for preventing hyperkalemia or steroidal MRA discontinuation, but none has evaluated whether this strategy impacts clinically important endpoints such as cardiovascular events. Due to this and other limitations related to cost, clinical availability, pill burden and patient selection, alternative potential strategies to mitigate hyperkalemia may be more practical. Conservative strategies include increased monitoring and use of loop or thiazide diuretics to increase urinary potassium excretion. Non-steroidal MRAs may have a lower risk of hyperkalemia than steroidal MRAs and have stronger anti-inflammatory and anti-fibrotic effects with resultant reduced risk of kidney disease progression. Sodium-glucose cotransporter-2 inhibitors also decrease hyperkalemia risk in patients on MRAs and decrease cardiovascular events and kidney disease progression. These may be better first-line interventions to obviate the need for potassium binders and offer additional benefits.

INTRODUCTION

Renin–angiotensin–aldosterone system inhibition (RAASi) with an angiotensin-converting enzyme inhibitor (ACEi) or angiotensin receptor blocker (ARB) is considered standard of care to reduce proteinuria in patients with chronic kidney disease (CKD) or provide afterload reduction in patients with heart failure. However, in some patients, despite treatment with an ACEi or ARB, circulating aldosterone (due to aldosterone escape phenomenon) can still contribute to profibrotic and proinflammatory gene expression, causing worsening fibrosis of the kidneys and heart, and contributing to unfavorable patient outcomes [1]. Consequently, addition of a mineralocorticoid receptor antagonist (MRA) is often indicated to prevent kidney and heart fibrosis, for example as is included in guideline-directed medical therapy (GDMT) for patients with heart failure with reduced ejection fraction (HFrEF) [2].

One of the primary adverse effects of MRAs is hyperkalemia, which can limit their use in the clinical setting, particularly for patients who have decreased urinary potassium excretion due to CKD or concomitant RAASi prescription [3]. Introduction of modern potassium binders that prevent absorption of potassium from the gastrointestinal tract have been studied as a potential strategy to prevent hyperkalemia and enable prescription of RAASi and MRAs at optimal doses. This review describes the mechanism by which MRAs lead to hyperkalemia, summarizes existing data regarding the use of potassium binders to reduce the risk of hyperkalemia in patients on MRAs or RAASi or enable optimization of these treatments, and discusses alternative mitigation strategies for hyperkalemia.

EFFECT OF MRAs ON POTASSIUM

Mechanism of hyperkalemia with MRAs

Elevated aldosterone in the presence of high salt intake leads to urinary potassium wasting, and conversely, hyperkalemia is one of the most common adverse effects of blocking the action of aldosterone with MRAs. In the absence of MRAs, aldosterone binds to cytosolic mineralocorticoid receptors (MR), which then bind coactivators and move to the nucleus, where these complexes impact gene transcription. The coactivators are thought to be associated with transcription of pathways involved in the profibrotic and proinflammatory downstream effects of aldosterone [4]. One of the effects of aldosterone is to prevent ubiquitination of the epithelial sodium channel (ENaC), thus preventing its endocytosis from the apical membrane of principal cells in the collecting duct [1]. Maintaining ENaC in the apical membrane contributes to increased sodium reabsorption in the collecting duct (Fig. 1A). This increased sodium reabsorption generates a luminal electronegativity as well as increased intracellular potassium concentration through augmented activity of the basolateral sodium-potassium adenosine triphosphatase. These factors drive potassium secretion through the renal outer medullary potassium channel (ROMK) and large conductance calcium-activated potassium channels (BK) [5]. Increased ROMK activity may also contribute to greater potassium secretion in the presence of aldosterone [6, 7].

Figure 1:

Effects of aldosterone and MRAs on urinary potassium excretion in the distal nephron. In the absence of MRAs, aldosterone diffuses across the cell membrane of principal cells in the collecting duct of the nephron, where it binds to cytosolic MR. These complexes bind to coactivators and move to the nucleus, where they activate transcription of profibrotic and proinflammatory pathways, as well as inhibit the ubiquitination and endocytosis of ENaC from the apical membrane. This stimulates sodium reabsorption, which generates luminal electronegativity and elevated intracellular potassium concentration via the basolateral sodium-potassium adenosine triphosphatase. These factors drive potassium excretion through ROMK and BK (A). Steroidal MRAs bind to the ligand-binding site of the MR in such a way that allows for the binding of coactivators, which are involved in the increased transcription of proinflammatory and profibrotic pathways. Unlike aldosterone, steroidal MRAs do not prevent the endocytosis of ENaC from the apical membrane, thus diminishing the sodium reabsorption that is the driving force for potassium secretion. Other factors, such as changes in membrane trafficking of ROMK, may also contribute to diminished potassium excretion but are incompletely understood (B). Non-steroidal MRAs bind to the MR in a way that prevents the binding of coactivators and leads to a unique pattern of target gene expression, with decreased transcription of proinflammatory and profibrotic genes compared to steroidal MRAs (C).

