Hemodynamic Benefit of Positive End-expiratory Pressure during Acute Descending Aortic Occlusion

Animals were handled according to the guidelines approved by the American Physiologic Society and the Guide for the Care and Use of Laboratory Animals, published by the National Institutes of Health (publication 85–23, revised 1996). The Institutional Animal Care and Use Committee of the University of Texas Medical Branch approved this study. Animals were allowed ad libitum  water intake up to the time of surgery.

General anesthesia was induced using ketamine (20 mg/kg intramuscularly) in 40 mature, female Yorkshire pigs weighing 30.6 ± 2.9 kg (range, 25–36 kg). After tracheal intubation, each animal was ventilated with a tidal volume of 15 ml/kg using an air and oxygen mixture.

The respiratory rate was adjusted to maintain normocapnia as measured by capnography (Datex Instrumentation, Helsinki, Finland). Anesthesia was deepened with a bolus dose of fentanyl (15 μg/kg) and diazepam (0.2 mg/kg) followed by a continuous infusion of fentanyl (10 μg · kg⁻¹· h⁻¹) and diazepam (0.3 mg · kg⁻¹· h⁻¹) through a catheter inserted in a lateral auricular vein. During the surgical preparation, isoflurane (0.5–1.5% inspired concentration) was administered as necessary to maintain a sufficient depth of anesthesia using standard clinical criteria. 17Muscle paralysis was provided by 0.1 mg/kg pancuronium administered intravenously as necessary to prevent shivering. In all animals, during aortic cross-clamping, isoflurane was maintained at 1.5% end-tidal concentration, which was then decreased to 0.5% end-tidal concentration immediately before and throughout declamping.

A 20-cm catheter was inserted into the femoral artery by cutdown and threaded intraabdominally into the common iliac artery. A thermodilution catheter was inserted through a side branch of the right internal jugular vein into the pulmonary artery. Core temperature was recorded using the thermistor on the pulmonary artery catheter.

Through a median sternotomy incision, fluid-filled catheters were inserted into the aorta via  the right internal mammary artery and into the superior vena cava via  the right internal mammary vein. A micromanometer-tipped catheter (MPC-500, Miller Instruments, Houston, TX) was inserted into the aorta via  the left internal mammary artery and was recalibrated according to the fluid-filled catheter in similar location. Over the anterior free wall of the left ventricle, two 3-mm diameter holes were made in the pericardium, and a pair of ultrasonic dimension transducers was implanted into the LV subendocardium. The ultrasonic signals were received and analyzed instantaneously to measure LV dynamic distances (model 301 Sonmicrometer, Triton Technology, San Diego, CA). The median sternotomy incision was tightly reapproximated. Through a midabdominal incision, the descending aorta was isolated 1 cm above the origin of the celiac artery and loosely encircled with an umbilical tape snare. Heparin (300 U/kg) was administrated to prolong the activated clotting time (ACT) more than 250 s. Repeat ACT measurements were taken hourly, and additional doses of heparin (150 U/kg) were added as necessary.

Animals were assigned using a random numbers table to one of four groups: zero end-expiratory pressure (ZEEP; n = 10); ZEEP + volume expansion (n = 10); 15 cm H²O PEEP (n = 10); 15 cm H²O PEEP + volume expansion (n = 10). PEEP therapy was initiated in appropriate animals throughout the aortic occlusion and was discontinued at the time of reperfusion. A basal infusion of 0.9% saline was administered in all animals at 5 ml · kg⁻¹· h⁻¹during instrumentation and throughout the experimental protocol. In the animals receiving volume expansion, 6% hetastarch solution was administered intravenously at 10 ml/kg within the period of aortic occlusion. No additional fluid was administered during cross-clamp release.

After baseline measurements were recorded, the descending aorta was acutely occluded by tightening the tape through the rubber snare for 30 min. Repeat measurements, including proximal (internal mammary artery) and distal (femoral artery) aortic pressures, were recorded at 1, 5, and 30 min during aortic occlusion. Next, the occlusion was released, and measurements were acquired 1, 5, and 30 min during reperfusion. No vasoactive drugs or sodium bicarbonate were administered throughout the study. During aortic occlusion, the isoflurane concentration was maintained at 1.5% end-expiratory concentration, which was then reduced to 0.5% end-expiratory concentration immediately before and throughout reperfusion. At the conclusion of each experiment, the heart was arrested with a lethal injection of potassium chloride solution.

