Detailed safety analysis of DIAMOND trial: ‘primum non nocere’?

This commentary refers to ‘Patiromer for the management of hyperkalemia in heart failure with reduced ejection fraction: the DIAMOND trial’, by J. Butler et al., https://doi.org/10.1093/eurheartj/ehac401 and the discussion piece ‘Carat, clarity, colour, and cut: grading the DIAMOND trial’, by J. Butler et al., https://doi.org/10.1093/eurheartj/ehad469. Mineralocorticoid receptor antagonists (MRAs) are beneficial in heart failure with reduced ejection fraction (HFrEF), but implementation is suboptimal due to the perceived risk of hyperkalaemia. To address this issue, the DIAMOND study1 investigated the effects of patiromer, a potassium binder, on potassium levels in HFrEF patients who had a potassium >5.0 mmol/L or had reduced or discontinued Renin Angiotensin Aldosterone System inhibitors (RAASi) or MRA due to hyperkalaemia. The authors found that patiromer lowered potassium compared to placebo. Although the trial was initially designed to assess ‘hard’ cardiovascular (CV) outcomes, it ultimately evaluated surrogate endpoints such as the difference in mean change in potassium or MRA discontinuation between groups. In the accompanying editorial,2 several concerns were raised regarding the clinical impact of this strategy, estimating a number needed to treat (NNT) > 400 for CV death or heart failure (HF) hospitalization. However, it is crucial to perform a detailed safety analysis when evaluating a drug that treats the side effects of another drug. Therefore, it is important to examine the incidence of hypokalaemia, a side effect specifically caused by patiromer. DIAMOND revealed an absolute increased risk of 4.3% (odds ratio 1.49, 95% CI 0.99–2.2) in investigator-reported hypokalaemic events (not based on laboratory cut-offs). This was not surprising, given that in the (PEARL)-HF3 study, which studied patiromer in a less severe HF population [mean left ventricular ejection fraction (LVEF) 40%], the drug was associated with a significant increase in hypokalaemia (6% vs. 0% for <3.5 mmol/L and 47% vs. 6% for <4.0 mmol/L). In DIAMOND, most hypokalaemic episodes were considered ‘mild’. However, unlike the PEARL-HF study, there is no information on the incidence of potassium levels <4.0 mmol/L or <3.5 mmol/L in this challenging population. Hypokalaemia is linked to major CV events in HFrEF and is possibly more dangerous than hyperkalaemia, especially in those who are ‘protected’ from bradycardia by implantable cardioverter-defibrillator (ICD) and/or cardiac resynchronization therapy (CRT) devices.4 Potassium slightly below 4.0 mmol/L is associated with poor outcomes in patients with severe LV dysfunction (as in DIAMOND, with a mean LVEF of 33.5%). Additionally, in DIAMOND, few patients were using angiotensin receptor-neprilysin inhibitor that are associated with a lower risk of hyperkalaemia compared to angiotensin-converting enzyme inhibitors5; also, patients were up titrated to 50 mg of MRA, a dose not commonly used in a real-world setting and that might reduce the incidence of patiromer-induced hypokalaemia. Therefore, the actual risk of hypokalaemia may be underestimated, and the authors are encouraged to provide a detailed analysis of these data. Finally, it is important to note that DIAMOND showed a numerical increase in all-cause death and CV hospitalizations in the intervention group (49 vs. 39 events), but no data are available regarding sudden cardiac death, arrhythmic events, or ICD shocks. In summary, patiromer is useful for managing potassium levels in chronic kidney disease due to the clinical significance of hyperkalaemia. However, in HFrEF patients, the ability to regulate potassium levels in those taking MRAs should be weighted considering the potential for inducing hypokalaemia. The DIAMOND trial was underpowered to evaluate hard outcomes, but do we have enough safety data to proceed with this strategy in patients with severe LV dysfunction?