RESEARCH. Hemodynamic Changes and Fluid Shifts After Large-Volume Fluid Infiltration. Results From a Porcine Model

Transcription

1 RESEARCH Hemodynamic Changes and Fluid Shifts After Large-Volume Fluid Infiltration Results From a Porcine Model Selahattin Ozmen, MD,* Krzysztof Kusza, MD, PhD, Betul G. Ulusal, MD, Landon Pryor, MD, Maria Siemionow, MD, PhD, and James E. Zins, MD, FACS Abstract: While certain parameters such as blood loss and serum lidocaine levels following liposuction have been well studied, fluid shifts between the intravascular and extravascular space have not. With the advent of large volume liposuction, prudent fluid management has become obligatory. Hence, the reason for our study. To test the impact of large volume infiltration on intercompartmental fluid shifts, we measured urine output and hemodynamic changes in 10 anesthetized female Yorkshire pigs weighing between 50 and 85 kg. Eight pigs were infused with 5 to 10 L of tumescent fluid. Two pigs were anesthetized, received no wetting solution, and served as controls. Hemodynamic variables were recorded before infusion and hourly for 48 hours. Animals were extubated after 4 hours of anesthesia. Plasma volume was measured using Evan s Blue Dye, and intravascular fluid shifts were calculated using Foldager s method. Total fluid shift into the intravascular space ranged between 511 and 1036 ml per animal with a mean of 767 ml in the first 3 hours. Higher volumes of fluid infiltration did not lead to fluid overload in the experimental group. Hemodynamic changes were characterized by significant increases in central venous pressure, cardiac output, pulmonary artery pressure, and heart rate consistent with the increase in intravascular volume. Hemodynamic parameters returned to baseline 20 hours following tumescent fluid infiltration. In this porcine model, animals were able to tolerate large fluid challenges delivered by clysis with statistically significant but only modest increases in hemodynamic parameters which gradually returned to baseline within 20 hours. Key Words: tumescence, hemodynamics, cardiac, pulmonary edema, fluid overload (Ann Plast Surg 2010;64: 83 88) Liposuction is currently the most commonly requested esthetic surgical procedure in the United States. 1 In early years, liposuction was performed with sharp catheters using dry techniques. However, with the advent of blunt catheter tips and wetting solu- Received July 29, 2008, and accepted for publication, after revision, January 4, From the *Department of Plastic Surgery, Gazi University Faculty of Medicine, Ankara, Turkey; Department of Anesthesiology and Intensive Care, University of Nicolaus Copernicus, Collegium Medicum of Ludwik Rydygier, Bydgoszcz, Poland; and Department of Plastic Surgery, Cleveland Clinic, Cleveland, OH. Presented at the Plastic Surgery Research Council; April 23 26, 2003; Las Vegas, NV and The Ohio Valley Society of Plastic Surgery; June 7 9, 2002; Pittsburgh, PA. Reprints: James E. Zins, MD, Department of Plastic Surgery, The Cleveland Clinic Foundation, Desk A Euclid Ave, Cleveland, OH zinsj@ccf.org. Copyright 2009 by Lippincott Williams & Wilkins ISSN: /10/ DOI: /SAP.0b013e31819adfc5 tions, liposuction has become a safe and commonly performed procedure. 2,3 In the 1980s, Illouz, Clayton et al, Hetter, and others described the wet technique, in which a solution of xylocaine, saline, epinephrine, and hyaluronidase was injected into the subcutaneous fat before liposuctioning. 2,4,5 In the late 1980s, Klein described the tumescent technique of liposuction in which a large volume of a mixture of lidocaine, epinephrine, sodium-bicarbonate, and normal saline solution was infiltrated into the subcutaneous fat compartment. 6 Klein subsequently documented that up to 35 mg/kg of lidocaine could be used safely in this solution when infiltrated into the subcutaneous space at the beginning of the procedure. 6 Tumescent infiltration allowed liposuction to be performed in larger volumes and under either conscious sedation, general or, in some cases, local anesthesia. 4 The superwet and tumescent techniques, currently the 2 most common liposuction methods, provide a ratio of infiltrate to fat aspirate of 1:1 to 1:3, respectively. 