Hypoxic Brain Tissue following Subarachnoid Hemorrhage

These studies received University of Illinois Institutional Clinical Review Board approval, and informed consent was received from patients or family members. Patients were recruited who were undergoing craniotomy for clipping of cerebral aneurysms. These included 15 control patients who had not bled and 15 patients who had SAH 1–13 days before surgery (table 1). Fisher scores were determined before surgery by computerized tomographic scan: 1 = no blood observed, 2 = diffuse clots or layer less than 1 mm thick, 3 = blood layer more than 1 mm thick. For this study, patients with Fisher grade 1 or 2 hemorrhaging were considered to have modest SAH (n = 8), and those with Fisher grade 3 to have severe SAH (n = 7).

Patients were anesthetized with 10–15 μg/kg fentanyl and 3–5 mg/kg thiopental and paralyzed with vecuronium, their tracheas were intubated, and their lungs ventilated with 4–6% desflurane in oxygen and air (inspired oxygen fraction = 0.4). Mean arterial pressure was measured continuously using a radial artery catheter, and esophageal temperature was allowed to decrease to 34°C during surgery. Arterial carbon dioxide tension was maintained at 30–35 mmHg, corrected to a temperature of 37°C.

After pterional craniotomy and dural retraction, a Neurotrend probe (Diametrics, St Paul, MN) that measures Pt^O2, Pt^CO2, pHt, and brain temperature was inserted 2 cm into cortex tissue supplied by the artery associated with the aneurysm. Probes placed for anterior cerebral artery aneurysms were close to the midline; probes for middle cerebral artery, posterior cerebral artery, and internal carotid artery aneurysms were more lateral, but all probes were placed in the frontal cortex. Probe placement for each artery was standardized and was not within the surgical field. The probe was covered to avoid extraneous light during the procedure, and care was taken to avoid retractor pressure in the region of tissue measurement. After a 30-min equilibration period with an end-tidal desflurane concentration of 4%, steady-state measurements of Pt^O2, Pt^CO2, p  H, tissue temperature, and mean arterial pressure were obtained. All parameters were collected and averaged every 10 s by a computer using a Labview program (National Instruments, Dallas, TX).

Nine percent desflurane was requested by the neurosurgeon before clipping of the aneurysm. 9The end-tidal desflurane level was maintained at 9% for 15 min, and blood pressure was maintained at 90–95 mmHg with an intravenous phenylephrine infusion before the second tissue measurement in each patient. During this time no additional surgery was performed.

Subarachnoid hemorrhages were separated prospectively into two groups: modest SAH (Fisher grade 1 or 2 [n = 8]) and severe SAH (Fisher grade 3 [n = 7]). Data are reported as the mean ± SD. Comparisons of cardiovascular and tissue measures among the three experimental groups at baseline and during 9% desflurane were made by analyses of variance. Baseline and 9% desflurane were compared within experimental groups using a two-way repeated measures analysis of variance. Tukey tests were used for post hoc  testing if appropriate. If normality and equal-variance tests indicated that the data were not parametric, then Kruskal–Wallis analysis of variance on ranks was performed. Analyses were performed using Sigma Stat (SPSS, Chicago, IL).

The most common aneurysm location was the middle cerebral artery, followed by the anterior cerebral artery in both experimental groups (table 1). The median time of surgery after SAH was 2 days, with a range of 1–13 days. During baseline anesthetized conditions, there was no difference in mean arterial pressure, arterial blood gases, or brain temperature between control patients and patients with SAH (table 2). During treatment with 9% desflurane, mean arterial blood pressure was maintained above 90 mmHg, and brain temperature decreased modestly.

During baseline conditions, Pt^O2and pHt were lower in patients with Fisher grade 3 hemorrhage compared with controls (table 2). At a 9% desflurane concentration, Pt^O2increased significantly in all patient groups. In patients with Fisher grade 3 hemorrhage, pHt increased with 9% desflurane to levels seen in controls. No aneurysm rupture occurred during treatment with 9% desflurane in any patient.

We found that patients with SAH undergoing surgery for clipping of cerebral aneurysm have lower Pt^O2and pHt levels compared with control patients, with a greater effect as the amount of hemorrhage increases. Control and SAH patients responded to 9% desflurane with an increase in Pt^O2. In patients with Fisher grade 3 SAH, tissue hypoxia and acidosis recovered to control levels. These results indicate that brain ischemia after SAH is significantly related to the amount of hemorrhage. Tissue hypoxia and acidosis were reversed by treatment with 9% desflurane.

