and Systolic Performance Following

Transcription

1 The Effect of PEEP on Left Ventricular Diastolic Dimensions and Systolic Performance Following Myocardial Revascularization Peter Van Trigt, M.D., Thomas L. Spray, M.D., Michael K. Pasque, M.D., Robert B. Peyton, M.D., Gary L. Pellom, M.S., Charles M. Christian, M.D., Lennart Fagraeus, M.D., and Andrew S. Wechsler, M.D. ABSTRACT To quantitate the alterations in left ventricular (LV) dimensions and performance at successive levels of positive end-expiratory pressure (PEEP), 16 patients undergoing coronary artery bypass grafting (CABG) underwent instrumentation with ultrasonic dimension transducers to measure the minor-axis diameter of the left ventricle. Matched micromanometers were placed to measure intracavitary LV pressure and intrathoracic pressure. LV pressure and dimension data were recorded and computer analyzed during continuous positivepressure ventilation at 0, 5, 10, and 15 cm H,O of PEEP 4 to 8 hours postoperatively. Preload was determined by the end-diastolic minor-axis diameter, cardiac output was measured by thermodilution, and indices of LV contractility assessed included the maximal velocity of minor-axis shortening and the slope of the end-systolic pressure-diameter relationship. PEEP produced a progressive increase in intrathoracic pressure associated with a fall in cardiac output; this was associated with a decrease in LV end-diastolic diameter and no significant change in the maximal velocity of minor-axis shortening or the slope of the end-systolic pressure-diameter relationship. Our results indicate that PEEP of 10 cm H,O or greater will produce a significant fall in cardiac output in patients following CABG, due to a decrease in preload rather than impaired LV contractility. In 1969 Ashbaugh and co-workers [ll reported the use of positive end-expiratory pressure From the Departments of Surgery and Anesthesiology, Duke University Medical Center, Durham, NC. Presented at the Twenty-eighth Annual Meeting of the Southern Thoracic Surgical Association, Palm Beach, FL, NOV 5-7, Address reprint requests to Dr. Van Trigt, PO Box 3235, Duke University Medical Center, Durham, NC (PEEP) to treat hypoxemia associated with the adult respiratory distress syndrome. By increasing the functional residual capacity of the lung, PEEP was found to increase arterial oxygen tension. Since that time, PEEP has become a commonly used adjunctive therapy for patients following coronary artery bypass grafting (CABG). In addition to its customary application for improvement of arterial oxygen tension, PEEP also has been found helpful in reducing postoperative mediastinal bleeding, presumably by a tamponade effect of overinflated lungs [2]. However, studies on animals 13, 41 and human beings [5, 61 have demonstrated that PEEP decreases cardiac output. When the fall in cardiac output is large, it may offset any increase in oxygen transport afforded by the increase in arterial oxygen content. The mechanism of the fall in cardiac output associated with PEEP has been studied, but is not entirely clear. Cournand and associates 151 in 1948, and later Braunwald and colleagues [7] in 1957 and Morgan and colleagues [8] in 1966, found that a decrease in venous return to the right heart produced by elevated intrathoracic pressure was responsible for the decrease in cardiac output with the application of PEEP. Other mechanisms that have been proposed to explain the decrease in cardiac output have been (1) depression of left ventricular (LV) contractility by a circulating humoral agent [9-131, (2) a neural reflex depression of cardiac function due to overdistention of the lungs and stimulation of pulmonary stretch receptors [14, 151, and (3) restriction of LV filling due to leftward displacement of the interventricular septum [ 161. The purpose of this study was to quantitate the depressive effect on LV function of progres by The Society of Thoracic Surgeons

