Comparative Efficacy of Sodium Zirconium Cyclosilicate and Sodium Polystyrene Sulfonate for Acute Hyperkalemia: A Retrospective Chart Review

Abstract

Background: Sodium polystyrene sulfonate (SPS) is a nonselective sodium-potassium exchange resin commonly used along with intravenous (IV) insulin, albuterol, furosemide, and/or calcium for the treatment of acute hyperkalemia. Sodium zirconium cyclosilicate (SZC) is a newer non-absorbed exchange resin that preferentially increases fecal potassium excretion from the gastrointestinal tract. Limited data exists on the efficacy of SZC for the treatment of acute hyperkalemia. Objectives: To assess the achievement of normokalemia (serum potassium level [K+] 3.5-5.2 mmol/L) within 24 hours after administration of SZC or SPS in combination with insulin regular IV push. Methods: A multicenter, retrospective chart review (2020-2021) using electronic medical records at an academic health system. The study population included adult patients receiving one or more doses of SZC or SPS in combination with IV insulin for acute hyperkalemia (K+ >5.2 mmol/L). Patients receiving dialysis were excluded. Serum chemistries were assessed at baseline and an additional 2 values within 24 hours to determine normokalemia and hypokalemia at each follow-up. Results: Of 141 patients included, 51 received SZC and 90 received SPS. Normokalemia at the first follow-up was achieved in 51.0% of patients receiving SZC and 46.7% of patients receiving SPS (P = .622) and was sustained in 35.3%versus 44.4% (P = .289) of patients within 24 hours. Mean serum potassium differences from baseline to first follow-up were similar between SZC and SPS groups (0.9 mmol/L vs 1.0 mmol/L). Hypokalemia within 24 hours of administration occurred in 4 patients—1 in SZC, 3 in SPS. Conclusion: Both SZC and SPS yielded similar rates of normokalemia achievement with IV insulin for the treatment of acute hyperkalemia. Further prospective studies are needed to confirm these findings.

Background

Hyperkalemia is a common electrolyte abnormality seen especially in patients with chronic kidney disease (CKD) and those on specific medications, such as angiotensin-converting enzyme inhibitors (ACEi), angiotensin receptor blockers (ARBs), aldosterone antagonists, and potassium supplementation. The prevalence of hyperkalemia in the general population is nearly 1.6%, with higher numbers seen in chronic disease patients. ¹ In patients with CKD and/or heart failure, the annual prevalence can be as high as 6.35% ¹ ; among hospitalized patients, the incidence of hyperkalemia ranges from 3% to 10%.^2,3

While patients with mild hyperkalemia (potassium [K+] 5.3-5.7 mmol/L) can be asymptomatic, some may experience muscle weakness, fasciculations of upper and lower extremities, and paresthesia. Conversely, moderate to severe hyperkalemia (K+ greater than or equal to 5.8 mmol/L) can induce severe sequelae, such as cardiac conduction abnormalities and arrhythmias.^4,5 Therefore, the symptomatic hyperkalemic patient must be treated urgently to mitigate electrophysiologic cardiac effects. Treatment may include intravenous (IV) low-dose insulin regular (5-10 units) with or without dextrose, IV calcium salts, nebulized β-adrenergic agonists (eg, albuterol), IV loop diuretics, IV sodium bicarbonate, or dialysis. Intravenous calcium gluconate or calcium chloride stabilizes the cardiac membrane, while insulin lowers serum potassium by shifting potassium intracellularly. Enhanced potassium excretion can be achieved with diuretics or directly through hemodialysis, while β-adrenergic agonists and sodium bicarbonate can facilitate redistribution of serum potassium into the intracellular space. ⁵

Additionally, potassium binders/cation exchangers, such as sodium polystyrene sulfonate (SPS), sodium zirconium cyclosilicate (SZC), and patiromer, can be used as monotherapy or in conjunction with the above treatment to achieve or maintain the normokalemic state and remove excess potassium through enhanced gastrointestinal clearance. ⁶ SPS, a nonselective sodium-potassium exchange resin, has been used in the past for hyperkalemia treatment and carries a risk for intestinal necrosis. However, no large, randomized trials have investigated this indication. ⁷ A retrospective cohort study by Noel et al ⁷ compared patients receiving SPS in an outpatient setting to non-users and concluded an increased risk of hospitalizations for adverse effects associated with SPS.