Steroidal MRAs (spironolactone and eplerenone) were derived from the structure of progesterone. They are small, relatively flat molecules that bind to the ligand-binding site of the MR in such a way that coactivators can still bind (Fig. 1B) [8]. These complexes do not preclude the endocytosis of ENaC. However, there can still be some downstream proinflammatory and profibrotic pathway activation due to partial agonist activity of these medications for coactivator binding. Decreased sodium reabsorption through ENaC diminishes the driving forces for potassium secretion. Other factors likely also influence urinary potassium excretion; for example, it has been hypothesized that ROMK membrane trafficking, possibly affected by altered activity of With No Lysine (WNK) kinases, may also contribute to the development of hyperkalemia in the presence of steroidal MRAs [9].

Novel non-steroidal MRAs include finerenone, esaxerenone and aparerenone. These agents are known as ‘bulky’ antagonists, which bind to the ligand-binding site of the MR in such a way that displaces one helix of the receptor, preventing the binding of coactivators to the site (Fig. 1C) [8]. Antagonizing coactivator binding accounts for the stronger anti-inflammatory and anti-fibrotic effects of non-steroidal MRAs compared with steroidal MRAs. It is thought that the coactivator binding profiles and target gene regulation differ between individual agents, which may contribute to differences in the clinical effects seen with these medications [10].

The exact downstream effects of steroidal versus non-steroidal MRAs on potassium excretion remain to be fully elucidated. It has been postulated that non-steroidal MRAs may affect potassium handling differently from steroidal MRAs due to unique transcription targets, shorter half-life, lack of active metabolites, and distribution in both the heart and kidneys, compared with steroidal MRAs which accumulate predominantly in the kidneys [4].

Hyperkalemia risk between MRAs

MRAs are known to cause hyperkalemia compared with placebo, with one systematic review and meta-analysis showing a risk ratio (RR) of 2.17 [95% confidence interval (CI) 1.47, 3.22] for developing hyperkalemia with an MRA compared with placebo or standard of care among 3001 participants in 17 trials [3]. Factors associated with a higher risk of hyperkalemia in patients treated with steroidal or non-steroidal MRAs include higher baseline serum potassium level [11, 12], lower baseline estimated glomerular filtration rate (eGFR) [11, 12], more severe albuminuria [12], lower hemoglobin [11], type 2 diabetes mellitus (T2DM) [11] and concomitant RAASi prescription [11]. Prescription of diuretics and sodium-glucose cotransporter-2 (SGLT2) inhibitors were associated with decreased risk of hyperkalemia [12].

The risk of hyperkalemia may differ between individual MRAs. In one study, eplerenone led to lesser increases in serum potassium level from baseline compared with spironolactone [13]. In a few other small trials serum potassium level was found to be similar between groups treated with spironolactone or eplerenone, but these studies enrolled patients who were not at high risk of hyperkalemia and may have been underpowered to detect a difference (Table 1) [14–17].

Table 1:

Effect on potassium in head-to-head randomized trials MRAs.

