Secondary Logo

Journal Logo

Original Article

The effect of sevoflurane-induced hypotension in combination with acute hypervolaemic haemodilution on middle cerebral artery flow velocity in surgical patients

Fukusaki, M.*; Kanaide, M.*; Inadomi, C.*; Takada, M.*; Terao, Y.*; Sumikawa, K.

Author Information
European Journal of Anaesthesiology: August 2008 - Volume 25 - Issue 8 - p 657-661
doi: 10.1017/S0265021508003955



Preoperative acute normovolaemic haemodilution (ANH) or controlled hypotension are both effective for blood conservation during hip surgery [1]. Acute hypervolaemic haemodilution (HHD) without removing blood is a simple as well as time- and cost-saving alternative to ANH [2]. Hypotensive anaesthesia induced with sevoflurane using higher inspired concentrations is a simple technique to induce hypotension. The combination is more useful for avoiding homologous transfusion; however, it may reduce oxygen-carrying capacity and perfusion pressure resulting in the impairment of tissue oxygenation [3,4]. HHD has been well shown to increase cerebral blood flow (CBF) [5] but the effect of sevoflurane at higher concentrations on autoregulation is controversial [6,7].

Haemodilution activates compensatory mechanisms, i.e. increases in tissue flow through increased cardiac output and increased oxygen extraction. However, the combination of hypotensive anaesthesia with higher concentrations of sevoflurane may inhibit the compensatory cardiovascular responses so that the haemodilution results in reduced cerebral oxygen supply.

Although middle cerebral artery (MCA) flow velocity (Vmca), as measured by transcranial Doppler ultrasonography (TCD), is not a direct measurement of CBF, changes in Vmca have been shown to correlate reliably with changes in CBF because the MCA maintains a constant vessel diameter [8]. TCD is now widely used as a surrogate measure of CBF. Bruder and colleagues [9] reported that TCD accurately assessed the effect of isovolaemic haemodilution on the cerebral circulation. Fukusaki and colleagues [10] showed that prostaglandin E1-induced hypotension combined with ANH would not impair MCA flow using TCD.

As it has not been clarified whether HHD preserves Vmca in spite of marked reduction in arterial pressure due to sevoflurane, this study was carried out to clarify the effect of sevoflurane-induced hypotension in combination with HHD on Vmca using TCD in patients undergoing hip surgery.


Thirty ASA physical status I or II total hip arthroplasty patients (age 51–70 yr, weight 48–72 kg) were studied. The protocol was approved by the Nagasaki Rosai Hospital Institutional Human Committee, and written informed consent was obtained from each patient. Exclusion criteria were a history of hypertension, ischaemic heart disease, stroke, renal dysfunction and anaemia (haemoglobin level < 11 g dL−1). Patients who did not have adequate TCD measures of Vmca on the preoperative day were also excluded from the study.

Patients were continuously monitored with pulse oximetry (Oxypal OLV-1200, Nihon Kohden Co, Ltd, Tokyo, Japan) and 3-lead electrocardiography (Bedside monitor BSM-8500, Life Scope12, Nihon Kohden Co, Ltd). A radial arterial catheter was inserted for continuous monitoring of arterial pressure and for obtaining blood samples. Anaesthesia was induced with intravenous (i.v.) thiamylal 5 mg kg−1 and fentanyl 2 μg kg−1, and maintained with 60% nitrous oxide in oxygen supplemented with 1.0 minimum alveolar concentration (MAC) end-tidal sevoflurane concentration using a circle system with 5 L min−1 total flow. Tracheal intubation was facilitated with vecuronium 0.1 mg kg−1 i.v. Fentanyl 1–2 μg kg−1 i.v. and vecuronium 0.05 mg kg−1 i.v. were injected during surgery as required. End-tidal carbon dioxide tension (etCO2) was continuously monitored, and ventilation was controlled to maintain arterial carbon dioxide partial pressure (PaCO2) at approximately 40 mmHg (Nelcor N-1000; Nelcor CMI, Kyoto, Japan). Ringer's acetate (RA) solution containing 5% glucose was infused at an amount of 10 mL kg−1 before surgery over a 4-h period. RA solution was continued at a rate of 6 mL kg−1 h−1 during surgery. The same solution was infused in amounts corresponding to three times the blood loss. Rectal temperature was maintained at 36.0°C to 36.5°C using a circulating water blanket.

