Tissue Engineering to Repair Diaphragmatic Defect in a Rat Model.

Tissue engineering is an emerging strategy for repairing damaged tissues or organs. The current study explored using decellularized rat diaphragm scaffolds combined with human amniotic fluid-derived multipotent stromal cells (hAFMSC) to provide a scaffold, stem cell construct that would allow structural barrier function during tissue ingrowth/regeneration. We created an innovative cell infusion system that allowed hAFMSC to embed into scaffolds and then implanted the composite tissues into rats with surgically created left-sided diaphragmatic defects. Control rats received decellularized diaphragm scaffolds alone. We found that the composite tissues that combined hAFMSCs demonstrated improved physiological function as well as the muscular-tendon structure, compared with the native contralateral hemidiaphragm of the same rat. Our results indicate that the decellularized diaphragm scaffolds are a potential support material for diaphragmatic hernia repair and the composite grafts with hAFMSC are able to accelerate the functional recovery of diaphragmatic hernia.

Transforming growth factor-beta 3 alters intestinal smooth muscle function: implications for gastroschisis-related intestinal dysfunction.

BACKGROUND: Gastroschisis (GS) is a congenital abdominal wall defect that results in the development of GS-related intestinal dysfunction (GRID). Transforming growth factor-ß, a pro-inflammatory cytokine, has been shown to cause organ dysfunction through alterations in vascular and airway smooth muscle. The purpose of this study was to evaluate the effects of TGF-ß3 on intestinal smooth muscle function and contractile gene expression. METHODS: Archived human intestinal tissue was analyzed using immunohistochemistry and RT-PCR for TGF-ß isoforms and markers of smooth muscle gene and micro-RNA contractile phenotype. Intestinal motility was measured in neonatal rats ± TGF-ß3 (0.2 and 1 mg/kg). Human intestinal smooth muscle cells (hiSMCs) were incubated with fetal bovine serum ± 100 ng/ml of TGF-ß 3 isoforms for 6, 24 and 72 h. The effects of TGF-ß3 on motility, hiSMC contractility and hiSMC contractile phenotype gene and micro-RNA expression were measured using transit, collagen gel contraction assay and RT-PCR analysis. Data are expressed as mean ± SEM, ANOVA (n = 6-7/group). RESULTS: GS infants had increased immunostaining of TGF-ß3 and elevated levels of micro-RNA 143 & 145 in the intestinal smooth muscle. Rats had significantly decreased intestinal transit when exposed to TGF-ß3 in a dose-dependent manner compared with Sham animals. TGF-ß3 significantly increased hiSMC gel contraction and contractile protein gene and micro-RNA expression. CONCLUSION: TGF-ß3 contributed to intestinal dysfunction at the organ level, increased contraction at the cellular level and elevated contractile gene expression at the molecular level. A hyper-contractile response may play a role in the persistent intestinal dysfunction seen in GRID.

Effects of traumatic brain injury on intestinal contractility.

BACKGROUND: Patients with traumatic brain injury (TBI) often suffer from gastrointestinal dysfunction including intolerance to enteral feedings. However, it is unclear how TBI affects small intestinal contractile activity. The purpose of this study was to determine if TBI affects intestinal smooth muscle function. METHODS: Sprague-Dawley rats were subjected to controlled cortical impact injury (TBI). Sham animals underwent a similar surgery but no injury (SHAM). Animals were sacrificed 1, 3, and 7 days after TBI and intestinal smooth muscle tissue was collected for measurement of contractile activity and transit, NF-kB activity, and cytokine levels. Brains were collected after sacrifice to determine volume loss due to injury. KEY RESULTS: Contractile activity decreased significantly in ileum, but not jejunum, in the TBI group 7 days after injury compared with SHAM. Brain volume loss increased significantly 7 days after injury compared with 3 days and correlated significantly with the contractile activity 1 day after injury. In the intestinal smooth muscle, NF-kB activity increased significantly in the TBI group 3 and 7 days after injury vs SHAM. Wet to dry weight ratio, indicating edema, also increased significantly in the TBI group. Interleukin-1α, -1ß, and -17 increased significantly in the TBI group compared with SHAM. CONCLUSIONS & INFERENCES: Traumatic brain injury causes a delayed but significant decrease in intestinal contractile activity in the ileum leading to delayed transit. The decreased intestinal contractile activity is attributed to secondary inflammatory injury as evidenced by increased NF-kB activity, increased edema, and increased inflammatory cytokines in the intestinal smooth muscle.

