Experience of a Bespoke Hyperkalaemia Clinic to Facilitate Prescribing of Renin-Angiotensin-Aldosterone System Inhibitors in Patients with Heart Failure with Reduced Ejection Fraction

Renin-angiotensin-aldosterone system inhibitors (RAASi) represent a key group of pharmacological therapies for patients with heart failure with reduced ejection fraction (HFrEF) [1] given that angiotensin-converting enzyme inhibitors (ACEi) [2], angiotensin receptor blockers (ARB) [3], mineralocorticoid antagonists (MRA) [4], and angiotensin receptor-neprilysin inhibitors [5] improve patient survival. Achieving maximal dosing of RAASi is therefore recommended in international guidelines [1] but is often hindered by hyperkalaemia (typically defined as potassium >5.0 mmol/L), which results in either submaximal RAASi dosing or RAASi discontinuation, actions which consequently contribute to poorer long-term prognosis [6]. The risk of hyperkalaemia in heart failure is further increased in those with diabetes and chronic kidney disease (CKD) [7‒9], and episodes of hyperkalaemia can often be recurrent in these high-risk patients [10].

The advent of oral potassium binders, patiromer (Veltassa), and sodium zirconium cyclosilicate (SZC; Lokelma) have heralded a promising ability to better manage hyperkalaemia in patients with heart failure receiving RAASi [11‒15]. For instance, the 4-week PEARL-HF [11] trial recruited 105 patients with heart failure, of whom 41% had experienced RAASi discontinuation due to hyperkalaemia. Compared with placebo, patiromer significantly reduced potassium levels and increased the proportion of patients who could safely tolerate a dose increase of spironolactone (91 vs. 74%). Furthermore, 105 patients with heart failure in a subgroup of the AMETHYST-DN trial [12] also maintained normokalaemia with patiromer for up to 52 weeks. A 12-month study of SZC [15] in 758 patients showed that after initial normalization of potassium with SZC, a mean serum potassium of ≤5.1 mmol/L was achieved in 88% of patients, and among the patients taking RAASi, 87% continued them or had a dose increase. There were however only 15% with heart failure in the maintenance phase of this study, and thus, further data are awaited to extend the evidence base for using SZC in patients with heart failure [16].

However, while guidance now supports the use of potassium binders [17], there is a lack of real-world experience of how potassium binders change RAASi prescribing in patients with heart failure. Encouraged by the appeal of the new potassium binders, especially with patiromer, we established a hyperkalaemia clinic at our centre for the purpose of prescribing patiromer to facilitate RAASi optimization in patients with HFrEF. This current work reports on the outcomes of the first 21 months since inception of this clinic with specific focus placed on changes to RAASi prescribing, serum potassium levels, and estimated glomerular filtration rate (eGFR) in our patient cohort during follow-up.

The study population comprised patients with HFrEF who were referred from local general hospital heart failure services to a bespoke hyperkalaemia clinic at the Northern Care Alliance NHS Foundation Trust between July 2019 and December 2020. Patients with symptomatic heart failure for whom initiation or up-titration of RAASi was limited by the occurrence of hyperkalaemia were the targets for referral. These included patients whose potassium had climbed >5.3 mmol/L while receiving a RAASi and where the clinical team overseeing heart failure care felt further dose increases were felt not possible or had led to RAASi reduction or discontinuation.

Since its inception in July 2019, the clinic has run on a monthly basis led by a consultant nephrologist. The main intervention has been to oversee the initial and continued prescribing of patiromer, but it has also provided specialist renal care to those with advanced CKD. Prior to commencing patiromer, patients are educated on the importance of achieving normokalaemia, provided with low-potassium dietary information and given information about patiromer including directions for administration, common side effects, and the importance of taking the drug 3 h before, or after, other prescribed medications [18]. Upon commencing patiromer, recommendations for early checking of potassium and magnesium levels after patiromer initiation and during follow-up are provided to the referring team with subsequent monitoring taking place every 3–4 months in the bespoke hyperkalaemia clinic. RAASi up-titration is also deferred to the referring heart failure teams and is recommended when repeat potassium values are in the normal range (3.5–5.3 mmol/L). The referring team also has responsibility for electrolyte monitoring upon any RAASi dose titration. SZC is chosen at the discretion of the consultant nephrologist if patiromer is not tolerated. Patients are reviewed in the hyperkalaemia clinic every 3–4 months at the discretion of the consultant nephrologist for assessment of (i) tolerance to therapy, (ii) the need to continue a potassium binder, and (iii) patient’s potassium control. Patients are discharged from clinic if they discontinue patiromer due to drug intolerance or if patiromer is judged to be no longer indicated.

