http://prf.sagepub.com/Perfusion

http://prf.sagepub.com/content/early/2012/12/20/0267659112469643The online version of this article can be found at:

 DOI: 10.1177/0267659112469643

published online 20 December 2012PerfusionSimpkins

M Armbruster, E Grimley, J Rodriguez, D Nacionales, P Efron, L Moldawer, K Papadopoulos, R Ungaro, A Cuenca and CSoybean oil: a potentially new intravascular perfusate

  

Published by:

http://www.sagepublications.com

can be found at:PerfusionAdditional services and information for    

  http://prf.sagepub.com/cgi/alertsEmail Alerts:

 

http://prf.sagepub.com/subscriptionsSubscriptions:  

http://www.sagepub.com/journalsReprints.navReprints:  

http://www.sagepub.com/journalsPermissions.navPermissions:  

What is This? 

- Dec 20, 2012OnlineFirst Version of Record >>

by guest on December 22, 2012prf.sagepub.comDownloaded from

Perfusion0(0) 1 –7

© The Author(s) 2012Reprints and permission: sagepub.

co.uk/journalsPermissions.navDOI: 10.1177/0267659112469643

prf.sagepub.com

Introduction

Soybean oil micelles were demonstrated in this study to be superior to Ringer’s lactate in raising the blood pres-sure after rapid removal of 55% of mouse blood volume. No previous investigations have reported on the utility of an oil-based perfusion fluid. Soybean oil has proper-ties that would make it ideal for reperfusion after blood loss. Free radical scavenging capabilities of soybean oil have been directly measured and reported.1 Emulsions of soybean oil, which have been used in parenteral nutri-tion since the early 1960s, recently have been shown to mitigate reperfusion injury. In a model of ischemia/reperfusion injury in the isolated rat heart, Liu and col-leagues demonstrated smaller infarct size, lower levels of creatinine kinase and lactate dehydrogenase, and reduc-tion in apoptosis of cardiac myocytes when the heart tis-sue was re-perfused with a commercial 20% soybean oil

emulsion after 40 – 45 minutes of ischemia.2,3 Similarly, Stadler and colleagues found that a 20% soybean oil

Soybean oil: a potentially new intravascular perfusate

M Armbruster,1,2 E Grimley,3 J Rodriguez,3 D Nacionales,1,4 P Efron,1,4 L Moldawer,1,4 K Papadopoulos,2 R Ungaro,1,4 A Cuenca1,4 and C Simpkins5

AbstractBackground: Given that micelles of lipids are colloids, the hypothesis was generated that the rapid administration of large volumes of soybean oil micelles would be an effective perfusion fluid. We also hypothesized that oxygen loading would be enhanced due to the greater solubility of oxygen in lipids compared to water. Methods: A 100% lethal mouse model of blood loss was used to compare the ability of soybean oil micelles to that of Ringer’s lactate, blood and other fluids, with respect to raising and maintaining the blood pressure for one hour. Oxygen on- and off-loading of various concentrations of soybean oil micelles was determined using mass spectroscopy. Nitric oxide uptake by micelles was also determined in a similar fashion. Results: A 20% soybean oil emulsion was superior to Ringer’s lactate in raising and maintaining blood pressure. A 20% soybean oil emulsion with 5% albumin added was superior to shed blood as well as solutions comprised of 5% albumin added to either normal saline or Ringer’s lactate. There was a linear relationship between oxygen content and micelle concentration between 10% and 30%. Off-loading of oxygen from the micelles was nearly as fast as off-loading from water. Nitric oxide also loaded preferentially onto soybean oil micelles. Conclusions: (1) Soybean oil emulsions were superior to other fluids in restoring and maintaining the blood pressure; (2) oxygen-carrying ability of soybean oil micelles exceeds that of water and follows Henry’s law between 10% and 30% w/v oil content; (3) nitric oxide was carried by the micelles; (4) animals receiving soybean oil micelles did not exhibit fat embolization; (5) colloids comprised of soybean oil-containing micelles may be used to replace blood loss and may be used to deliver oxygen and other potentially therapeutic gases such as nitric oxide to tissues.

