COMPARATIVE EFFECTIVENESS AND SAFETY OF POTASSIUM BINDERS FOR HYPERKALEMIA MANAGEMENT IN DIALYSIS PATIENTS: A SYSTEMATIC REVIEW AND NETWORK META-ANALYSIS

Hyperkalemia is a frequent and serious complication in patients with End-Stage Renal Disease (ESRD) despite receiving maintenance haemodialysis. Hyperkalemia has been associated with arrhythmias and mortality due to adverse cardiovascular events, yet management options remain challenging as therapy must correct serum potassium abnormalities while aligning with the overall management of ESRD. Recently, gastrointestinal potassium binders have emerged as adjunct therapy, but evidence in dialysis populations remains limited and inconsistent. Hence, we conducted a network meta-analysis (NMA) to evaluate the comparative effectiveness and safety of patiromer (PAT), sodium zirconium cyclosilicate (SZC), calcium polystyrene sulfonate (CPS), and sodium polystyrene sulfonate (SPS) in dialysis patients.

This NMA (PROSPERO CRD420251161103) followed PRISMA-NMA 2015 guidelines. PubMed, Embase, Cochrane CENTRAL, Web of Science, CINAHL and Scopus were searched to October 8, 2025. Randomized and non-randomized studies evaluating PAT, SZC, SPS or CPS in adult dialysis patients with hyperkalemia were included. Risk of bias was assessed using RoB-2, and certainty of evidence using GRADE-NMA. Outcomes measured were mean change in serum potassium, proportion achieving normokalemia, and total adverse events (TAEs). A frequentist random-effects NMA in R (v4.4.2), estimated pooled MDs or ORs (95% CIs). Consistency was assessed using node-splitting and design-by-treatment interaction models. Treatment rankings were based on the surface under the cumulative ranking curve (SUCRA) probabilities.

Nine randomized controlled trials (RCTs) were included, comparing six treatment strategies: Placebo/Standard of Care (SOC), SZC, PAT, SPS, CPS, and SZC combined with Dialysate potassium 3mEq/L(DK3), with sample sizes ranging from 15 to 1348 per arm. For normokalemia achievement(8 RCTs), SPS achieved the highest odds (OR 14.13, p = 0.013), followed by SZC (OR 6.26; p = 0.0003) and PAT (OR 5.40; p = 0.009), whereas SZC + DK3 was less effective (OR 0.12; p = 0.018). Moderate heterogeneity was observed (τ² = 0.52, I² = 69%), with no between-design inconsistency. The network plot is shown in Figure 1. The league table summarising all pairwise comparisons is presented in Figure 2. For TAEs (9 RCTs), PAT and SPS showed a higher risk than Placebo/SOC (OR 2.49, p = 0.010; OR 3.90, p = 0.0003, respectively), while SZC demonstrated the most favourable safety (OR 0.98, p=0.82). No significant heterogeneity or inconsistency was noted (I² = 0%). For mean potassium change(9 RCTs). SZC showed the largest numerical reduction (MD −0.47, p = 0.054); however, considerable heterogeneity existed (τ² = 0.2013, I² = 97.5%), and none were statistically significant. Figure 3 presents the P-score rankings across the three outcomes.

Fig. 1: Network Geometry Plot for proportion achieving  normokalemia

Fig. 2: League Table (OR [95% CI], Random-Effects Model) for proportion achieving  normokalemia

Fig. 3: P-score rankings of interventions across three outcomes

In dialysis patients, SPS and SZC were the most effective at achieving normokalemia, while SZC was the safest. However, findings for mean potassium change were heterogeneous and inconsistent, limiting their generalisability for clinical practice. This highlights the need for further evidence in dialysis patients to strengthen these findings.