Demographic data included age, gender, ethnicity, systolic, and diastolic blood pressure. Comorbidities, in addition to heart failure, included diabetes mellitus, hypertension, and CKD. Medications included ACEi, ARB, MRA, angiotensin receptor-neprilysin inhibitor, diuretics, beta-blockers, sodium bicarbonate, and insulin. Laboratory measurements included eGFR, calculated using the CKD-EPI equation, serum bicarbonate (in mmol/L), and serum potassium (in mmol/L).

The primary outcome was the percentage of patients receiving a potassium binder for whom RAASi prescribing was increased, which was defined as initiating a RAASi, a dose increase in prescribed RAASi, or the initiation of an additional RAASi agent, either prescribed alone or as part of a combination therapy. Secondary outcomes included the proportion of patients with a serum potassium in the normal laboratory range and change in eGFR from the point of referral to patients’ last clinic visit. Outcome data were evaluated until 1 May 2021.

Continuous variables are presented as mean and standard deviation and categorical data as absolute numbers with percentages. The paired Student’s t test was used to analyse changes to patients’ potassium and eGFR levels during follow-up. A p value of <0.05 was considered statistically significant. Statistical analyses were conducted using R, version 4.0.2 (the R Foundation for Statistical Computing Platform).

A total of 34 patients with HFrEF were referred to the hyperkalaemia clinic from July 2019 until December 2020. In this cohort, the mean age was 71.6 years (±10.6 years), and patients were predominantly female (71%), and almost exclusively Caucasian (94%), as shown in Table 1. There was a high prevalence of significant comorbidities, including hypertension (53%), diabetes mellitus (56%), and CKD stages 3a-5 (71%); the mean eGFR was 56 mL/min/1.73 m² (±21 mL/min/1.73 m²). Most patients were already established on a RAASi (88%) at the first clinic visit with 76% receiving an ACEi or ARB, either alone or in combination with an MRA. Two patients were receiving an angiotensin receptor-neprilysin inhibitor. The mean potassium at first clinic visit was 4.9 mmol/L (±0.5 mmol/L). At review in clinic, most patients (68%) had a potassium in the normal range, while 9 patients (26%) had a value between 5.3 and 6.0 mmol/L. The serum bicarbonate was normal in all patients with a mean of 27.2 mmol/L (±2.2 mmol/L).

All 34 patients were prescribed patiromer 8.4 g daily. Over a follow-up of 10.5 months (±6.4 months), 13 patients (38%) discontinued patiromer (Fig. 1). Of this latter group, 6 did so due to gastrointestinal side effects. Other reasons for discontinuation included patient reluctance to increase RAASi: a concern for patient compliance in the setting of vascular dementia and improved echocardiographic findings of heart function that led to patient discharge from the heart failure services (Table 2). In those that discontinued patiromer, 1 patient was successfully switched to SZC and remained under follow-up, while the remaining 12 patients were discharged. Two patients died having had only one clinic visit (1 died from sepsis and the other from an intracerebral haemorrhage).

A total of 20 patients were analysed for the study outcomes: 19 patients were receiving patiromer 8.4 g daily, and 1 patient received SZC 5 g daily. The mean follow-up time in this patient group was 13.4 months (±5.8 months). All patients had had at least 2 hyperkalaemia clinic visits. The minimum time that a patient received patiromer was for at least 3 months, with 12 (63%) receiving it for more than 12 months and 4 patients (21%) receiving it for 20 months. The patient receiving SZC received it for 7 months up until their last clinic visit. Seventeen of the 20 patients (85%) receiving a potassium binder were able to have an increase in their RAASi medication by the end of the study follow-up (Table 3): 3 patients were initiated on a RAASi, 8 had their RAASi dose increased, 4 had an additional agent added, and 2 had both an increase in their RAASi and an addition of another agent. Four patients (20%) attained maximal dosing of a RAASi by the end of the study follow-up. We also reviewed changes to patients’ diuretic therapy and beta-blocker use in those receiving potassium binders: 6 (30%) patients received a slight up-titration of their beta-blocker, and 2 (10%) received a slight reduction, whereas there were minimal changes to diuretic therapy as 1 patient had their dose reduced, another discontinued, and another replaced with an MRA.

