Introduction

The incidence of hyperkalemia is one of the most common complications of kidney disease. Its incidence increases in patients who previously experienced hyperkalemia, with successively shorter time between the episodes [1].

Patiromer and Sodium Zirconium Cyclosilicate (SZC), new exchange ions polymer resin and exchange ions microporous resin, were recently developed reducing adverse events and improving palatability compared to old potassium binders [2–4].

Furthermore, many studies in conservative CKD demonstrated that these new drugs reduce hyperkalemia in patients treated with RAASIs [5].

Many systematic reviews compared new potassium binders with old potassium binders or placebo, but no one performed it in dialysis patients [6].

The main objective of our metanalysis is to evaluate the difference in serum potassium levels after treatment with SZC and Patiromer compared to placebo, sodium polystyrene sulfonate or calcium polystyrene sulfonate. Furthermore, the safety needs to be evaluated among these potassium binders in this population, due to the different pharmacokinetics that improve them.

 

Methods

Data source and search strategy

This meta-analysis followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [7] and was conducted according to a pre-published protocol (CRD42024608049) [8, 9]. The literature search was designed and performed by two Authors (V.C. and A.P.). We performed a systematic, highly sensitive search in MEDLINE, PubMed, CINAHL and EMBASE for English-language articles without time or sex restriction up to November 1, 2024. Grey literature was screened through Google Scholar, SCOPUS and clinicaltrials.gov.

Study selection and data extraction

We included any randomized controlled trials (RCTs) testing the effects of potassium binders without sex restriction. Studies were included if provided information on outcomes of interest, such as 1) Differences in serum potassium concentration, 2) Adverse events (AEs), and 3) Mortality.

Studies were excluded if they included oncological patients, acute dialysis or CKD in conservative treatment.

Articles were screened by titles and abstracts by two independent investigators (V.C. and A.F.), excluding studies not pertinent to the topic and, subsequently, assessed the full texts to determine eligibility according to the pre-specified inclusion/exclusion criteria. Any disagreement on study judgments was discussed with a third author (V.M.) who was not involved in the selection process.

Reviews, letters, case reports and studies performed on children (age<18) were excluded from analyses but screened for potential additional references. Ongoing, unpublished trials were searched on the clinicaltrials.gov website. Authors of some of the included studies were contacted for additional information about study methods and unreported data.

Data analysis

Primary analysis will compute serum potassium differences between a network meta-analysis comparing each potassium binder/placebo. Secondary analysis will consist of a comparison between SZC/Patiromer and placebo/sodium polystyrene sulfonate or calcium polystyrene sulfonate.

All data will be analyzed with fixed-effect model or random-effect model based on the heterogeneity of the studies. Mean differences, and 95% confidence interval (CI), will be calculated for continuous outcomes. For dummy outcomes, the Odds Ratio (OR), computing 95% confidence interval (CI), will be computed. Data were pooled using the fixed-effects model and also analyzed with the random-effects method to guarantee the strength of the model. We plan to test for heterogeneity using the χ² statistic related to the degrees of freedom, with a p value of 0.05 used as the cut-off value to determine statistical significance. In addition, the degree of heterogeneity will be investigated by calculating the l² statistics. We will consider l² low if <25%, moderate if 25-50%, moderate-high if 50-75% and very high if >75%. In case of high heterogeneity, we will perform sensitivity analyses to explore sources of heterogeneity, such as study quality, year of publication, intervention or control variables, participants characteristics, and risk of bias. In addition, sub-group analyses will be conducted.

We will assess funnel plot asymmetry and the contour-enhanced funnel plot to explore publication bias, if the number of studies allows it. GRADE System will be used to evaluate the certainty of the evidence and to summarize the study conclusions. We planned to construct a summary table via the GRADEpro-GDT (GRADEpro GDT 2015) [10], reporting a summary of the available outcomes’ findings and the quality of the evidence supporting each outcome.

Statistical analyses were performed using Review Manager (RevMan; Version 5.4.) software and R4.4.0 software to perform the Network meta-analysis of all outcomes.

Quality and risk of bias assessment

The quality of RCTs was assessed by using the checklist developed by the Cochrane Renal Group which evaluates the presence of potential selection bias (random sequence generation and allocation concealment), performance bias (blinding of investigators and participants), detection bias (blinding of outcome assessors), attrition bias (incomplete outcome data), reporting bias (selective reporting) and possible other sources of bias.

