The maintenance of a constant cerebral blood flow (CBF) during changing cerebral perfusion pressure (CPP) is termed cerebral autoregulation. In many conditions in which cerebral autoregulation is impaired, it can be restored by hyperventilation. For example, hyperventilation in humans can restore cerebral autoregulation that has become impaired because of stroke or tumor (1), head injury (2), meningitis (3), and hepatic encephalopathy (4). In animal studies, it has been found that cerebral autoregulation impaired by subarachnoid hemorrhage (5), hypoxia (6), and the post-ictal state (7) can also be restored by hypocapnia.
Isoflurane, in common with most other inhaled anesthetics, increases CBF and impairs cerebral autoregulation in a dose-related manner (8,9). The aim of this study was to determine whether hyperventilation can restore cerebral autoregulation in humans during anesthesia with isoflurane. Only one previous study has examined the effect of hypocapnia on cerebral autoregulation during administration of inhaled anesthetics. In 1976, Miletich et al. (10) investigated the effects of halothane and enflurane on the control of CBF in goats. With either anesthetic, they found that cerebral autoregulation was absent at 1.0 minimum alveolar anesthetic concentration (MAC) and it remained absent with hyperventilation.
After approval by our institutional Ethics Review Committee, 12 healthy (ASA physical status I or II) patients gave written informed consent. Exclusion criteria included a history of neurological disease, any cardiovascular disease including hypertension, or any active respiratory disease. The patients comprised 7 women and 5 men, aged 21–59 yr, presenting for elective surgery requiring general anesthesia and mechanical ventilation. Anesthesia was induced with an IV injection of propofol (2 mg/kg). An infusion of remifentanil was commenced and, after paralysis with rocuronium (0.6 mg/kg), the trachea was intubated. Subjects’ lungs were ventilated with isoflurane in 100% oxygen.
Routine intraoperative monitoring included measurement of end-tidal carbon dioxide and end-tidal isoflurane partial pressures. The following additional monitors were applied: a radial artery catheter was placed for invasive mean arterial blood pressure (MAP) monitoring and sampling of arterial blood, surface electrodes were placed to monitor the frontal electroencephalogram (EEG), and bilateral transcranial Doppler (TCD) probes were placed to monitor middle cerebral artery blood flow velocity (Vmca). The Doppler probes were held in constant position by a frame. The software within the TCD monitor (Multidop-T; DWL, Sipplingen, Germany) detects the peak flow velocity and outputs an analog voltage signal representing the peak velocity waveform. A personal computer running a digital to analog converter and multichannel recording software (ADInstruments, Sydney, Australia) was used to continuously record the MAP waveform, the outline of the Vmca waveform, the airway isoflurane concentration, and the capnograph.
The dose of isoflurane was increased until we noted short periods of electrical silence on the EEG. The inspired isoflurane concentration was then reduced to maintain an end-tidal concentration 0.1%–0.2% less than that required to induce short periods of EEG silence. This method of individualizing the isoflurane dose was chosen in order to produce the maximal effect on cerebral autoregulation without inducing periods of electrical silence (burst-suppression). The onset of a period of electrical silence is associated with a sudden reduction in Vmca (11) and this would have confounded the assessment of cerebral autoregulation.
In each subject, cerebral autoregulation was tested twice in random order: once with normocapnia (Paco2 38–43 mm Hg) and once with hypocapnia (Paco2 27–34 mm Hg). Paco2 was manipulated by adjusting respiratory rate and/or tidal volume and/or bypassing the CO2 absorber. After the end-tidal CO2 had stabilized, and before each autoregulation test, arterial blood was analyzed to confirm the Paco2. Autoregulation testing was performed either before surgical incision or intraoperatively during a time of hemodynamic stability (as judged by constant MAP, heart rate, and Vmca). The patients in whom testing was performed intraoperatively were undergoing nephrectomy for living-donor renal transplant.
Cerebral autoregulation was tested in the following manner. If necessary, remifentanil was infused to achieve a MAP <80 mm Hg. The remifentanil infusion rate was then left unchanged for the duration of the autoregulation test. An IV infusion of phenylephrine was then commenced and titrated to increase MAP to >100 mm Hg over a period of 1–2 min.