2 586 The Annals of Thoracic Surgery Vol 33 No 6 June 1982 sive levels of PEEP in patients following CABG, and to clarify the mechanism by which this occurs using a pressure-dimension analysis. Prior studies using pressure and output data alone have not discerned the effects of augmented intrathoracic pressure, altered venous filling, and changed transmural cardiac chamber pressures on ventricular dimensions. Material and Methods This series comprises 16 patients undergoing aortocoronary bypass grafting (mean, 3.4 vessels) between October, 1980, and April, The indication for CABG was chronic, disabling angina. Informed consent for instrumentation and postoperative monitoring was obtained prior to operation in all patients using a protocol approved by the institutional human experimentation committee. LV hypertrophy was not present in any of the patients studied, and all patients were free from valvular dysfunction. Operations were performed using cardiopulmonary bypass with moderate total-body hypothermia (30" to 32"C), using narcotic anesthesia with fentanyl. The distal coronary anastomoses were fashioned during a single period of cardiac arrest induced by a hypothermic, hyperkalemic cardioplegic solution in conjunction with topical hypothermia. The ultrasonic technique for measurement of LV dimensions in clinical settings has been described elsewhere [17, 181 and will be only summarized here. Following saphenous vein grafting and during the rewarming period prior to weaning from cardiopulmonary bypass, a pair of 6 mm hemispherical ultrasonic dimension transducers [19] was implanted onto the epicardium of the anteroposterior minor-axis diameter of the left ventricle in all patients. In addition, the dimension of anterior wall thickness was measured in 6 patients by placing through an oblique needle tract a 1.2 mm crystal transducer in the subendocardial region opposite the anterior epicardial crystal. Figure 1 illustrates the transducer placement. In the configuration used for this study, the anterior epicardial crystal was employed as the transmitting transducer, being excited electrically by the sonomicrometer so as to send an ultrasonic wave through the myocardium and across the Fig 1. Placement of dimension and pressure transducers for use in the postoperative period. LV chamber to the opposing transducer. The transit time of the sonic pulse is proportional to the separation of the transducers (1.56 mm per microsecond), and the resolution of this system has been determined to be a fraction of the wave length of the energy transmitted, or less than 0.08 mm. Following placement of the dimension transducers, a balanced 6F micromanometer by Medical Measurements was placed into the right superior pulmonary vein and through the mitral valve into the middle of the LV chamber. A second, matched, micromanometer was placed in the mediastinum external to the left ventricle, and the pericardium was loosely approximated with two to three interrupted sutures. After cessation of cardiopulmonary bypass and prior to placement of sternal wires, all dimension and pressure transducer leads were exteriorized anteriorly, lateral to the mediastinal chest tubes. Postoperative LV pressure and dimension data were collected 4 to 8 hours postoperatively in the surgical acute care unit. All patients were ventilated with Bennet MA-1 volume respirators with tidal volume set at 15 ml per kilogram of body weight in an intermittent mandatory ventilation cycle. PEEP was adjusted from 0 to 15 cm HzO in all patients. Mediastinal chest tubes were clamped to minimize differences between patients since a varying number of tubes subjected to varying negative pressures were used. This did not affect the patterns of systolic performance or diastolic filling which were subsequently measured at successive levels of PEEP. LV minor-axis

3 587 Van Trigt et al: Effect of PEEP on LV Diastolic Dimensions and Systolic Performance diameter, LV pressures, intrathoracic pressure, and when measured, LV wall thickness were recorded on FM magnetic tape after a steady hemodynamic state had been reached at 0, 5, 10, and 15 cm H,O of PEEP. In addition, thermodilution cardiac output and pulmonary artery pressures were obtained simultaneously with the pressure-dimension data. Any inotropic or afterload reducing agent required by the patient postoperatively was kept at a constant dose at each level of PEEP. The dimension transducers and pressure catheters were removed percutaneously without complication 24 to 30 hours postoperatively prior to removal of the mediastinal chest tubes. After removal, the pressure catheters were rebalanced to check for electronic drift. All 16 patients were extubated successfully 12 to 24 hours postoperatively, and there were no postoperative deaths. Data Andysis, Statistics All analog pressure-dimension data were digitized at a sampling rate of 200 per second and analyzed by a PDP microprocessor computer. Results are expressed as mean k standard error of the mean. A Student s test for paired data was used to compute statistical significance between the measured hemodynamic variables at each level of PEEP. Results Analog recordings obtained in the postoperative period from a representative patient are shown in Figure 2, in which the minor-axis diameter and wall-thickness dimensions are shown along with the corresponding intracavitary LV pressure and intrathoracic pressure. Calculated variables of LV contractility derived from the pressure-dimension data include the peak derivative of minor-axis shortening and minor-axis excursion from enddiastole to end-systole. As a measure of global LV performance, the area of the pressuredimension loop (dynes per centimeter) also was calculated at each level of PEEP. In the 6 patients in whom anterior wall thickness was measured, LV end-diastolic volume was approximated by modeling the ventricle to a thick-walled sphere. LV transmural pressure was calculated as LV intracavitary pressure minus mediastinal pressure. The effect of 15 cm H,O of PEEP on LV dimensions and pressure is shown in Figure 3, with minor-axis diameter showing a decrease as wall thickness increased. For the entire group of patients, the effect of progressive levels of PEEP on the measured LV dimensions and measured LV pressures is summarized in Table 1. PEEP applied at levels of 10 cm HzO or greater resulted in a significant decrease in LV preload as reflected by the decrease in minor-axis diameter and the transmural LV end-diastolic pressure. In the patients with wallthickness measurements, calculated LV enddiastolic volume decreased by 6.0 f 1.8% and 10.6 f 2.3% at 10 and 15 cm HzO of PEEP, respectively ( p < 0.05 for each). The effect of progressive levels of PEEP on LV systolic performance is summarized in Table 2. Cardiac output and the area of the pressurediameter work loop, both heavily dependent on LV diastolic filling and preload, were significantly depressed at 10 and 15 cm HzO of PEEP (p < 0.05). The maximum extent of minor-axis shortening, which is related to stroke volume and thus is dependent on preload, was significantly decreased at the highest level of PEEP. However, a variable more closely related to LV contractile function, the maximal velocity of minor axis shortening, was not significantly altered with the application of PEEP. A more sensitive index of LV contractility that has been proposed to be independent of preload and afterload [201, the end-systolic pressure-diameter relationship, was calculated over the range of PEEP applied (Fig 4). At any given level of PEEP, LV end-systolic pressure and diameter had a close and linear relationship (slope = 10.36, r = 0.96), indicating that LV contractility was not altered by applying different levels of PEEP. To evaluate the influence of elevated intrathoracic pressure as a cause for decreased venous return to the heart, 7 patients underwent application of PEEP intraoperatively off cardiopulmonary bypass after transducer placement but with the sternum still divided and widely retracted. The hemodynamic data collected in this setting are summarized in