Study	Participants	Drug	Comparator	Reported potassium-related outcome	Result (proportions expressed as drug vs comparator)
Spironolactone vs eplerenone
Weinberger, et al. 2002 [13]	N = 409 with mild-to-moderate hypertension	Eplerenone at various doses	Spironolactone	Mean change in serum potassium level at 8 weeks	P < .05 for eplerenone 50 mg per day or 100 mg per day in 1 or 2 divided doses vs spironolactone 50 mg twice daily; no difference for eplerenone 400 mg per day vs spironolactone
Yamaji et al. 2010 [14]	N = 107 with stable mild heart failure	Spironolactone	Eplerenone	Serum potassium level at 4 months	Mean (SD) 4.44 (0.1) vs 4.48 (0.04), P = .47
Karashima et al. 2016 [15]	N = 54 with primary hyperaldosteronism	Spironolactone	Eplerenone	Serum potassium level at the end of the study	Mean (SD) 4.3 (0.3) vs 4.2 (0.3) mmol/L, P > .05
Sehgal et al. 2020 [16]	N = 105 with decompensated cirrhosis	Spironolactone	Eplerenone	Difference in potassium levels at 1, 2 and 3 months	No differences
Nabati et al. 2021 [17]	N = 85 with new onset dilated cardiomyopathy	Spironolactone	Eplerenone	Drug discontinuation due to increased serum potassium level	None in either group
Non-steroidal vs steroidal MRAs
ARTS, 2013 [19]	N = 458 with HFrEF and eGFR 30 to <90	Finerenonea	Spironolactone	Between groups effect on K	At visit 6/7 P < .05 for mean change from baseline in K between finerenone at any dose and spironolactone groups
				Investigator-reported hyperkalemia events	4.5% vs 11.1%
ARTS-HF, 2016 [20]	N = 1066 with HFrEF with T2DM and/or CKD stage 3 on RAASi	Finerenonea	Eplerenone	K >5.5 mmol/L	4.2% vs 4.7%
				K >6.0 mmol/L	0.5% vs 0.5%
ARTS-HF Japan, 2016 [21]	N = 72 with HFrEF with CKD and/or T2DM	Finerenonea	Eplerenone	Change in serum K from baseline	No differences
				Any hyperkalemia event	2 events (3.4%) vs 0
ESAX-HTN, 2020 [22]	N = 1001 with hypertension	Esaxerenonea	Eplerenone	‘Blood K increased’	2.2% vs 0.9%
				K ≥5.5 mmol/L twice consecutively	0.4% vs 0%
				K ≥ 6.0 mmol/L	0.3% vs 0%

^aMultiple doses of non-steroidal MRA were tested. The data noted here represent the proportion with the indicated outcome in the pooled non-steroidal MRA groups.

eGFR is in mL/min/1.73 m²; SD, standard deviation; T2DM, type 2 diabetes mellitus.

Non-steroidal MRAs may have lower risk of hyperkalemia compared with steroidal MRAs. In a rat model of CKD, one study compared the effect of the non-steroidal MRA PF-03 882 845 versus eplerenone on the ratio of the half maximal effective concentration required to raise serum potassium over that required to lower albuminuria [18]. The non-steroidal MRA demonstrated a 57-fold more favorable ratio against hyperkalemia compared with eplerenone. In humans, a few head-to-head trials have compared the risk of hyperkalemia between non-steroidal and steroidal MRAs, with mixed results (Table 1). The minerAlocorticoid Receptor Antagonist Tolerability Study (ARTS) phase II trial randomized 392 individuals with HFrEF and eGFR of 30 to <60 mL/min/1.73 m² to receive either finerenone at various doses, placebo or open-label spironolactone for 4 weeks [19]. They showed that at visit 6/7 (22 or 29 days after randomization), the mean change from baseline in serum potassium level was significantly lower for finerenone at any dose compared with spironolactone. They further showed that investigator-reported hyperkalemia events occurred in 4.5% of those randomized to finerenone at any dose, compared with 11.1% of those randomized to spironolactone. ARTS-Heart Failure (ARTS-HF) and ARTS-HF Japan randomized patients with HFrEF with T2DM and/or CKD stage 3 on RAASi to receive either finerenone or eplerenone [20, 21]. They showed that hyperkalemia events were similar between the two drugs. Most recently, A Double Blind Study of CS-3150 to Evaluate Efficacy and Safety Compared to Eplerenone in Patients with Essential Hypertension (ESAX-HTN) randomized 1001 patients with hypertension to receive either esaxerenone or eplerenone and found that hyperkalemia ≥5.5 mmol/L on two consecutive visits occurred in 0.4% of the esaxerenone group compared with 0% in the eplerenone group, and any value ≥6.0 mmol/L in 0.3% versus 0%, respectively [22]. Despite this, they reported ‘blood K increased’ in 2.2% versus 0.9%, suggesting that these events were mostly either isolated serum potassium values between 5.5 and 6.0 mmol/L or values <5.5 mmol/L which are generally not considered clinically significant and are frequently amenable to conservative management. Thus, hyperkalemia risk in this study was similar between esaxerenone and eplerenone. A systematic review and meta-analysis showed that the risk of hyperkalemia was similar and between finerenone and eplerenone, RR 0.89 (95% CI 0.45, 1.77) among 1023 participants from one study (ARTS-HF), but the certainty of the evidence was rated as low [3, 20]. In summary, some studies suggested a lower risk of hyperkalemia with non-steroidal compared with steroidal MRAs, but estimates are imprecise due to low numbers of outcome events in some trials.

