The cardiovascular side effects of volatile anesthetics are one of the chief causes of postoperative complications in children, and infants seem to be at the greatest risk for this. This study compared cardiovascular changes at equipotent concentrations of sevoflurane and halothane in infants.
Thirty infants classified as American Society of Anesthesiologists physical status I or II who required elective surgery were randomized to receive either halothane or sevoflurane for inhalation induction. Cardiovascular and echocardiographic data were recorded in both groups at baseline and at end-tidal concentrations of 1 and 1.5 minimum alveolar concentration (MAC).
Sevoflurane did not alter heart rate or cardiac index at all concentrations compared with awake values. Sevoflurane significantly decreased blood pressure and systemic vascular resistance compared with awake values at all concentrations. Shortening fraction and rate-corrected velocity of circumferential fiber shortening decreased at 1.5 but not at 1 MAC. Myocardial contractility assessed by stress-velocity index and stress-shortening index decreased significantly at all concentrations, but did not fall into the abnormal range at any concentration. Halothane caused a greater decrease in heart rate, shortening fraction, stress-shortening index, velocity of circumferential fiber shortening, stress-velocity index, and cardiac index at all concentrations than did sevoflurane.
Sevoflurane causes a lesser decrease in cardiac output than does halothane in infants.
Sevoflurane is a new halogenated volatile anesthetic that provides rapid induction and emergence from anesthesia compared with halothane and isoflurane.  It has the potential to be an ideal inhalational induction agent for children or infants because of its pleasant smell and nonirritating property. [2,3]
The cardiovascular side effects of volatile anesthetics are one of the chief causes of postoperative complications in children, and infants seem to be at the greatest risk for this.  Indeed, one of the main problems of using halothane for induction was a deleterious cardiovascular effect in infants. [5,6] For sevoflurane, dose-related depression of left ventricular function or cardiac output has been reported in adults  and children,  but no data were available in infants.
This study compared dose-response cardiovascular changes at equipotent (equal minimum alveolar concentration [MAC]) concentrations of halothane and sevoflurane in infants.
Materials and Methods
Thirty infants classified as American Society of Anesthesiologists physical status I or II who required elective surgery were studied after the protocol was approved by the Human Studies Committee and informed written parental consent was obtained. The infants were were allowed to eat or drink nothing for 4 h before operation and were premedicated 30 min before induction of anesthesia with 0.3 mg/kg rectal midazolam. After baseline cardiovascular measurements, infants were randomly assigned to receive either halothane (HALO) or sevoflurane (SEVO) for induction and maintenance of anesthesia. An inhaled induction of anesthesia was started with allocated gas in a mixture of oxygen and nitrous oxide (50:50) through an open circuit without soda lime absorber. In both groups, the inspired concentration was increased every five breaths in the following order: 1%, 2%, 3%, and 3.5% in the HALO group, and 2%, 4%, 6%, and 7% in the SEVO group. Infants were breathing spontaneously during induction until endotracheal intubation. Anesthetic gas concentrations and carbon dioxide concentrations were measured from gas samples continuously aspirated from an elbow connector subsequently attached to the endotracheal tube. Inspired and expired gas concentrations (carbon dioxide, oxygen, nitrogen, halothane, or sevoflurane) were recorded every minute.
After placement of an intravenous line, the trachea was intubated and the lungs were fitted with a ventilator to maintain normocapnia. Inspired MAC anesthetic concentrations were adjusted to maintain end-expired concentrations at 1 MAC and 1.5 MAC (adjusted for age). [9–11] When the inspired and expired concentrations of halothane or sevoflurane were stable for 10 min at the two concentrations, two new sets of cardiovascular data were collected during a 2-min period.
At each study period, the following measurements were recorded by the same observer: heart rate (beats/min), blood pressure (mmHg) measured using an automated blood pressure cuff, pulsed-Doppler, and two-dimensional transthoracic echocardiographic data (Hewlett-Packard Sonos 1000; Andover, MA). The echocardiographic data obtained in each patient included the measurement of aortic diameter and shortening fraction (SF). In the long axis view of the left ventricular outflow tract (using the M mode), the end-systolic internal aortic diameter ([null set] Ao) was measured, immediately distal to the aortic valves, and the aortic sectional area (SaAo) was calculated as (SaAo = pi ([null set] Ao) 2/4). The SF was measured by M mode from the parasternal long axis view of the left ventricle at the junction of mitral valve leaflets and the papillary muscle. The left ventricular end-diastolic diameter (LVDD) was measured at the point of maximal diameter and the left ventricular end-systolic diameter (LVSD) was measured at the point of peak of upward deflection of the posterior wall. Posterior wall thickness (PWes) was measured at end-systole. The SF was calculated as (LVDD-LVSD)/LVDD and expressed as a percentage. Rate-corrected velocity of circumferential fiber shortening (VCFc) was calculated using the formula VCFc = SF/rate-corrected ejection time. Rate-corrected ejection time = ejection time divided by the square root of the R-R interval (to correct to a heart rate of 60 beats/min). Left ventricular end-systolic wall stress (ESWS) was calculated using the formula ESWS =(1.35 x MBP x LVSD)/[(1 + PWes/LVSD) x (PWes) x 4]. 