Keywordssoybean oil; micelles; perfusion fluid; blood loss; oxygen carrying; nitric oxide carrying; Ringer’s lactate; albumin; blood pressure

1University of Florida College of Medicine, Gainesville, FL, USA 2 Department of Chemical and Biomolecular Engineering, Tulane University, New Orleans, LA, USA

3Department of Physics, Centenary College of Louisiana, Shreveport, LA, USA 4 Department of Surgery, University of Florida College of Medicine, Gainesville, FL, USA

5Vivacelle Bio, Shreveport, LA, USA

Corresponding author:Cuthbert Ormond Simpkins II, M.D. Vivacelle Bio3060 Nottingham Drive Shreveport Louisiana 71115, USA Email: irtscience@gmail.com

469643 PRF0010.1177/0267659112469643PerfusionRodriguez et al.2012

by guest on December 22, 2012prf.sagepub.comDownloaded from

2 Perfusion 0(0)

emulsion inhibited reperfusion injury in an ex-vivo liver preparation.4 Another advantage of an oil emulsion is that the micelles can act as a colloid and, thereby, provide a volume that will remain in the intravascular space. The micelles of a commercially available soybean oil emul-sion, Intralipid ® (Baxter, Deerfield, IL, USA), have a mean droplet size distribution of 300 – 400 nm, which is greater than that of albumin (diameter of 7 nm).5 A fur-ther advantage of an oil emulsion is that oxygen and other lipophilic gases are more soluble in oil than in water. Therefore, a greater amount of oxygen could be delivered to tissues by an oil emulsion than by aqueous fluids alone. Moreover, many of the mediators of shock are lipophilic. Therefore, it is possible that these harmful factors could be removed by the infusion of an oil emul-sion. Given these theoretical possibilities, we were moti-vated to determine the feasibility of using a fluid that contained a significant hydrophobic component.

Methods

Soybean Oil Micelles

Soybean oil-in-water emulsions were used in this study. Oil-in-water emulsions consist of micelles with a hydrophobic inner core encapsulated by an amphiphilic emulsifier agent surrounded by a polar liquid carrier. The architecture and charge of the micelle provides stability against phase separa-tion. In this study, the hydrophobic contribution was from soybean oil (SIGMA, St. Louis, MO, USA), the emulsifier was either egg yolk phospholipid (Sigma-Aldrich, St. Louis, MO, USA), soy lecithin (SIGMA-Aldrich) or alpha-phos-phatidylcholine (Avanti Polar Lipids, Alabaster, AL, USA), and the polar liquid was de-ionized water.

Commercial Source. Commercial emulsions were used in animal experiments. Intralipid consisting of a 20% w/v soybean oil-in-water emulsion was employed. In addi-tion to the 20% w/v soybean oil content, the Intralipid used in this study contained 1.2% w/v egg phospholipid (without protein), 2.25% w/v anhydrous glycerol, water, and sodium hydroxide for adjustment of the pH to approximately 8. In some experiments, human albumin (SIGMA-Aldrich) was added, to a concentration of 5% w/v to 20% w/v Intralipid, normal saline, or Ringer’s lac-tate containing only L-lactate. Albumin was chosen as an adjuvant since prior studies have shown it to be capable of scavenging reactive oxygen species.6

A second commercial emulsion, 20% Liposyn® (Hospira, Lake Forest, IL, USA) was used for nitric oxide solubility measurements. Twenty percent Liposyn® consists of 20% w/v soybean oil, 2.5% w/v glycerin, 1.2% w/v egg phospha-tides, NaOH to adjust pH to 8.3, and water to 100 mL.