From the recorded data, Vmca was analyzed at two times: when MAP was approximately 80 mm Hg and when MAP was approximately 100 mm Hg. For each data point, the MAP and Vmca were averaged over a full ventilatory cycle in order to account for respiratory swings in MAP and CBF. Vmca data from the two sides were averaged. The autoregulation index (ARI) was then calculated according to a previously published formula1 (9). An ARI of 1.0 indicates perfect autoregulation with no change in Vmca despite the increase in MAP. An ARI of zero indicates totally pressure-passive CBF, i.e., Vmca increasing in direct proportion to MAP. As in a previous report, autoregulation was taken to be significantly impaired when the ARI was ≤0.4 (12).
The ARI with normocapnia was compared with the ARI with hypocapnia using the Mann-Whitney test for nonparametric data. The number of subjects with significantly impaired cerebral autoregulation was compared using Fisher’s exact test. Statistical analysis was performed with the assistance of GraphPad Prism software (GraphPad Software, San Diego, CA).
The end-tidal isoflurane concentration was 1.6% ± 0.2% (mean ± sd) corresponding to an age-adjusted MAC multiple of 1.4 ± 0.2 (13). The Paco2 during normocapnia was 41 ± 2 mm Hg and during hypocapnia was 31 ± 2 mm Hg. At a MAP of 80 ± 2 mm Hg, the Vmca was 74 ± 16 cm/s with normocapnia and 42 ± 9 cm/s with hypocapnia. Thus, a 10 mm Hg reduction in mean Paco2 led to a 43% reduction in mean Vmca.
Table 1 presents the data for Paco2, MAP, Vmca, and ARI. The median (interquartile range) of the ARI was 0.3 (0.2–0.6) with normocapnia and 0.8 (0.7–0.8) with hypocapnia (P < 0.005). Figure 1 illustrates the relationship between Paco2 and ARI. When normocapnic, cerebral autoregulation was judged to be significantly impaired (ARI ≤0.4) in 8 of the 12 subjects but no subject had significantly impaired autoregulation when hypocapnic (P = 0.001). The combined effect of MAP and Paco2 on Vmca in an individual subject is illustrated in Figure 2. The data points in Figure 2 were calculated by averaging the recorded Vmca for each mm Hg increment in MAP from 80 to 100 mm Hg.
We investigated the influence of hypocapnia on cerebral autoregulation in humans during administration of the volatile anesthetic, isoflurane. The finding that hypocapnia can reverse the effect of isoflurane on cerebral autoregulation may be of interest given that hyperventilation is often used during anesthesia for neurosurgical procedures; a setting in which maintenance of cerebral autoregulation may be desirable. However, these findings do not necessarily indicate that hyperventilation should be routinely used during isoflurane anesthesia; the possible benefits of preserving cerebral autoregulation would need to be balanced against the possible risks of hypocapnia such as excessive cerebral vasoconstriction. In clinical situations in which preservation of autoregulation is thought to be important, an anesthetic that does not impair cerebral autoregulation at normocapnia [such as propofol (9) or sevoflurane (14)] could also be considered.
Previous investigations demonstrating that isoflurane impairs cerebral autoregulation were usually performed at normocapnia. Using a similar dose of isoflurane and a similar test of autoregulation to this study, Strebel et al. (9) reported a highly significant impairment of cerebral autoregulation. Individual results were not reported so it is not clear whether, as in this current study, a proportion of subjects retained cerebral autoregulation while normocapnic. Tiecks et al. (15) also performed autoregulation testing with a similar dose of isoflurane. However, they also administered nitrous oxide. Judged by the same criteria used in this study, 7 of their 10 subjects had impaired autoregulation.
The results of the present study emphasize the importance of considering Paco2 when interpreting tests of cerebral autoregulation. During isoflurane anesthesia in dogs, McPherson and Traystman (8) reported that the autoregulatory response to reduced CPP is preserved with 1 MAC isoflurane but abolished at twice that dose. The average Paco2 in the animals receiving 1 MAC isoflurane ranged from 32 to 35 mm Hg. In light of our current results, it is possible that hypocapnia contributed to the maintenance of autoregulation with 1 MAC isoflurane in that study. In 2 of our subjects (subjects 4 and 5), the Paco2 was 33 and 34 mm Hg during hyperventilation and this modest degree of hypocapnia was sufficient to restore cerebral autoregulation. In an animal model of subarachnoid hemorrhage, hyperventilation to a Paco2 of 35 mm Hg was sufficient to restore autoregulation in 7 of 8 rats (5).