With respect to potassium control in the 20 patients who remained under follow-up, 45% of patients (9/20) had normokalaemia at the time of referral to the clinic, 85% (17/20) at their first clinic visit, and 75% (15/20) at their last clinic visit. The mean serum potassium was 5.5 mmol/L (±0.3 mmol/L) at the point of referral, 4.9 mmoL/L (±0.4 mmol/L) at the first clinic visit (approximately 6 weeks on average after referral), and 4.9 mmol/L (±0.4 mmol/L) at patients’ last clinic visit (Fig. 2); p < 0.001 for the difference between the potassium at the time of referral and at their last clinic visit. No patient experienced hyperkalaemia of >6.0 mmol/L or hypokalaemia of <3.5 mmol/L at any point during follow-up. Furthermore, no patients received magnesium supplementation during the follow-up period; mean magnesium level was 0.76 mmol/L (±0.08 mmol/L) at the last clinic visit.

With respect to changes in eGFR, 8 patients (40%) experienced a decline in eGFR. The mean eGFR at the time of referral was 57 mL/min/1.73 m² (±22 mL/min/1.73 m²), 60 mL/min/1.73 m² (±22 mL/min/1.73 m²) at the first clinic visit and 55 mL/min/1.73 m² (±25 mL/min/1.73 m²) at the last clinic visit; p = 0.37 comparing eGFR at time of referral and at their last clinic visit.

This study provides for the first time, to the best of our knowledge, a real-world experience of prescribing potassium binders in a nephrology-led hyperkalaemia clinic. While patiromer was not tolerated in a quarter of patients, we show that for those who continued treatment, and SZC in one case, RAASi optimization was facilitated in patients with HFrEF by controlling potassium. We also highlight the safe continuation of patiromer beyond 12 months in clinical practice.

The role for patiromer prescribing was suited to nephrology given our experiences of managing ambulatory hyperkalaemia in patients with CKD, in particular ensuring acidosis is corrected and patients receive information to maintain a low-potassium diet. While there was a lack of treatment algorithms in the literature for managing hyperkalaemia in heart failure at the time of establishing our clinic, we devised a relatively simple management approach: if the serum potassium is in the normal range while on patiromer, up-titration of RAASi should be undertaken for improved heart failure symptom control and maximum prognostic benefit. This proved to be achievable through the collaborative efforts of the referring heart failure teams as 85% of patients taking a potassium binder long term experienced an increase in RAASi prescription. The majority of patients experienced an increase in their RAASi dose (e.g., from ramipril 2.5 mg daily to 5 mg daily or sacubitril/valsartan 49/51 mg daily, increased to 97/103 mg twice daily; Table 3, patients 8 and 13, respectively), but some experienced both an up-titration of their initial RAASi and an additional agent (e.g., ramipril 1.25 mg daily increased to ramipril 10 mg with the addition of spironolactone 25 mg daily; Table 3, patient 5). Importantly, 4 patients (20%) reached maximal doses of at least one RAASi by the end of the study follow-up.

Current guidance from the National Institute for Health and Care Excellence (NICE) recommends that patiromer or SZC be prescribed for outpatients with heart failure who are taking a suboptimal dose of RAASi if they have had a serum potassium of ≥6.0 mmol/L [19]. We employed a modified approach: all our patients received patiromer at 8.4 g daily even where the potassium was in the normal range. This is because patients had been referred by heart failure teams who had encountered elevated potassium levels when RAASi doses had been previously increased, and which had hindered further RAASi increases or had led to its discontinuation. While the confidence to continue RAASi will vary between healthcare professionals, there was a reluctance to initiate or up-titrate RAASi when serum potassium was in the range 5.3–6.0 mmol/L. This is most likely due to concerns for the potential for levels to rise beyond 6.0 mmol/L, which often necessitates hospitalization for acute potassium-lowering management due to the risk of life-threatening cardiac arrhythmias. Our strategy ensured no patient experienced potassium levels of ≥6.0 mmol/L.

We now welcome the recent publication of a proposed treatment algorithm based on expert consensus for managing RAASi-associated hyperkalaemia in patients with heart failure [20]. The algorithm details a methodology for prescribing patiromer, importantly recommending its use when potassium levels are ≥5.1 mmol/L. It also provides advice on the frequency of electrolyte monitoring, while a patient is receiving a potassium binder and after RAASi doses are adjusted. We envisage this will support the development of clinical practice guidelines for our heart failure services and may permit them to execute both patiromer prescribing and RAASi titration under a single service framework as opposed to our current model. This will be an important step to help increase the accessibility of patiromer and afford prompt RAASi titration.

