Cerebral vasoreactivity in HeartMate 3 patients.

BACKGROUND

While rates of stroke have declined with the HeartMate3 (HM3) continuous- flow (CF) left ventricular assist device (LVAD), the impact of non-pulsatile flow and artificial pulse physiology on cerebrovascular function is not known. We hypothesized that improved hemodynamics and artificial pulse physiology of HM3 patients would augment cerebrovascular metabolic reactivity (CVR) compared with HeartMate II (HMII) CF-LVAD and heart failure (HF) patients.

METHODS

Mean, peak systolic and diastolic flow velocities (MFV, PSV, MinFV, respectively) and cerebral pulsatility index were determined in the middle cerebral artery (MCA) before and after a 30 sec breath-hold challenge in 90 participants: 24 healthy controls; 30 HF, 15 HMII, and 21 HM3 patients.

RESULTS

In HM3 patients, breath-holding increased MFV (Δ8 ± 10 cm/sec, p < .0001 vs baseline) to levels similar to HF patients (Δ9 ± 8 cm/sec, p > .05), higher than HMII patients (Δ2 ± 8 cm/sec, p < .01) but lower than healthy controls (Δ13 ± 7 cm/sec, p < .05). CF-LVAD altered the proportion of systolic and diastolic flow responses as reflected by a differential cerebral pulsatility index (p = .03). Baseline MFV was not related to CVR (r2 = 0.0008, p = .81). However, CF-LVAD pump speed was strongly inversely associated with CVR in HM II (r2 = 0.51, p = .003) but not HM3 patients (r2 = 0.01, p = .65).

CONCLUSIONS

Compared with HMII, HM3 patients have a significantly improved CVR. However, CVR remains lower in HM3 and HF patients than in healthy controls, therefore suggesting that changes in cerebral hemodynamics are not reversed by CF-LVAD therapy. Further research on the mechanisms and the long-term impact of altered cerebral hemodynamics in this unique patient population are warranted.