Decreased myosin phosphatase target subunit 1(MYPT1) phosphorylation via attenuated rho kinase and zipper-interacting kinase activities in edematous intestinal smooth muscle.

BACKGROUND: Intestinal edema development after trauma resuscitation inhibits intestinal motility which results in ileus, preventing enteral feeding and compromising patient outcome. We have shown previously that decreased intestinal motility is associated with decreased smooth muscle myosin light chain (MLC) phosphorylation. The purpose of the present study was to investigate the mechanism of edema-induced decreases in MLC in a rodent model of intestinal edema. METHODS: Intestinal edema was induced by a combination of resuscitation fluid administration and mesenteric venous hypertension. Sham operated animals served as controls. Contractile activity and alterations in the regulation of MLC including the regulation of MLC kinase (MLCK) and MLC phosphatase (MLCP) were measured. KEY RESULTS: Contraction amplitude and basal tone were significantly decreased in edematous intestinal smooth muscle compared with non-edematous tissue. Calcium sensitivity was also decreased in edematous tissue compared with non-edematous intestinal smooth muscle. Although inhibition of MLCK decreased contractile activity significantly less in edematous tissue compared with non-edematous tissue, MLCK activity in tissue lysates was not significantly different. Phosphorylation of MYPT was significantly lower in edematous tissue compared with non-edematous tissue. In addition, activities of both rho kinase and zipper-interacting kinase were significantly lower in edematous tissue. CONCLUSIONS & INFERENCES: We conclude from these data that interstitial intestinal edema inhibits MLC phosphorylation predominantly by decreasing inhibitory phosphorylation of the MLC targeting subunit (MYPT1) of MLC phosphatase via decreased ROCK and ZIPK activities, resulting in more MLC phosphatase activity.

Evaluating the effects of immediate application of negative pressure therapy after decompression from abdominal compartment syndrome in an experimental porcine model.

PURPOSE: The purpose of this large-animal study was to assess the safety and effects of negative pressure therapy (NPT) when used as temporary abdominal closure in the immediate post-decompression period after abdominal compartment syndrome (ACS). METHODS: Using a hemorrhagic shock/resuscitation and mesenteric venous pressure elevation model, ACS was physiologically induced in 12 female Yorkshire swine. At decompression, animals were allocated to either NPT (n = 6) or Bogota bag (n = 6) as temporary abdominal closure and studied for a period of 48 h or until death. Outcomes measured included morbidity and mortality, as well as hemodynamic parameters, ventilator-related measurements, blood gases, coagulation factors, and organ (liver, kidney, lung, and intestinal) edema and histology at the time of death/sacrifice. RESULTS: All animals developed ACS. Early application of NPT was associated with decreases in mesenteric venous and central venous pressure, and significantly increased drainage of peritoneal fluid. In addition, there was no increase in the incidence of mortality, recurrent intra-abdominal hypertension/ACS, or any deleterious effects on markers of organ injury. CONCLUSIONS: Early application of NPT in this porcine ACS model is safe and does not appear to be associated with an increased risk of recurrent intra-abdominal hypertension. The results of this animal study suggest that the application of NPT following decompression from ACS results in greater peritoneal fluid removal and may translate into augmented intestinal edema resolution secondary to more favorable fluid flux profiles.

A murine model for the study of edema induced intestinal contractile dysfunction.

BACKGROUND: We have published extensively regarding the effects of edema on intestinal contractile function. However, we have found the need to expand our model to mice to take advantage of the much larger arsenal of research support, especially in terms of transgenic mouse availability and development. To that end, we have developed and validated a hydrostatic intestinal edema model in mice. METHODS: Male C57 Black 6 mice were subjected to a combination of high volume crystalloid resuscitation and mesenteric venous hypertension in an effort to induce hydrostatic intestinal edema. Wet to dry ratios, myeloperoxidase activity, mucosal injury scoring, STAT-3 nuclear activation, phosphorylated STAT-3 levels, NF-κB nuclear activation, myosin light chain phosphorylation, intestinal contractile activity, and intestinal transit were measured to evaluate the effects of the model. KEY RESULTS: High volume crystalloid resuscitation and mesenteric venous hypertension resulted in the development of significant intestinal edema without an increase in myeloperoxidase activity or mucosal injury. Edema development was associated with increases in STAT-3 and NF-κB nuclear activation as well as phosphorylated STAT-3. There was a decrease in myosin light chain phosphorylation, basal and maximally stimulated intestinal contractile activity, and intestinal transit. CONCLUSION & INFERENCES: Hydrostatic edema in mice results in activation of a signal transduction profile that culminates in intestinal contractile dysfunction. This novel model allows for advanced studies into the pathogenesis of hydrostatic edema induced intestinal contractile dysfunction.