After induction of anaesthesia, HHD was produced by prior to surgery by infusion of 1000 mL of 6% hydroxyethylstarch (HES; molecular weight 70 000) solution without removing blood. HES was infused at a rate of approximately 50 mL min−1 using a rapid infusion pump. RA and HES were infused at approximately 36.5°C after warming by a medical warmer. Patients were randomly divided by sealed envelope assignment into two groups. Group A (n = 15) was the no-controlled hypotension group and Group B (n = 15) was the controlled hypotension group. In Group A, mean arterial pressure (MAP) was maintained at approximately 95 mmHg during surgery. In Group B, MAP was maintained at approximately 55 mmHg for 80 min by increasing the inspired concentration of sevoflurane. Autologous blood was collected by a cell saver (Haemolite, Haemonetics Corp., Boston, MA, USA) during and after surgery, and then re-transfused into each patient. The volume of blood loss was estimated during surgery by weighing swabs and measuring blood collected from the wound drainage.

Vmca was measured continuously using a pulsed 2-MH TCD (TC2-64B; EME, Uberlingen, Germany). The measurement methods and calculated formula in Vmca and pulsatility index (PI) used were done according to the previous report [5,7]. The Doppler probe was positioned at the right temporal scalp surface and was fixed at the site of best insonation. TCD signals of the MCA were identified at a depth of 45 to 50 mm. Systolic and diastolic velocity values were obtained manually by manipulating the cursor to read the average value from three cardiac cycles, excluding the effect of electrical artefacts. The time-mean Vmca was calculated from the systolic and diastolic velocities using the formula: Vmca = (systolic velocity − diastolic velocity) / 3 + diastolic velocity. The PI was also calculated from systolic, diastolic and mean velocities using the formula: PI = (systolic velocity − diastolic velocity) / mean velocity. The Vmca and PI values were obtained only during end-expiration to avoid respiratory fluctuations.

Arterial blood gas (ABG) was analysed by an ABG analyser (ABL-4, Radiometer Corp., Odense, Denmark). Measurements included Vmca, PI, ABG and haematocrit (Hct). An equivalent of cerebral O2 transport (TEO2) was calculated as arterial oxygen content (CaO2) × Vmca. Measurements were made before haemodilution (T0) in Groups A and B, after haemodilution (T1) in Groups A and B, 80 min after starting hypotension (T2) in Group B (80 min after starting surgery in Group A), and 60 min after recovery from hypotension (T3) in Group B (60 min after the end of surgery in Group A. Sampling times and their relation to the various phases of the investigation were similar in the two groups.

A study-blinded anaesthesiologist and neurosurgeon performed the neurological assessment on postoperative days 1 and 7.

Data are expressed as mean ± SD. Statistical comparison was made between Group A and Group B. Categorical data were compared by using the χ2 test, and continuous data were compared by using the two-tailed t-test. Haemodynamic, blood gas, Vmca, TEO2 and PI changes were analysed by two-way repeated-measure analysis of variance followed by Bonferroni's test. The U-test was used for evaluation of differences between the groups, followed by the Wilcoxon rank sum test where necessary. A statistical application, Stat View Version 5 (SAS Institute, Cary, NC, USA), was used for statistical analysis. A P value less than 0.05 was considered statistically significant.


The two groups were similar in patient characteristics (gender, age and weight), operative period, infusion volume and intraoperative blood loss (Table 1). PaO2 values were more than 180 mmHg and there was no apparent acidaemia or alkalaemia in either group throughout the study.

Table 1
Table 1:
Patient characteristics.

The calculated minimum alveolar anaesthetic concentration (MAC)-hour of sevoflurane (the mean value for MAC over the course of an hour) was 1.2 ± 0.3 in Group A and 3.2 ± 0.2 in Group B (P < 0.05, Group A vs. Group B).

In Group A, Vmca value significantly increased by 28% at T1 (P < 0.05, vs. T0), 24% at T2 (P < 0.05 vs. T0) and 22% at T3 (P < 0.05 vs. T0). In Group B, Vmca value significantly increased by 30% at T1 (P < 0.05 vs. T0), whereas it decreased to the baseline value at T2 (P < 0.05 vs. T1 and Group A) and returned to pre-hypotensive value at T3 (P < 0.05 vs. T0 and T2). In Group B, calculated TEO2 showed no change at T1 (vs. T0), whereas it significantly decreased by 33% at T2 (P < 0.01, vs. T1). PI value showed no significant change throughout the time course in the two groups (Table 2). Changes in haemodynamic, ABG data and calculated TEO2 are presented in Table 3.