In a study of goats receiving 1 MAC of either enflurane or halothane, hyperventilation to a Paco2 of 22 mm Hg did not restore autoregulation (10). At 0.5 MAC of either anesthetic, cerebral autoregulation was evident but the lower limit of autoregulation was higher than in awake animals. There was evidence of an interaction between Paco2 and the smaller dose of anesthetics in that the lower limit of autoregulation was further increased by hypercapnia and, conversely, was decreased toward awake values by hypocapnia.
The effect of hypocapnia on cerebral autoregulation demonstrated in the current study may have been caused by antagonism of the vasodilatory effect of isoflurane. It seems that any process leading to cerebral vasodilatation tends also to obtund the mechanism of autoregulation. Carbon dioxide is a potent cerebral vasodilator and hypercapnia impairs cerebral autoregulation (6). We have previously shown that when autoregulation is intact during anesthesia with sevoflurane or propofol, it becomes impaired with hypercapnia (12). The likelihood of impaired autoregulation during general anesthesia is related to the cerebral vasodilating properties of the anesthetic. Propofol is not a cerebral vasodilator and it does not impair cerebral autoregulation (9). There is evidence that sevoflurane has some intrinsic cerebrovascular dilating properties (16) but doses up to 1.5 multiples of MAC are not associated with impaired autoregulation (14). However, the rapid phase of the autoregulatory response is marginally impaired at that dose of sevoflurane (17) and we found that autoregulation is more likely to be impaired at mild hypercapnia with sevoflurane compared with propofol (12). Isoflurane and desflurane are more potent cerebral vasodilators and they are associated with significant impairment of autoregulation at clinical doses (9).
The method for testing cerebral autoregulation used in this study is sometimes called the static rate of autoregulation (15) because CPP is changed slowly, in contrast with dynamic autoregulation tests that use more rapid changes in CPP. Dynamic tests are influenced by changes in the latency of the initial rapid phase of the autoregulatory response (15). Significant impairment of the static rate of autoregulation indicates near-complete loss of autoregulation and is arguably more clinically important than an increased latency of otherwise-intact autoregulation. One limitation of the method used in this study is that the autoregulatory response was only tested over a narrow range of MAP, i.e., between 80 and 100 mm Hg. For example, in subjects judged to have intact autoregulation by our protocol, the lower limit of autoregulation may have been around 50 mm Hg or it may have approached 80 mm Hg. The test used in this study would not discriminate between these situations. Another limitation is that autoregulation was assessed by comparing Vmca at only two values of MAP. It is possible that some of the subjects judged to have impaired autoregulation were able to maintain a constant CBF, but only within a range significantly >80 mm Hg MAP. In that case, the observed improvement in the ARI with hypocapnia would represent a left-shift of the lower limit of cerebral autoregulation rather than restoration of completely abolished autoregulation.
Testing cerebral autoregulation does not require measurement of actual CBF; it is sufficient to be able to measure proportional changes in CBF. For the following reasons, the interventions used in this study should not invalidate the use of TCD for assessment of cerebral autoregulation. The remifentanil infusion rate and the Paco2 were not changed during each autoregulation test. Changes in Paco2 have not been found to significantly alter middle cerebral artery diameter (18). Remifentanil at the doses used in this study (0–0.15 μg · kg−1 · min−1) does not alter Vmca (19) and changes in Vmca correlate well with other measures of CBF during administration of remifentanil (20). Isoflurane is not known to alter middle cerebral artery diameter and the isoflurane end-tidal concentration in each subject was kept constant throughout the study.
It is possible that the propofol used for induction of anesthesia could have influenced our results. However, a minimum of 45 minutes elapsed between injection of propofol and the first autoregulation test, by which time we expect that the residual effects of a single bolus of propofol would be negligible. Any time effect, caused for example by decreasing propofol concentration, should be equally distributed between the states of hypocapnia and normocapnia because the order of the two levels of Paco2 were randomized.
In conclusion, hypocapnia restores cerebral autoregulation during isoflurane anesthesia at 1.4 MAC. This study was confined to a small sample of normal individuals; it remains to be determined whether the same effect occurs in patients with neurosurgical conditions.