Twenty percent Intralipid and 20% Liposyn® have an osmolality comparable to human plasma. The pub-

lished micelle diameters for Intralipid and Liposyn® are both less than 450 nm, which is within the United States Pharmacopeia (USP) safety guidelines for par-enteral emulsions. USP guidelines restrict the mean droplet diameter of parenteral lipid emulsions to no more than 500 nm and the volume-weighted distribu-tion of droplets larger than 5 micrometers must be 5% or less.7,8

Sonication. For the determination of the change in oxygen content with respect to lipid concentration, lipid emul-sions were prepared by changing the soybean oil, alpha-phosphatidylcholine, and glycerol concentrations with respect to water while maintaining weight ratios between the three (1g oil to 0.06g alpha-phosphatidylcholine to 0.12 g glycerol). This mixture was put into a 15-ml vial and emulsified in an ice bath using an ultrasound cell disruptor. These preparations were used for the determination of the relationship between micelle concentration and oxygen content. Emulsion viability, as determined through lack of observable phase separation, was maintained throughout the measurement procedure for all samples.

Homogenization and microfluidization. In order to deter-mine the long-term stability of lipid emulsions pre-pared in large volumes, a coarse mixture, consisting of oil, emulsifier and, in some cases, human albumin (Sigma-Aldrich) or NaCl, was homogenized for five minutes with a Silverson® high-shear rotor/stator homogenizer (Silverson, New Orleans, LO, USA). The coarse emulsion produced by the Silverson apparatus was then processed into the final sample by microflu-idization with an M-110T® microfluidizer (Microfluid-ics, Newton, MA, USA). To ensure long-term stability against phase separation across a range of oil contents, the mass of the emulsifying agent used in the emulsion was approximately 6% of the mass of the oil content. Human albumin was included in some emulsions at a concentration of 5% w/v while NaCl was added to some others to a concentration of 154 mM.

Mouse Models

Hemorrhagic shock model used in fluid resuscitation experi-ments. Commercially obtained emulsions were used in the animal experiments because they had been prepared using sterilization procedures. The Animal Care and Use Committee of the Louisiana State University Health Sci-ences Center, Shreveport approved these animal methods. The purpose of these experiments was to assess the ability of various fluids to increase and sustain arterial blood pres-sure for one hour following resuscitation from hemorrhagic shock. Healthy mice weighing between 25 and 47 grams were obtained as a gift from the LSU Health Sciences Center Animal Resource Facility. Ketamine/xylazine anesthesia was

by guest on December 22, 2012prf.sagepub.comDownloaded from

Armbruster et al. 3

administered subcutaneously. After the attainment of effec-tive anesthesia, the jugular vein and/or the ipsilateral carotid artery were cannulated. Over the course of 3 minutes, 55% of the blood volume was removed via the carotid artery. After the removal of blood, test resuscitation fluids were infused over 2 minutes via the carotid artery instead of the jugular vein. The artery was chosen for infusion because, in preliminary experiments using this severe blood loss model and Ringer’s lactate as the perfusate, none of the 5 mice that were infused via the internal jugular vein regained a blood pressure, while 6 out of 6 mice infused via the carotid artery did regain a blood pressure. In order to mimic the circum-stances in the field, the mice were not warmed nor were their respirations supported. In the absence of intervention, the mortality rate was 100% within 10 minutes. The base-line test resuscitation fluid was Ringer’s lactate (RL), with a composition of 109 mM NaCl, 28 mM Na L-Lactate, 4mM KCl, and 1.5 mM CaCl2.

Comparison of the effect on blood pressure was made between mice resuscitated with Ringer’s lactate and 20% Intralipid. In another set of experiments, the effect of albumin was tested. Human albumin (SIGMA-Aldrich) was added, to a concentration of 5% w/v, to 20% Intralipid, Ringer’s lactate, and normal saline and the mice were resuscitated with these fluids and shed blood.

The blood pressure was observed for one hour after resuscitation, after which the mice were euthanized by exsanguination.

