Impact of increased plasma potassium levels on biventricular pacing: a prespecified analysis from a randomised clinical trial (POTCAST)

Background

Optimal cardiac resynchronization therapy (CRT) requires near-continuous biventricular pacing. Loss of biventricular pacing is associated with an increased risk of heart-failure hospitalization and death. Low-normal potassium levels promote atrial and ventricular ectopy and may impair pacing delivery.

Purpose

To assess whether an intervention aimed at increasing plasma potassium levels toward high-normal levels prevents episodes of suboptimal biventricular pacing <90%.

Methods

In this prespecified analysis of the randomised POTCAST trial, participants with a CRT-defibrillator (CRT-D) were included. POTCAST randomised participants in a 1:1 ratio to potassium-targeted therapy in addition to standard care or standard care alone. The intervention included dietary guidance and a) mineralocorticoid receptor antagonists (MRAs), b) potassium supplements, or c) a combination of both, targeting a plasma potassium level between 4.5 and 5.0 mmol/L.

The primary outcome was time to first drop in biventricular pacing <90%, analysed using Cox proportional hazards regression. The secondary outcome was the association between measured plasma potassium levels and risk of biventricular pacing <90%, analysed using generalized additive logistic models applied to paired plasma potassium and biventricular pacing measurements obtained at biannual follow-up visits.

Results

Among 230 participants (mean age 65 ± 10 years, 22% female) treated with CRT in the POTCAST trial, 119 were assigned to the intervention group and 111 to the control group. Baseline characteristics, including plasma potassium levels, were well balanced between groups (Table 1). At biannual visits, mean plasma potassium levels in the intervention group were 0.20-0.27 mmol/L higher than in controls (all p<0.015).

The risk of suboptimal biventricular pacing was significantly reduced in the intervention group compared with controls (HR 0.48, 95% CI 0.25-0.92; p=0.026) (Figure 1).

Across plasma potassium levels between 3.4 and 5.0 mmol/L, higher potassium levels were associated with a lower probability of biventricular pacing <90% (p=0.0004).

Conclusions

In patients with CRT, increasing plasma potassium levels toward high-normal concentrations prevented loss of biventricular pacing.Table 1