Table 2
Table 2:
Changes in PI and Vmca in Groups A and B.
Table 3
Table 3:
Changes in haemodynamics, blood gas and TEO2 in Groups A and B.

No patient had any central neurological problems (cranial nerve deficits symptoms or findings of cerebral infarction) after surgery.


Our findings indicate that moderate HHD significantly increases MCA flow, while the combination of hypotensive anaesthesia to a MAP of approximately 55 mmHg induced with sevoflurane reduces the flow towards the pre-haemodilution condition.

In the present study, preoperative infusion volume and rate of HES in HHD were decided according to the report of Mielke and colleagues [2]. HHD during sevoflurane anaesthesia was performed safely without the risk of volume overload. HHD markedly decreased CaO2 but significantly increased Vmca. Bruder and colleagues [9] reported that isovolaemic haemodilution increased Vmca in anaesthetized (N2O–O2–isoflurane) patients with normocapnia, normotension and normothermia. There is an inverse relationship between CBF and Hct in the augmentation of CBF with haemodilution. The relationship may be explained by two mechanisms, i.e. an autoregulatory mechanism for maintaining constant oxygen delivery to the tissues and a direct haemorrheologic effect due to a decrease in blood viscosity. Although Hct value (25–26%) in our study was lower than that (30%) in Bruder and colleagues [9], our haemodilution also showed no significant change in calculated TEO2, and thus, the relationship between CBF and Hct could be maintained. Oxygen transport has been shown to be well maintained in spite of a decrease in Hct to about 20% as long as normovolaemia is maintained [11]. The accuracy of TCD is dependent on the change of the MCA diameter. The effect of haemodilution on large cerebral diameters is controversial, and Tu and Liu [12] reported that CBF augmentation during haemodilution might be due partially to vasodilation. However, the increased Vmca in our study is more likely attributable to the reduction in blood viscosity because PI values, an indicator of vascular resistance, showed no change after haemodilution indicating no vasodilatative effect. It is still possible that there was variation in cerebral vascular resistance and also the autoregulatory response in smaller arterial vessels and arterioles and not in the larger MCA where we measured flow rate.

It has been shown that the addition of N2O to sevoflurane anaesthesia in adults increases Vmca [13,14]. High concentrations of N2O have generally increased CBF and brain metabolism when given alone or with volatile anaesthetic. Cho and colleagues [13] used TCD in their study and found that sevoflurane reduced Vmca but the cerebrovascular CO2 reactivity was well maintained during sevoflurane with N2O anaesthesia at 1.2 MAC. Cardiac output may indirectly affect CBF via the baroreceptor reflex inducing an increase in cerebral vascular resistance. The mechanisms by which volatile anaesthetics impair CBF regulation in response to cerebral perfusion pressure (CPP) changes are still unclear. A lowered metabolism by sevoflurane will induce vasoconstriction in relation to the decrease in the metabolic demand whilst impaired autoregulation by sevoflurane will have the opposite effect. Although Werner and colleagues [15] reported that sevoflurane dose-dependently increased brain tissue nitric oxide (NO)2 and impaired CBF autoregulation during haemorrhagic hypotension in rats and suggested that sevoflurane impaired the autoregulatory capacity secondary to an increase of the perivascular NO availability, these findings would be unclear in hypervolaemic situation.

In the present study, the combination of hypotensive anaesthesia with sevoflurane reduced Vmca and TEO2 towards the pre-haemodilution condition. However, O2 supply could be maintained within normal ranges because the decreased TEO2 was similar in level of non-hypotension group. Higher concentrations of sevoflurane during hypotension did not alter the compensatory cardiovascular responses that result from the hypervolaemic haemodilution and in turn resulted in preservation of cerebral perfusion.

The mechanisms of the decrease in Vmca may be one of three possibilities or a combination: an increase in cerebral arteriolar tone and thereby cerebrovascular resistance; vasodilation with some loss of autoregulation; and a decrease of oxygen demand. First, cerebral arteriolar tone and cerebrovascular resistance probably did not increase because PI remained unchanged during the study. Second, Molnar and colleagues [16] found a direct vasodilatative effect on the MCA but sevoflurane maintains autoregulation, at least up to 1.5 [17]. Third, the reduced oxygen demand induced by the large doses of sevoflurane that was used for induction of hypotension would have resulted in an associated reduction in flow if coupling remained intact. Unfortunately, there are no human studies of autoregulation or oxygen consumption at >3 MAC and so we can only speculate.