POTASSIUM BINDERS

Mechanism of action of potassium binders

Patiromer and sodium zirconium cyclosilicate (SZC) are potassium binders that have been studied for the effect of their chronic use on hyperkalemia in patients on RAASi or MRAs. Potassium binders entrap potassium ions in the gastrointestinal tract, thus increasing total body potassium excretion in the stool. In the absence of potassium binders, free potassium ions are passively absorbed in the small intestine by solvent drag, and the colon net secretes potassium ions through both active and passive mechanisms [23]. Potassium binders achieve increased gastrointestinal potassium excretion by two primary mechanisms: (i) preventing the absorption of dietary potassium and (ii) decreasing the concentration of free potassium ions in the bowel lumen, thus augmenting the driving forces for potassium secretion [23].

Patiromer is a nonabsorbable cation exchange polymer that contains sorbitol as a purgative and calcium ions that exchange for potassium ions. It primarily acts in the colon, with a time to onset of approximately 7 h (Table 2) [24]. It is associated with gastrointestinal side effects and can also cause hypomagnesemia by non-selectively binding magnesium in the gastrointestinal tract.

Table 2:

Pharmacokinetic and pharmacodynamic properties of novel potassium binders.

Characteristic	Patiromer	Sodium zirconium cyclosilicate (SZC)
Starting dose (adults)	8.4 g once daily	10 g three times per day for up to 48 h, followed by 10 g once daily
Maximum dose (adults)	25.2 g per day	15 g daily
Mechanism of action	Nonabsorbable cation exchange polymer that exchanges calcium for potassium ions and contains sorbitol as a laxative, increasing stool potassium excretion	Nonabsorbable non-polymer that selectively entraps potassium and ammonium ions in exchange for sodium and hydrogen ions, increasing stool potassium excretion
Primary site of action	Colon	Throughout the gastrointestinal tract, as early as the duodenum
Time to onset	Initial response: 7 h, peak response: 48 h	Initial response: 1 h
Duration	24 h	48 h in patients treated with SZC 10 g three times daily for up to 48 h
Adverse effects and tolerability	Gastrointestinal discomfort, constipation, diarrhea, flatulence, nausea	Peripheral edema, vomiting, constipation
Effect on magnesium	Can cause hypomagnesemia in 5.3%–9%	Negligible
Additional effects		Increased serum bicarbonate level (within the normal range) due to SZC binding ammonium

SZC is a selective cation exchanger. It contains negatively charged pores that selectively entrap potassium and ammonium ions in exchange for sodium and hydrogen ions [25]. It acts throughout the gastrointestinal tract as early as the duodenum, with a time to onset as low as 1 h (Table 2) [26–28]. The primary adverse effect associated with SZC is peripheral edema. It is highly selective for potassium, so unlike patiromer it has negligible effects on magnesium excretion. It can cause gastrointestinal side effects, but less so than patiromer based on meta-analysis of studies conducted in differing patient populations [29]. There have been no head-to-head trials comparing adverse effects between patiromer and SZC.

Efficacy trials of potassium binders in patients on RAASi or MRAs

Three placebo-controlled clinical trials have evaluated whether oral potassium binders can be used to effectively enable uptitration or maintenance of optimally dosed RAASi or MRAs without developing severe hyperkalemia (Table 3). In 2011, the Evaluation of Patiromer in Heart Failure Patients (PEARL-HF) trial enrolled patients with heart failure and hyperkalemia that led to discontinuation of RAASi, or with eGFR <60 mL/min/1.73 m² who would be at risk for hyperkalemia limiting RAASi therapy [30]. After randomization to receive patiromer or placebo for 4 weeks, the change in potassium from baseline to study end was lower in the patiromer group, fewer participants randomized to patiromer experienced a potassium level >5.5 mmol/L, and participants in the patiromer group were more likely to receive spironolactone at a dose of 50 mg/day at study end. The Study Evaluating the Efficacy and Safety of Patiromer for the Treatment of Hyperkalemia (OPAL-HK) enrolled patients with CKD stages 3–4 on RAASi with hyperkalemia in an initial treatment phase with patiromer for 4 weeks, and then randomized those who became normokalemic to either continue patiromer or change to placebo for 8 weeks [31]. Compared with placebo recipients, those who continued patiromer had less rise in potassium level, fewer episodes of hyperkalemia and a higher proportion remaining on RAASi at the end of the study. The Patiromer versus placebo to enable spironolactone use in patients with resistant hypertension and chronic kidney disease (AMBER) trial randomized individuals with eGFR 25 to ≤45 mL/min/1.73 m² and resistant hypertension to receive spironolactone with either patiromer or placebo for 12 weeks [32]. Participants receiving patiromer were more likely to remain on spironolactone at the end of the study. The Patiromer in the Treatment of Hyperkalemia in Patients with Hypertension and Diabetic Nephropathy (AMETHYST-DN) open label phase II trial showed that patiromer at various doses led to decreases from baseline in serum potassium level among individuals with hyperkalemia on RAASi [33]. Another open-label study of patiromer and spironolactone without a control group among patients with CKD, heart failure and a potassium level of 4.3–5.1 mmol/L showed that a normal serum potassium level was achieved in 90.5% of participants [34]. There are other trials, both completed and ongoing, that evaluate the impact of potassium binders on hyperkalemia, acidosis and other important outcomes but are outside the scope of this review. The results of these trials suggest that potassium binders enable the prescription of RAASi and steroidal MRAs in patients at risk for developing hyperkalemia.