Hemorrhagic shock model used in histology. The Institu-tional Animal Care and Use Committee of the Univer-sity of Florida approved these animal procedures. The purpose of these procedures was to examine the lung tissue for evidence of fat emboli following resuscitation from hemorrhagic shock with 20% Intralipid. Observa-tions of blood pressure were not made following resus-citation in this model. Male C57BL/6J mice aged 8-12 weeks were anesthetized using isoflurane, restrained in a supine position, and catheterized via the femoral artery. Laparotomy was performed. Anesthesia was discontin-ued and the catheter was attached to a blood pressure (BP) analyzer. The mice were bled via the femoral artery to a mean BP of 30 mmHg and kept stable for 90 min-utes. The mice were then resuscitated via the femoral artery: three mice were resuscitated with shed blood and three mice were resuscitated with 20% Intralipid, using a volume equal to the volume of blood removed. The mice were euthanized by CO2 asphyxiation at 24 hours post-resuscitation and lung tissue was obtained from two of the three mice resuscitated with 20% Intralipid and two of the three mice resuscitated with blood. Sam-ples were fixed in 10% neutral buffered formalin and sent to the Molecular Pathology and Immunology Core of the University of Florida to be paraffin embedded,

sectioned and mounted for H&E staining. We then examined the sections for evidence of fat emboli.

Gas loading & measurement in laboratory-prepared emulsions

Oxygen loading in test fluids. Two hundred microliters of either water or sonication-prepared soybean oil emul-sion was placed in a 15-mL vial and purged with pure oxygen gas for two minutes. Gas was delivered right above the sample without bubbling. The vial was then rapidly sealed and laid horizontally for 60-90 minutes in order to reach equilibrium. Oxygen loading was per-formed at room temperature (25 – 26oC).

Nitric Oxide loading in test fluids. In order to prevent loss of nitric oxide from the interaction with oxygen, helium was chosen for delivery of nitric oxide to the samples. Five hundred microliters of 20% Liposyn® were placed in a 15-mL vial and nitric oxide at 100 ppm in helium was bubbled into the emulsion for two minutes. For nitric oxide loading in water, two different trials were performed. In the first trial, nitric oxide at 20 kPa and 100 ppm in helium was bubbled into de-ionized water for one minute while, in the second trial, bubbling occurred for two minutes.

Separate procedures were used for loading oxygen and nitric oxide into the sample fluids. The results of prelimi-nary oxygen-loading trials (not reported) found that the bubbling technique led to a slight over-estimation of oxy-gen saturation (approximately 10%) in water and oil emul-sion, compared to what is seen when the fluids are allowed to equilibrate for 60 minutes with head gas. Therefore, the latter procedure was used.

Quantification of dissolved gas in a sample fluid. A novel mass spectroscopy technique was used for the quantifica-tion of nitric oxide and oxygen loaded in sample fluids.9

For the measurement of loaded oxygen in laboratory-prepared micelles and water, 100 microliters of the sam-ple fluid were injected through a septum into the purge vessel of the apparatus. This portion of the apparatus was maintained at 37oC, with a continuous purge stream of pure helium gas and contained a 40-ml solution, con-sisting of deionized water and, when needed, 1 mL of antifoaming solution (1/10 dilution of Antifoam B Emulsion from SIGMA-Aldrich). To prepare the anti-foam solution, 2 mL of 10% anti-foam B emulsion were added to 38 mL of de-ionized water. The purge rate through this vessel was maintained at a constant rate of 100 cc/min with a mass flow controller.

The sample gas was released quickly from the fluid and was transported as a bolus from the purge vessel toward a continuous sampling quadrupole mass spectrometer.

by guest on December 22, 2012prf.sagepub.comDownloaded from

4 Perfusion 0(0)

Flow to the mass spectrometer was limited to no more than 4 cc/min. Signals generated by sample gas contact with the detector were integrated using Peakfit® (Systat Software Inc., Chicago, IL, USA) and compared with saturation values obtained with de-ionized water.

