Secondary Logo

Journal Logo

Esmolol-Induced Regional Wall Motion Abnormalities Do Not Affect Regional Ventricular Elastances

Strum, David P. MD, FRCP(C); Pinsky, Michael R. MD, CM

doi: 10.1213/00000539-200002000-00004
CARDIOVASCULAR ANESTHESIA
Free
SDC

The effect of regional wall motion abnormalities (RWMA) on regional and global left ventricular (LV) elastances has not been defined. To induce RWMA, we infused esmolol (9 mg) into the left anterior descending coronary artery in eight anesthetized open-chest canine preparations. Global and regional stroke volumes and end-systolic pressure-volume relationships (LV conductance catheter) and pressure-length relationships (sonomicrometer crystals) were measured in dysfunctional (apical) and normal (basilar) LV regions at baseline, during esmolol infusion, and after treatment with a systemic dobutamine infusion (4 μg · kg−1 · min−1) and combined dobutamine-esmolol. Esmolol induced apical dyskinesis, as evidenced by reduced effective apical stroke volumes and stroke work, a parallel right-shift of the apical regional, global end-systolic pressure-volume relationships, and increased regional and global LV volumes (P < 0.05). However, global and regional elastances remained unchanged. Dobutamine increased global and apical regional elastances, but did not increase regional volumes. During the infusion of combined esmolol-dobutamine, apical elastance and volumes increased compared with baseline (P < 0.05), but apical regional stroke work decreased. Thus, esmolol-induced RWMA were associated with cardiac dilation but not with decreased regional or global elastances.

Implications Regional and global elastances and maximal stroke volumes may not identify esmolol-induced left ventricular regional dysfunction in dogs. The primary effect of asynchrony of regional contraction is global cardiac dilation. Systemic dobutamine infusion increases regional and global left ventricular elastances but does not reverse regional wall motion abnormality-induced cardiac dilation.

Cardiopulmonary Research Laboratory, Department of Anesthesiology and Critical Care Medicine, University of Pittsburgh, Pittsburgh, Pennsylvania

September 28, 1999.

This work was supported in part by the Veterans Administration, Pittsburgh, PA.

Address correspondence and reprint requests to David P. Strum, MD, Department of Anesthesiology, 4301 West Markham St., Slot 515, Little Rock, AR 72205. Address e-mail to dpstrum@life.uams.edu.

Regional myocardial dysfunction is a common clinical finding usually manifesting as regional wall motion abnormalities (RWMA). RWMA reduce the efficiency of ventricular pump function by decreasing both the regional amplitude of contraction and global synchrony of ventricular contraction among regions of the myocardium (1). Previous investigators have produced RWMA by pacing the heart at various points in the ventricle (2) or by coronary occlusion (3–5) and measured left ventricular (LV) volumes with sonomicrometer crystals. In those studies, RWMA resulted in a rightward shift of the LV end-systolic pressure-volume relationship (ESPVR), but the slope of the ESPVR was unaltered. The slope of the LV ESPVR is referred to as “end-systolic elastance” (Ees). Thus, the volume intercept of the ESPVR appears to increase as contractile asynchrony increases. However, the relationship between global and regional LV volumes and their resultant stroke work (SW) in the setting of RWMA is unknown. Echocardiographic analysis of ventricular function is now commonplace and uses measures of LV areas to estimate volume when assessing ventricular function. However because echocardiography measures LV function in only a single plane, it is dependent on the ability of regional LV function to describe global LV performance. It is unclear, however, the extent to which asynchrony of contraction in an observed or remote region of the myocardium will alter either global LV performance or its assessment in any one region. We hypothesized that the effect of regional myocardial paresis would not be well quantified by measures of LV Ees, such that profound alterations in contractile synchrony (RWMA) would not be appreciated by this commonly used measure of LV contractility. Accordingly, we studied the effect of esmolol-induced apical RWMA on the interaction between global and regional volumes and Ees, as well as the effect of treatment with systemic dobutamine on these interactions in an acute canine preparation.