Table 3:

Efficacy outcomes of clinical trials of chronic use of potassium binders for preventing hyperkalemia in patients on RAASi or MRAs.

Published reference	Participants	Intervention, control, duration	Outcome	Results
PEARL-HF, 2011 [30]	N = 105 with history of heart failure and hyperkalemia resulting in discontinuation of RAASi or eGFR <60	Patiromer, placebo, 4 weeks	Change in K from baseline to week 4	K lower in patiromer group, –0.45 mmol/L, P < .001
			Any K >5.5	24.5% in placebo group vs 7.3% in patiromer group, P = .015
			Proportion titrated to spironolactone 50 mg/day	74% in placebo group vs 91% in patiromer group, P = .019
AMETHYST-DN, 2015 [33]	N = 306 with T2DM, eGFR 15 to <60 and serum K >5.0 on RAASi	Patiromer at various doses stratified by baseline K >5.0 to 5.5 or >5.5 to <6.0, 52 weeks	Mean change in serum K from baseline to week 4 or prior to patiromer dose titration	P < .001 for all doses of patiromer in both strata by baseline K level
			Mean change in serum K from baseline to Week 8 or prior to patiromer dose titration	P < .001 for all doses of patiromer in both strata by baseline K level
			Mean change in serum K from baseline to 52 weeks	P < .001 in both strata by baseline K level
OPAL-HK, 2015 [31]	N = 237 with CKD stage 3–4 and serum K 5.1 to <6.5 on RAASi	Initial treatment phase: patiromer, no control, 4 weeks	Change in K from baseline to Week 4 of initial treatment phase	K decreased from baseline, mean (95% CI) –1.01 (–1.07, –0.95) mmol/L, P < .001
	N = 107 with CKD stage 3–4 on RAASi, baseline K >5.5, normal K on patiromer during initial treatment phase	Randomized withdrawal phase: patiromer, placebo, 8 weeks	Between groups difference in change in K over first 4 weeks of randomized withdrawal phase	K increased more in placebo than patiromer group, median (95% CI) 0.72 (0.46, 0.99) mmol/L, P < .001
			Any K ≥5.5 over 8 weeks	60% (95% CI 47, 74) in placebo group vs 15% (95% CI 6, 24) in patiromer group, P < .001
			Proportion remaining on RAASi	44% in placebo group vs 94% in patiromer group
Patiromer-204, 2018 [34]	N = 63 with CKD, heart failure, and K 4.3–5.1	Open-label patiromer and spironolactone, no control, 8 weeks	K 3.5–5.5 at 8 weeks	Achieved by 90.5% of participants
AMBER, 2019 [32]	N = 295 with eGFR 25 to ≤45 and uncontrolled resistant hypertension	Patiromer, placebo, all received spironolactone, 12 weeks	Between-group difference in proportion on spironolactone at 12 weeks	66% in placebo group vs 86% in the patiromer group, between-group difference 19.5% (95% CI 10.0, 29.0), P < .0001
DIAMOND (ClinicalTrials.gov ID NCT03888066) [36]	N = 878 with HFrEF started or continued on MRA, RAASi and patiromer during run-in phase	Continued patiromer, placebo	Between groups difference in change in K from baseline; secondary endpoint: all-cause death, CV hospitalization or use of comprehensive heart failure medication	TBD; outcomes were refocused on change in potassium level from initially planned CV outcomes due to lower than expected recruitment and safety concerns due to the COVID-19 pandemic
LIFT (ClinicalTrials.gov ID NCT05004363) [38]	N = 130 with HFrEF and CKD stage 3–5 and K 5.0–5.5	SZC, placebo, maximizing RAASi in all participants, 16 weeks	Number of participants achieving maximum dose of RAASi + MRA	TBD
MorphCKD (EnduraCT ID: 2020-001595-15) [37]	N = 140 with eGFR 25–60, UACR >500 mg/g if no T2DM or >200 mg/g if T2DM, and K >4.5 who develop K >5.5 with RAASi intensification/MRA initiation in run-in phase	Patiromer, standard care, 12 months	Between-groups difference in change in UACR from randomization to study completion	TBD
OPRA-HF (ClinicalTrials.gov ID NCT04789239)	N = 230 with HFrEF	SZC, placebo, 6 months	Optimization of MRA usage	TBD
REALIZE-K (ClinicalTrials.gov ID NCT04676646)	N = 265 on RAASi but not MRA or on low dose MRA and K 5.1–5.9 or risk of hyperkalemia	SZC, placebo, all on spironolactone, 8 months	Occurrence of K 3.5–5.0, are on spironolactone ≥25 mg daily, and did not use rescue therapy for hyperkalemia	TBD
STABILIZE-CKD (ClinicalTrials.gov ID NCT05056727)	N = 1500 with eGFR 25–59 with hyperkalemia risk precluding RAASi initiation or uptitration	SZC, placebo, all on lisinopril or valsartan 27 months	Total eGFR slope and chronic eGFR slope	TBD