The solubility of nitric oxide in water and Liposyn® was determined with a procedure that minimized the loss of the analyte due to reactivity with oxygen. For the measurement of loaded nitric oxide in Liposyn® and water, 100 microliters of sample fluid were injected into the purge vessel of the apparatus. After each measure-ment of a sample fluid bolus, 100 microliters of 100 ppm nitric oxide gas in helium was then injected into the purge vessel and measured by the mass-sensitive detec-tor. As with the oxygen measurements, the water in the purge vessel was maintained at 37°C and contained anti-foam solution when needed.

Size determination of micelles prepared with alpha-phosphatidylcholine emulsifier

Micelle preparations were sought with droplet diameters that did not exceed 5 micrometers. The micelle diameter was measured by passing the emulsion through a particle size detector, the Mastersizer 2000 (Malvern Instruments Ltd., Malvern, Worcestershire, UK).

Statistical analysis

Data were analyzed using a two-tailed unpaired Student’s t test or one-way ANOVA.

Results

Gas loading & phase stability in soybean oil emulsions

Total oxygen content of laboratory-prepared micelles. Over-all oxygen content was shown to be a linear function of the oxygen dissolved in the water and oil phases of the emulsion (Figure 1). Ten measurements were made at each concentration.

Release of oxygen from laboratory-prepared micelles. Mass spectrometry also revealed that oxygen is released from water and lipid emulsions at approximately the same rate, as determined from the temporal width of the mass spectrometric peaks. Differences in the off-loading time between water and lipid emulsion were not significant (p = 0.13727).

Phase stability of laboratory-prepared micelles. Emulsions were prepared with soybean oil concentrations of 10%, 20%, 30%, and 40% w/v, using homogenization and microfluidization. The average micelle diameter,

measured within 24 hours of preparation, was 1-2 µm. The emulsions were refrigerated and all except the 40% preparation remained stable over a 30-day observation period.

Nitric oxide-carrying capacity of Liposyn®. Measurements of NO absorbed by 20% Liposyn® and water were obtained and compared. Ten measurements were made in each group. The mean values (arbitrary units) and standard errors were 3.19×10-3 ± 0.19×10-3 and 2.12×10-3 ± 0.17×10-3 moles per liter for 20% Liposyn® and water, respectively, showing that the solubility of NO in 20% Liposyn® was 1.5 times that of water .

Effect of soybean oil emulsion on blood pressure after severe hemorrhage

Soybean oil micelles vs. Ringer’s lactate. Following hemor-rhage, a group of six mice were resuscitated by rapid infusion of 20% w/v Intralipid (LM) and another group of six mice received Ringer’s lactate (RL). The volume of LM or RL given equaled the hemorrhage volume. Mice were observed after resuscitation for one hour. All mice that received LM survived until they were euthanized. Two of the mice that received RL died within 10 minutes of resuscitation.

Systolic and diastolic pressures were recorded for both groups immediately after hemorrhage and immediately before infusion, 1 minute post-resuscitation, 30 minutes post-resuscitation, and 60 minutes post-resuscitation. These results are reported in Table 1. The post resuscitation pressures in Table 1 show systolic and diastolic pressures minus the pressures immediately before resuscitation. Differences between the LM and RL groups were ana-lyzed using the two-tailed unpaired Student’s t-test. LM was shown to restore both systolic and diastolic pressures to a higher level than RL and this difference was signifi-cant at each time point after resuscitation.

R2 = 0.9955

1.2

1.4

1.6

1.8

2

2.2

0.1 0.15 0.2 0.25 0.3Frac�on by Weight Soybean Oil

Oxy

gen

Carr

ying

Cap

acit

y

Figure 1. Oxygen content relative to water (Y-axis) vs. micelle concentration (X-axis).