Back to Top | Article Outline

Methods

Surgical Preparation

After approval of our Animal Care and Use Committee, we studied eight mongrel dogs (24.6 ± 0.5 kg, mean ± SEM). Anesthesia was induced with IV morphine sulfate (0.5 mg/kg) and sodium pentobarbital (30 mg/kg) and maintained with continuous infusion of sodium pentobarbital (1.0 mg · kg−1 · hr−1) with intermittent boluses if needed for light anesthesia and a maintenance infusion of 0.9% NaCl at 100 mL/hr. All animals were tracheally intubated and their lungs ventilated with 100% oxygen. Blood gas tensions were measured before and after the experimental protocol and more frequently if hemodynamic instability developed. Acid-base status was adjusted with intermittent boluses of bicarbonate solution as needed to maintain arterial blood pH between 7.35 and 7.45, and the ventilator was adjusted to maintain arterial blood PCO2 between 35 and 45 mm Hg. Neuromuscular blockade was established with IV boluses of pancuronium (1.0 mg IV as needed). Body temperature was maintained at 36° to 38°C by using a thermostatically controlled heating blanket.

A fluid-filled arterial catheter with multiple side holes was inserted into the thoracic aorta via the left femoral artery to measure arterial pressure. A double-lumen, 12F catheter was inserted into the left external jugular vein to administer IV fluids. A 7.5F balloon-tipped pulmonary artery catheter (List #5328, Model #P7110-EP8-H, 9510 animal model; Baxter-Edwards, Santa Ana, CA) was introduced through the right external jugular vein to measure cardiac outputs. To record LV volume-time relationships, a 6F, 11-pole, multi-electrode, dual-excitation conductance catheter (Webster Laboratories, Irvine, CA) was inserted through the aortic valve and into the LV via the right internal carotid artery. Intermittent fluoroscopy and continuous inspection of the regional volume-time signals confirmed the position of the conductance catheter. To induce rapid but reversible decreases in LV volumes necessary for construction of ESPVR, as described below, we performed inferior vena cava occlusions using an intraluminal Fogarty 43-mL balloon occlusion catheter (model 62-080-8/22F; Baxter Healthcare Corporation, McGraw Parker, IL) inserted to the level of the supradiaphragmatic inferior vena cava via a femoral venous site.

To gain surgical access, a thoracotomy was performed through the fourth left intercostal space and a pericardotomy was performed over the anterior surface of the heart. To administer intracoronary esmolol, a 22-gauge, 1.25-in. IV catheter was placed into the left anterior descending (LAD) coronary artery. The intracoronary catheter was heparin-locked to prevent thrombus formation, and care was taken to prevent air embolism. Two independent pairs of piezoelectric dimension crystals (Triton, San Diego, CA) were introduced into different coronary artery perfusion zones (LAD and circumflex). Both pairs of crystals were aligned parallel with the horizontal hoop fibers of the LV. The crystals of each pair were inserted using rubber bumpers to a depth of 1.5 mm. The crystals of each pair were positioned 1 cm apart. To document regional dysfunction, one pair of crystals was positioned in the LAD coronary arterial perfusion zone distal to the insertion of the 22-gauge catheter (near the LV apex). To document function of normal myocardium, a second pair of control crystals was inserted in the perfusion zone of the left circumflex coronary artery.

A 5F, high-fidelity pressure transducer (Micro-tip catheter model SPC-350; Millar, Houston, TX) was inserted 2 cm through an apical puncture in the LV to record LV pressure. Immediately before termination of the experiment, 2 mL of 1% crystal violet dye was injected into the LAD coronary artery catheter to document the size of the perfusion zone. The dog was then killed, while still under general anesthesia, by using an IV injection of 10 mL of saturated KCl, and the heart was removed to document the position of the conductance catheter in the LV. The dyed area (LAD coronary artery perfusion zone) was dissected out and weighed postmortem. The remaining LV myocardium was dissected free and weighed.

Back to Top | Article Outline

Calibration of the Conductance Catheter

The conductance catheter method of Baan et al. (6) for measuring ventricular volume has been described and validated previously (6–8). Parallel conductance artifact was determined by using the saline dilution method (9) using 5 mL of 10 N saline injected into the right atrium.

Back to Top | Article Outline

Esmolol-Induced Regional Dysfunction

Nine milligrams of esmolol (10 mg/mL) was administered by bolus through the LAD catheter to produce transient pharmacological RWMA, as confirmed in real time by observation of the associated regional pressure-dimension loop. This dosage of esmolol was used to induce marked reversible RWMA without significant systemic effects. This treatment is referred to as esmolol. After approximately 30 s, the bolus infusion of esmolol induced stable regional dysfunction lasting at least 6 min. Data were collected 4–5 min after the induction of stable apical regional dysfunc- tion.

