Does hypothermia impair cerebrovascular autoregulation in neonates during cardiopulmonary bypass?

Autoregulation monitoring has been proposed as a means to identify optimal arterial blood pressure goals during cardiopulmonary bypass, but it has been observed that cerebral blood flow is pressure passive during hypothermic bypass. When neonates cooled during cardiopulmonary bypass are managed with vasodilators and controlled hypotension, it is not clear whether hypothermia or hypotension were the cause of impaired autoregulation.

We sought to measure the effect of both arterial blood pressure and hypothermia on autoregulation in a cohort of infants cooled for bypass, hypothesizing a collinear relationship between hypothermia, hypotension, and dysautoregulation.

Cardiopulmonary bypass was performed on 72 infants at Texas Children's Hospital during 2015 and 2016 with automated physiologic data capture, including arterial blood pressure, nasopharyngeal temperature, cerebral oximetry, and a cerebral blood volume index derived from near infrared spectroscopy. Cooling to 18°C, 24°C, and 30°C was performed on 33, 12, and 22 subjects, respectively. The hemoglobin volume index was calculated as a moving correlation coefficient between mean arterial blood pressure and the cerebral blood volume index. Positive values of the hemoglobin volume index indicate impaired autoregulation. Relationships between variables were assessed utilizing a generalized estimating equation approach.

Hypothermia was associated with hypotension, dysautoregulation, and increased cerebral oximetry. Comparing the baseline temperature of 36°C with 18°C, arterial blood pressure was 44 mm Hg (39-52) vs 25 mm Hg (21-31); the hemoglobin volume index was 0.0 (-0.02 to 0.004) vs 0.5 (0.4-0.7) and cerebral oximetry was 59% (57-61) vs 88% (80-92) (Median, 95% CI of median; P<.0001 for all three associations by linear regression with generalized estimation of equations with data from all temperatures measured).

Arterial blood pressure, temperature, and cerebral autoregulation were collinear in this cohort. The conclusion that hypothermia causes impaired autoregulation is thus confounded. The effect of temperature on autoregulation should be delineated before clinical deployment of autoregulation monitors to prevent erroneous determination of optimal arterial blood pressure. Showing the effect of temperature on autoregulation will require a normotensive hypothermic model.