4,7,8 Further, administering intravenous fluids in the tumescent technique may not be necessary, even for fat removal of up to 5 L. 4,9 Rohrich et al, 10,11 in 2 recent publications, have documented that intraoperative fluid ratios (defined as the volume of superwet solution and intraoperative intravenous fluids divided by total aspirate volume) of 2.1 (small volume) and 1.4 (large volume) were safe and resulted in no fluid overload. Before the use of wetting solutions, third-space losses, hypovolemia, and blood loss were frequent concerns during liposuction. 12 Recent studies have suggested that when current wetting solutions are used, large amounts of third space fluid collection do not occur and that most infiltrate is not removed by suctioning. 9,13 With large volume liposuction becoming more commonplace, the focus has shifted from preventing hypovolemia to preventing fluid overload. While many aspects of the liposuction procedure have been well studied, the physiology of fluid shifts into the extracellular and intravascular space at the time of large volume liposuction is less well understood. 10,11,14,15 In a porcine model and later in a clinical study, Kenkel et al described the effect of both ultrasound-assisted and traditional liposuction on a variety of hemodynamic parameters. 16,17 The purpose of our study was to (1) confirm the findings of Kenkel et al, (2) investigate the effect of significantly increasing the degree and amount of the fluid challenge on the pig, (3) define sequentially when and where this fluid went in the extracellular space, and (4) dissect out the effect of the fluid challenge per se from the effect of the fluid challenge plus liposuction. MATERIALS AND METHODS The study was approved by the Institutional Review Board of the Cleveland Clinic Foundation. All animals used in this study received humane care in compliance with the Guide for the Care and Use of Laboratory Animals, published by the National Institutes of Health. Annals of Plastic Surgery Volume 64, Number 1, January

2 Ozmen et al Annals of Plastic Surgery Volume 64, Number 1, January 2010 The Experimental Model Ten female, Yorkshire pigs, weighing between 50 and 85 kg were sedated with ketamine (20 mg/kg, IM) and xylazine (1.5 mg/kg IM), intubated orotrachealy, and mechanically ventilated under isoflurane anesthesia. Of 10 pigs, 8 were infused with 5 or 10 L of fluid. In the first experimental animals, 5 L was chosen since this mirrored the expected fluid challenge if large volume liposuction were being performed using the superwet technique (1:1 fluid to aspirate). However, when minimal hemodynamic changes were noted in the first few animals, the protocol was amended and the fluid challenge was increased to 10 L. Tumescent infiltration of 5 to 10 L in our 50 to 80-kg animals corresponds to approximately 1.5 to 2.5 times their intravascular volume. A Swan-Ganz thermodilution catheter was introduced through the right external jugular vein, its distal tip positioned in the pulmonary artery, and its proximal tip in the superior vena cava for continuous monitoring of cardiac output, central venous pressure (CVP), pulmonary artery pressure (PAP), and pulmonary capillary wedge pressure (PCWP). Nearly continuous cardiac output monitoring was provided by the Swan-Ganz thermodilution catheter. This catheter is fitted with a distal heated filament, which allows automatic thermodilution measurement by heating the blood and measuring the resultant thermodilution trace. 