Back to Top | Article Outline

Dobutamine-Induced Increased Inotropy

Dobutamine HCl (4 g · kg−1 · min−1) was infused into a large central vein and induced a stable hemodynamic response within approximately 5 min (dobutamine). This dosage of dobutamine was used because it reproducibly increased global contractility under baseline conditions without inducing changes in heart rate or mean arterial pressure. Data were collected during dobutamine infusion 10–15 min after beginning the infusions. This treatment is referred to as dobutamine. In the treatment referred to as dobutamine-esmolol, once a stable dobutamine infusion response was established, esmolol was infused as a bolus into the LAD to induce RWMA, and data were collected 4–5 min later, once a new hemodynamic steady state was achieved. This treatment was referred to as dobutamine-esmolol.

Back to Top | Article Outline

Protocol

Regional myocardial function was compared sequentially over three treatments and four baseline conditions: baseline, dobutamine, baseline two, esmolol, baseline three, dobutamine-esmolol, and baseline four. Before each experimental treatment, baseline values for measured variables for regional function were allowed to return to baseline. Each of the three experimental treatments was alternated in a randomized block design to control for the order of the treatments and any possible time effects.

Back to Top | Article Outline

Data Processing

We recorded these experimental variables: arterial blood pressure, LV pressure, LV dP/dt, intercrystal dimensions for the piezoelectric crystals, total and regional LV impedance (volume), and lead II electrocardiogram. All data were recorded on an eight-channel physiologic recorder (Gould, Cleveland, OH). Additionally, all data were digitized and recorded on magnetic disc for subsequent analysis as described below.

Signals from the two pairs of sonomicrometer dimension crystals were processed simultaneously on two synchronized channels using a four-channel sonomicrometer (model 120; Triton San Diego, CA). dP/dt was derived from the LV pressure signal by using a differentiator (Gould) with a 100-Hz high-cutoff and a 1-V/s slope. Signals were collected and processed using a conductance catheter data processor and signal conditioner (Leycom Sigma 5DF; Leycom Sigma Leyden, Netherlands). A four-electrode calibration chamber was used to determine blood conductivity (α value). Cardiac output was determined in triplicate using a 10-mL iced 0.9% saline bolus thermodilution technique (model 9520 cardiac output computer; Baxter-Edwards) before starting the protocol to confirm SV derived from the conductance catheter. The signal acquisition and analysis system consisted of an Apollo DN4000 Unix workstation (Hewlett-Packard, Palo Alto, CA) and an analog to digital subsystem (SignifiCAT® RTS-132, Buffalo, NY).

Two pressure-dimension loops for the sonomicrometer crystals, four regional pressure-volume loops, and a global pressure-volume loop were compared simultaneously on a large video screen to aid in defining steady-state conditioning. Post hoc analysis of myocardial function was done by replaying digitized data stored on magnetic hard disc. Regional and global ESPVR were calculated by the iterative technique by using data from inferior vena caval occlusion runs and the method of least squares as previously described (3). The slope of the ESPVR was taken as Ees, whereas the position of the ESPVR on the volume axis was taken to reflect shifts in ESPVR independent of Ees. LV regional or global LV SW was calculated as the area within the regional or global pressure-volume loop, respectively, over one cardiac cycle. To quantify the contribution of regional stroke volume (SV) to total LV ejection, we determined regional ejection by two different methods. Maximal SV was defined as the difference of the maximal and minimal regional volume (or segmental length) for each cardiac cycle independent of the phase of regional contraction during apnea. Effective SV was measured for each region using maximums and minimums of the total LV volume as gated markers of the time when regional contraction would contribute to total LV ejection. If regional systole is not synchronous with global systole, then part of the ejection of some regions may occur either before or after total end-ejection. Accordingly, not all of the regional SV may contribute to total LV SV.

All variables were examined graphically by using standardized normal probability plots to ensure that the data were normally distributed. All results were reported as mean ± SEM. Regional and global pressure-volume relationships Ees, maximal and minimal volumes, stroke force, and SW were compared across treatments and regions of the myocardium by using repeated measures analysis of variance and 95% confidence interval testing (10). Post hoc analysis was accomplished using Tukey’s comparisons. Significance reports a difference corresponding to a P value < 0.05.