 

Results

Search results

The selection process has been summed up in the flowchart (Figure 1). Two hundred and twenty-two references were retrieved. Among these, 95 were duplicates. By screening titles and abstracts, a total of 16 citations were selected for full-text evaluation. Among these, 6 articles were excluded because: 1) dealing with other populations or not reporting outcomes pertinent to the topic (n=7), 2) wrong intervention or no comparator (n=2), 3) duplicate (n=0), 4) various reasons (n=2). One study has been retrieved by Clinicaltrial.gov. Six articles referring to 5 full studies were reviewed in detail and included in the review [11, 15].

Study characteristics

All RCTs employed a parallel design. All studies were published after 2019. The final population analyzed in this review included 3155 patients, but the study sample sizes were variable, ranging from 33 [11] to 2690 [13]. The percentage of male participants varied from 49.3% [14] to 73% [15]. The mean age of patients ranged from 54 [11] to 66 [15] years. Three studies compared SZC to placebo, one study compared Patiromer to placebo, and one crossover study compared Patiromer to both placebo and SPS. The main characteristics of the eight RCTs reviewed are summarized in Table 1.

Trial reference	Study population and its characteristics	Intervention	Comparator	Study duration	Outcome(s)	Results	Notes
NCT03781089	Thrice weekly HD, adults Exclusion: Pregnancies -N=33 -Age (yr)= ~54 – Men (%)= 54.5% – Hispanic = 18.2% Countries= United States (single centre)	Patiromer (8.4mg/die)   N=17	SPS (15g)   N=16	7 weeks	Delta K	Not reported	Patients were checked up at 3, 5, 10, 14 and 27 weeks after initiation of treatment
					hypekalemia events*	13 in patiromer group vs 41  in SPS group
					AEs	Death: 0 in patiromer group vs 0 SAE: 2 in patiromer group vs 2 (Arrhythmia: 0 in patiromer group vs 1 Infection: 0 in patiromer group vs 1 FAV thrombosis: 1 in patiromer group vs 0 Hospitalization: 1 in patiromer group vs 0)
NCT03 303521	Thrice weekly HD, adults Exclusion: Hemoglobin <9 g/dl Pregnancies -N= 196 -Age (yr)=58 -Men (%)= 58.7% -Hispanic (%)= 52% -Countries= Japan, Russia, US, UK	SZC (5-10 mg)   N=97	Placebo   N=99	11 weeks	Delta K*	Mean difference between SZC and Placebo groups: − 0.74 mmol/L (95% CI − 0.97/− 0.52)	It is a post-Hoc analysis
					hyperkalemia events	Not reported
					AEs	Hypokalemia (k<2 mmol/L) was not registered in either SZC group or the placebo group
NCT03303521	Thrice weekly HD, adults Exclusion: Hemoglobin <9 g/dl Pregnancies -N= 196 -Age (yr)=58 -Men (%)= 58.7% -Hispanic (%)= 52% -Cauntries= Japan, Russia, US, UK	SZC (5-10 mg)   N=97	Placebo   N=99	11  weeks	Delta K	Not reported
					hyperkalemia events*	6/97 in SZC group vs 13/99 in placebo group
					AEs	Total AEs: 40/97 vs 46/99 in placebo group GI disorders: 19/97 vs 17/99 in placebo group Infection: 12/97 vs 9/99 in placebo group SAEs: 7/9 vs 8/99 in placebo group
NCT04847232	Recurrent serum K > 5.5 mmol/L in adults thrice weekly HD patients. Exclusion: -Pregnancies -Arrhythmias within 7 days before screening N= 2690 -Age (yr)=56 -Men (%)= NS Countries=Argentina, Austria, Brazil, Bulgaria, Canada, China, Czechia, Germany, Hungary, Italy, Japan, Malaysia, Mexico, Peru, Poland, Russian Federation, Slovakia, Spain, Taiwan, Thailand, Turkey, Ukraine, the UK, the USA and Vietnam	SZC (5-10mg)   N= 1349	Placebo   N=1341		Delta K	Not reported	Although the RCT results were completed, data are published but available partially on clinicaltrial.gov
					hyperkalemia events*	125 in SZC group vs 280 in placebo group
					AEs	GI disorders: 251 vs 260 in placebo group Infection: 295 vs 252 in placebo group SAEs: 7/9 vs 8/99 in placebo group
NCT04217590	China adults thrice weekly HD patients with preHD K >5.4 mmol/L Exclusion: Pregnancies -N= 134 -Age (yr)=55 -Men (%)= 49.3%	SZC (5-10 mg)   N=67	Placebo   N=67	11 weeks	Delta K	Mean difference between SZC and Placebo groups: –0.65 mmol/L (95% CI, –0.81 to –0.48; P < 0.001).
					hyperkalemia events*	40/67  in SZC group vs 56/67 in placebo group
					AEs	No severe hyperkalemia has been registered hypokalemia 2/67 in SZC group vs 0/67 in placebo group AEs 42/67 in SZC group vs 44/67 in placebo group SAEs 6/67 in SZC group vs 8/67 in placebo group Infection 0 in SZC group vs 1/67 in placebo group Death 0 in SZC group vs 1/67 in placebo group GI 6/67 in SZC group vs 9/67 in placebo group
SNCTP000003912	Swiss adults thrice weekly HD patients Exclusion: Pregnancies -N= 51 -Age (yr)=66 -Men (%)= 73% – Caucasian= 82% – Countries= Switzerland (multicentric)	Patiromer (8.4mg/die)   N=51	Placebo   N=51	8 weeks	Delta K	0.15+0.16 mmol//l in patiromer group vs 0.32+0.06 in placebo group	unblinded two-arm crossover RCTs
					hyperkalemia events*	17% in patiromer group vs 34% in placebo group
					AEs
SNCTP000003912	Swiss adults thrice weekly HD patients Exclusion: Pregnancies -N= 51 -Age (yr)=66 -Men (%)= 73% – Caucasia= 82% -Countries= Switzerland (multicentric)	Patiromer (8.4mg/die)     N=51	SPS (15g)   N=51	8 weeks	Delta K	0.15+0.16 mmol//l in patiromer group vs  -0.3+0.05 in SPS group	unblinded two-arm crossover RCTs
					hyperkalemia events*	17% in patiromer group vs  12% in SPS group
					AEs	GI 26% in Patiromer group vs 24% in SPS group