by guest on December 22, 2012prf.sagepub.comDownloaded from

Armbruster et al. 5

The effect of albumin on resuscitation with LM, RL, and normal saline. In a separate set of experiments, human albu-min was added, to a concentration of 5% w/v, to LM (LMA), RL (RLA), and normal saline (NSA). Mice were hemorrhaged then resuscitated with LMA, LRA, NSA, or shed blood. Each group consisted of six to seven mice. The volume of resuscitation fluid equaled the hemor-rhage volume. There were no differences among the post-hemorrhage blood pressures between the groups. Systolic and diastolic pressures were measured at 0, 5, 15, 30, and 60 minutes post-transfusion. The 0 time on the graphs below is the blood pressure immediately after infusion of the test solution or emulsion. Data were nor-malized to the pre-hemorrhage pressure and are shown as percentages. Differences were analyzed using two-tailed, one-way ANOVA and Tukey’s test. Results are shown Figures 2 and 3.

LMA was found to be superior to the other fluids in the restoration and maintenance of systolic (p<0.0008) and diastolic pressure (p<0.0007). Using Tukey’s test, we found both systolic and diastolic pressures to be signifi-cantly greater than RLA (p<0.001) NSA (p<0.01) and even shed blood (p<0.05). One mouse died in the shed blood group and three died in the RLA group within the observation period. All mice in the LMA and NSA groups survived the observation period.

Lung histology in mice after resuscitation

At 24 hours post-resuscitation, lung tissue was obtained from two of the three mice resuscitated with LM and two of the three mice resuscitated with shed blood, stained with H&E and examined under light microscopy. Representative samples are shown in Figure 4. At 24

Table 1. Systolic and diastolic pressures in mice receiving LM and RL Effect of LM and RL on Blood Pressure after Severe Hemorrhagic Shock

LM Systolic, mmHg RL Systolic, mmHg

Pre-infusion 12.83 ± 2.96 11.0 +/-3.36 p = 0.6907Immediately post-infusion 51.33 ± 4.29 28.17 +/- 3.38 p = 0.001730 minutes after infusion 45.33 ± 6.21 18.83 ± 6.23 p = 0.013060 minutes after infusion 50.67 ± 5.43 18.50 ± 6.10 p = 0.0028

LM Diastolic, mmHg RL Diastolic, mmHg

Pre-infusion 8.17 ± 3.32 8.0 ±3.32 p = 0.9703Immediately post-infusion 41.67 ± 1.98 9.83 ± 1.66 p<0.0130 minutes after infusion 26.83 ± 6.71 8.50 ± 3.84 p<0.0560 minutes after infusion 24.5 ± 4.60 6.17 ± 3.18 p<0.01

Systolic

Minutes

% Pr

e-Hem

orrha

ge P

ressu

re

0 20 40 60 800

50

100

150

200EmulsionBloodNaClRinger's

Figure 2. Systolic pressures ± SE normalized to the pre-hemorrhage pressure in mice receiving LMA, RLA, NSA, and shed blood

Diastolic

Minutes

% Pr

e Hem

orrha

gic P

ressu

re

0 20 40 60 800

50

100

150

200EmulsionBloodNaClRinger's

Figure 3. Mean diastolic pressures ± SE normalized to pre-hemorrhage pressure in mice receiving LMA, RLA, NSA, and shed blood

by guest on December 22, 2012prf.sagepub.comDownloaded from

6 Perfusion 0(0)

hours post-resuscitation, there was no evidence of fat emboli in the mice resuscitated with LM. The lung parenchyma of the mice resuscitated with LM appears very similar to that of the mice resuscitated with shed blood. Furthermore, all mice in both the LM and shed blood resuscitation groups survived until euthanized at 24 hours post-resuscitation.

Discussion

Our experiments support the rapid infusion of an oil-based emulsion for replacement of the loss of a large per-centage of blood volume. In the first 60 minutes following resuscitation, blood pressure was restored and main-tained after the loss of 55% of blood volume by both LM and LMA. None of the mice that received the soybean oil emulsion, with or without albumin, died before the end

of this 60 minute observation period while animals that were given Ringer’s lactate or blood did. LM was supe-rior to Ringer’s lactate and LMA was superior to blood or albumin in Ringer’s lactate or in normal saline in raising and maintaining the blood pressure in the first 60 min-utes following resuscitation. Furthermore, our model used for histology not only demonstrated absence of fat emboli in mice resuscitated with LM, but also demon-strated long-term survival following resuscitation with LM through the survival of all of these mice until eutha-nized at 24 hours for lung tissue acquisition. No oxygen was added to the emulsion during the blood replacement experiments. We also demonstrated the ability of the emulsion to carry increased amounts of oxygen and nitric oxide.