Back to Top | Article Outline

Results

The esmolol bolus infusion affected the apical regional myocardium, averaging 23% of the total LV mass (perfused zone 29.8 ± 2.63 g; left ventricle 127.9 ± 13.5 g). The experimental protocol compared baseline with systemic dobutamine, regional esmolol, and regional esmolol-systemic dobutamine. Intermediate baseline values did not differ from one another; thus, only initial baseline values were reported.

Back to Top | Article Outline

Systemic Hemodynamics

Systemic hemodynamics, with a few exceptions, remained unchanged for all treatment protocols (Table 1). Heart rate was unaffected by either esmolol or dobutamine. Dobutamine increased LV end-systolic pressure 13% from baseline (114 ± 6 to 129 ± 3 mm Hg, P < 0.05), increased maximal LV dP/dt 46% (1804 ± 113 to 2627 ± 267 mm Hg/s, P < 0.05), and decreased maximal negative LV dP/dt 63% (−2209 ± 80 to −3606 ± 401 mm Hg/s, P < 0.05). Esmolol decreased mean arterial pressure 9% (92 ± 2 to 84 ± 3 mm Hg, P < 0.05), increased LV end-diastolic pressure 125% (1.8 ± 0.5 to 4.2 ± 0.8 mm Hg, P < 0.05), and decreased maximal negative LV dP/dt 12% (−2209 ± 88 to −1938 ± 105 mm Hg/s, P < 0.05). Dobutamine-esmolol reversed the changes in mean arterial pressure and maximal negative LV dP/dt associated with esmolol.

Table 1

Table 1

Back to Top | Article Outline

Regional Volumes

Regional maximal SV was not reduced by either esmolol or dobutamine-esmolol (Figure 1A). In fact, basal SV increased under both conditions as compared with control (P < 0.05). However, effective SV decreased markedly for the apical region during esmolol (73 ± 12% of control, P < 0.05), and this was not improved during dobutamine-esmolol (75 ± 10% of control, P < 0.05) (Figure 1B). Sonomicrometer crystal data demonstrated that the involved apical segment became dyskinetic during esmolol (elongating during systole) and remained hypokinetic during dobutamine-esmolol. Thus, esmolol induced a profound regional dyskinesis.

Figure 1

Figure 1

Systemic dobutamine alone did not alter regional or global maximal or minimal volumes. Intracoronary esmolol, however, increased both maximal and minimal volumes in the affected and normal regions (Table 2). Esmolol increased apical, chordal, basal, and global maximal volumes and increased apical, basal, and global minimal volumes (P < 0.05). Dobutamine-esmolol did not reverse the esmolol-induced increases in apical regional volumes. Compared with baseline, dobutamine-esmolol increased apical and papillary regional volumes (P < 0.05), but not global volumes. Sonomicrometer crystal dimensions were unchanged by treatment with dobutamine and dobutamine-esmolol. Esmolol decreased apical regional SW by 45%–67% (P < 0.05) and decreased SW in the otherwise normal basal region (P < 0.05) (Table 3).

Table 2

Table 2

Table 3

Table 3

Back to Top | Article Outline

ESPVR

Esmolol induced a parallel shift of the ESPVR to the right (with volume on the x axis) for the apical and chordal regions and the global LV, such that affected LV regional and global volumes increased but neither regional nor global LV Ees were affected. The global LV pressure-volume relationships during baseline and esmolol administration for one animal are illustrated during inferior vena caval occlusion in Figure 2, which also illustrates that this parallel shift in LV ESPVR without a change in Ees resulted in an increased LV end-diastolic volume butdoes not affect LV SV. Data from sonomicrometers (Figure 3) and regional volumes (Figure 4) demonstrate changes qualitatively similar to those described for global LV volume. All regional Ees increased with systemic dobutamine (Table 3), but only the apical and global LV Ees reached statistical significance (P < 0.05). Regional Ees were unaltered by esmolol, but dobutamine-esmolol increased apical regional Ees compared with esmolol alone. Sonomicrometer-derived Ees were not altered by any treatment in the protocol. Although sonomicrometer data (Figure 3A) show collapse of the pressure-length loop, when apical regional volume (which includes paresed and normal myocardium) is analyzed, only a parallel shift of the ESPVR to the right is observed (Figure 4A).