Fluid-filled transducers were zeroed against atmospheric pressure at the anterior axillary level and calibrated with a mercury manometer. Hemodynamic variables recorded at end-expiration (Model ES-1000, Gould, Inc., Valley View, OH) included mean aortic pressure, mean pulmonary artery pressure, central venous pressure, pulmonary artery wedge pressure, and heart rate. Arterial blood gases (model 1306, Instrumentation Laboratory, Lexington, MA) and serum hemoglobin (Cooximeter model 482, Instrumentation Laboratory) were repeatedly measured throughout the experiment. Mean cardiac output was determined using thermodilution by averaging two 5-ml injections of iced saline solution injected at end-expiration. Stroke volume (ml/beat) was calculated as (mean cardiac output × 1,000)/heart rate. LV end-diastolic segment length was measured under the QRS complex of the electrocardiogram immediately before the major aortic pressure inflection; LV end-systolic segment length was measured at the dicrotic notch of aortic pressure. 18Segmental shortening (%) was calculated as ([end-diastolic length]−[end-systolic length]/end-diastolic length) × 100. Systemic vascular resistance (dyn · s · cm⁻⁵) was calculated as ([mean aortic pressure − central venous pressure]/cardiac output) × 79.9. LV stroke work (g × m) was calculated as ([stroke volume]×[mean aortic pressure − pulmonary artery wedge pressure]) × 0.0136.

Data, expressed as mean ± SD, were analyzed using analysis of variance for a three-factor factorial experiment with repeated measures. The three factors were PEEP (0 and 15 cm H²O), volume expansion (6% hetastarch, 0 and 10 ml/kg), and time (five measurement intervals). Effects and interactions were assessed at significance level of 0.05. We adjusted for multiple comparisons using Fisher's least significant difference procedure with P < 0.005 for a comparison-wise error rate.

The study was successfully completed in all animals. Acute aortic occlusion caused systemic vascular resistance to double in all groups (table 1). As seen in figure 1, after 1 and 5 min of aortic occlusion, mean aortic blood pressure significantly increased in both ZEEP groups but not in animals treated with 15 cm H²O PEEP. Accordingly, mean aortic pressure was 25–35% lower with PEEP than ZEEP in respective groups. Cardiac output (table 1) and stroke volume (fig. 2) remained constant in both ZEEP groups. In contrast, the application of PEEP in combination with aortic cross-clamp reduced cardiac output and stroke volume by nearly 50%. Aortic cross-clamping caused a significant increase in LV stroke work in ZEEP animals, whereas PEEP reduced stroke work by 40% (table 1). Although LV segmental shortening remained unchanged with aortic occlusion, end-diastolic and end-systolic segment lengths increased in ZEEP animals, whereas PEEP attenuated this response. Filling pressures (central venous pressure and pulmonary artery wedge pressure) increased with aortic cross-clamp in all groups. Aortic pressures measured distal to the cross-clamp decreased similarly in all groups during aortic occlusion (table 1).

Volume expansion during cross-clamp application increased cardiac output in PEEP and ZEEP groups (table 1). Stroke volume and stroke work tended to decrease 30 min after cross-clamp without volume expansion, but each was maintained constant in animals receiving Hespan. Similarly, end-diastolic segment length decreased significantly in both groups without volume expansion while remaining unchanged in Hespan-treated animals. After 30 min of aortic cross-clamp, serum hemoglobin increased without volume expansion but decreased with volume expansion (table 2).

Acute aortic declamping markedly reduced systemic vascular resistance and filling pressures in all animals (table 1). Mean aortic blood pressure decreased significantly with clamp removal in all animals, with the largest decrease seen in both ZEEP groups irrespective of volume expansion (fig. 1). Cardiac output decreased in the ZEEP animals with or without volume expansion but increased significantly in both PEEP groups, particularly in animals receiving Hespan. Thirty minutes after cross-clamp removal, cardiac output remained greater in both volume-expanded groups and was not different from baseline values. Stroke volume and stroke work were significantly greater in the PEEP + volume group 5 min after declamping (fig. 2). Declamping caused serum hemoglobin to increase in all groups as end-diastolic segment length increased only after removal of 15 cm H²O PEEP. Segmental shortening was significantly lower than baseline values in both ZEEP groups with or without volume expansion.