The effect of transcatheter injections on cell viability and cytokine release of mononuclear cells.

BACKGROUND AND PURPOSE: Several studies suggest that various types of cellular therapies enhance recovery after stroke in animal models. IA-based delivery of cells to the brain is under investigation for stroke, but it is unknown whether cells are injured as a result of being injected through a catheter or exposed to iodinated contrast medium or solutions containing heparin. MATERIALS AND METHODS: We assessed the effect of catheterization with the Excelsior SL-10 catheter or exposure to heparin or iodine contrast on human bone marrow MNCs. Viability and cell injury were assessed by trypan blue exclusion, caspase-3 activity, and lipid peroxidation. Cellular function of MNCs was assessed by their production and release of VEGF, IL-10, and IGF-1. RESULTS: Flow rates of 10 million cells from 0.5 to 2 mL/min did not alter MNC viability; however, 5 mL/min of MNCs did reduce viability by 19%. Iodine and low-dose heparin exposure did not affect cell viability; however, high-dose heparin was cytotoxic. Catheter delivery at 2 mL/min did not affect levels of VEGF, IL-10, or IGF-1. CONCLUSIONS: MNCs do not appear to be damaged by heparin, iodine contrast, and the Excelsior SL-10 catheter at flow rates up to 2 mL/min. However, higher flow rates did reduce viability, and high-dose heparin did cause cell death.

Current trends in cell therapy for pediatric acquired brain injury.

Pediatric traumatic brain injury (TBI) represents a major burden on healthcare worldwide. In the United States, TBI accounts for 435000 Emergency Department visits, 37000 hospital admissions, and approximately 2500 deaths each year. While aggressive early rehabilitation has shown some functional improvement, the acute care of TBI with focus on controlling intracranial pressure while maintaining adequate cerebral perfusion has not shown the ability to reverse neuronal injury on either a cellular or subcellular level. Preliminary investigation into the use of cell therapeutics has shown promise for the treatment of TBI in animal models. While progenitor cells may potentially act via altering the intracerebral milieu (modulation of inflammatory response and trophic factor secretion), the exact mechanism remains controversial. In addition, traditional delivery vehicles (intravenous, intra-arterial, intrathecal injections, and direct implantation) have shown significant barriers to translation coupled with inconsistent results. Therefore, investigation into novel delivery vehicles such as nanofiber scaffolds and hydrogels could enhance transplant cell viability, engraftment, and efficacy. Overall, a large amount of preclinical work remains to clearly define optimal progenitor cell type, dosage, and delivery vehicle. The optimal therapeutic benefit may be seen using a combination of therapies (controlled hypothermia, hypertonic therapy, and/or cellular therapeutics) to attack the complex pathophysiology of TBI at multiple points.

Regulation of endothelial nitric oxide synthase (eNOS) in myocardium subjected to cardioplegic arrest.

BACKGROUND: Nitric oxide (NO) production by both coronary endothelial cells and cardiomyocytes is thought to play a significant role in myocardial pathophysiology following ischemia/reperfusion (I/R). METHODS: In thirteen pigs subjected to 1 hour cardioplegic arrest (CA) on CPB, left ventricular (LV) biopsies were collected prior to CPB (baseline), at 60 min CPA, at 15 and 30 min reperfusion on CPB, and at 120 min post CPB. LV specimens were immunocytochemically stained against phospho-eNOS (Ser1177), phospho-eNOS (Thr495), phosphorylated ERK1/2, and AKT/PKB. Four additional pigs without CA served as controls. Cardiomyocytes were quantitatively investigated using TV densitometry (gray units: U). RESULTS: After 60 min CA phosphorylation of eNOS (Ser1177) increased significantly and remained elevated until 30 min of reperfusion. In contrast, eNOS (Thr495) phosphorylation remained unchanged during CA and throughout reperfusion. In control animals, eNOS phosphorylation remained unchanged. Akt/PKB activity significantly increased after 60 min CA and decreased thereafter. ERK1/2 activity remained unchanged during ischemia but increased during reperfusion. CONCLUSIONS: ENOS activation during ischemia occurs through phosphorylation at Ser1177 mediated by Akt/PKB. ERK1/2 does not seem to be involved in myocardial eNOS regulation especially not via phosphorylation at eNOS (Thr495).