In conclusion, during 1.2 MAC sevoflurane anaesthesia, acute hypervolaemic haemodilution to a Hct of 26% increased MCA flow velocity while sevoflurane-induced hypotension to a MAP of 55 mmHg during acute hypervolaemic haemodilution reduced MCA flow to pre-haemodilution levels during normocapnia in surgical patients.


1. Barbier-Böhm G, Desmons JM, Couderc E et al. Comparative effects of induced hypotension and normovolemic haemodilution on blood loss in total hip arthroplasty. Br J Anaesth 1980; 52: 1039–1043.
2. Mielke LL, Entholzner EK, Kling M et al. Preoperative acute hypervolemic hemodilution with hydroxyethylstarch; an alternative to acute normovolemic hemodilution? Anesth Analg 1997; 84: 26–30.
3. Fukusaki M, Nakamura T, Miyoshi H et al. Splanchnic perfusion during controlled hypotension combined with acute hypervolemic hemodilution: a comparison with combination of acute normovolemic hemodilution – gastric intramucosal pH study. J Clin Anesth 2000; 12: 421–426.
4. Fukusaki M, Sumikawa K. The combination of hemodilution and controlled hypotension: physiology and clinical application. J Anesth 2000; 14: 194–203.
5. Gottstein U. Normovolemic and hypervolemic hemodilution in cerebrovascular ischemia. Bibl Haematol 1981; 47: 127–138.
6. Gupta S, Heath K, Matta BF. Effect of incremental doses of sevoflurane on cerebral pressure autoregulation in humans. Br J Anaesth 1997; 79: 469–472.
7. Ishikawa H, Iwasaki K, Shiozawa T et al. Dynamic cerebral blood flow autoregulation during sevoflurane anesthesia and TIVA (in Japanese with English abstract). Masui (Jpn J Anesthesiology) 2003; 52: 370–377.
8. Bishop CC, Powell S, Rutt D, Browse NL. Transcranial Doppler measurement of middle cerebral velocity: a validation study. Stroke 1986; 17: 913–915.
9. Bruder N, Cohen B, Pellissier D, Francois G. The effect of hemodilution on cerebral blood flow velocity in anesthetized patients. Anesth Analg 1998; 86: 320–324.
10. Fukusaki M, Kanaide M, Inadomi C et al. Human middle cerebral artery flow velocity during controlled hypotension combined with hemodilution – transcranial Doppler study. J Clin Anesth 2005; 17: 177–181.
11. Laks J, Pilon RN, Klovekorn WP et al. Acute hemodilution: its effect on hemodynamics and oxygen transport in anesthetized man. Ann Surg 1974; 180: 103–109.
12. Tu YK, Liu HM. Effects of isovolemic hemodilution on hemodynamics, cerebral perfusion, and cerebral vascular reactivity. Stroke 1996; 27: 441–445.
13. Cho S, Fujigaki T, Uchiyama Y et al. Effects of sevoflurane with and without nitrous oxide on human cerebral circulation. Transcranial Doppler study. Anesthesiology 1996; 85: 755–760.
14. Aono M, Sato J, Nishino T. Nitrous oxide increases normocapnic cerebral blood flow velocity but does not affect the dynamic cerebrovascular response to step changes in end-tidal P (CO2) in humans. Anesth Analg 1999; 89: 684–689.
15. Werner C, Lu H, Engelhard K, Unbehaun N, Kochs E. Sevoflurane impairs cerebral blood flow autoregulation in rats: Reversal by nonselective nitric oxide synthase inhibition. Anesth Analg 2005; 101: 509–516.
16. Molnár C, Settakis G, Sárkány P et al. Effect of sevoflurane on cerebral blood flow and cerebrovascular resistance at surgical level of anaesthesia: a transcranial Doppler study. Eur J Anaesthesiol 2007; 24: 179–184.
17. Summors AC, Gupta AK, Matta BF. Dynamic cerebral autoregulation during sevoflurane anesthesia: a comparison with isoflurane. Anesth Analg 1999; 88: 341–345.


© 2008 European Society of Anaesthesiology