This table includes studies that expressly enrolled participants receiving RAASi or MRA or unable to receive them due to hyperkalemia, had an intervention that involved prescription or uptitration of RAASi or MRA in addition to randomization to potassium binder/control, or had an outcome related to optimizing RAASi or MRA use. This is not a comprehensive list of all trials evaluating the impact of potassium binders on hyperkalemia.

Potassium values are expressed in mmol/L and eGFR values are expressed in mL/min/1.73 m² unless otherwise specified.

COVID-19, coronavirus disease 2019; CV, cardiovascular; K, potassium; SD, standard deviation; T2DM, type 2 diabetes mellitus; UACR, urine albumin to creatinine ratio; TBD, to be determined.

Knowledge gaps about potassium binders to enable RAASi or MRA use

The primary limitation of the existing literature is that no current trials evaluated the impact of potassium binders on long-term outcomes such as death, cardiovascular events or CKD progression. We are left to extrapolate from observational data that RAASi and MRA treatment discontinuation is associated with poorer outcomes [35], and therefore that strategies to optimize RAASi and MRA use by minimizing hyperkalemia are likely warranted.

The Patiromer for the Management of Hyperkalemia in Subjects Receiving RAASi Medications for the Treatment of Heart Failure (DIAMOND) trial was designed to fill this important knowledge gap. This randomized withdrawal trail enrolled 878 participants with HFrEF who were either started or continued on MRA, RAASi and patiromer during a run-in phase, and then randomized to either continue patiromer or change to placebo [36]. The planned primary outcome was time to cardiovascular death or first cardiovascular hospitalization. However, the outcomes were changed while the study was still blinded due to the impact of the COVID-19 pandemic on enrollment, safety concerns and event rates. The primary endpoint is now the change in serum potassium level from baseline, similar to the prior published studies detailed above. Of other ongoing trials, only A Study to Evaluate the Effect of SZC on CKD Progression in Participants with CKD and Hyperkalaemia or at Risk of Hyperkalaemia (STABILIZE-CKD) seeks to evaluate the effect of SZC on a clinically relevant endpoint of the change in eGFR among patients with CKD whose risk of hyperkalemia precluded RAASi use or uptitration (Table 3). Another will evaluate the surrogate outcome of change in albuminuria measured by spot urine albumin-to-creatinine ratio from randomization to study completion between those receiving patiromer or standard of care for 12 months [37]. The remaining ongoing studies will follow previously published trials and evaluate either optimization of RAASi or MRA use or hyperkalemia as the primary outcome (Table 3) [38].

Another limitation of the existing literature is that no currently published trials have evaluated the effect of SZC on hyperkalemia in patients on MRAs or evaluating an outcome of MRA maintenance. Several ongoing trials are currently enrolling participants to fill this knowledge gap (Table 3).

Published trials also offer sub-optimal guidance on which patients may most benefit from the use of a potassium binder. These treatments are expensive, not accessible to all and contribute to the already high pill burden in a patient population with multiple comorbidities, so indiscriminate prescription would be unwise. Most trials enrolled patients at high risk for hyperkalemia who were likely to benefit from treatment. For example, the OPAL-HK trial randomized only individuals who responded to patiromer during the run-in phase. While such a design is sensible for conducting an efficient clinical trial, this leaves practical questions about which patients may be the best candidates to receive therapy and what criteria would constitute treatment failure in clinical practice to merit discontinuation. Furthermore, most trials were of relatively short duration and did not evaluate the durability of efficacy over longer term use, which would be necessary for many patients to enable chronic MRA prescription.