Edemagenic gain and interstitial fluid volume regulation.

Under physiological conditions, interstitial fluid volume is tightly regulated by balancing microvascular filtration and lymphatic return to the central venous circulation. Even though microvascular filtration and lymphatic return are governed by conservation of mass, their interaction can result in exceedingly complex behavior. Without making simplifying assumptions, investigators must solve the fluid balance equations numerically, which limits the generality of the results. We thus made critical simplifying assumptions to develop a simple solution to the standard fluid balance equations that is expressed as an algebraic formula. Using a classical approach to describe systems with negative feedback, we formulated our solution as a "gain" relating the change in interstitial fluid volume to a change in effective microvascular driving pressure. The resulting "edemagenic gain" is a function of microvascular filtration coefficient (K(f)), effective lymphatic resistance (R(L)), and interstitial compliance (C). This formulation suggests two types of gain: "multivariate" dependent on C, R(L), and K(f), and "compliance-dominated" approximately equal to C. The latter forms a basis of a novel method to estimate C without measuring interstitial fluid pressure. Data from ovine experiments illustrate how edemagenic gain is altered with pulmonary edema induced by venous hypertension, histamine, and endotoxin. Reformulation of the classical equations governing fluid balance in terms of edemagenic gain thus yields new insight into the factors affecting an organ's susceptibility to edema.

Inhibition of intestinal transit by resuscitation-induced gut edema is reversed by L-NIL.

BACKGROUND: Post-resuscitation gut edema and associated gut dysfunction is a common and significant clinical problem that occurs after traumatic injury and shock. We have shown previously that gut edema without ischemia/reperfusion injury delays intestinal transit [1]. We hypothesized that gut edema increases expression of inducible nitric oxide synthase (iNOS) protein, and that selective iNOS inhibition using L-NIL reverses the delayed intestinal transit associated with gut edema. MATERIALS AND METHODS: One hour prior to laparotomy, rats were pretreated with 10 mg/kg body weight of intraperitoneal L-NIL or saline vehicle and underwent 80 ml/kg body weight of 0.9% saline + superior mesenteric venous pressure elevation (Edema) or sham surgery (Sham). A duodenal catheter was placed to allow injection of a fluorescent dye for the measurement of intestinal transit. At 6 h, the small bowel was divided and the mean geometric center (MGC) of fluorescent dye was measured to determine transit. Ileum was harvested for histological assessment of mucosal injury, evaluation of iNOS protein expression by Western blotting, and MPO activity. Tissue water was determined using the wet-to-dry weight ratio to assess gut edema. Data are expressed as mean +/- SEM, n = 3-6 and * = P <0.05 using ANOVA. RESULTS: Gut edema, expressed as increased wet-to-dry ratio, was associated with decreased intestinal transit and elevated iNOS protein expression. Pretreatment with l-NIL improved intestinal transit and decreased expression of iNOS protein without decreasing intestinal tissue water compared to edema animals. There was no difference in mucosal injury or MPO activity among groups. CONCLUSION: Gut edema delays intestinal transit via an iNOS-mediated mechanism.

Enhancement of coagulation control using the streaming current detector.

In water treatment processes, the individual unit operations are complex, highly non-linear and poorly understood. Whilst many models have been developed to improve process understanding, these are rarely in a form easily exploited by the control engineer. Attempts to improve the performance of water treatment works through the application of improved control and measurement have had variable success. This paper discusses investigations into the application of feedback control on the clarification process of a large-scale pilot plant using a streaming current detector (SCD). The application is aimed towards maximising the efficiency of the chemical coagulation process. To achieve this, a simple model of the interactions of process operating conditions on the SCD measurements must be made.

Surgery-associated complications in necrotizing enterocolitis: A multiinstitutional study.