Oxygenation variables are shown in table 2. As expected, arterial pH decreased significantly but similarly in all groups with early reperfusion. Thirty minutes after cross-clamp removal, pH values remained lower than baseline in all groups. Although carbon dioxide tension increased with declamping, arterial bicarbonate concentrations decreased and remained lower than baseline throughout reperfusion. Base deficit values increased in all groups during declamping, with only partial recovery by the conclusion of the experiment without intergroup differences.

In this porcine model, 15 cm H²O PEEP reduces the hypertensive response to acute aortic occlusion and allows concomitant volume infusion, which, in turn, reduces the degree of declamp hypotension by improving stroke volume. This study confirms the independent benefit of volume expansion during cross-clamp application as indicated by a greater stroke volume after declamping; others have reported similar results. 11However, PEEP alone produces a similar benefit on stroke volume during declamping as the ZEEP + volume animals demonstrated, indicating that PEEP, by attenuating venous return, may increase splanchnic venous capacitance during aortic cross-clamping. The clinical implication is that the use of PEEP during aortic occlusion in patients may allow rapid control of loading conditions to attenuate systemic hypertension while allowing simultaneous volume expansion. The discontinuation of PEEP with cross-clamp removal can lead to an improvement in stroke volume so that early declamp hypotension is attenuated.

The hemodynamic response to acute aortic occlusion is characterized by an increase in systemic vascular resistance and mean aortic blood pressure with variable effects on LV stroke volume and cardiac filling pressures. 6Multiple factors influence the magnitude of cardiac load and include (1) an increase in LV afterload secondary to greater aortic impedance 19and humoral changes in the renin–angiotensin system;20(2) alterations in preload from blood volume redistribution;21,22(3) a change in coronary hydraulic properties, which can reduce myocardial compliance;23and (4) underlying myocardial performance, specifically the sensitivity of LV pump function to acute alterations in preload and afterload. In patients with coexisting ventricular contractile dysfunction and limited preload-recruitable function, an acute increase in LV afterload could cause a marked decrease in stroke volume, in turn causing ventricular dilation 4and ensuing ischemic changes. 2In animal studies of supraceliac aortic cross-clamp, preload increases secondary to splanchnic venous collapse from elastic recoil and reduced venous volume. 21,22Our data with ZEEP animals confirm this finding, with a significant increase in LV end-diastolic segment length after aortic cross-clamping. The application of PEEP prevents an increase in end-diastolic segment length, although filling pressures were greater presumably the result of transmitted intrathoracic pressure. In fact, PEEP during aortic cross-clamp produces similar hemodynamic effects as described in the experimental model with simultaneous occlusion of the aorta plus inferior vena cava. 24,25Whereas occlusion of the descending thoracic aorta in dogs causes a net transfer of blood volume to the upper body, occlusion of aorta and inferior vena cava causes blood redistribution to the lower body without a concomitant increase in systolic arterial pressure. 24It is possible that, by impeding venous return to the right heart and increasing splanchnic venous volume, PEEP works in a similar yet more controllable manner.

Although perfusion distal to the cross-clamp appears proportionate to proximal aortic pressure, 10the increase in systolic pressure with cross-clamp application represents a true afterload experienced by the heart and can interfere with secure placement of the cross-clamp. Consequently, some control of proximal blood pressure may be required during aortic surgery. Therapeutic measures to reduce or minimize the hemodynamic consequences of acute aortic occlusion include the titration of vasodilator drugs, volatile anesthetic agents, and hemodilution. Volatile anesthetic agents effectively reduce aortic pressure but have the distinct disadvantage of depressing myocardial contractility with persistent effects after declamping. Vasodilators, such as nitroglycerin, have a predominant preload-reducing rather than afterload-reducing effect; PEEP produces a similar hemodynamic profile. However, clinical experience with nitroglycerin shows an increase in cardiac output and mixed venous oxygen saturation during cross-clamping, 8whereas the addition of PEEP in our study further reduces cardiac output by nearly 50%. Nevertheless, because nearly half the body is excluded by the aortic clamp, the decrease in cardiac output by PEEP does not adversely affect arterial pH or base deficit when compared with ZEEP animals. Vasodilators, such as nitroprusside, are effective to reduce blood pressure during aortic occlusion but can further reduce distal aortic pressure, visceral perfusion, and oxygen consumption. 10In our study, PEEP does not adversely affect distal perfusion pressure measured below the cross-clamp, although further study is required to examine regional visceral perfusion with the addition of PEEP.