SZC is a non-absorbed zirconium cyclosilicate that preferentially exchanges potassium for hydrogen and sodium ions, thus increasing fecal potassium excretion. It is FDA approved for the treatment of chronic hyperkalemia with initial dosing of 10 g orally 3 times a day for up to 48 hours, followed by a maintenance dose of 5 to 15 g daily. ⁸ Currently, minimal prospective data has been has been published describing SZC in the treatment of acute hyperkalemia. ⁹ A recently published retrospective analysis compared the use of a one-time dose of SZC to SPS and found no difference in the clinical efficacy of the 2 products. ¹⁰

One prospective study, ENERGIZE trial by Peacock et al, ¹¹ potentially supports the use of SZC for this indication. In this phase II exploratory trial, 70 adult ED patients with hyperkalemia (defined as K+ ≥5.8 mmol/L) were randomized to receive SZC or placebo up to 3 times during a 10-hours period in addition to a background of insulin IV. ¹¹ A numerically lower proportion of patients in the SZC group required additional adjunctive therapies at 0 and 4 hours. Additionally, phase 2 and phase 3 trials showed SZC effectively reduced potassium after 2 days of therapy. ¹² In phase 3, an open-label, single-arm trial by Roger et al, patients received SZC 10 g 3 times daily for 24 to 72 hours until the achievement of normokalemia, with 82% of patients achieving normokalemia within 24 hours. ¹³

Objectives

This retrospective study seeks to compare the use of SZC versus SPS for the treatment of acute hyperkalemia in an acute care setting. The primary objective of this study was the achievement of sustained normokalemia (3.5-5.2 mmol/L) within 24 hours after administration of SZC compared to SPS, in combination with insulin regular IV push. The study’s secondary objectives included the mean difference in serum potassium level between the baseline and the first follow-up measurement and documented adverse effects, including the incidence of hypokalemia.

Methods

This retrospective, multicenter, cohort study was conducted to assess the achievement of sustained normokalemia with SZC versus SPS in combination with insulin regular IV. All doses of SPS were of the oral suspension and sorbitol-free. Normokalemia was considered sustained if serum potassium remained within 3.5 to 5.2 mmol/L (the defined standardized range within this health network’s electronic medical records system, EMR) for the follow-up measurements within 24 hours post-drug administration. Hyperkalemia was defined as serum potassium levels greater than 5.2 mmol/L as coded by the EMR. Patients must have received insulin regular IV push as an inclusion criterion to demonstrate the acuity of the acute hyperkalemia episode. Patients could receive other agents for hyperkalemia treatment, including IV calcium, loop diuretic, sodium bicarbonate, or nebulized albuterol, all of which appear on an order panel of the EMR for this indication. The study design was approved by the Institutional Review Board (IRB) as expedited research. The study design, data collection, and analysis were conducted by the authors of this study.

Data Source

Using EMR documentation, data were collected from August 1, 2020 to August 31, 2021, via REDCap electronic data capture tools.^14,15 Baseline characteristics include age, weight, sex, race, comorbidities, and serum chemistries measured 3 times: at baseline on admission, a first reassessment (first follow-up) after medication administration (any time within 24 hours), and the second reassessment (closest to 24 hours after receiving the first study medication(s) without going over 24 hours). Data on the incidence of adverse effects (eg, hypokalemia, hypomagnesemia) were collected based on the EMR documentation.