Pulsed-Doppler scanning was used to measure flow in the ascending aorta from the apical four-chamber view of the heart.  Pulsed-Doppler studies were performed with the same scanner as in the other measurements using 5-MHz transducers with duplex imaging. Sample volume was positioned with duplex imaging in the aorta immediately distal to the aortic valves to measure maximal flow velocities. The audio signal intensity was used to confirm positioning for maximal aortic blood flow velocities. Because the angle between the estimated direction of blood flow and the Doppler beam was 15 degrees or less, no angle correction of the doppler signal was made. The mean aortic flow velocity (Vao) was calculated using the software of the ultrasound system as average velocity x flow period. The average of three consecutive flow velocity integrals was taken. Cardiac Index (CI) was calculated from the volumetric equation: CI (ml [center dot] kg sup -1 [center dot] min sup -1)= Vao (cm/s) x SaAo (cm2) x 60/body weight (kg). Systemic vascular resistance (SVR) was evaluated as the quotient of the mean blood pressure and CI without measurements of right atrial pressure.
The VCFc-ESWS relation or SF-ESWS relation can be used to determine contractility independently of loading condition. However, SF, VCFc, and ESWS were age and body surface area dependent. As described by Colan et al.,  Z scores rather than absolute values were used to normalize these calculations, with age- and growth-related means taken from values established for healthy children or infants. The stress-velocity index for each infant was determined relative to distribution of this index in healthy persons and calculated as a normal deviate (Z score). Stress velocity index =(X1 - M)/SD, X1 = measured VCFc, M = mean VCFc calculated for the measured ESWS, and SD = standard deviation of the regression group stress-velocity index. The stress-shortening index was correspondingly quantified as the normal deviate of fractional shortening for the given ESWS, obtained in a manner analogous to that described for the stress-velocity index. The normal value of Z score was =-2 to +2. All the echocardiographic data were calculated again by a second observer in a blinded manner.
Continuous data, parametrically distributed between groups, were analyzed using one-way analysis of variance. Analysis of variance for repeated measures was used for multiple treatments within the same group; to compare two groups with multiple treatments within groups, a two-way analysis of variance with repeated measures was used. Probability values for multiple comparisons were corrected with Bonferroni multiple comparison test. A probability value less than 0.05 was considered significant. Values are expressed as mean +/- SD.
Of the 30 infants enrolled in this protocol, 15 received halothane and 15 received sevoflurane. The treatment groups were comparable with regard to age, sex ratio, and weight (Table 1). They were also comparable with regard to baseline values of cardiovascular measurements (Figure 1, Figure 2(A and B), Table 2). End-tidal halogenated gas concentrations measured were, respectively, 1.04 +/- 0.08 MAC in the HALO group and 0.98 +/- 0.09 MAC in the SEVO group at the 1 MAC study level and 1.54 +/- 0.11 MAC in the HALO group and 1.46 +/- 0.04 MAC in the SEVO group at the 1.5 MAC study level.
Cardiovascular Effects of Halothane
Halothane significantly decreased systolic, and mean aortic blood pressures (SBP, MBP), SF, VCFc (Table 2), and CI (Figure 1), but it did not alter LVDD and SVR at all concentrations compared with awake values (Table 2). End-systolic wall stress decreased compared with awake values at 1.5 MAC, and heart rate decreased compared with awake values at all concentrations (Table 2). The stress-velocity index, or VCFc-ESWS relation, and the stress-shortening index, or SF-ESWS relation, decreased to less than the normal range at all concentrations (Figure 2(A and B)).
Cardiovascular Effects of Sevoflurane
Sevoflurane did not alter heart rate, LVDD (Table 2), or CI (Figure 1) compared with awake values. Sevoflurane significantly decreased SBP, MBP, and SVR compared with awake values at all concentrations. Shortening fraction and VCFc decreased at 1.5 but not at 1 MAC. End-systolic wall stress decreased compared with awake values at 1.5 MAC (Table 2). Stress-velocity index and stress-shortening index decreased significantly at all concentrations but did not fall into the abnormal range at any of the concentrations (Figure 2(A and B)).
Cardiovascular Effects of Sevoflurane Compared with Those of Halothane
When treatment groups were compared directly, halothane caused a greater decrease in heart rate, SF, stress-shortening index, VCFc, stress velocity index, and CI at all concentrations and in SBP at 1.5 MAC, than did sevoflurane (Figure 1and Figure 2(A and B);Table 2). No significant difference was found in LVDD and ESWS at any of the concentrations (Table 2).