4 588 The Annals of Thoracic Surgery Vol 33 No 6 June 1982 Fig 2. Analog left ventricular (LV) pressure and dimension data obtained postoperatively while the patient was on mechanical ventilation. Fig 3. Analog left ventricular (LV) pressure and dimension data showing the decrease in minor-axis diameter and increase in wall thickness associated with the application of 15 cm H,O of positive end-expiratory pressure (PEEP).

5 589 Van Trigt et al: Effect of PEEP on LV Diastolic Dimensions and Systolic Performance Table 1. Effect of PEEP on Dimensions and Pressures of the Left Ventriclea Systolic End-diastolic Intrathoracic Minor-Axis Wall PEEP Pressure Pressure Pressure Diameter Thickness (cm H2O) (mm Hg) (mm Hg) (mm Hg) (mm) (mm) Control 107 f f f f f * f * 0.5' 57.5 * f 3' 11.9 & 1.4b 1.8 * 1.0' 56.8 f 1.9' 7.9 f f 2.5h 9.8 f 1.5b 3.6 * 1.1' 56.5 f 1.5' 8.0 f 0.6' adata obtained in the closed-chest state 4 to 8 hours postoperatively. All ventricular pressures listed are transmural. "p < 0.05 vs control. Table 2. Effect of PEEP on Variables of Left Ventricular Systolic Function" Cardiac Area Pressure- PEEP output Diameter Loop Pk-dL/dt AL (cm H2O) (Llmin) [(dyneslcm) x lo?] (mmlsec) (mm) Control 5.2 k f ? zk f k k * 0.3' 34.1 f 3.6' 55.2 f k f 0.2' 33.1 f 3.9b 54.3 f k 0.5' "Left ventricular function data obtained 4 to 8 hours postoperatively. "p < 0.05 vs control. Pk-dLldt = peak derivative of minor-axis shortening; AL = maximum extent of minor-axis shortening. / 1 t I I I 80 I I I LV END-SYSTOLIC DIAMETER (mm) Fig 4. The effect of positive end-expiratory pressure (PEEP) on the end-systolic pressure-diameter relationship of the left ventricle. That the relationship remains linear with a constant slope implies contractility is not altered with the application of PEEP. (LV = left ventricular.) Table 3. With the sternum open and thus with little or no impairment to right heart filling, application of PEEP reduced neither cardiac output nor minor-axis diameter of the left ventricle. Comment We have evaluated the influence of varying levels of PEEP on cardiac dimensions and function in patients following CABG, and have demonstrated that PEEP equal to or greater than 10 cm H20 significantly reduces cardiac output. Depression of cardiac output is directly due to a decrease in LV end-diastolic volume; this conclusion is supported by the fact that the minoraxis diameter decreased significantly along with transmural LV end-diastolic pressure at levels of PEEP greater than 5 cm H,O. LV contractility is not primarily decreased during application of PEEP to levels commonly employed following CABG, since neither the maximal minor-axis velocity of shortening nor the slope of the end-systolic pressure-diameter relationship were altered over the range of PEEP studied. We conclude that PEEP, within the range

