Temperature monitoring during cardiopulmonary bypass--do we undercool or overheat the brain?

OBJECTIVE: Brain cooling is an essential component of aortic surgery requiring circulatory arrest and inadequate cooling may lead to brain injury. Similarly, brain hyperthermia during the rewarming phase of cardiopulmonary bypass may also lead to neurological injury. Conventional temperature monitoring sites may not reflect the core brain temperature (Tdegrees). We compared jugular bulb venous temperatures (JB) during deep hypothermic circulatory arrest and normothermic bypass with Nasopharyngeal (NP), Arterial inflow (AI), Oesophageal (O), Venous return (VR), Bladder (B) and Orbital skin (OS) temperatures.

METHODS: 18 patients undergoing deep hypothermia (DH) and 8 patients undergoing normothermic bypass (mean bladder Tdegrees-36.29 degreesC) were studied. For DH, cooling was continued to 15 degreesC NP (mean cooling time-66 min). At pre-determined arterial inflow Tdegrees, NP, JB and O Tdegree's were measured. A 6-channel recorder continuously recorded all Tdegree's using calibrated thermocouples.

RESULTS: During the cooling phase of DH, NP lagged behind AI and JB Tdegree's. All these equilibrated at 15 degreesC. During rewarming, JB and NP lagged behind AI and JB was higher than NP at any time point. During normothermic bypass, although NP was reflective of the AI and JB Tdegrees trends, it underestimated peak JB Tdegrees (P=0.001). Towards the end of bypass, peak JB was greater than the arterial inflow Tdegrees (P=0.023).

CONCLUSIONS: If brain venous outflow Tdegrees (JB) accurately reflects brain Tdegrees, NP Tdegrees is a safe surrogate indicator of cooling. During rewarming, all peripheral sites underestimate brain temperature and caution is required to avoid hyperthermic arterial inflow, which may inadvertently, result in brain hyperthermia.