In exsanguination or in the absence of the ability to increase the cardiac output sufficiently, it would be advantageous for the infusate to be able to carry increased oxygen. Figure 1 shows that the oxygen content of an emulsion is linearly related to the concentration of soy-bean oil. It is not known whether this is enough oxygen to support energy needs, but, based on measurements of oxygen content, we calculate an oxygen content of 55 ml/L for a 30% concentration of soybean oil micelles at 25oC after loading with 100% oxygen. At a hemoglobin concentration of 12 grams/100 ml and 100% oxygen, the oxygen content of blood is 179 ml/L. However, the con-tent of oxygen available to tissues from hemoglobin is actually 54-90 ml/L. In contrast to hemoglobin, 100% of the oxygen dissolved in soybean or other oil should be available. This possibility is supported by our observation of rapid oxygen off-loading during the measurements of oxygen content. Using mass spectroscopy, we also showed enhanced uptake of NO by 20% Liposyn®. This indicates the possibility that other lipophilic gases could be intro-duced into the bloodstream using an emulsion as a car-rier. Loading with NO could open the microcirculation that constricts in shock or be useful in opening the micro-circulation in peripheral vascular disease, traumatic brain injury or after ischemic stroke. In addition, other lipo-philic gases that have therapeutic potential, such as hydrogen sulphide, carbon monoxide or xenon, may be delivered by an emulsion. Hydrogen sulphide has been shown to promote suspended animation.10 Carbon mon-oxide has been shown to inhibit apoptosis11 and xenon is an anesthetic with many favorable characteristics.12

The concerns regarding parenteral infusion of oil emulsions include fat embolization, liver toxicity, fat overload syndrome, rapid clearance and immune sup-pression. However, recent studies have dispelled the issues of liver toxicity, fat overload syndrome, and rapid clearance.13,14,15 Our histological examination of lung tis-sue failed to demonstrate evidence of fat embolization at 24 hours after resuscitation with LM. The representative H&E sections in Figure 4 demonstrate the remarkable

Figure 4a. H&E-stained lung 24 hours after resuscitation with shed blood

Figure 4b. H&E stained lung 24 hours after resuscitation with 20% soybean oil emulsion (20% Intralipid). Lung parenchyma resembles that of mice resuscitated with shed blood. No evidence of fat emboli is observed.

by guest on December 22, 2012prf.sagepub.comDownloaded from

Armbruster et al. 7

similarity in lung parenchyma we observed for mice resuscitated with LM or shed blood.

Our findings open the door to the exploitation of a hydrophobic physicochemical platform from which to direct numerous novel interventions. Emulsions can refill the intravascular space with micellar colloids of a variety of sizes and compositions. Subsequent work will be directed toward investigating the compatibility and utility of this platform in artificial life support scenarios, such as extracorporeal membrane oxygenation using larger animals. In order to study larger animals, we would need large volumes of emulsions. For this reason, we employed homogenization and microfluidization to pre-pare large volumes of emulsions, with soybean oil con-centrations of 10, 20, 30 and 40% w/v. We found that all of the concentrations were stable over the thirty-day observation period except 40%. At 40%, we observed phase separation within this time frame.

In future experiments, we will employ oils from sources other than soybeans. Oil from chia beans or algae are of interest because of their significant anti-inflammatory potential due to their high content of omega-3 fatty acids. In addition, we will determine ways of achieving higher concentrations of stable emulsions at oil concen-trations of 40% and higher.