PURPOSE: This study was designed to evaluate the wound and stomal complication rate associated with surgical intervention in infants with necrotizing enterocolitis (NEC). METHODS: Comprehensive demographic and perioperative data were collected prospectively from 4 separate university hospitals on 51 infants with surgically treated NEC. The postoperative complication rate included wound (infection, dehiscence) and stomal (prolapse, retraction, necrosis, stricture) problems. For analysis, patients were grouped based on gestational age less than 28 weeks (group I, n = 30) and >/=28 weeks (group II, n = 21). Z-score analysis was used for intergroup evaluation. RESULTS: Significantly more infants in group I (21 of 30 [70%] versus group II, 6 of 21 [29%]; P <.001) were treated initially with Penrose drainage alone, but most eventually underwent laparotomy (group I, 28 of 30 [93%] versus group II, 19 of 21 [91%]; P value, not significant). The combined stomal/wound complication rate was significantly higher in group I (14 of 30 [47%]) versus group II (6 of 21 [29%]; P <.025). Of 51 patients, one operation was required in 23 (45%), 2 in 18 (35%), 3 in 8 (16%), and 4 in 2 (4%). CONCLUSIONS: Although the stomal/wound complication rate was significantly higher in group I, both groups had very substantial complication rates, emphasizing the vulnerability of this infant population. Parents, especially of very premature babies, should be advised that multiple operations are likely and that complications should be expected.

Effect of sialyl Lewis(x) selectin blockade on myocardial protection during cardioplegic arrest and reperfusion.

PURPOSE: Selectins play a crucial role in the neutrophil-mediated myocardial injury associated with ischemia/reperfusion. We investigated the effect of selectin inhibition on neutrophil-endothelial cell adhesion, myocardial water content, and left ventricular (LV) recovery after cardiopulmonary bypass (CPB) and cardioplegia. METHODS: Dogs were subjected to CPB and 60 minutes of hypothermic cardioplegia. A selectin inhibitor (SI) (25 mg/kg) was given five minutes prior to CPB and as a continuous infusion (5 mg/kg/h) throughout CPB (n = 6). Saline-treated controls (n = 6) received identical volumes. Preload recruitable stroke work (PRSW) was calculated by sonomicrometry and micromanometry. Myocardial water content was determined by microgravimetry. Myeloperoxidase (MPO) activity was measured to quantify polymorphonuclear neutrophil (PMN) infiltration. RESULTS: SI did not attenuate PRSW as well as post-CPB MPO tissue activity. While we found no difference in myocardial water gain between groups 120 minutes post-CPB, there was better edema resolution with SI. CONCLUSIONS: Selectin antagonism does not reduce CPB-associated myocardial injury, and contractile recovery is not enhanced.

Effects of selectin-sialyl Lewis blockade on mesenteric microvascular permeability associated with cardiopulmonary bypass.

OBJECTIVES: Cardiopulmonary bypass is associated with an inflammatory response that is associated with a neutrophil-mediated microvascular barrier injury. We studied the effects of blocking neutrophil-endothelial tethering on microvascular permeability and edema formation during cardiopulmonary bypass. Using a selectin antagonist that prevents interactions with their ligands, we hypothesized that there would be less neutrophil infiltration into the tissue and a reduction in microvascular permeability and edema formation. METHODS: A canine mesenteric lymphatic fistula was created to measure Starling forces and to determine microvascular permeability. Normothermic, atrial-femoral cardiopulmonary bypass was initiated (70-90 mL. kg(-1). min(-1)). Intestinal tissue water was determined with microgravimetry. Ileal tissue myeloperoxidase was measured as an index of neutrophil tissue infiltration. One experimental group received the selectin antagonist TBC 1269 before the initiation of bypass, and the control group received saline solution. RESULTS: There was a modest increase in microvascular permeability in both groups, as evidenced by significantly increased transvascular protein clearance and a trend toward a decrease in reflection coefficient. There were no differences in the experimental group compared with the control group. Ileal tissue myeloperoxidase levels were lower in the experimental group than in the control group. CONCLUSIONS: The selectin antagonist TBC 1269 reduces neutrophil infiltration into the ileum without altering ileal microvascular permeability or edema associated with cardiopulmonary bypass.

Emergent abdominal decompression with patch abdominoplasty in the pediatric patient.