Progressive decreases in cerebral blood flow have been reported in patients with SAH for 3 weeks after hemorrhage, and the decrease in blood flow is related to a worsening in clinical grade. 1,2Others report that cerebral blood flow is not altered within 7 days after SAH, but cerebral pressure autoregulation and cerebrovascular carbon dioxide reactivity are abnormal, and this dysfunction is correlated with clinical grade and development of cerebral vasospasm. 11,12However, cerebrovascular carbon dioxide reactivity is intact in other patients with SAH, and it has been questioned as a measure of pathology related to subsequent brain injury. 13Although there is evidence of sequential ischemic cerebrovascular changes initiated by SAH, these results are controversial. 14This may be related to the time of evaluation or the amount of blood associated with SAH, which are variable between studies.

Jugular-bulb oxygen saturation has been tested as an early indicator of cerebral ischemia after SAH. In 26 patients with SAH, jugular oxygen saturation was normal (0.56–0.74%) 90% of the time. 6If jugular oxygen desaturation was seen, it was significantly correlated with increased intracranial pressure. Other investigators did not find a significant relation between low jugular oxygen saturation and the incidence of cerebral vasospasm. 7However, SAH patients with low jugular-bulb oxygenation had a higher incidence of ischemic injury and death. These results indicate that jugular-bulb oxygenation can be intermittently low after SAH, and that this hypoxic effect is related to subsequent brain injury.

The measurement of Pt^O2provides a more direct indication of local brain oxygenation than does jugular-bulb saturation. In studies of comatose patients with head injury and SAH, hypoxic episodes with Pt^O2less than 10 mmHg were associated with unfavorable neurologic outcome. 5In patients with SAH or head injury, brain tissue hypoxia was observed intermittently, even with normal cerebral perfusion pressure. 15Interventions such as increasing cerebral perfusion pressure and decompressive craniotomy consistently increased brain oxygenation. Our results agree that low tissue oxygenation is present after SAH, although cerebral perfusion pressure is normal.

Previous studies suggest that the ability of 9% desflurane to increase Pt^O2is related to an increase in cerebral blood flow and a decrease in brain metabolism. 16,17High-dosage desflurane increases cerebral blood flow in dogs. 16Increased Pt^O2occurs simultaneously with an increase in laser Doppler measured brain blood flow during 9% desflurane in patients. 17High-dosage desflurane abolishes autoregulation, and if blood pressure is not maintained above 90 mmHg, Pt^O2decreases. This indicates that there is no direct effect of desflurane on the sensor to produce a falsely elevated Pt^O2level. Other anesthetics that produce electroencephalographic burst suppression have different effects on Pt^O2: Thiopental does not change tissue oxygenation, and etomidate is associated with a significant decrease in Pt^O2. 9,17The cerebrovasodilating effect of inhalation anesthetics may be important for the increase in Pt^O2because we observed in dogs that a 3% end-tidal isoflurane concentration increases brain tissue oxygenation (unpublished results, July 1999).

We observed that 9% desflurane increased Pt^O2in all groups, but an increase in pHt was more dependent on baseline acid–base status. This agrees with previous results that if pHt is 7.1 or higher, changes induced by 9% desflurane are small. 17This is probably an indication that if normal acid–base balance is attained in brain tissue, an increase in cerebral perfusion with desflurane is of little consequence. In contrast, Pt^O2does not appear to be controlled in the same manner as pHt, and oxygenation can increase with 9% desflurane even if Pt^O2appears to be normal. Previously it was shown that Pt^O2is more closely related to tissue perfusion than to pHt. 18This is probably an indication that tissue oxygenation is dependent on oxygen delivery and capillary partial pressure of oxygen (P^O2). In contrast, the diffusion coefficient of carbon dioxide from tissue to blood is 20 times greater than oxygen, and, therefore, acid–base balance can be maintained at a relatively low perfusion state. However, at ischemic tissue perfusion levels, metabolic acidosis produced by critically low Pt^O2is accompanied by impaired carbon dioxide clearance. The ability of 9% desflurane to reverse hypoxia and acidosis in patients with SAH may be an indication that tissue perfusion is increased toward normal levels.

One concern in this study is that differences in tissue gases were related to the cortex region measured; for example, the anterior cerebral artery or the middle cerebral artery arterial supply. However, control patients and patients with SAH were well-matched with respect to aneurysm location. In addition, we saw no evidence for differences in Pt^O2with respect to aneurysm location in control patients. Dings et al.  19reported that Pt^O2was 10 mmHg higher at the surface of the cortex, compared with 1 cm deeper. This should not be a factor in our study because the probe always was placed at the same depth in each cortex region. This supports the conclusion that differences in Pt^O2and pHt between control patients and those with Fisher grade 3 hemorrhage were related to SAH rather than to probe location.

In conclusion, these results show that patients with SAH have tissue hypoxia and acidosis during aneurysm clipping related to the severity of the bleed. At a median time of 2 days after SAH, significant tissue hypoxia and acidosis were seen in patients with Fisher grade 3 SAH compared with patients without SAH. Nine percent desflurane, given before clipping of an aneurysm, increased Pt^O2in all patients and increased pHt in patients with Fisher grade 3 hemorrhage to the level of that seen in controls.