Figure 2

Figure 2

Figure 3

Figure 3

Figure 4

Figure 4

Back to Top | Article Outline

Discussion

This study demonstrates that global and regional Ees and maximal SV can be insensitive measures of regional myocardial dysfunction, even when measured in regions of the heart involving RWMA. At the same time, segmental pressure-length relationships within the involved regions of esmolol-induced paresis demonstrated profound dysfunction. These observations are clinically important because commonly available echocardiographic techniques also measure regional LV function. Thus, if significant RWMA occur, their presence may not be indicated by analysis of regional Ees alone.

Theoretically, esmolol-induced asynchrony would result in LV dilation solely on a mechanical basis. Furthermore, treatment with dobutamine, an inotrope commonly used to treat contractile dysfunction, although it also increased global and regional Ees in our model, did not reverse esmolol-induced LV dilation because the asynchrony was not reversed. Therefore, although regional Ees may be a sensitive measure of contractility, it does not identify LV dilation or impaired SW created by esmolol-induced RWMA.

Dilation of the affected regions without change in Ees will result in a decreased ejection fraction. However, the apparent response of the affected region to changes in afterload appears to be depressed because of the rightward shift of the regional ESPVR (Figures 2–4). Accordingly, the parallel shift of regional and global ESPVR without measurable changes in regional or global Ees suggest that asynchrony of contraction plays a primary role in determining overall LV volume in our model.

The extrapolation of these data to coronary ischemic conditions is questionable. End-systolic pressure-length relationships by using ultrasonic crystals are difficult to interpret in the presence of coronary occlusion (11). In two previous studies, coronary occlusion increased the slope of the ESPVR (12,13). Our data, however, were similar to those of Sunagawa et al. (3) and Lawrence et al. (5) who showed that the slope of the global ESPVR was unaltered by coronary occlusion and that only the volume intercept shifts to the right. Similar studies, using nonischemic RWMA, also indicate that regional Ees are particularly insensitive measures of regional RWMA (2,14).

We chose not to measure the extrapolated zero pressure intercept of the ESPVR. Early studies of LV systolic function assumed that a constant LV volume intercept existed at zero pressure when calculating LV ESPVR (15). This assumption was subsequently shown to be incorrect, and the leveraged intercept of the ESPVR with the volume axis was shown to be highly variable (16). In order to minimize variability in our volume measures, we chose to directly measure end-diastolic and end-systolic volumes.

We demonstrated that echocardiographic imaging techniques may overestimate effective regional SW if not gated to global LV ejection (8). The regional pressure-volume loop area can also be separated into total SW (area within the pressure-dimension loop) and effective SW (that portion of the work contributing to ejection of blood from the ventricle) (17,18). The difference between total and effective regional work reflects asynchrony of regional contraction associated with RWMA. In the present study, total regional SW was reduced 45%–67% by esmolol infusion. Because asynchrony occurred (Figure 1B), if we had calculated effective regional SW, we would have seen an even greater decrease (Figures 3 and 4).

Recently, attention has focused on parallel shifts of the ESPVR rather than changes in the Ees, and these shifts have been inferred to reflect changes in LV contractility (19). Our data support this concept by demonstrating that, when a portion of contracting myocardium is rendered dysfunctional by esmolol infusion, parallel shifts in the ESPVR can occur without measurable changes in Ees. Such parallel shifts in the ESPVR may reflect asynchrony among regions, suggesting that regional asynchrony may be an important mechanism in determining global contractility.

Back to Top | Article Outline

Limitations

Some of the regional dysfunction that we observed may be explained on the basis of normally occurring regional asynchrony and not solely by esmolol-induced RWMA. LV contraction is normally heterogeneous (19). Although we compared regional volumes and ESPVR along the LV longitudinal axis, this regional separation is neither anatomically correct nor reflective of the distribution of the induced RWMA. Specifically, esmolol-induced RWMA affected only a portion of the apical and chordal regions with some myocardium in both regions unaffected, as demonstrated by the crystal violet dye distribution. Finally, in support of our method of analysis, systemic dobutamine infusion affected all regions, whereas intracoronary esmolol infusion only affected apical regions.

We previously confirmed the accuracy of relative conductance volumes in our laboratory (17), as verified by others both in vivo and in vitro (20,21). Thus, it is unlikely that observed parallel shifts of the LV ESPVR reflect treatment-specific parallel conduction artifacts.