18 An arterial catheter was placed in the right external carotid artery for monitoring mean arterial blood pressure (MAP). Heart rate (HR) was calculated from the EKG readings. Bilateral auricular veins were catheterized for blood sampling and IV drug administration. A urinary catheter was placed, and the bladder was emptied before infiltration. Thereafter, urine volume was measured hourly during the 4 hours of general anesthesia. Of the 10 animals, 8 experimental animals were infused with tumescent solution, whereas the 2 animals who served as controls were not. The wetting solution was prepared with 1 ml epinephrine (1 ml of 1:1000), 500 mg lidocaine (50 ml of 1%), and 12.5 meq NaHCO3 in 1 L of Lactated Ringer s solution. Two incisions, each 0.5 cm long, were made in the midabdomen. A 3-mm diameter, blunt-tipped cannula was used for infiltration. The tumescent solution was infiltrated into the deep subcutaneous tissue of the abdomen and medial thighs at a constant rate of 100 ml/min. Care was taken to equally distribute the whole amount throughout the area. The incisions were closed with 4/0 Mersilene sutures at the end of the infiltration. No liposuction was performed in any of the animals. No extra intravascular fluid was given to the animals during or after the operation other than routine fluid to keep vascular lines open (KVO). Data Collection Hemodynamic variables were recorded before infiltration (baseline) and at hourly intervals for the next 48 hours. Arterial blood gases were collected before surgery (baseline) and at 2-hour intervals for the following 48 hours. Plasma Volume Measurement A venous blood sample obtained from the left auricular vein before infiltration provided baseline values. Following this, 5 ml (22.5 mg) of Evan s Blue Dye was injected into the right auricular vein. After 15 minutes, a second blood sample was taken from the left auricular vein just prior to infiltration. Blood samples were then collected every hour for 3 hours, (the maximum time for reliable assessment of data after Evan s Blue Dye injection). 19 The intravascular volume change was calculated as described by Foldager and Blomqvist. 19 Collected blood samples were centrifuged at 4500 rpm for 7 minutes. The concentration of the dye was measured in the plasma with a spectrophotometer (Beckman Coulter, DU-640 spectrophotometer), at a wavelength of 620 nm against a blank value. The obtained values were fit to a logarithmic decay curve of plasma dye concentration with respect to time using linear regression analysis. Standard decay curves were created for each animal. Complete blood count was performed at 2-hour intervals. Plasma lidocaine levels were measured at 4, 8, 12, and 16 hours. Animals were extubated at 4 hours postintubation. Upon extubation, the animals were caged individually and transported to the intensive care unit. Mild sedation was achieved with midazolam infusion, when needed. Pain therapy was supplied with buprenorphine ( mg IV, every 4 hours). At the end of the 48-hour follow-up period, the animals were euthanized with pentobarbital (40 ml, IV) and autopsied. Kidneys, lungs, liver, and subcutaneous tissue samples were taken for histologic evaluation after H&E staining. Statistical Methods The results of each outcome measure were summarized descriptively using the mean level for experimental and control animals. Individual measures for each animal were plotted over time along with the group means. Repeated measure ANOVA models were fit to the data from the first 4 hours, and the linear trend was tested for each measure. An autoregressive correlation between observations from the same animal was assumed, meaning that observations closer in time to one another were assumed to be more similar. The difference between baseline and 4 hours was calculated for each animal and Wilcoxon Signed Rank tests were performed to assess whether a significant change had occurred. A significant level of 0.05 was assumed for all tests. RESULTS Plasma Volume Measurements The mean intravascular volume increases were 355 ml, 271 ml, and 130 ml at 1, 2, and 3 hours postinfusion, respectively. The total volume of fluid shifted into the intravascular space ranged during this time between 511 and 1036 ml per animal. As expected, the lowest total fluid volume was in the animal that received the lowest tumescent infiltrate (5 L) and that had the lowest intracompartment hydrostatic pressure. In 6 animals, the highest increase in plasma volume was at hour 1, and in 2 at hour 2. Urine output mirrored increases in intravascular volume with maximum mean output of greater than 3 ml/h at 1 hour. Urine output gradually decreased to baseline by 2 hours. Hemodynamic Variables Most animals showed changes in