Additionally, this work may lead to the development of colloids that could be used to perfuse organs for trans-plantation during transit. Future work may also lead to safer and more efficacious exchange transfusion proce-dures for the treatment of acquired coagulopathies. These hydrophobic colloids can absorb lipophilic mediators or be used to deliver gaseous, liquid or solid modulators, intravascularly, in the interstitium or intra-cellularly, for the production of numerous salutary effects in a wide range of different disease states. One final use could be to minimize the hypotension and inflammatory response of routine dialysis.

FundingThis research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Conflict of interest statementThe only author with a possible conflict of interest is Dr. Simpkins. Three US patents and one Chinese patent have been issued. Dr. Simpkins is the sole owner of these patents. He has formed a company, Vivacelle Bio, LLC that is working to bring the emulsion to market.

References 1. Espín JC, Soler-Rivas C, Wichers HJ. Characterization of

the total free radical scavenger capacity of vegetable oils and oil fractions using 2,2-diphenyl-1-picrylhydrazyl rad-ical. J Agric Food Chem 2000; 48: 648–656.

2. Liu SL, Wang Y, Wang RR, et al. Protective effect of Intralipid on myocardial ischemia/reperfusion injury in isolated rat heart. Zhongguo Wei Zhong Bing Ji Jiu Yi Xue 2008; 1110: 227-230.

3. Liu SL, Liu J. Apoptosis mechanism of intralipid postcon-ditioning to reduce ischemia reperfusion injury of iso-lated rat hearts. Sichuan Da Xue Xue Bao Yi Xue Ban. 2007 38: 663-66.

4. Stadler M, Nuyens V, Boogaerts JG. Intralipid minimizes hepatocyte injury after anoxia-reoxygenation in an ex vivo rat liver model. Nutrition 2007; 23: 53-61.

5. Armstrong JK, Wenby RB, Meiselman HJ, Fisher TC. The hydrodynamic radii of macromolecules and their effect on red blood cell aggregation. Biophys J 2004; 87: 4259-4270.

6. Simpkins CO, Little D, Brenner A Griswold JA. Heterogeneity in the effect of albumin and other resusci-tation fluids on intracellular oxygen free radical produc-tion. J Trauma 2004; 56: 548-558

7. Driscoll DF. The pharmacopeial volution of intralipid injectable emulsion in plastic containers: from a coarse to a fine dispersion. Internat J Pharmaceutics 2009; 368: 193-198.

8. Globule size distribution in lipid injectable emulsions. United States Pharmacopeia & National Formulary. Rockville, MD: United States Pharmacopeia, 2007, pp. 3968-3970

9. Cornelius J, Tran T, Turner N, et al. Isotope tracing enhancement of chemiluminescence assays for nitric oxide research. Biol Chem 2009; 390: 181-189.

10. Szabo C. Hydrogen sulphide and its therapeutic potential. Nat Rev Drug Discov 2007; 6: 917-935

11. Rochette L, Cottin Y, Vergely C. Carbon monoxide: mechanisms of action and potential clinical implications. Pharmacol Ther 2012 Sep 29 epub ahead of print http://www.ncbi.nlm.nih.gov/pubmed/23026155t

12. Jordan BD, Wright EL. Xenon as an anesthetic agent. AANA J 2010; 78: 387-392.

13. Nishimura M, Yamaguchi M, Naito S, Yamauchi A. Soybean oil fat emulsion to prevent TPN-induced liver damage: possible molecular mechanisms and clinical implications. Biol Pharm Bull 2006; 29: 855-862.

14. Masterton GS, Plevris JN, Hayes PC. Review article: omega-3 fatty acids - a promising novel therapy for non-alcoholic fatty liver disease. Aliment Pharmacol Ther 2010; 31: 679-692.

15. Lindholm M, Rössner S. Rate of elimination of the intralipid fat emulsion from the circulation in ICU patients. Crit Care Med 1982; 10: 740-746.

by guest on December 22, 2012prf.sagepub.comDownloaded from