Table 1. Summary of main characteristics and findings of the RCTs reviewed.

Study quality and risk of bias

Random sequence generation and allocation concealment were detailed in all trials. Three RCTs were double-blind, and one was open-label [15]. Blinding of participants, investigators and outcome assessors was specified in all studies. Attrition bias was low in all studies. The overall dropout rate was lower than 10 % for each study. Reporting bias was high in all studies. No other potential sources of bias were observed in the majority of studies (Figure 2).

Outcome data

Mortality

The three studies that compared SZC to placebo reported 114 deaths in the placebo group and 120 in the SZC group, and data on mortality were available.

Differences in potassium level

Data on pre-HD potassium value as a continuous variable were reported by 3 studies [12, 14, 15]. a pooled analysis involving 432 patients, both SPS and SZC showed a significant reduction in serum potassium levels compared to placebo, with a mean difference of –0.6200 (95% CI: [–0.89 / –0.35]; p < 0.0001) and –0.68 (95% CI: [–0.81 / –0.55]; p < 0.0001), respectively. Conversely, no differences between Patiromer and placebo were found (mean difference -0.17, 95%CI [-0.41/0.07], p=0.16). This collective analysis was not affected by heterogeneity (I²=0%) (Figure 3). The GRADE quality of this analysis was low after downgrading for the small number of the included studies.

Similarly, a meta-analysis comparing the new agent to the old agent/placebo showed a significant reduction of serum potassium in the new binders group (-0.47, 95%CI [-0.91/-0.02], p=0.04), but it revealed the highest heterogeneity (I²=94%).

Hyperkalemia events

Hyperkalemia has been reported in all studies. In this case, differences have been found in each group compared to placebo (Patiromer -0.791, 95%CI:[-1-0.078], p=0.030; SPS -0.998, 95%CI [-1-0.250], p=0.009; SZC  -0.620, 95%CI [-1-0.220], p=0.002). According to these results, all treatments can avoid pre-dialysis hyperkalemia (K>5.5 mmol/L) compared to placebo (Figure 4). However, it revealed the highest heterogeneity (I²=99%).