hemodynamic parameters, but only one had severe signs of hemodynamic instability. During preoperative assessment, this animal was active and seemed to be healthy. After the induction of anesthesia, however, baseline PaCO 2 was 65 mm Hg and PaO 2 was 70 mm Hg. After 21 hours, the animal was euthanized and autopsied. An old, right-sided empyema and interstitial pneumonia was confirmed at autopsy. This animal was, therefore, excluded from experimental group. After extubation at 4 hours and transfer to the intensive care unit, measurement of the hemodynamic parameters continued for all experimental animals. Although the animals were sedated, level of arousal varied widely, animals were agitated, and catheters occasionally dislodged. Hourly hemodynamic parameters even within the same animal varied widely. It was, therefore, felt that the data gathered once the animals had awakened from anesthesia was unreliable. Therefore, only the 4-hour data while the animals were under general anesthesia will be reported Lippincott Williams & Wilkins

3 Annals of Plastic Surgery Volume 64, Number 1, January 2010 Large-Volume Fluid Infiltration TABLE 1. Changes in Hemodynamic Parameters During the First 4 Hours Postinfiltration* Average Hourly Change 4-Hour Change (4 h Baseline) Measure Mean (SE) P Median (25th, 75th) P Cardiac output 1.24 (0.26) (2.2, 5.1) Central venous pressure 0.85 (0.31) (1.0, 5.0) Hematocrit 0.38 (0.10) ( 1.6, 0.9) Hemoglobin 0.11 (0.04) ( 0.5, 0.2) Heart rate 6.23 (2.07) (0.0, 34.0) Mean arterial pressure 5.86 (1.51) (5.0, 31.0) Pulmonary artery pressure 1.93 (0.56) (1.0, 12.0) Pulmonary capillary wedge pressure 0.40 (0.25) (0.0, 3.0) *The left side of the table shows the average hourly change with a measure of significance, whereas the right side of the table shows the median and quartiles of change between baseline and 4 hours. The P value is from a Wilcoxon Signed Rank Test. FIGURE 1. Cardiac output from time 0 to 4 hours while animals were under general anesthesia for 4 hours following infusion. Individual animal results are shown with dotted lines, whereas mean levels by group are shown by solid lines. Experimental animals in red and control animals in blue. Modest but statistically significant changes during the first 4 hours following fluid infusion were observed with most hemodynamic parameters (Table 1). Significant increases were observed hourly as well as between baseline and 4 hour measurements in cardiac output, CVP, HR, mean arterial pressure, and PAP (Figs. 1 3). Significant decreases were observed in hemoglobin and hematocrit. However, no significant change in PCWP was observed (Fig. 4). Normal histology was found in all animals surviving for 48 hours. Again in the one animal with acute signs of hemodynamic decompensation, evidence of pneumonia was confirmed by interstitial lung edema and this animal was excluded from the study. DISCUSSION Early recommendations for fluid replacement during liposuction were based on the volume of the aspirated fat. 2,5,20 Clinical experience and experimental studies have documented that only small volumes of intravenous fluid are necessary when tumescent or superwet techniques are used, 14,15,21 even for large volumes of fat removal. While most patients can tolerate a significant intraoperative fluid challenge when cardiac, pulmonary, hepatic, and renal status is normal, unrecognized systemic disease can significantly reduce this margin of safety This, indeed, was the case in one of our pigs who had an undiagnosed empyema and interstitial lung changes, and this led to the animals operative death. Pitman reported the possibility of cardiac and pulmonary decompensation after large-volume fluid infiltration and therefore recommended stopping intravenous fluid delivery when subcutaneous infusion volume exceeds twice the aspiration volume. 21 Rohrich et al 10,11 introduced the concept of the intraoperative fluid ratio defined as the volume of superwet solution and intraoperative intravenous fluid divided by the total aspiration volume to guide volume replacement. They demonstrated that intraoperative fluid ratios of 2.1 for small volume and 1.4 for large volume liposuction were safe and did not cause volume overload in 2 studies Similarly, Klein suggests using minimal or no intravenous fluid replacement during tumescent liposuction procedures. 