ALTERNATIVE MITIGATION STRATEGIES FOR HYPERKALEMIA

Although potassium binders are likely effective for preventing hyperkalemia in patients on RAASi or steroidal MRAs, limitations of their use include cost, clinical availability, pill burden, unclear patient selection criteria and absent efficacy data for longer term use. Alternative strategies to mitigate hyperkalemia are thus important to achieve optimal patient outcomes.

Conservative strategies

Clinical trials of MRAs have used various strategies to address hyperkalemia. Generally, these included cautious, staged uptitration of MRAs, frequent monitoring of serum potassium levels and protocolized methods of reintroducing MRAs in patients with resolved hyperkalemia [11, 12, 39]. They applied protocolized study drug discontinuation or downtitration in response to hyperkalemia and close monitoring in the setting of resumption of study drug. For example, in the FInerenone in reducing kiDnEy faiLure and dIsease prOgression in Diabetic Kidney Disease (FIDELIO-DKD) and Finerenone in Reducing Cardiovascular Mortality and Morbidity in Diabetic Kidney Disease (FIGARO-DKD) trials, investigators suspended the study drug if potassium was >5.5 mmol/L [12, 39]. Laboratory measures were rechecked every 72 h until potassium was ≤5.0 mmol/L, at which point drug was resumed. Study drug was permanently discontinued if a participant developed recurrent hyperkalemia attributed to the intervention. The result of such a strategy is to avoid permanent discontinuation of drug in patients for whom hyperkalemia was transient. It should be noted, however, that participants in these trials had a baseline serum potassium level <4.8 mmol/L, so these strategies may not be realistic in the general CKD population. While conservative measures such as increased monitoring to guide medication resumption are appealing from a cost perspective, the pace of clinical medicine precludes this frequency of visits and blood draws when applied to the scale of real-world practice. Nonetheless, clear guidance on criteria for reinitiation of MRAs may be an important strategy to decrease their permanent discontinuation in clinical practice [40].

Recommendation of adherence to a low potassium diet is becoming increasingly controversial. Given that the diet is a major source of potassium intake, restricting dietary potassium has been frequently recommended to manage hyperkalemia in patients with impaired potassium excretion, such as those on RAASi. However, potassium-rich foods such as many fruits and vegetables have important health benefits, including better blood pressure control and improved cardiovascular risk [41]. They also are frequently high in fiber and other nutrients, and a low potassium diet may be associated with diminished quality of life [41, 42]. It has been proposed that potassium binders may be useful for enabling a liberalized diet, as major clinical trials of these medications did not restrict dietary potassium intake [25]. Consequently, the general health benefits of a diet rich in fruits and vegetables may outweigh the risk of mild hyperkalemia.

One of the most frequently employed strategies for addressing hyperkalemia to enable RAASi or MRA prescription is to augment kaliuresis with loop or thiazide diuretics, either alone or in combination. Although no randomized trials have assessed the efficacy of this strategy, in FIDELIO-DKD baseline prescription of a loop or thiazide diuretic was associated with lower risk of hyperkalemia [hazard ratio (HR) 0.76 (95% CI 0.66, 0.87), P < .0001] [12]. Given that many patients with CKD or heart failure require these agents for management of elevated extracellular volume or blood pressure, this remains an inexpensive, accessible and often otherwise clinically indicated intervention to prevent hyperkalemia in the context of RAASi or MRA use.

Use of non-steroidal MRAs

Alternative treatment strategies with lower hyperkalemia risk offer additional long-term benefits for patients over enabling steroidal MRA use with a potassium binder. Use of non-steroidal MRAs with potentially lower risk of hyperkalemia than steroidal MRAs may obviate the need for potassium binders for some patients. In one head-to-head clinical trial, finerenone at various doses caused less hyperkalemia than spironolactone at 25 or 50 mg daily [19], indicating that hyperkalemia may be a lesser problem with newer agents that also have stronger anti-inflammatory and anti-fibrotic effects. However, in this trial the maximum studied dose of finerenone was 10 mg daily, which is less than the optimal dose, leaving questions about whether higher doses finerenone may still be associated with lower risk of hyperkalemia than spironolactone.