Systolic blood pressure decreased less in the sevoflurane group than in the halothane group, with no change in left ventricular cardiac output. Indeed, blood pressure is one of the main parameters to determine hemodynamic status in children. However, to evaluate the possible differences between the blood pressure effects of the volatile anesthetics, a careful examination of the two key determinants of blood pressure, cardiac output and peripheral resistance, is required. 
Effects of Halothane and Sevoflurane on Contractility
Sevoflurane appeared to have a direct myocardial depressant effect [16–18]; however, it has been shown that sevoflurane causes less depression of the inotropic state than does halothane in children. 
In a human study, the pattern of unchanged echocardiographic ejection-phase indices of left ventricular function with unchanged preload and decreased afterload also indicates decreased myocardial contractility with sevoflurane.  The ejection-phase indices would be expected to increase after the decreasing afterload if contractility was maintained.  In this study, we observed effects similar to those that Malan et al.  found in the sevoflurane group, indicating a decrease in contractility. The SF and VCFc (indices of left ventricular function) decreased at 1.5 MAC, whereas SVR and ESWS (indices of afterload) decreased significantly at all concentrations and at 1.5 MAC, respectively.
The relations between VCFc and ESWS (stress velocity index) and between SF and ESWS (stress-shortening index) were sensitive measures of contractility; these indices can distinguish the changes in contractile state from the alteration in loading condition.  Nevertheless, normal values were different according to age [22,23]; thus we reported, as Holzman et al.  did, the contractility indices as Z-scores rather than as absolutes values. Using these indices, contractility decreased significantly in both groups in this study and fell into the abnormal range at all concentrations in the halothane group, unlike sevoflurane.
The decrease in contractility that we observed probably was not exclusively dependent on volatile anesthetics (halothane or sevoflurane), because both anesthetics were used in a mixture of oxygen and nitrous oxide (50:50) and the direct negative inotropic effect of nitrous oxide was previously shown by Pagel et al.  This may partly explain why both agents had more cardiovascular depressant effects than those measured in a study of children, in whom nitrous oxide was discontinued during induction and echocardiographic measurements.  However, Malan et al.  showed that sevoflurane administered in 60% nitrous oxide:40% oxygen caused a smaller decrease in mean arterial pressure at 1 MAC than did an equipotent concentration of sevoflurane in 100% oxygen, although neither CI nor SVR independently were higher with sevoflurane:nitrous oxide than with sevoflurane alone.  On the other hand, the direct negative inotropic effect of nitrous oxide was demonstrated by Pagel et al.  with isoflurane, and the direct extrapolation with sevoflurane was critical. Indeed, with or without nitrous oxide, sevoflurane appeared to cause less contractility than did halothane in children and in infants.
Effects of Halothane and Sevoflurane on Afterload
Stroke volume (cardiac output) was not primarily dependent on contractility but also on loading conditions. The effects of sevoflurane on SVR are somewhat more controversial. [18,25] Previous investigations have shown that sevoflurane did not alter calculated SVR in experimental animals. [25,26] In contrast, other studies have implied that sevoflurane reduced SVR with a decrease in arterial blood pressure. [8,27,28]
In this study, the lower decrease in contractility combined with the greater decrease in SVR may be explained by the fact that CI remained unchanged at all concentrations in the sevoflurane group but not in the halothane group. Nevertheless, ESWS decreased significantly only at 1.5 MAC, and no difference appeared between groups. Thus prevention of a decrease in cardiac output did not depend on a major reduction in afterload in sevoflurane, but mainly on the stability of the heart rate when increasing the alveolar concentration of sevoflurane.
Effects of Halothane and Sevoflurane on Heart Rate
Heart rate is one of the main factors of cardiac output in infants. Indeed, the increase or the absence of change in heart rate can play an important role in maintaining a correct level of cardiac output in pediatric anesthesia. [5,29] Halothane anesthetic induction decreased heart rate in infants, from 18% to 30%.  In an echocardiographic study, Murray et al.  showed that halothane at 1 and 1.5 MAC significantly decreased heart rate, stroke volume, and cardiac output in infants. However, heart rate and cardiac output increased without change in stroke volume or in ejection fraction after 0.02 mg/kg atropine given to infants.
In this study, heart rate was stable in the sevoflurane group, as previously reported in children. [8,30] Sarner et al.  reported that children receiving halothane tended to have a decrease in heart rate, whereas children receiving sevoflurane maintained or increased heart rate. The increase in heart rate with sevoflurane during induction of anesthesia has been previously reported by Piat et al.  and by Mori and Suzuki  in children, but it was mainly dependent on age. Lerman et al.  showed that heart rate was unchanged at 1 MAC sevoflurane compared with awake values in infants and children as old as 3 yr but increases > 10% above awake values in children older than 3 yr.
In conclusion, sevoflurane decreased cardiac output less than halothane did in infants. The minor decrease in contractibility was compensated by a greater decrease in SVR and mainly by no change in heart rate compared with halothane in infants. Therefore, a wide epidemiologic study is needed to determine if sevoflurane appears safer than halothane for inducing anesthesia in infants.