6 590 The Annals of Thoracic Surgery Vol 33 No 6 June 1982 Table 3. Effect of PEEP on Left Ventricular Dimensions and Hemodynamics in the Open-Chest Stateazb Minor- Systolic End-diastolic Axis Cardiac PEEP Pressure Pressure Diameter output (cm H2O) (mm Hg) (mm Hg) (mm) (Llmin) Control 110 k f f k f f f f f * k f k t f t 0.5 adata obtained prior to closure of the sternotomy, with mediastinal pressure = 0 mm Hg. bnone of the results differed significantly from control studied, decreases cardiac output and stroke volume primarily by decreasing preload as determined by LV end-diastolic volume. Manny and colleagues [121 suggested that a humoral factor decreases myocardial contractility and cardiac output during the application of PEEP. In those studies, LV function curves were plotted from LV end-diastolic volume (measured with a latex balloon) and peaksystolic pressures in isolated canine hearts perfused by donor dogs. When PEEP of 15 cm H20 was applied to the donor dogs, a decrease in myocardial contractility occurred on the basis of a downward shift in the Starling curves. Such an alteration in myocardial contractility was not present in our studies in patients whose chest was closed following CABG. Other studies employing ejection-phase indices of contractility [211 or the end-systolic pressuredimension relationship as an index of contractility [16] also failed to show that a decrease in contractility is the mechanism responsible for reduced cardiac output with PEEP. These findings do not support a humoral mechanism in intact dogs or patients. Cassidy and co-workers [14] suggested that hyperinflation of the lungs during PEEP produces a reflex depression of ventricular function that decreases cardiac output. However, this appears to be only a transient effect, as Glick and associates 1151 showed that an airway pressure of 20 mm Hg in dogs produced an immediate decrease in heart rate and contractility which returned to control levels in fifteen to twenty seconds. It is doubtful that such a reflex phenomenon influenced our data, since mea- surements were not made until a hemodynamic steady state had been reached after approximately five minutes of PEEP at each level. Furthermore, other investigators [22] have demonstrated that bilateral cervical vagotomy, which divides the majority of the afferent pathways from the lungs, does not alter the cardiovascular response to PEEP. From these data and previous studies mentioned, it appears that PEEP decreases cardiac output primarily by reducing LV end-diastolic volume. The mechanism most frequently cited as responsible for decreased diastolic filling is a decreased venous return to the right heart resulting from elevated intrathoracic pressure which compresses the pericardium and great veins [7]. Another alternative mechanism which has been proposed includes a decrease in LV diastolic compliance with the application of PEEP, attributed to ventricular interaction or a shift of the interventricular septum into the LV chamber, thereby effecting a decreased enddiastolic volume at a given end-diastolic pressure [16, 231. Finally, it has been proposed that the application of positive airway pressure decreases venous return to the left heart by increasing the impedance to right ventricular ejection (increased right side afterload) [241. Alteration of diastolic compliance of the left ventricle at the levels of PEEP applied in our study is doubtful, since end-diastolic diameters and transmural pressures both decreased in a similar fashion as PEEP was applied. A leftward displacement of the interventricular septum due to elevated right heart pressures with PEEP could theoretically shift the LV diastolic