Inclusion/Exclusion Criteria

The study included any adults (age greater than or equal to 18) receiving at least 1 dose of SZC or SPS, in combination with IV insulin regular, for the treatment of acute hyperkalemia. Patients were excluded if no doses of SZC or SPS were administered or if they were not administered concurrently with insulin. Other exclusion criteria included baseline serum potassium level less than 5.3 mmol/L; no documented baseline serum potassium level; no documented serum potassium level within 24 hours after administration of the treatment; receipt of dialysis during the 24-hour period after potassium binder administration; administration of both SZC and SPS; administration of patiromer concomitantly with SZC or SPS; or hemolysis of the laboratory specimen.

Statistical Analysis

SPSS statistical software, version 22.0, and SAS software were used for data analysis. Descriptive statistics, measures of central tendency, spread, and frequency were used to analyze the baseline characteristics of the treatment groups. Based on the distribution of the variables, parametric (t-test, Pearson correlation) and/or non-parametric (Wilcoxon rank sum, Chi square, Fishers exact) analysis was performed to assess the mean differences between the groups. The correlational statistics and regression analysis with the significance level of alpha = .05 was used to evaluate relationships among the study variables. Logistic regression analysis was conducted to determine covariates that increased the odds of the primary outcome.

Results

A total of 141 patients met the inclusion criteria for this study—90 in the SPS group and 51 in the SZC group. Baseline characteristics were similar between the 2 groups (Table 1). Patients included for analysis were 80.9% White and 12.8% African American, 44.0% females, had a mean age of 71.9 years, and an average BMI of 29.0. Congestive heart failure was observed in nearly a quarter of the study participants, and 63 patients (44.7%) had chronic kidney disease. Additionally, 71% of subjects had documentation of acute kidney injury in the medical chart. Finally, the average serum potassium on presentation were 6.24 and 6.10 mmol/L for SPS and SZC, respectively (Table 1).

Table 1.

Study Participant Baseline Characteristics Prior to Administration of Potassium Binder.

Patient characteristic, mean (SD)	SZC (n = 51)	SPS (n = 90)	P-value
Age, y	71.69 (13.26)	72.06 (15.88)	.89
Height, cm	169.52 (11.11)	168.8 (10.63)	.70
Weight, kg	82.28 (24.43)	83.38 (23.45)	.79
BMI, kg/m2	28.58 (7.99)	29.22 (7.2)	.63
Female gender, n (%)	20 (39.2)	42 (46.7)	.39
Race, n (%)			.51
White	43 (84.3)	71 (78.9)
African American	5 (9.8)	13 (14.4)
Asian	0	3 (3.3)
American Indian/Alaska Native	0	1 (1.1)
More than 1 race	1 (2.0)	1 (1.1)
Unknown/not reported	2 (3.9)	1 (1.1)
Comorbidities, n (%)
Congestive heart failure	13 (25.5)	24 (26.7)	.88
CKD Stage I	0	0	n/a
CKD Stage II	1 (2.0)	2 (2.2)	1.00
CKD Stage IIIa	17 (33.3)	19 (21.1)	.11
CKD Stage IIIb	3 (5.9)	8 (8.9)	.75
CKD Stage IV	6 (11.8)	10 (11.1)	.91
SARS-CoV-2	7 (13.7)	11 (12.2)	.79
Diabetic ketoacidosis	1 (2.0)	4 (4.4)	.65
Rhabdomyolysis	1 (2.0)	3 (3.3)	1.00
Tumor lysis syndrome	0	0	n/a
Acute kidney injury	35 (68.6)	65 (72.2)	.65
Admission medications, n (%)
Angiotensin converting enzyme inhibitor (ACEi)	10 (19.6)	12 (13.3)	.32
Angiotensin receptor blocker (ARB)	8 (15.7)	21 (23.3)	.28
Angiotensin receptor and neprilysin inhibitor (ARNI)	0	3 (3.3)	.55
Mineralocorticoid receptor antagonist (MRA)	4 (7.8)	3 (3.3)	.25
Potassium supplement	7 (13.7)	10 (11.1)	.65
Baseline laboratory values, mean (SD)
Serum creatinine, mg/dL	3.65 (2.87)	3.03 (2.12)	.19
eGFR, a mL/min/m2	23.23 (15.02)	22.89 (13.05)	.89
Serum potassium, mmol/L	6.10 (0.53)	6.24 (0.66)	.22
Serum bicarbonate, mmol/L	21.04 (5.00)	19.68 (5.21)	.14
Serum phosphorus, b mg/dL	5.92 (1.57)	5.81 (1.99)	.84
Serum magnesium, mg/dL	2.16 (0.38)	2.10 (0.42)	.57