27,28 Although only a few cases of pulmonary edema during tumescent liposuction have been reported in the surgery literature, the true complication rate is unknown. 22,24,26 Many of these procedures are performed in outpatient centers and office settings making the possibility of under-reporting quite probable. 26 As a result of the pharmacological properties of tumescent solution and the hydrostatic pressure created by a large amount of infiltrate, intraoperative and postoperative fluid shifts are complex. 9,11 Few studies have directly investigated these fluid shifts. In one such study, 16 Kenkel et al described hemodynamic parameters during both ultrasound-assisted and traditional liposuction. In their 2009 Lippincott Williams & Wilkins 85

4 Ozmen et al Annals of Plastic Surgery Volume 64, Number 1, January 2010 FIGURE 2. Central venous pressure measurements from time 0 to 4 hours while the animals were under general anesthesia for 4 hours following infusion of wetting solution. Individual animal results are shown with dotted lines, whereas the mean levels by groups are shown in solid lines. Experimental animals in red and control animals in blue. FIGURE 3. Pulmonary artery pressure during the first 4 hours is shown. Individual animal results are shown with dotted lines, whereas mean levels by groups are shown in solid lines. Experimental animals are shown in red and controls in blue. study, mean total infiltration volume was ml and total lipoaspirate was ml, using pigs weighing 250 to 400 pounds. This volume of lipoaspirate, therefore, does not meet large volume liposuction criteria set by the American Society of Plastic Surgeons. 29 We have attempted to extend Kenkel et al s findings by more closely simulating large volume liposuction. In our study, 50 to 80-kg pigs received a mean of 7000 ml, corresponding to 1.5 to 2.5 times their intravascular volume. In spite of the significant increase in tumescent fluid challenge, our hemodynamic data were similar to that found by Kenkel et al. 16 Like their animals, our pigs tolerated these large fluid challenges surprisingly well with only temporary increases in cardiac output, CVP, PAP, mean arterial pressure, and HR followed by a relatively rapid return to baseline levels. As stated in our introduction, one of the purposes of this study was to attempt to define where in the extracellular space this clysis fluid load went and at what rate. We were able to show using our Evan s blue dye technique mean increases in intravascular volume of 355, 271, and 130 ml at hours 1 to 3 postinfusion, respectively. While this technique allows us to follow intravascular volume changes for short periods only, the rapid return to baseline of hemodynamic parameters as documented by our invasive monitoring techniques suggests that these intravascular volume changes continued to mitigate with time. While all experimental animals received a large volume of tumescent fluid, in none of these animals was liposuction actually performed. We tried to simplify our experimental model by minimizing the variables that could potentially affect intercompartmental fluid shifts, ie, we attempted to dissect out the effect of the fluid challenge per se on the hemodynamic fluid shifts that occur with a large volume challenge. Our reasoning was as follows: It seemed most sensible to define and objectify the hemodynamics of hypodermoclysis without liposuction first. Once this model is understood, Lippincott Williams & Wilkins