A quarter of patients discontinued patiromer due to an intolerance, the most common cause being gastrointestinal side effects. This is a well-recognized issue and occurred in 18% of our cohort, similar to figures observed in trial data: 17% of patients in the OPAL-HK trial [12] and 21% of patients in the PEARL-HF trial [11].

Hypomagnesaemia can also arise with patiromer treatment, but mean serum magnesium levels were within normal limits at follow-up in our cohort and no patients required supplementation. In the patient taking SZC, peripheral swelling or oedema did not arise. While the eGFR did fall in 8 patients (40%) who continued with a potassium binder, these patients all experienced an increase in their RAASi prescription during follow-up and there was no statistical difference between the eGFR at referral and at patients’ last clinic visit.

A key limitation of the clinic is the dependence on the referring team to make the necessary RAASi adjustments and ensure appropriate blood monitoring is undertaken. While the majority of patients received a RAASi dose increase, we question whether more rapid titration could have occurred, and a greater proportion of patients could have been in receipt of maximal doses during the follow-up period. A closer multidisciplinary approach may improve this aspect of patient care. In the long term, we foresee the service being entirely delivered by the cardiology teams with support from ourselves if concerns regarding CKD arose. Secondly, we lacked baseline serum magnesium levels, but follow-up magnesium levels showed no patient required supplementation. Thirdly, the early blood monitoring undertaken by the district hospitals upon initiation of a potassium binder was not collected in this study. However, no issues regarding these tests were directly raised by the referring teams and follow-up results in the clinic highlighted the efficacy of patiromer in maintaining potassium levels <6.0 mmol/L. Fourthly, not all patients who discontinued patiromer were directly offered SZC, which may have been a suitable and effective alternative. However, during the early phase of the clinic’s operation, there was little familiarity with its use in heart failure and so it was not prescribed. We have subsequently communicated with the referring heart failure teams to raise the question as to whether they may wish to trial SZC in patients who are intolerant to patiromer. Finally, we report on the experience of a small cohort of patients from a single centre and lack long-term data on patients’ symptoms or major clinical endpoints. The ongoing DIAMOND trial [21], seeking to recruit over 2,000 patients, will hopefully provide further evidence on this front.

Our work also has strengths. We highlight, through the delivery of a bespoke hyperkalaemia clinic, that the use of patiromer was efficacious at controlling potassium and facilitating RAASi administration over an extended period of time. We show that patiromer can be used safely for more than 12 months, extending the evidence base for its safe use beyond that reported in clinical trials [13]. Despite our cohort being small, it was characterized by a high-risk patient group that was most likely to benefit from RAASi therapy. Our work also has the advantage of providing a granular inspection of RAASi prescribing, and we suggest there is scope for this to be improved to align more closely with international guidelines [17].

Ultimately, we hope our real-world clinic experience, in tandem with the recently published treatment algorithm for managing hyperkalaemia in heart failure, will provide confidence to multidisciplinary specialists to consider incorporating potassium binders more readily into routine clinical care to negate the impact of RAASi-associated hyperkalaemia in patients with HFrEF. In conclusion, in the setting of a nephrology-led hyperkalaemia clinic, patients with HFrEF who tolerated novel potassium binders, largely patiromer, benefited from potassium control that subsequently facilitated RAASi optimization. While RAASi doses were increased for the majority of patients, we await larger clinical studies to evaluate the impact of potassium binders on major outcomes such as cardiovascular and all-cause mortality.

All patients referred to the hyperkalaemic clinic were enrolled into the Salford Kidney Study (SKS) [22] an ongoing prospective epidemiological study, which since 2002 has been recruiting patients with non-dialysis-dependent chronic kidney disease. The SKS gained ethical approval from the North West Greater Manchester South Research Ethics Committee (REC15/NW/0818). Written informed consent was obtained from all patients.

Philip A. Kalra has received honoraria for lectures and advisory board activity from Vifor Pharma and AstraZeneca as well as research funding from Vifor Pharma. Paul Kalra has received grants from AstraZeneca, Pharmacosmos, and Vifor Pharma and honoraria from AstraZeneca, Bayer, Boehringer Ingelheim, Napp, Novartis, Pharmacosmos, Servier, and Vifor Pharma.

No fundings were received.

Ibrahim Ali and Philip A. Kalra conceived the study. Philip A. Kalra led the hyperkalaemia service. Ibrahim Ali undertook data collection, performed statistical analyses, and wrote the initial draft. Darren Green, Paul Kalra, and Philip A. Kalra critically revised the article. Ibrahim Ali, Darren Green, Paul Kalra, and Philip A. Kalra approved the final manuscript.

Data are available on reasonable request.