Note. BMI = body mass index; CKD = chronic kidney disease; eGFR = estimated glomerular filtration rate; mmol = millimole; SARS-CoV2 = severe acute respiratory syndrome coronavirus 2; SPS = sodium polystyrene sulfonate; SZC = sodium zirconium cyclosilicate.

^aData available for 47 SZC and 78 SPS patients.

^bData available for 16 SZC and 31 SPS patients.

Study medication utilization is presented in Table 2. Of note, a higher number of doses of SZC were administered per patient compared to SPS (P = .044). Other adjunctive antihyperkalemics, including intravenous calcium salts, loop diuretic, sodium bicarbonate, and nebulized albuterol, were frequently coadministered in both groups, the most common of which were calcium gluconate followed by sodium bicarbonate injection (Table 3); utilization of adjunctive agents was similar between the SPS and SZC groups.

Table 2.

Utilization of Study Medications.

Characteristic, mean (SD)	SZC (n = 51)	SPS (n = 90)	P-value
Initial dose, g	10.00 (0)	21.17 (8.07)	N/A
Total dose, g	18.33 (14.06)	30.00 (21.45)	N/A
Number of doses	1.76 (1.24)	1.38 (0.72)	.044

Note. SPS = sodium polystyrene sulfonate; SZC = sodium zirconium cyclosilicate.

Table 3.

Utilization of Adjunctive Agents for Hyperkalemia.

Adjunctive agent, n (%)	SZC (n = 51)	SPS (n = 90)	P-value
Albuterol	14 (27.5)	17 (18.9)	.2382
Calcium chloride	0 (0.0)	2 (2.2)	.5350
Calcium gluconate	40 (78.4)	61 (67.8)	.1775
Furosemide	9 (17.7)	8 (8.9)	.1249
Sodium bicarbonate	29 (56.9)	41 (45.6)	.1969

Note. SPS = sodium polystyrene sulfonate; SZC = sodium zirconium cyclosilicate.

There were no statistically significant differences between the normokalemia achievement rate in the SPS and SZC groups at first follow-up (P = .622) nor sustained up to 24 hours after initiation of the study medications (P = .289). A sensitivity analysis with normokalemia defined as 3.5 to 5.5 mmol/L yielded higher achievement of normokalemia but no difference between groups (Table 4). Patients received SZC on average 1 hour later, compared to SPS, after baseline laboratories were drawn (P = .001); however, there was no significant difference in time from drug administration to the first (P = .293) or second follow-up (P = .157) measurements. Furthermore, no difference (P = .988) was noted in the first follow-up serum potassium values between SPS and SZC; both treatment groups resulted in a mean reduction of approximately 1.0 mmol/L. Interestingly, recipients of SZC had higher mean serum potassium values at the second follow-up measurement. Hypokalemia was a rare event encountered in only 4 patients—3 in the SPS group and 1 in the SZC group (Table 4). Serum magnesium levels were similar between groups at the first and second follow ups (Table 4).

Table 4.

Primary and Secondary Study Outcomes.

