The brain is a complex organ and critically dependent on its oxygen metabolism and blood supply. The purpose of the present study was twofold: firstly to design, evaluate, and adopt a physiologically stable experimental model for repeated measurements of cerebral cortical microvascular perfusion by a new laser Doppler technique (experimental part (I) - intrinsic model). Secondly, to examine the complication rates of a brief period of hypothermia for protection of the brain in patients who  were operated on for subarachnoid aneurysms (clinical part (II) - extrinsic model).

Method. Part I. In contrast to conventional laser Doppler flowmetry, the laser Doppler perfusion imager (LDPI) creates a two-dimensional map of the tissue perfusion in an area of up to 120 x 120 mm, comprising 4096 measurement points. Measurements were made in a closed cranial window in ventilated anaesthetised pigs (n=25). Provocations that altered cerebral blood flow (CBF) were those that can occur during anaesthesia and critical care, including varying arterial concentration of carbon dioxide and mean arterial blood pressure (MABP), and those caused by different types (isoflurane and sevoflurane) and concentrations (minimum alveolar concentration, MAC, 0.3-1.2) of volatile anaesthetics.

Part II. The feasibility and complication rates of rapidly-induced moderate hypothermia in 359 operations were examined prospectively. The complications that we sought were: altered haemostasis, infections, haemodynamic instability, and increased need for postoperative ventilatory support. Hypothermia was produced by cold (4°C) intravenous infusions and convective cooling. This lowered the body temperature at a mean (SD) 4 (0.4)^oC h¹ to 32.5 (0.4)^oC. The body temperature returned to normal in all patients by 5 (2) hours postoperatively.

Results. Part I. Cortical CBF was distributed highly heterogeneously and it was strongly dependent on MABP, as well as concentrations of CO₂ During decreasing MABP the lower limit of autoregulation varied locally in the cortex and the 'classic' autoregulatory pressure-flow relation was present in only a few areas. Alterations in P_aC0₂ concentration did not affect the pressure-flow relation at low perfusion pressures, whereas at normal or above normal values, hypercapnia increased CBF considerably (p < 0.001). CBF was not affected by the anaesthetic (isoflurane or sevoflurane) or dose (MAC 0.3-1.2) used.

Part II. Perioperative and postoperative complications itcluded: circulatory instability (10%), arrhythmias (5%), abnormalities of coagulation, need for blood transfusions (47%), infections (8%), and puhnonary complications including infiltrates or oedema while on ventilatory support (27%). There was no correlation between the extent of hypothermia and any of the complications. However, there was a strong correlation (p < 0.001) between the incidence of complications and the severity of the underlying neurological disease as assessed by the Hunt and Hess score.

Conclusion Part I. The LDPI method was useful and relevant for measurement of cortical CBF in this experimental model. 'Classic' autoregulatory patterns were found only when all values sampled wereclustered together, whereas the autoregulatory capacity is often lacking at the local level in the cortex. Volatile anaesthestics (isoflurane and sevoflurane) in the doses 0.3 - 1.2 (MAC) did not affect the regulation of CBF in this model. Part II. Moderate hypothermia can be achieved within 1 hour of induction of anaesthesia and seems to be safe as far as the risks of perioperative and postoperative complications are concerned.