BACKGROUND/PURPOSE: Abdominal compartment syndrome (ACS) is the cardiac, pulmonary, and renal dysfunction that occurs as a result of elevated intraabdominal pressure. The authors present their experience with patch abdominoplasty (PA) in pediatric patients as a means to treat and prevent ACS. METHODS: The charts of patients who underwent PA were reviewed retrospectively. ACS was defined as the increased oxygen requirements and elevation of peak inspiratory pressures (PIP) associated with abdominal distension and worsening renal and or cardiac function. RESULTS: A total of 23 patients (13 boys) were treated (average age, 23 months). Diagnoses included necrotizing enterocolitis (NEC, n = 13), trauma (n = 3), Hirschsprung's enterocolitis (n = 2), perforated bowel (n = 4), and bilateral Wilms' tumor with bowel obstruction (n = 1). Oxygen requirements decreased after patch abdominoplasty (mean preoperative FIO2, 0.87 +/- 24, mean postoperative, 0.67 +/- 24 [P = .01]). The PIP decreased significantly in the 13 patients who survived (mean preoperative PIP, 33 +/- 8, mean postoperative PIP, 27 +/- 7 [P = .01]). These PIPs failed to respond in the 8 nonsurvivors (mean preoperative PIP, 35 +/- 10, mean postoperative PIP, 33 +/- 14 [P value not significant]). Six of the 8 nonsurvivors had NEC. Complications of intraabdominal abscess and enterocutaneous fistula were seen in 5 patients, all of who had NEC. CONCLUSIONS: Patch abdominoplasty effectively decreases airway pressures and oxygen requirements associated with ACS. Complications with PA occur primarily in patients with NEC. Failure to respond with a decrease in PIP and FIO2 requirements is an ominous sign.

Extracorporeal circulation exacerbates microvascular permeability after endotoxemia.

BACKGROUND: Extracorporeal life support without prior inflammatory stimuli results in a modest increase in microvascular permeability. Initiating ECLS after shock, sepsis, or hypoxia results in marked increases in interstitial fluid and total body water. We sought to determine whether an inflammatory stimulus prior to initiating ECLS increases microvascular permeability to protein. METHODS: An anesthetized canine lymphatic fistula model was used to study Starling forces. LPS + ECLS received 1 mg/kg LPS 1 h prior to initiating ECLS, and the group ECLS received saline vehicle. To determine mesenteric microvascular permeability, mesenteric venous pressure was elevated to 32 +/- 1 mm Hg to reach a minimal lymph protein concentration (C(L)). With simultaneous measurement of plasma protein concentration (C(P)), the reflection coefficient, sigma, was calculated using the formula sigma = 1 - C(L)/C(P). Transvascular protein clearance and filtration coefficient (K(f)) were calculated from the measured Starling variables. After a steady state was achieved, normothermic right atrial-to-femoral artery ECLS was initiated and continued for 2 h and then discontinued. Measurements were continued for 30 min after ECLS was discontinued. Measurements were continued for 30 min after ECLS was discontinued. Within-group comparisons were made with ANOVA and Fisher's LSD, and between-group comparisons were made with Student's t tests where appropriate. RESULTS: The reflection coefficient (sigma) decreased significantly from 0.77 +/- 0.02 to 0.53 +/- 0.07 with LPS + ECLS and was lower compared to ECLS alone (0.77 +/- 0.02 to 0.65 +/- 0.03). Transvascular protein clearance increased to a greater extent in LPS + ECLS from 266 +/- 46 to 819 +/- 125 microl/min compared to ECLS alone (284 +/- 49 to 819 +/- 125 microl/min) (P = 0. 06). K(f) increased in both groups after initiation of ECLS, but to a significantly greater extent in LPS + ECLS at 60 min (30.7 +/- 2.2 microl/min/mm Hg/g in ECLS and 50.0 +/- 8.9 microl/min/mm Hg/g in LPS + ECLS). Ileal tissue water increased in both groups, but there were no differences between groups. CONCLUSIONS: Initiation of ECLS after exposure to LPS increases ileal microvascular permeability to protein and water to a greater extent than ECLS alone.

Impact of hetastarch on the intestinal microvascular barrier during ECLS.

The effects of hetastarch on microvascular fluid flux were determined in anesthetized dogs undergoing extracorporeal life support (ECLS) with a roller pump and membrane oxygenator. ECLS with a lactated Ringer priming solution resulted in a decrease in microvascular protein reflection coefficient and an increase in transvascular protein clearance. Use of a 6% hetastarch priming solution attenuated the decrease in microvascular protein reflection coefficient and blunted the increase in transvascular protein clearance. Ileal tissue water increased in the group treated with the lactated Ringer priming solution compared with the group treated with 6% hetastarch. The effective plasma-to-interstitial colloid osmotic pressure gradient was greater in the group treated with hetastarch than in the group treated with lactated Ringer solution. Hetastarch decreases the edema associated with ECLS. The reduction in edema is due to the maintenance of the plasma-to-interstitial colloid osmotic pressure gradient and the reduction in the microvascular permeability to protein.