Can cerebral autoregulation be assessed without measuring cerebral blood flow?

Abstract

Cerebral autoregulation is becoming more popular as a useful parameter in clinical practice. It has been demonstrated that dynamic autoregulation assessed using transcranial Doppler ultrasongraphy is associated with outcome following severe head injury, worsens with impairment of cerebral blood inflow in patients with carotid artery stenotic disease and when vasospasm arises in patients after subarachnoid haemorrhage, etc 1. Although current techniques for calculation of dynamic autoregulation do not require pharmacological changes in arterial pressure, the measurement in clinical practice is still difficult and depends on continuous insonation of the Middle Cerebral Artery- subject to artifacts mainly caused by the minor displacement of the ultrasound probes. We introduced Pressure Reactivity Index (PRx) derived from continuous monitoring of arterial pressure (AP) and intracranial pressure (ICP) to clinical practice in 1990 s. Previously PRx has been positively correlated with outcome 1, PET-CBF derived autoregulation 2 and cerebral metabolism rate for oxygen 3. Our objective was to correlate PRx with Transcranial Doppler dynamic autoregulation index (Mx) and and outcome in a large group of head injured patients. Data from intermittent bedside monitoring of ICP, AP, cerebral perfusion pressure (CPP=ABP-ICP) and of transcranial Doppler middle cerebral artery blood flow velocity (FV) in 237 patients were analysed retrospectively. Indices describing cerebral autoregulation and pressure reactivity were calculated as correlation coefficients between 'slow waves' of mean FV and CPP (Mx) and ABP and ICP (PRx) over moving three minute periods. Data were averaged over multiple recordings made in the same patients and compared to clinical outcome at 6 months after head injury. All patients were sedated, paralysed and ventilated during the period when brain variables were monitored. Both PRx and Mx were well correlated with each other (R=0.58;p<0.00001). The absolute difference between Mx and PRx increased with greater ICP (R=0.27;p<0.0001) and with worse outcome (R=0.21;p<0.003). Both indices correlated positively with rising ICP and negatively with decreasing CPP, but the associations were stronger for PRx than Mx. Both indices were significantly correlated with outcome and were able to differentiate between fatal and non-fatal outcome with comparable power. However, where multiple regression was applied to outcome prediction, only mean ICP, age and PRx were retained as independent outcome predictors (R=0.35;p<0.00001). Although significantly inter-correlated in a group analysis, pressure-reactivity may be, in individual case, different from dynamic autoregulation. Pressure-reactivity seems to be more robust in head injured patients when continuous ICP monitoring is required.