Continuous monitoring of cerebral blood flow during general anaesthesia in infants

Abstract

Background: General anaesthesia is associated with neurocognitive deficits in infants after noncardiac surgery. Disturbances in cerebral perfusion as a result of systemic hypotension and impaired autoregulation may be a potential cause. Our aim was to study cerebral blood flow (CBF) velocity continuously during general anaesthesia in infants undergoing noncardiac surgery and compare variations in CBF velocity with simultaneously measured near-infrared spectroscopy (NIRS), blood pressure, and heart rate.

Methods: NeoDoppler, a recently developed ultrasound system, was used to monitor CBF velocity via the anterior fontanelle during induction and maintenance of general anaesthesia until the start of surgery, and during recovery. NIRS, blood pressure, and heart rate were monitored simultaneously and synchronised with the NeoDoppler measurements.

Results: Thirty infants, with a median postmenstrual age at surgery of 37.6 weeks (range 28.6-60.0) were included. Compared with baseline, the trend curves showed a decrease in CBF velocity during induction and maintenance of anaesthesia and returned to baseline values during recovery. End-diastolic velocity decreased in all infants during anaesthesia, on average by 59%, whereas peak systolic- and time-averaged velocities decreased by 26% and 45%, respectively. In comparison, the reduction in mean arterial pressure was only 20%. NIRS values were high and remained stable. When adjusting for mean arterial pressure, the significant decrease in end-diastolic velocity persisted, whereas there was only a small reduction in peak systolic velocity.

Conclusions: Continuous monitoring of CBF velocity using NeoDoppler during anaesthesia is feasible and may provide valuable information about cerebral perfusion contributing to a more targeted haemodynamic management in anaesthetised infants.

Keywords: brain injury; cerebral perfusion; general anaesthesia; infants; multimodal monitoring.

Fig. 1

Continuous trend curves displaying multimodal monitoring during anaesthesia. The figure shows characteristic patterns of cerebral blood flow velocity (CBF velocity) and blood pressure, near-infrared spectroscopy (NIRS) and heart rate in one infant with gastrointestinal atresia, gestational age 37 weeks. There is a decrease in CBF velocity during induction (P2). After intubation (P3), the CBF velocity stabilises at a level below baseline and remains stable until surgery commences (P4). During this period, from P3 to P4, NIRS is stable at high values (95%), whereas invasive mean arterial pressure (MAP) measurements vary between 25 and 43 mm Hg. CBF velocities return to baseline values during recovery, whereas blood pressure stabilises at a higher level than baseline. During recovery, NIRS is still high, but lower than during anaesthesia (P3–P4), and comparable with baseline values. NIRS sensor for infants and neonates, INVOS™ 5100c OxyAlert™ (Medtronic Parkway, Minneapolis, MN, USA). Synchronising of the monitoring variables was done using pyMIND (https://pymind.readthedocs.io/en/latest/), an open-source Python-based software designed to acquire and integrate multi-modal scientific data from medical devices. The software was modified in-house and extended to capture NIRS data in addition to invasive measurements (e.g. Philips IntelliVue). On the receiver side a common time axis was maintained to which all measurements were synchronised. The highest available temporal resolution was captured for all signals and all data were saved into HDF format for offline processing. EDP, end-diastolic pressure; EDV, end-diastolic velocity; HR, heart rate; MAP, mean arterial pressure; NIRS, near-infrared spectroscopy; P1, baseline; P2, induction of anaesthesia; P3, early anaesthesia; P4, late anaesthesia; P5, early recovery; P6, late recovery; PI, pulsatility index; PSP, peak systolic pressure; PSV, peak systolic velocity; RI, resistive index; SpO₂, arterial oxygenation; TAV, time-averaged maximum velocity.

Fig. 2

Changes in Doppler waveforms before, during and after anaesthesia. The figure shows representative examples of Doppler waveforms from the same infant as Figure 1 during the different periods. Detailed morphology of the brain by a two-dimensional greyscale image is not obtained, instead the depth-versus-time colour M-mode is used to place the sample volume where there is a strong arterial signal. The colour M-mode in the upper panel shows that the measurements are obtained in the same depth, 20–28 mm, during the different periods. The time scale is in seconds. The Doppler signals suggest a decrease in velocities during anaesthesia compared with baseline and return to baseline velocities during recovery. EDV, end-diastolic velocity; HR, heart rate; PI, pulsatility index; PSV, peak systolic velocity; Q, quality; RI, resistive index; TAV, time-averaged maximum velocity.

Fig. 3

Changes of the cerebral blood flow velocity from baseline to late recovery. Changes from baseline are shown as natural log transformed values, including 95% confidence interval, and show a decrease in cerebral blood flow velocity unadjusted (a) and adjusted for mean arterial blood pressure (b). Data from n=27. EDV, end-diastolic velocity; P1, baseline; P2, induction of anaesthesia; P3, early anaesthesia; P4, late anaesthesia; P5, early recovery; P6, late recovery; PSV, peak systolic velocity; TAV, time-averaged maximum velocity.

Figs1

Patient inclusion diagram.