Cerebral Hemodynamics With TCD in ECMO

First Author(s)

• 

  GIORGIO CATUREGLI, BS,

  n/a

  Disclosure information not submitted.

  Slides and Audio

Co-Author(s)

• CC

  Chun (Dan) Woo Choi, MD

  Physician
  Johns Hopkins Hospital, United States

  Disclosure information not submitted.
• RG

  Romergryko Geocadin, MD

  Professor
  Johns Hopkins Hospital, United States

  Disclosure information not submitted.
• GW

  Glenn Whitman, MD

  Professor
  Johns Hopkins Hospital, United States

  Disclosure information not submitted.
• SC

  Sung-Min Cho, DO, MHS

  Assistant Professor
  Johns Hopkins Hospital, United States

  Disclosure information not submitted.
• WZ

  Wendy Ziai, MD, FRCPC

  Professor
  Johns Hopkins University
  Baltimore, Maryland, United States

  Disclosure information not submitted.

Title: Assessing Cerebral Hemodynamics by TCD in ECMO: Lack of Pulsatility Implicated in Neurologic Injury

Introduction: Extracorporeal membrane oxygenation (ECMO) is an increasingly common method of mechanical circulatory support for medically refractory pulmonary and/or cardiac failure. Acute brain injury, (stroke and brain death) is a likely underestimated but common (7-15%) and highly fatal (68-89%) complication. Transcranial Doppler (TCD) ultrasound describes real-time regional cerebral blood flow velocities and microembolic signals (MES), but their clinical significance in ECMO remain limited.

Methods: Adult ECMO patients at a tertiary care center from 11/2017 to 2/2021 were included. Demographics, medical history, clinical outcomes, and physiologically relevant hemodynamic and laboratory findings were collected. Serial TCD studies (day 1, 3-5, and 7-10) assessed mean flow velocities (MFV) and pulsatility indices (PI). The middle cerebral arteries were monitored for MES for 30 minutes of each TCD.

Results: 100 ECMO patients underwent 192 TCDs; 71 (71%) patients received VA-ECMO while 29 (29%) received VV-ECMO. The study cohort was 58% male with median age of 55 years (IQR: 42 - 66). Compared to a published normal cohort, ECMO patients exhibited similar MFVs in all vessels except the internal carotid artery, where MFV was slightly higher in ECMO patients. No significant differences in MFV were observed in any vessel by ECMO type or use of intra-aortic balloon pump. MFVs were not significantly associated with Fick’s cardiac output, cardiac index, ECMO flow or speed. Presence of a recordable PI in any insonated vessel was strongly associated with ECMO mode: 43% in VA (without additional support) vs. 97% in VV (p < 0.001). Absence of recordable PI was associated with intracerebral hemorrhage: 11% vs. 0% (p=0.057). Presence of MES was strongly associated with ECMO mode (40% in VA vs. 6% in VV, p< 0.001) but neither presence nor number of MES was significantly correlated with clinical outcome. MES were more frequently observed in the presence of fibrin, but not clot, in the ECMO circuit: 35% vs. 12%, p=0.02.

Conclusions: TCD reliably captures MFV in both VA- and VV-ECMO, with similar values to published normal ranges. MES remain as yet of uncertain clinical significance. PIs are recorded less frequently in VA-ECMO, and the absence of pulsatile blood flow is associated with a higher rate of intracerebral hemorrhage.