7 591 Van Trigt et al: Effect of PEEP on LV Diastolic Dimensions and Systolic Performance pressure-volume curve to the left, indicating decreased compliance to filling. However, Jardin and co-workers [16], employing twodimensional echocardiography in patients to measure ventricular cross-sectional area and radius of curvature of the interventricular septum, found such a septal shift to occur only with high levels of PEEP (3 20 cm HzO), whereas the cross-sectional area of the left ventricle was significantly reduced at 10 cm H20 of PEEP. Although in our patients the pericardium was only loosely approximated, thereby decreasing the amount of septal shifting that might have occurred, we believe that at the levels of PEEP most frequently used clinically, an alteration in the diastolic compliance of the left ventricle does not contribute significantly to a decrease in LV preload, and thus, cardiac output. In summary, we have evaluated the influence of PEEP at levels commonly used clinically on LV dimensions and hemodynamics in patients following CABG. A significant depression in cardiac output occurred at levels of 10 cm HzO of PEEP or greater, largely due to a decrease in preload from impaired venous return. Employing PEEP at levels of 10 cm HzO or greater may decrease oxygen transport unless additional volume is administered to maintain adequate filling pressures. This study provides a scientific basis for the empiric observation that augmentation of intravascular volume may correct the impaired cardiac performance induced by PEEP. It does not support the hypothesis that PEEP impairs the intrinsic myocardial contractile state, either by a reflex or humoral mechanism. We thank Mr. Lorenz Preyz for technical contributions involving the sonomicrometer used for this study. We also gratefully acknowledge the assistance of the Duke University ACU Nursing Staff, and Mrs. Ruby Griffin for her work in the preparation of the manuscript. References 1. Ashbaugh DG, Petty TL: Positive end-expiratory pressure. J Thorac Cardiovasc Surg 65:165, Ilabaca PA, Ochsner JL, Mills NL: Positive endexpiratory pressure in the management of the patient with a postoperative bleeding heart. Ann Thorac Surg 30: , Summer WR, Permutt S, Sagawa K, et al: Effects of spontaneous respiration on canine left ventricular function. Circ Res 45:719, Prewitt RM, Wood LDH: Effect of positive endexpiratory pressure on ventricular function in dogs. Am J Physiol236:H534, Cournand A, Motley ML, Werko L, Richards DW: Physiological studies on the effect of intermittent positive-pressure breathing on cardiac output in man. Am J Physiol 152:162, Powers SR, Manual R, Neclino M, et al: Physiologic consequences of positive end-expiratory pressure (PEEP) ventilation. Ann Surg 178:265, Braunwald E, Bivian JT, Morgan WL, Sarnoff SJ: Alterations in central blood volume and cardiac output induced by positive-pressure breathing and counteracted by metaraminol. Circ Res 5:670, Morgan BC, Martin WE, Horenbein TF, et al: Hemodynamic effects of intermittent positive pressure respiration. Anesthesiology 27:584, Patten MT, Liebman PR, Manny J, et al: Humorally mediated alterations in cardiac performance as a consequence of positive end-expiratory pressure. Surgery 84:201, Manny J, Grindlinger G, Mathe AA, Hechtman HB: Positive end-expiratory pressure, lung stretch, and decreased myocardial contractility. Surgery 84:127, Liebman PR, Patten MT, Manny J, et al: The mechanism of depressed cardiac output on positive end-expiratory pressure (PEEP). Surgery 83:594, Manny J, Patten MT, Liebman PR, Hechtman HB: The association of lung distention, PEEP and biventricular failure. Ann Surg 187:151, Grindlinger GA, Manny J, Justice R, et al: Presence of negative inotropic agents in canine plasma during positive end-expiratory pressure. Circ Res 45:460, Cassidy SS, Robertson CH, Pieru AK, Johnson RL: Cardiovascular effects of positive endexpiratory pressure in dogs. J App Physiol 44:743, Glick G, Wechsler AS, Epstein SE: Reflex cardiovascular depression produced by stimulation of pulmonary stretch receptors in the dog. J Clin Invest 48:467, Jardin F, Farcot JC, Boisante L, et al: Influence of positive end-expiratory pressure on left ventricular performance. N Eng J Med 304:387, Chitwood RW, Hill RC, Sink JD, et al: Measurement of global ventricular function in man using sonomicrometry. J Thorac Cardiovasc Surg 80: 724, 1980

8 592 The Annals of Thoracic Surgery Vol 33 No 6 June Kleinman CH, Hill RC, Chitwood RW, et al: Regional myocardial dimensions following coronary artery bypass grafting in patients. J Thorac Cardiovasc Surg 77:13-23, Van Trigt P, Bauer BJ, Olsen CO, et al: An improved transducer for measurement of cardiac dimensions with sonomicrometry. Am J Physiol 240:H , Grossman W, Braunwald E, Mann T, et al: Contractile state of the left ventricle in man as evaluated from end-systolic pressure-volume relations. Circulation 56: , Fewell JE, Abendschein DR, Carlson JC, et al: Continuous positive-pressure ventilation decreases right and left ventricular end-diastolic volumes in the dog. Circ Res 46:125, Scharf SM, Caldini P, Ingram RH Jr: Cardiovascular effects of increasing airway pressure in the dog. Am J Physiol232:H35-43, Scharf SM, Brown R, Tow DE, Parisi AF: Changes in ventricular volume with high lung volumes and negative pleural pressure in man. J Appl Physiol47:257, Qvist JH, Pontoppidan H, Wilson RS, et al: Hemodynamic response to mechanical ventilation with PEEP. Anesthesiology 42:45, 1975