Neurological Dysfunction During Contemporary Magnetically Levitated Left Ventricular Assist Device Support for...

The burden of neurological dysfunction during contemporary magnetically levitated (Mag-Lev) left ventricular assist device (LVAD) support for advanced heart failure is uncertain.

The objectives were to determine the incidence, prognosis, and risk factors of stroke and nonstroke neurological adverse events (NAEs) during Mag-Lev LVAD support.

A retrospective cohort study using the multicenter U.S.-based Society of Thoracic Surgeons INTERMACS database was performed from January 1, 2017, to March 31, 2023. The study population comprised patients with heart failure on Mag-Lev LVAD support. The primary outcome was survival on Mag-Lev LVAD support. Multivariable Cox regression was used, with stroke and nonstroke NAE status as a time-varying covariate to determine the risk of death in comparison to no NAE. Separate multivariable Cox regression models were built to determine factors associated with death after stroke and nonstroke NAEs and to determine baseline clinical variables associated with the occurrence of stroke and nonstroke NAE.

Within 11,739 patients, 879 (8%) had a stroke, 972 (8%) experienced a nonstroke NAE, and 9,888 (84%) did not have a NAE over a median of 1.4 years (Q1-Q3: 0.5-2.7 years). At 2 years, survival on Mag-Lev LVAD support after stroke and nonstroke NAE was 45% and 56%, respectively (P < 0.001). In comparison to no NAE, the risk of death was elevated after stroke (HR: 7.3) and nonstroke NAE (HR: 5.1). Older age (HR: 1.02 per year), pre-LVAD implantable cardioverter-defibrillator (HR: 1.32), higher pre-LVAD blood urea nitrogen level (HR: 1.01 per 1 mg/dL), and pre-LVAD dialysis (HR: 1.91) were significantly associated with death after stroke, whereas higher pre-LVAD body mass index (HR: 1.02 per 1 unit), pre-LVAD serum creatinine (HR: 1.36 per 1 mg/dL), pre-LVAD international normalized ratio (HR: 1.14 per 1 unit), and lower pre-LVAD lower albumin (HR: 0.98 per 1 g/dL) were significantly related to mortality after nonstroke NAE. Pre-LVAD extracorporeal membrane oxygenation (stroke HR: 1.70; nonstroke HR: 1.45) and concomitant surgery (stroke HR: 1.23; nonstroke HR: 1.17) were associated with the occurrence of NAEs.

Stroke and nonstroke NAEs are incurred by a similar proportion of patients, and both are associated with substantially reduced survival. Consequently, stroke and nonstroke NAEs should be given comparable weight with universal nomenclature when evaluating the safety of durable LVADs. Pre-LVAD extracorporeal membrane oxygenation and concomitant surgery are potentially modifiable targets to reduce neurological dysfunction during Mag-Lev LVAD therapy.