The authors acknowledge AstraZeneca who provided funding for the purchase of the analog to digital recording equipment used in this study.
1. Paulson OB, Olesen J, Christensen MS. Restoration of autoregulation of cerebral blood flow by hypocapnia. Neurology 1972;22:286–93.
2. Newell DW, Weber JP, Watson R, et al. Effect of transient moderate hyperventilation on dynamic cerebral autoregulation after severe head injury. Neurosurgery 1996;39:35–43.
3. Moller K, Skinhoj P, Knudsen GM, Larsen FS. Effect of short-term hyperventilation on cerebral blood flow autoregulation in patients with acute bacterial meningitis. Stroke 2000;31:1116–22.
4. Strauss G, Hansen BA, Knudsen GM, Larsen FS. Hyperventilation restores cerebral blood flow autoregulation in patients with acute liver failure. J Hepatol 1998;28:199–203.
5. Ma X, Willumsen L, Hauerberg J, et al. Effects of graded hyperventilation on cerebral blood flow autoregulation in experimental subarachnoid hemorrhage. J Cereb Blood Flow Metab 2000;20:718–25.
6. Haggendal E, Johansson B. Effects of arterial carbon dioxide tension and oxygen saturation on cerebral blood flow autoregulation in dogs. Acta Physiol Scand Suppl 1965;258:27–53.
7. Monin P, Stonestreet BS, Oh W. Hyperventilation restores autoregulation of cerebral blood flow in postictal piglets. Pediatr Res 1991;30:294–8.
8. McPherson RW, Traystman RJ. Effects of isoflurane on cerebral autoregulation in dogs. Anesthesiology 1988;69:493–9.
9. Strebel S, Lam AM, Matta B, et al. Dynamic and static cerebral autoregulation during isoflurane, desflurane, and propofol anesthesia. Anesthesiology 1995;83:66–76.
10. Miletich DJ, Ivankovich AD, Albrecht RF, et al. Absence of autoregulation of cerebral blood flow during halothane and enflurane anesthesia. Anesth Analg 1976;55:100–9.
11. Lam AM, Matta BF, Mayberg TS, Strebel S. Change in cerebral blood flow velocity with onset of EEG silence during inhalation anesthesia in humans: evidence of flow-metabolism coupling? J Cereb Blood Flow Metab 1995;15:714–7.
12. McCulloch TJ, Visco E, Lam AM. Graded hypercapnia and cerebral autoregulation during sevoflurane or propofol anesthesia. Anesthesiology 2000;93:1205–9.
13. Mapleson WW. Effect of age on MAC in humans: a meta-analysis. Br J Anaesth 1996;76:179–85.
14. Gupta S, Heath K, Matta BF. Effect of incremental doses of sevoflurane on cerebral pressure autoregulation in humans. Br J Anaesth 1997;79:469–72.
15. Tiecks FP, Lam AM, Aaslid R, Newell DW. Comparison of static and dynamic cerebral autoregulation measurements. Stroke 1995;26:1014–9.
16. Matta BF, Heath KJ, Tipping K, Summors AC. Direct cerebral vasodilatory effects of sevoflurane and isoflurane. Anesthesiology 1999;91:677–80.
17. Summors AC, Gupta AK, Matta BF. Dynamic cerebral autoregulation during sevoflurane anesthesia: a comparison with isoflurane. Anesth Analg 1999;88:341–5.
18. Giller CA, Bowman G, Dyer H, et al. Cerebral arterial diameters during changes in blood pressure and carbon dioxide during craniotomy. Neurosurgery 1993;32:737–41.
19. Paris A, Scholz J, von Knobelsdorff G, et al. The effect of remifentanil on cerebral blood flow velocity. Anesth Analg 1998;87:569–73.
20. Klimscha W, Ullrich R, Nasel C, et al. High-dose remifentanil does not impair cerebrovascular carbon dioxide reactivity in healthy male volunteers. Anesthesiology 2003;99:834–40.
1 ARI = [(R2 − R1)/R1]/[(MAP2 − MAP1)/MAP1] where R = MAP/Vmca. R is an index of cerebral vascular resistance (ignoring the contribution of intracranial pressure to CPP). Subscripts 1 and 2 refer to measurements at the lower and higher MAP respectively.