Adverse Events

Adverse events cannot be statistically analyzed since they are not reported in each study and, where reported, as non-aggregable data. However, no difference has been found for total AEs in the included studies. The same can be assessed either for gastrointestinal events, for infectious or for SAEs. They are detailed in Table 1.

 

Discussion

Our analysis showed a significant difference in the management of hyperkalemia for each potassium binder compared to the only dialytic treatment with a lower pre-HD potassium concentration for SZC and SPS, without significant hypokalemia recurrence.  At the same time, the recurrence of AEs not seem to be higher than only dialytic treatment.

Intradialytic hyperkalemia has been related to mortality and comorbidity, as well as hyperkalemia in conservative CKD [17–19]. Furthermore, the worst management of potassium worsens the intradialytic cardiovascular risk due to the speediest correction, almost obliged in patients with kalemia higher than 6.5 mmol/L. Although diet education, disionemia is not easy to manage in dialysis patients, often due to the lack of alimentary compliance [20].

The higher reduction of serum potassium in SZC patients than in patiromer patients can be easily explained by the speedier efficacy of the first. Indeed, SZC results are speedier in the first 48h than patiromer, and serum potassium in dialysis patients needs management into the 48/72h without dialysis [21]. This difference in speed could be crucial in patients with acute hyperkalemia or those requiring rapid potassium normalization for urgent medical procedures.

As reported in the published protocol, highlighting possible gaps in actual knowledge was one of the aims and this systematic review revealed an important gap: the lack of sufficient RCTs. The crucial relevance of knowing the efficacy and the safety of these new drugs for clinicians should be juxtaposed with enough structured and dialysis-based RCTs. In particular, the lack of studies directly comparing the different potassium binders with each other makes it difficult to establish which is the optimal choice for dialysis patients. Future studies should focus on direct comparisons to provide more robust evidence.

The lack of a sufficient number of RCTs has been reported since the first decade of this century, highlighting the phenomenon of the “Invisible trial” [22]. Even though the Restoring Invisible & About palatability and adherence, no data available from RCTs exist even though the APPETIZE study ended in 2022 [23]. Abandoned Trials (RIAT) initiative on Cochrane evidence products showed a possible solution to this problem, trials for smaller populations are often not enough to generate evidence. Furthermore, the variability of dialysis patient populations, with different comorbidities and dialysis regimens, makes it difficult to generalize the results of existing studies. A personalized approach to hyperkalemia management is needed, taking into account the specific needs of each patient.

Furthermore, some studies compared the cost-effectiveness of new and old potassium binders, with evidence of slightly lower cost in the new K binders, often due to the reduction of hospitalisation [24, 25].

Nowadays, the small number of performed trials on dialysis patients represents the major limit of this systematic review. Indeed, the lack of RCT on hemodialysis did not allow us to give a clear opinion about the use of these drugs on dialysis patients, preventing us from fully understanding if there is one potassium binder more efficient or secure. It is a serious gap in our acknowledgement that needs to be solved. Also, hypokalemia, one of the major adverse events of these drugs, is not reported enough to compute a real risk. The lack of detailed data on adverse events, particularly hypokalemia, limits our ability to fully assess the safety profile of these drugs. Future studies should include a systematic assessment of adverse events, with particular attention to hypokalemia and its clinical consequences. One of the limitations of this systematic review is the highest heterogeneity, perhaps due to the small sample, the small number of included studies and the heterogeneity of the comparisons (SPS and Placebo).

 

Conclusion

In conclusion, according to our analysis, all potassium binders seem to have more efficacy than placebo and seem to have high safety. However, the knowledge gap cannot be solved due to the need for more RCTs, paying attention that all outcomes are evaluated and reported. Furthermore, it is essential that future studies focus on evaluating the long-term impact of these drugs on the quality of life of dialysis patients and on cardiovascular outcomes. The management of hyperkalemia must be integrated into a comprehensive approach to the care of the dialysis patient, taking into account their specific needs and comorbidities.