5 Annals of Plastic Surgery Volume 64, Number 1, January 2010 Large-Volume Fluid Infiltration FIGURE 4. Pulmonary capillary wedge pressure during the first 4 hours following intubation is shown. Individual animal results are shown with dotted lines, whereas mean levels by groups are shown in solid lines. Experimental animals are shown in red and controls in blue. the variable effects of liposuction could then be additionally studied. We expected that the trauma of liposuction and the associated fat removal would lead to more complex fluid shifts and third space fluid losses. Interestingly, this may not be the case since our hemodynamic findings were so similar to those of Kenkel et al. 16 Autopsy findings in the experimental and control animals were similar. In no animals were the lung changes suggestive of pulmonary edema. The only pig to demonstrate any lung abnormalities had total consolidation of the right lung due to an old empyema and interstitial pneumonia. This animal became unstable with the induction of anesthesia prior to any tumescent infiltration. The pathologic findings clearly explained the animal s death. Our pathologic findings differ from those of Kenkel et al 16 who found clear evidence of pulmonary hemorrhage and fat emboli in both the lungs and kidneys at autopsy, although their pigs showed no evidence of clinical compromise. As the authors noted, this suggested subclinical pulmonary and fat emboli secondary to liposuction. Although we found statistically significant but modest increases in cardiac output, CVP, HR, mean arterial pressure, and PAP, there was no statistically significant increase in the PCWP. This discrepancy remains unclear and unexplained. Our results indicate that the infusion of a large volume of wetting solution can be infiltrated subcutaneously in the Yorkshire pig with surprisingly modest increases in hemodynamic parameters as long as the animal is healthy. These results are consistent with recent data reported by Rohrich et al in a human study but also speak to the importance of careful preoperative evaluation to uncover unsuspected cardiopulmonary or systemic disease. 10,11 Therefore, careful preoperative evaluation of patients undergoing large volume liposuction is obligatory. Effect of General Anesthesia on Hemodynamic Changes The isoflurane used in this experiment acted not only as an anesthetic but also as a modifying agent of the cardiovascular system. The effect of isoflurane on systemic vascular resistance, MAP, CVP, heart contractility, and vascular network capacity depends on the minimal alveolar concentration As a rule, isoflurane causes peripheral vasodilation leading to a decrease in peripheral vascular resistance. 33,34 This was evidenced in our control animals by a consistent decrease in PCWP over hours 1 to 4. Isoflurane decreases mean arterial pressure by decreasing diastolic blood pressure and to a lesser extent by decreasing systolic blood pressure. CONCLUSIONS In this study, 50 to 80-kg Yorkshire pigs tolerated large fluid challenges delivered by clysis equal to 1 1/2 to 2 1/2 times their blood volume surprisingly well. This large-volume, wetting solution caused hemodynamic changes consistent with a documented increase in intravascular volume as evidenced by Evans blue dye injection. In spite of significant increases in intravascular volume in all experimental animals, surprisingly modest changes were seen in most hemodynamic parameters measured. Statistically significant but modest increases were seen in cardiac output, CVP, HR, mean arterial pressure, and PAP. No statistically significant change was seen in PCWP. It, therefore, follows that in this animal model the infusion of large volumes of wetting solution is safe as long as the animal is healthy. Finally, isoflurane, the general anesthetic used, may be an advantageous agent in large-volume liposuction procedures. Isoflurane has a benefit of decreasing peripheral systemic resistance and may be theoretically beneficial in preventing or controlling marginal fluid overload. ACKNOWLEDGMENTS The authors thank Nilgun M. Ertas for her contribution to the study, Jennifer Mule for her contribution in the data analysis, and Darlene Lyons for preparation of the manuscript. REFERENCES 1. Cosmetic Surgery Quick Facts: 2005 ASAPS Statistics January 3, 2000 Highlights of the ASAPS 2005 Statistics on Cosmetic Surgery Available at: 2. Illouz YG. Body contouring by lipolysis: a 5-year experience with over 3000 cases. Plast Reconstr Surg. 1983;72: Coleman WP III. The history of liposuction and fat transplantation in America. Dermatol Clin. 1999;17: , v. 4. Clayton DN, Clayton JN, Lindley TS, et al. Large volume lipoplasty. Clin Plastic Surg. 1989;16: Lippincott Williams & Wilkins 87

