Transvenous Phrenic Nerve Stimulation is Safe and Effective in Patients with Heart Failure and Cheyne-stokes Respiration

Abstract

Introduction The most common type of central sleep apnea (CSA) breathing pattern in patients with heart failure (HF) is Cheyne-Stokes respiration (CSR). CSR manifests as a cyclical pattern of apnea and hyperventilation, with absence of airflow and respiratory effort followed by hyperventilation in a crescendo-decrescendo pattern. CSR may be a predictor of negative clinical outcomes in HF patients. Purpose Examine the response to transvenous phrenic nerve stimulation (TPNS) in patients with HF and different amounts of CSR. Methods Patients with CSA and HF who participated in the remedē System Pivotal Trial were divided into clinically relevant subgroups based on percentage of sleep in CSR (%CSR: 50%) during their baseline polysomnogram. The subgroups were analyzed for efficacy and safety by randomized group (TPNS vs control [no TPNS]). Epworth sleepiness scale (ESS) was also collected. Results At baseline, higher percentage CSR subgroups had lower left ventricular ejection fraction (median 45% for CSR 50%). Baseline apnea-hypopnea index (AHI) indicated severe CSA in all subgroups. All CSR subgroups receiving TPNS therapy had more than 50% of patients achieve ≥50% reduction in AHI at 6 months (see Table); control response rates were ≤11% in each CSR subgroup. The AHI decreased by ≥21 events/hour in each CSR subgroup on therapy while Control subgroups had little or no improvement. The central apnea index decreased to median ≤3 events/hour in each TPNS CSR subgroup; each control subgroup had median 22-24 events/hour remaining at 6 months. Baseline daytime sleepiness was worse in lower %CSR subgroups, however improved in all subgroups. No apparent relationship existed between %CSR and cardiovascular death in patients using TPNS therapy through 6 months, with the only death being in the 20-50% CSR subgroup; the death was determined not related to device, therapy or procedure by the clinical events committee. Further, 2/22 control subjects with CSR>50% experienced a cardiovascular death compared to 0/12 in the TPNS treated subgroup. Conclusions Transvenous phrenic nerve stimulation can treat CSA in HF patients regardless of the percentage of sleep in CSR and no signal of increased mortality was observed.