 

Bibliography

1. Thomsen RW, Nicolaisen SK, Hasvold P, Sanchez RG, Pedersen L, Adelborg K, Egstrup K, Egfjord M, Sørensen HT. Elevated potassium levels in patients with chronic kidney disease: occurrence, risk factors and clinical outcomes-a Danish population-based cohort study. Nephrol Dial Transplant. 2018 Sep 1;33(9):1610-1620. https://doi.org/10.1093/ndt/gfx312. PMID: 29177463.
2. Vifor Pharma. Patiromer US Prescribing Information 2016.
3. PubChem. Sodium zirconium cyclosilicate. Available at: https://pubchem.ncbi.nlm.nih.gov/compound/91799284#section=Top (Accessed October 2019).
4. Stavros F, Yang A, Leon A, Nuttall M, Rasmussen HS. Characterization of structure and function of ZS-9, a K+ selective ion trap. PLoS One. 2014 Dec 22;9(12):e114686. https://doi.org/10.1371/journal.pone.0114686. PMID: 25531770; PMCID: PMC4273971.
5. Palmer BF, Carrero JJ, Clegg DJ, Colbert GB, Emmett M, Fishbane S, Hain DJ, Lerma E, Onuigbo M, Rastogi A, Roger SD, Spinowitz BS, Weir MR. Clinical Management of Hyperkalemia. Mayo Clin Proc. 2021 Mar;96(3):744-762. https://doi.org/10.1016/j.mayocp.2020.06.014.
6. Carvalho PEP, Veiga TMA, Lacerda H, Tofanelli MR, Gewehr DM, Nunes MCP, Simões E Silva AC. Safety and efficacy of new potassium binders on hyperkalemia management in patients with heart failure: a systematic review and meta-analysis of randomized controlled trials. Clin Res Cardiol. 2023 Jul;112(7):991-1002. https://doi.org/10.1007/s00392-023-02215-2.
7. Page, M. J. et al. PRISMA 2020 explanation and elaboration: updated guidance and exemplars for reporting systematic reviews. BMJ 372, n160 (2021). https://doi.org/10.1136/bmj.n160.
8. Maressa V, Farina A, Cernaro V, Gembillo G, Longhitano E, Taverna G, Santoro D, Calabrese V. Comparison Among Potassium Binders on the Management of Hyperkalemia on Chronic Dialysis Patients: A Protocol for Systematic Review. G Ital Nefrol. 2025 Feb 28;42(1):2025-vol1. https://doi.org/10.69097/42-01-2025-02. PMID: 40026221.
9. Available from https://www.crd.york.ac.uk/prospero/display_record.php?RecordID=608049
10. Hamilton: McMaster University (developed by Evidence Prime, Inc.). Gradepro GDT. GRADEpro Guideline Development Tool [www.guidelinedevelopment.org]. Hamilton: McMaster University (developed by Evidence Prime, Inc.), 2015
11. Available from https://clinicaltrials.gov/study/NCT03781089 (last accessed on 26 March 2025)
12. Fishbane S, Ford M, Fukagawa M, McCafferty K, Rastogi A, Spinowitz B, Staroselskiy K, Vishnevskiy K, Lisovskaja V, Al-Shurbaji A, Guzman N, Bhandari S. A Phase 3b, Randomized, Double-Blind, Placebo-Controlled Study of Sodium Zirconium Cyclosilicate for Reducing the Incidence of Predialysis Hyperkalemia. J Am Soc Nephrol. 2019 Sep;30(9):1723-1733. https://doi.org/10.1681/ASN.2019050450. Epub 2019 Jun 14. PMID: 31201218; PMCID: PMC6727265.
13. Fishbane S, Jadoul M, Dember L, Kovesdy CP, Al-Shurbaji A, Lisovskaja V, Sekar P, Katona B, Guzman N, Herzog C. Evaluation of the effect of sodium zirconium cyclosilicate on arrhythmia-related cardiovascular outcomes in patients receiving chronic haemodialysis with hyperkalaemia: protocol for the multicentre, randomised, controlled DIALIZE-Outcomes study. BMJ Open. 2023 May 25;13(5):e071309. https://doi.org/10.1136/bmjopen-2022-071309. PMID: 37230521; PMCID: PMC10230973.