Aortic declamping causes significant hypotension secondary to the precipitous reduction in vascular resistance, sequestration of blood volume in the distal capillary beds, and negative inotropic effect from acidosis and myocardial depressant factors. 6,11,14The potential for hypovolemia may be worsened by intravascular fluid loss caused by capillary damage with reperfusion. 6,10The amount of ensuing hypovolemia during declamping can be attenuated by volume expansion before clamp release;11–14volume expansion after aortic cross-clamp release is less effective. However, limitations on volume expansion include arterial hypertension during cross-clamp application and the potential to prolong postoperative recovery. 26Similar benefit is found with volume expansion in ZEEP animals, although higher aortic blood pressures are seen during cross-clamp application.

Because of its rapid onset, short half-life, and ability to be titrated, PEEP could offer particular advantages with the abrupt swings in loading conditions inherent to major aortic surgery. PEEP produces multiple effects on ventricular loading conditions. In experimental animals, PEEP reduces venous return and right ventricular preload, although the exact mechanisms remain somewhat unclear. 15,27PEEP increases intrathoracic pressure, which could reduce transmural filling pressure. However, the net pressure gradient for venous return to the right heart does not change with PEEP because of simultaneous increases in intraabdominal and systemic back pressures. 27,28Other potential mechanisms include inferior vena caval compression by lung hyperinflation 27and an increase in splanchnic venous resistance by mechanical compression from diaphragmatic descent. 28Similarly, PEEP can reduce LV transmural ejection pressure, defined as the difference between LV developed pressure and pericardial pressure, and, consequently, LV afterload. 16PEEP can also reduce LV compliance. Specifically, in the splanchnic circulation, PEEP causes passive venous distention with blood sequestration;29these effects could counteract the splanchnic venous collapse and blood redistribution associated with supraceliac aortic occlusion. 21,22In our study, intravascular volume appears to be preserved better with PEEP, which produces the same hemodynamic benefit during declamping as volume expansion in ZEEP animals. Accordingly, the greatest benefit occurred with PEEP + volume expansion, where stroke volume with declamping was double that seen with ZEEP alone.

A potential drawback to this study is the use of animals having normal ventricular function. Positive airway pressure produces greater hemodynamic benefit in patients with left heart failure and afterload sensitivity than in healthy patients. 30Anecdotally, we have used 7.5–12.5 cm H²O PEEP in several patients undergoing aortic aneurysmectomy who have a reduced ejection fraction resulting from ischemic heart disease. Our clinical experience with these patients indicates that PEEP allows more precise hemodynamic control and permits volume expansion and the administration of inotropic drugs (e.g ., dobutamine) during aortic occlusion with less declamp hypotension after clamp removal. Further study is necessary in an animal model of congestive failure to substantiate this clinical impression of PEEP. In addition, other factors should be considered when using PEEP therapy in patients. Our animals did not have underlying coronary artery disease, which could independently affect the response to PEEP. PEEP may worsen left 31and right 32ventricular ischemic injury, and this possibility should be considered when using PEEP in patients with vascular disease. PEEP can increase pulmonary vascular resistance and right ventricular oxygen demands, 32which may be detrimental in patients with severe underlying pulmonary disease. Second, the changes in stroke volume in our study do not correlate well with LV free wall segment lengths. Besides any obvious discrepancy that may exist between right- and left-sided events associated with PEEP and aortic cross-clamp, a limited number and position of dimension transducers were used in our study to minimize alterations in pericardial integrity. Examining a greater number of dimensions may allow a more detailed and sensitive reflection of changes in LV volume, particularly with the noncongruent changes in ventricular geometry reported with PEEP. 33 

In conclusion, 15 cm H²O PEEP attenuates the hypertensive response to acute aortic occlusion in pigs and allows concomitant volume expansion. During subsequent declamping, the abrupt discontinuation of PEEP causes less hypotension than with ZEEP. Although volume expansion during cross-clamp directly improves stroke volume with declamping, the greatest improvement in stroke volume and aortic blood pressure was found in PEEP + volume–treated animals. The use of PEEP may be beneficial in patients during acute aortic clamping and declamping.

The authors thank Tatsuo Uchida, M.S. (Senior Biostatistician, Office of Biostatitics, The University of Texas Medical Branch, Galveston, Texas), for statistical review; Irela Salinas, A.A. (Editorial Assistant, Department of Anesthesiology, The University of Texas Medical Branch, Galveston, Texas), for manuscript preparation; and Jordan Kicklighter, B.A. (Editor, Department of Anesthesiology, The University of Texas Medical Branch, Galveston, Texas), for editorial review.