Concomitant SGLT2 inhibitors and MRAs

SGLT2 inhibitors are included in GDMT for HFrEF and are often indicated for patients with CKD, so they are likely to be concomitantly prescribed in patients receiving RAASi or MRAs. These agents have profound long-term cardiovascular and kidney protective effects [43–47], and several studies also showed that they decrease the risk of hyperkalemia in patients taking MRAs. In FIDELIO-DKD, baseline use of an SGLT2 inhibitor was associated with a lower risk of developing any serum potassium level >5.5 mmol/L [HR 0.45 (95% CI 0.27, 0.75), P = .002] [12]. A secondary analysis of the Dapagliflozin and Prevention of Adverse Outcomes in Heart Failure (DAPA-HF) trial stratified by baseline steroidal MRA prescription showed that dapagliflozin halved the risk of hyperkalemia among those taking MRAs [HR 0.50 (95% CI 0.29, 0.85), P = .01] [48]. In the Empagliflozin Outcome Trial in Patients with Chronic Heart Failure (EMPEROR)-Reduced and EMPEROR-Preserved trials, empagliflozin reduced the risk of hyperkalemia or initiation of potassium binders [6.5% of the empagliflozin group compared with 7.7% of the placebo group; HR 0.82 (95% CI 0.71, 0.95), P = .01] [49]. This benefit was also seen in the subgroup with baseline use of an MRA [HR 0.77 (95% CI 0.63, 0.93)]. The empagliflozin group also had a lower risk of potassium >5.5 mmol/L [HR 0.85 (95% CI 0.74, 0.97)] or >6.0 mmol/L [HR 0.62 (95% CI 0.48, 0.81)]. In EMPEROR-Reduced, SGLT2 inhibitor use was associated with less discontinuation of MRA [HR 0.78 (95% CI 0.64, 0.96)] [50]. Whether this represents a consequence of preserved kidney function or an off-target effect of SGLT2 inhibitors remains to be determined.

The ongoing COmbinatioN effect of Finerenone anD EmpaglifloziN in participants with chronic kidney disease and type 2 diabetes using an UACR Endpoint study (CONFIDENCE) will directly assess whether SGLT2 inhibition decreases the risk of hyperkalemia with finerenone by randomizing 807 participants to receive either finerenone + placebo, empagliflozin + placebo or finerenone + empagliflozin [51]. They will compare dual therapy to finerenone alone and empagliflozin alone for a primary outcome of change in urine albumin-to-creatinine ratio, and will also evaluate hyperkalemia events and change from baseline in serum potassium level as safety endpoints.

CONCLUSIONS

In summary, given the known benefits of RAASi and MRAs, enabling their prescription by preventing hyperkalemia would improve outcomes for patients with CKD or heart failure. It further stands to reason that increasing gastrointestinal potassium excretion with a potassium binder should benefit patients who develop therapy-limiting hyperkalemia with a steroidal MRA. Indeed, clinical trials have shown that more participants taking a potassium binder have remained on RAASi or MRA at trial completion. However, alternative strategies to mitigate hyperkalemia may be more practical to implement prior to prescribing potassium binders. In cases where cost and clinical availability are primary barriers to prescription, augmented kaliuresis with loop or thiazide diuretics or guidance on when to reinitiate MRAs held in response to hyperkalemia may maximize a patient's ability to continue these treatments. Newer agents may have decreased risk of hyperkalemia in addition to offering important clinical benefits: non-steroidal MRAs have anti-inflammatory and anti-fibrotic effects, and SGLT2 inhibitors decrease the risk of cardiovascular events and CKD progression. Such strategies may diminish the clinical need to enable the use of steroidal MRAs by prescribing potassium binders, but these remain an important therapeutic option for patients with treatment-limiting hyperkalemia.

FUNDING

L.P.G. is supported by VA CSR&D Career Development Award (IK2CX002368) and the Houston VA Health Services Research & Development Center for Innovations grant (CIN13-413). S.D.N. is supported by research funding from the Department of Veterans Affairs Health Services Research & Development (1I01HX002917) and K24 HL161414- 01A1. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH or Veterans Administration.

DATA AVAILABILITY STATEMENT

Not applicable.

CONFLICT OF INTEREST STATEMENT

S.D.N. reported receiving personal fees from AstraZeneca (Data safety monitoring board), ACI clinical, Bayer, Boehringer Ingelheim and Eli Lilly and Co., Vertex and Vifor; receiving grants from Keryx; and receiving research funding from the Department of Veterans Affairs Health Services Research & Development outside the submitted work.