14. Ni Z, Lu R, Xu X, Bian X, et al; DIALIZE China Study Group. DIALIZE China: A Phase IIIb, Randomized, Placebo-Controlled Study to Reduce Predialysis Hyperkalemia With Sodium Zirconium Cyclosilicate in Chinese Patients. Clin Ther. 2023 Jul;45(7):633-642. https://doi.org/10.1016/j.clinthera.2023.04.014. Epub 2023 Jun 27. PMID: 37385905.
15. Jaques DA, Stucker F, Ernandez T, Alves C, Martin PY, De Seigneux S, Saudan P. Comparative efficacy of patiromer and sodium polystyrene sulfonate on potassium levels in chronic haemodialysis patients: a randomized crossover trial. Clin Kidney J. 2022 May 10;15(10):1908-1914. https://doi.org/10.1093/ckj/sfac129. PMID: 36158152; PMCID: PMC9494516.
16. Palmer BF, Clegg DJ. Physiology and Pathophysiology of Potassium Homeostasis: Core Curriculum 2019. Am J Kidney Dis. 2019 Nov;74(5):682-695. https://doi.org/10.1053/j.ajkd.2019.03.427. Epub 2019 Jun 19. Erratum in: Am J Kidney Dis. 2022 Nov;80(5):690. https://doi.org/10.1053/j.ajkd.2022.09.002. PMID: 31227226.
17. Calabrese V, Cernaro V, Battaglia V, Gembillo G, Longhitano E, Siligato R, Sposito G, Ferlazzo G, Santoro D. Correlation between Hyperkalemia and the Duration of Several Hospitalizations in Patients with Chronic Kidney Disease. J Clin Med. 2022 Jan 4;11(1):244. https://doi.org/10.3390/jcm11010244. PMID: 35011985; PMCID: PMC8746076.
18. Calabrese V, Tripepi GL, Santoro D. Impact of hyperkalemia in length of hospital stay in dialysis-dependent patients. Ther Apher Dial. 2022 Oct;26(5):1050-1051. https://doi.org/10.1111/1744-9987.13847. Epub 2022 Apr 11. PMID: 35366057.
19. Calabrese V, Gembillo G, Longhitano E, La Placa C, Ferio C, Cernaro V, Santoro D. Timing of the CKD Complications: A Longitudinal Analysis. G Ital Nefrol. 2024 Dec 23;41(6):2024-vol6. https://doi.org/10.69097/41-06-2024-10. PMID: 39931968.
20. Bayram R, Budak S, YIldIz H. The effect of non-compliance with diet and liquid restriction on fatigue in dialysis patients. Rev Esc Enferm USP. 2024 May 24;58:20230251. https://doi.org/10.1590/1980-220X-REEUSP-2023-0251en. PMID: 38808906; PMCID: PMC11135046.
21. Meaney CJ, Beccari MV, Yang Y, Zhao J. Systematic Review and Meta-Analysis of Patiromer and Sodium Zirconium Cyclosilicate: A New Armamentarium for the Treatment of Hyperkalemia. Pharmacotherapy. 2017 Apr;37(4):401-411. https://doi.org/10.1002/phar.1906. Epub 2017 Mar 10. PMID: 28122118; PMCID: PMC5388568.
22. Feldman AM. Publishing “Invisible” and “Abandoned” clinical trials: a commitment for CTS. Clin Transl Sci. 2013 Aug;6(4):251-3. https://doi.org/10.1111/cts.12087. PMID: 23919357; PMCID: PMC5350827.
23. Available from https://clinicaltrials.gov/study/NCT04566653?cond=appetize&rank=1 (last accessed on 15 May 2025).
24. Tian L, Fu S, Li M, Zhao X, Li H. Cost-effectiveness analysis of sodium zirconium cyclosilicate for treating hyperkalemia among Chinese patients. Front Public Health. 2023 Dec 7;11:1196789. https://doi.org/10.3389/fpubh.2023.1196789. PMID: 38145082; PMCID: PMC10740179.
25. Fischer B, Serra A, Telser H. Cost-Effectiveness of Treating Patients with Chronic Kidney Disease and Prior Hyperkalemia with Renin-Angiotensin-Aldosterone System Inhibitor and Patiromer: A Swiss Public Healthcare Perspective. Adv Ther. 2022 Jun;39(6):2717-2730. https://doi.org/10.1007/s12325-022-02123-3. Epub 2022 Apr 13. PMID: 35416597; PMCID: PMC9122858.