	SZC (n = 51)	SPS (n = 90)	P-value
Normokalemia (3.5-5.2 mmol/L), n (%)
First follow up	26 (51.0)	42 (46.7)	.622
Sustained by 24 h	18 (35.3)	40 (44.4)	.289
Normokalemia (3.5-5.5 mmol/L), n (%)
First follow up	35 (68.6)	62 (68.9)	.974
Sustained by 24 h	40 (78.4)	78 (86.7)	.204
Serum K+ values, mean (SD)
First follow up (mmol/L)	5.2 (0.61)	5.2 (0.87)	.988
Second follow up (mmol/L)	5.1 (0.69)	4.7 (0.75)	.021
Difference in serum K+ from baseline, mean (SD)
At first follow up (mmol/L)	0.9 (0.7)	1.0 (0.8)	.309
At second follow up (mmol/L)	2.2 (2.2)	2.7 (2.0)	.066
Hypokalemia (<3.5 mmol/L), n (%)
First follow up	0 (0.0)	3 (3.3)	.188
Anytime within 24 h	1 (2.0)	3 (3.3)	.637
Time to event (h), mean (SD)
Baseline measurement to drug administration	3.9 (2.2)	2.9 (1.8)	.001
Drug administration to first follow up	6.4 (5.3)	5.6 (5.1)	.293
Drug administration to second follow up	16.2 (3.9)	17.1 (4.9)	.157
Serum Mg2+ values, mean (SD)
First follow up (mg/dL) a	2.2 (0.3)	2.0 (0.3)	.078
Second follow up (mg/dL) b	2.0 (0.3)	2.1 (0.4)	.540

Note. SPS = sodium polystyrene sulfonate; SZC = sodium zirconium cyclosilicate.

^aData available for 14 SZC and 24 SPS patients.

^bData available for 24 SZC and 34 SPS patients.

Logistic regression analysis was performed to analyze the achievement of normokalemia against 5 predictor variables; however, the model poorly fit the data as only 6.6% of the variability in the response variable was explained by the model (Nagelkerke R squared = .066). None of the analyzed predictor variables significantly impacted the achievement of normokalemia.

Discussion

Initial phase 3 studies of SZC for chronic hyperkalemia management were conducted in a community-based setting. ¹² Given the favorable safety profile for SZC and the onset of potassium reductions within 1 hour of administration, prescribers have grown curious about the role SZC may play in acute hyperkalemia management. ⁸ However, limited prospective data exist describing efficacy in this context. As such, this retrospective review of potassium binder recipients provides timely information to help guide clinical practice. We observed no difference in sustained normokalemia achievement over a 24-hour period between SZC and SPS, and serum potassium levels were nearly identical at the first follow-up between both treatment groups.

Our retrospective review aligns with the results of another chart review by Lin et al ⁹ that found 44% of patients who received SZC versus 40% who received SPS achieved normokalemia within 24 hours of administration. Similar to the ENERGIZE, we also found a few instances of hypokalemia with either SZC or SPS. ¹¹ Clinicians can look forward to a prospective randomized trial, actively recruiting at the time of this manuscript’s publication, to compare potassium-lowering with SPS, SZC, patiromer, or a nonspecific laxative (polyethylene glycol 3350) in the emergency department. ¹⁶ Similar results were also demonstrated in recently published retrospective reviews by Hasara et al ¹⁷ and Joyce and Corpman. ¹⁸

Our study design is subject to several limitations. First, the retrospective design led to variation in the timing of laboratory assessments, unlike structured follow-up measurements that would be possible in a controlled trial. To mitigate this variable, we analyzed the time to laboratory measurements which were similar between groups. Next, our institutional EMR flags hyperkalemia at serum potassium level greater than 5.2 mmol/L as opposed to 5.5 mmol/L; we conducted a sensitivity analysis that yielded a similar result using both thresholds. Residual confounding may exist, as noted by higher average doses of SZC recorded compared to SPS, though no significant differences were seen in rates of adjunctive treatment use. This could be due to the EMR build for SZC featuring options to select “once” or “three times a day” with a defaulted duration of 2 days compared to 1-time doses for SPS. Receipt of dialysis was also an exclusion criterion which prevented assessment of the efficacy of either potassium binder in this patient population. The study may have been underpowered to assess multiple outcomes and increased the chance of type II error as a result.

Conclusions

Both SZC and SPS yielded similar rates of normokalemia achievement with IV insulin for the treatment of acute hyperkalemia. A similar decrease in serum potassium was noted at the first follow-up in both groups. These results must be tempered by limitations in sample size and study design. Future prospective studies should be conducted to confirm these findings and to determine the clinical utility of potassium binders in acute hyperkalemia management.