The authors would like to thank John Melick, Brian Ondulick, and John Lutz for their expert technical assistance.

Back to Top | Article Outline

References

1. Brutsaert DL. Nonuniformity: a physiologic modulator of contraction and relaxation of the normal heart. J Am Coll Cardiol 1987; 9:341–8.
2. Park RC, Little WC, O’Rourke RA. Effect of alteration of left ventricular activation sequence on the left ventricular end-systolic pressure-volume relation in closed-chest dogs. Circ Res 1985; 57:706–17.
3. Sunagawa K, Maughan L, Sagawa K. Effect of regional ischemia on the left ventricualr end-systolic pressure-volume relationship of isolated canine hearts. Circ Res 1983; 52:170–8.
4. Little WC, O’Roarke RA. Effect of regional ischemia on the left ventricular end-systolic pressure-volume relation in chronically instrumented dogs. J Am Coll Cardiol 1985; 5:297–302.
5. Lawrence WE, Maughan WL, Kass DA. Mechanism of global functional recovery despite sustained postischemic regional stunning. Circulation 1992; 85:816–27.
6. Baan J, Van Der Velde ET, Steendijk P, Koops J. Calibration and application of the conductance catheter for ventricular volume measurement. Automedica 1989; 11:357–65.
7. Kay HR, Afshari M, Bararash P, et al. Measurement of ejection fraction by thermodilution techniques. J Surg Res 1983; 34:337–46.
8. Gorcsan J, Strum DP, Mandarino W, et al. Quantitative assessment of alterations in regional left ventricular contractility by color-coded tissue Doppler echocardiography: comparison with sonomicrometry and pressure-volume relations. Circulation 1997; 95:2423–33.
9. Baan J, Van Der Velde ET, De Bruin H, et al. Continuous measurement of left ventricular volume in animals and man by conductance catheter. Circulation 1984; 70:812–23.
10. Wilkinson L. SYSTAT: the system for statistics. Version 5.2.1. Evanston, IL: SYSTAT Inc., 1992.
11. Foëx P, Francis CM, Cutfield GR, Leone B. The pressure-length loop. Br J Anaesth 1988; 60:65S–71S.
12. Goldberg AH, Diedericks J, Hopwood MB, Foëx P. Canine end-systolic pressure-length relationships: depressed by diltiazam, invalidated by ischemia. J Cardiothorac Vasc Anesth 1992; 6:593–9.
13. Victory JGG, Foëx P, Leone BJ. End-systolic pressure-length relations in halothane anesthesized dogs: effects of regional ischemia and inotropic agents. Cardiovasc Res 1991; 25:110–7.
14. Foëx P, Leone BJ. Pressure-volume loops: a dynamic approach to the assessment of ventricular function. J Cardiothorac Vasc Anesth 1994; 8:84–96.
15. Suga H, Sagawa K, Shoukas AA. Load independence of the instantaneous pressure-volume ratio of the canine left ventricle and effects of epinephrine and heart rate on the ratio. Circ Res 1973; 32:314–22.
16. Crozatier B, Caillet D, Thuillez C, et al. Preload changes modify systolic pressure-diameter relations in the conscious dog. Am J Physiol Heart Circ Physiol 1981; 240:H354–60.
17. Gorcsan J, Denault AY, Mandarino WA, Pinsky MR. LV pressure-volume relations using transesophageal automated border detection: comparison with conductance catheter technique. Am Heart J 1996; 131:544–52.
18. Safwat A, Leone BJ, Norris RM, Foëx P. Pressure-length loop area: its components analyzed during graded myocardial ischemia. J Am Coll Cardiol 1991; 17:790–6.
19. Diedericks J, Leone BJ, Foëx P. Regional differences in left ventricular wall motion in the anesthetized dog. Anesthesiology 1989; 70:82–90.
20. McKay RG, Spears JR, Aroesty JM, et al. Continuous measurement of left and right ventricular stroke volume and pressure-volume relationships with an impedance catheter. Circulation 1984; 69:703–10.
21. Baan J, Van Der Velde ET, Koops J, et al. Continuous registration of relative left ventricular volume in man. Circulation 1982; 66:II-277.
© 2000 International Anesthesia Research Society