6 Ozmen et al Annals of Plastic Surgery Volume 64, Number 1, January Hetter GP. Blood and fluid replacement for lipoplasty procedures. Clin Plast Surg. 1989;16: Klein JA. Tumescent technique for regional anesthesia permits lidocaine doses of 35 mg/kg for liposuction. J Dermatol Surg Oncol. 1990;16: Hanke CW, Bernstein G, Bullock S. Safety of tumescent liposuction in 15,336 patients: National survey results. Dermatol Surg. 1995;21: Fodor PB. Wetting solutions in aspirative lipoplasty: a plea for safety in liposuction. Aesthetic Plast Surg. 1995;19: Trott SA, Beran SJ, Rohrich RJ, et al. Safety considerations and fluid resuscitation in liposuction: an analysis of 53 consecutive patients. Plastic Reconstr Surg. 1998;102: Rohrich RJ, Beran SJ, Fodor PB. The role of subcutaneous infiltration in suction-assisted lipoplasty: a review. Plast Reconstr Surg. 1997;99: ; discussion Rohrich RJ, Leedy JE, Swamy R, et al. Fluid resuscitation in liposuction: a retrospective review of 89 consecutive patients. Plast Reconstr Surg. 2006; 117: Samdal F, Amland PF, Bugge JF. Blood loss during liposuction using the tumescent technique. Aesthetic Plast Surg. 1994;18: Pitman GH, Aker JS, Tripp ZD. Tumescent liposuction: a surgeon s perspective. Clin Plast Surg. 1996;23: ; discussion de Jong RH, Grazer FM. Perioperative management of cosmetic liposuction. Plast Reconstr Surg. 2001;107: Commons GW, Halperin B, Chang CC. Large-volume liposuction: a review of 631 consecutive cases over 12 years. Plast Reconstr Surg. 2001;108: ; discussion Kenkel JM, Brown SA, Love EJ, et al. Hemodynamics, electrolytes, and organ histology of larger-volume liposuction in a porcine model. Plast Reconstr Surg. 2004;113: Kenkel JM, Lipschitz AH, Luby M, et al. Hemodynamic physiology and thermoregulation in liposuction. Plast Reconstr Surg. 2004;114: ; discussion Engoren M, Barbee D. Comparison of cardiac output determined by bioimpedance, thermodilution, and the Fick method. Am J Crit Care. 2005;14: Foldager N, Blomqvist CG. Repeated plasma volume determination with the Evans Blue dye dilution technique: the method and a computer program. Comput Biol Med. 1991;21: Fisher A, Fisher G. First surgical treatment for molding body s cellulite with three 5 mm incisions. Bull Int Acad Cosmet Surg. 1976;3: Pitman GH. Tumescent technique for local anesthesia improves safety in large volume liposuction (Discussion). Plast Reconstr Surgery. 1993;92: Dohner VA. Iatrogenic pulmonary edema: a review of side effects of fluids, drugs and procedures. Rocky Mt Med J. 1971;68: Muir AL, Flenley DC, Kirby BJ, et al. Cardiorespiratory effects of rapid saline infusion in normal man. J Appl Physiol. 1975;38: Gilliland MD, Coates N. Tumescent liposuction complicated by pulmonary edema. Plast Reconstr Surg. 1997;99: Marmer MJ. Iatrogenic complications. Int Anesthesiol Clin. 1972;10: Adriani J, Zepernick R, Harmon W, et al. Iatrogenic pulmonary edema in surgical patients. Surgery. 1967;61: Klein JA. The tumescent technique. Anesthesia and modified liposuction technique. Dermatol Clin. 1990;8: Klein JA. Tumescent technique for local anesthesia improves safety in large-volume liposuction. Plast Reconstr Surg. 1993;92: ; discussion Iverson RE, Lynch DJ; the ASPS Committee on Patient Safety. Practice advisory on liposuction. Plast Reconstr Surg. 2004;113: Covino B, Fozzard H, Rehder K, et al. Effects of anesthesia: cardiovascular effects of general anesthesia. Am Physiol Soc. 1985;12: Eger EI II, Smith NT, Stoelting RK, et al. Cardiovascular effects of halothane in man. Anesthesiology. 1970;32: Tonkovic-Capin M, Gross GJ, Bosnjak ZJ, et al. Delayed cardioprotection by isoflurane: role of K(ATP) channels. Am J Physiol Heart Circ Physiol. 2002;283:H61 H Malan TP Jr, DiNardo JA, Isner RJ, et al. Cardiovascular effects of sevoflurane compared with those of isoflurane in volunteers. Anesthesiology. 1995; 83: Park KW, Dai HB, Lowenstein E, et al. Isoflurane and halothane attenuate endothelium-dependent vasodilation in rat coronary microvessels. Anesth Analg. 1997;84: Lippincott Williams & Wilkins