Intracerebral steal phenomenon associated with global hyperemia in moyamoya disease during revascularization surgery

Abstract

Object. The collateral vessels in moyamoya disease appear to retain their ability to constrict during hypocapnia but not to dilate during hypercapnia. It has been claimed that hypercapnia, as well as hypocapnia, decreases the blood flow in regions perfused by collateral vessels, presumably because of intracerebral steal. If this holds true, the decrease in blood flow may be proportional to the global hyperemia in the brain. To establish appropriate hemodynamic control during revascularization surgery, the authors monitored the jugular bulb oxygen saturation (SjO2) intraoperatively, a method that could sensitively detect global hyperemia.Methods. A total of 17 patients, most of whom presented with transient ischemic attacks or fluctuating neurological deficits, underwent intraoperative monitoring of their SjO2 and end-tidal carbon dioxide (ETCO2) after a state of anesthesia had been induced with isoflurane (Group 1) or propofol (Group 2). In eight of these patients, the regional cerebral blood flow (rCBF) of the collateral vessel territory was also monitored by laser Doppler flowmetry during the period of cortical exposure, and a total of 113 data sets (averaged values during 2.5-minute intervals) was collected. There was fluctuation in the ETCO2 levels ranging from 36 to 44 mm Hg. The mean SjO2 level was clearly greater (p < 0.01) in Group 1 (71.8 ± 2.2%) than in Group 2 (63.3 ± 2.1%). An episodic fall in rCBF was observed in association with a transient increase in SjO2. Such an event was not uncommon in Group 1 and there was a greater risk of rCBF decreasing when SjO2 exceeded a cutoff level of 76% (p < 0.01). This level could sometimes be reached at a broad range of ETCO2 readings (37–44 mm Hg). In Group 2, similar events sometimes occurred when SjO2 increased beyond 70%. However, this level could be reached only with a higher ETCO2 (42–44 mm Hg). The rCBF level was negatively correlated to SjO2 (p < 0.01), but not always to ETCO2, indicating that the episodic fall in rCBF was closely related to global hyperemia rather than the absolute level of hypercapnia.Conclusions. The observed association between a fall in rCBF and global hyperemia supports the intracerebral steal hypothesis and indicates that it is prudent to avoid excessive global hyperemia. The optimal range of CO2 for isoflurane is more restricted than that for propofol, presumably because isoflurane induces hyperemia by itself. Monitoring of SjO2 appears to represent the most practical technique for detecting global hyperemia as well as global ischemia, both of which may cause ischemic complications in moyamoya disease.