Bilateral exudative pleural effusions following intravenous ethchlorvynol administration

A 26-year-old man had bilateral alveolar infiltrates and exudative pleural effusions following self-administration of intravenous ethchlorvynol (ECV). The effusions and pulmonary edema resolved by 72 h with supportive therapy only. As no other etiology was established, we concluded that the pathogenesis of the pleural fluid was the transvisceral pleural leak of the increased extravascular lung water induced by ECV. Current experimental and clinical evidence support the concept that pleural effusions probably develop in most states of permeability pulmonary edema.     

Salicylate pulmonary edema: the mechanism in sheep and review of the clinical literature.

Several clinical reports of salicylate-induced pulmonary edema led us to investigate the mechanism in a chronic unanesthetized sheep preparation. We infused an aspirin-buffer solution intravenously at rates up to 1,200 mg of aspirin per hour and compared effects on lung lymph flow and lymph protein concentration to those seen after mechanical elevation of pulmonary vascular pressures. Aspirin had little effect on lung vascular pressures but caused lung lymph flow to increase an average of greater than twice baseline. Because lymph protein concentrations were higher for a given lymph flow with aspirin than during mechanical pressure elevation, lymph protein (lymph flow X lymph to plasma protein concentration) increased much more with aspirin. Thus, aspirin appears to cause increased permeability to fluid and protein in the pulmonary vascular bed. Aspirin caused arterial PO2 to decrease from 83 +/- 3 SE mm Hg to 74 +/- 3 mm Hg (P less than 0.05) and caused postmortem extravascular lung water to increase. These findings are supported by a review of the clinical literature, indicating that salicylate pulmonary edema in humans is noncardiac in origin and may occur at doses considered therapeutic for some diseases as well as after overdose.     

Initial treatment of pulmonary edema: a physiological approach.

An understanding of the physiological principles involved in lung fluid balance is useful in the initial treatment of pulmonary edema. Normally, a very small volume of fluid is filtered from the pulmonary vasculature into the interstitial space. This interstitial fluid enters the pulmonary lymphatics and is transferred to mediastinal lymphatics at an estimated rate of 20 ml/hr. Under abnormal circumstances, fluid filtration may occur at such a rapid rate that it overwhelms the lymphatics and interstitial space and results in alveolar flooding. This may occur as a result of increased pulmonary vascular pressure or increased vascular permeability. The two general goals of initial therapy are (1) to relieve hypoxemia and (2) to reduce pulmonary capillary pressure. Relieving hypoxemia may require the use of supplemental oxygen by nasal prongs or mask, continuous positive airway pressure (CPAP) mask, or even endotracheal intubation and mechanical ventilation. Measures to decrease preload and thereby reduce pulmonary capillary pressure include sitting the patient up, administering a loop diuretic or morphine intravenously, and in some circumstances using sublingual nitroglycerin. After initial treatment is underway, a search for and specific management of the underlying cause of pulmonary edema can proceed.     

Pleural effusions associated with ethchlorvynol lung injury result from visceral pleural leak

Traditional thinking suggests that pleural fluid develops on the basis of systemic venous hypertension or a primary pleural process. Recent investigations, however, indicate that both acute lung injury and pulmonary venous hypertension can be important in the pathogenesis of pleural effusions. To evaluate the role of acute lung injury in the formation of pleural effusions, we developed a model of acute, reversible lung injury in NZW rabbits. Intravenous ethchlorvynol (ECV), known to produce permeability edema in humans, was used to produce permeability pulmonary edema in rabbits. The injury was examined over 14 days with bronchoalveolar lavage, pleural fluid analysis, and morphologic analysis. Ethchlorvynol injection (40 mg/kg) produced a PMN-predominant, exudative alveolitis (2 h), alveolar hemorrhage (6 to 10 h), and pleural effusions by 2 h (peak, 10 h). Pathologic findings included a patchy, subpleural, hemorrhagic PMN inflammatory response, which peaked by 24 h, and an acute PMN vasculitis of small arterioles and capillaries; these changes resolved in 5 to 7 days. No parietal pleural abnormalities were observed. We conclude that ECV induces an acute, reversible parenchymal lung injury resulting in a capillary leak and that fluid moves from the interstitium of the lung into the pleural space along a pressure gradient through a relatively permeable mesothelium. The data support the concept that diffuse or localized lung injury can result in pleural effusions.     

Effects of differently composed liposomes on pulmonary arterial pressure in sheep--involvement of pulmonary intravascular macrophages]

The authors previously reported that liposomes, when injected intravenously, produce transitory pulmonary hypertension with increased secretions of thromboxane A2 from activated intravascular macrophages that phagocytize liposomes in sheep. In the present study, we attempted to determine whether such responses were modified by the lipid compositions of the liposomes. Five different types of liposomes were prepared by reverse-phase evaporation. The liposomes used were composed of phosphatidylcholine (PC), cholesterol (CHOL), and either phosphatidylglycerol (PG-liposomes), phosphatidylserine (PS-liposomes), phosphatidylethanolamine (PE-liposomes), stearlyamine (SA-liposomes) or none (PC-liposomes). The net charges of PG and PS-liposomes were negative, SA-liposomes were positive, PE- and PC-liposomes were neutral. Each liposome was injected intravenously to obtain a pulmonary arterial pressure response. Arterial blood was sampled before and after liposome injections to measure thromboxane B2 concentrations. All liposomes, but not PC-liposomes, produced pulmonary arterial hypertension associated with increased arterial thromboxane B2 concentrations, irrespective of the net surface charge of the liposome. PG and PS-liposomes, both of which were negatively charged, showed different dose-response curves, the two different types of neutral liposomes showed different responses, and PC-liposomes produced a small increase in pulmonary arterial pressure. PE-liposomes produced marked increases in the pulmonary arterial pressure. From these results, the authors concluded that pulmonary arterial pressure responses to the liposomes are modified by the lipid compositions of the liposomes, and that this is not caused by the difference in the net charge of each liposome.     

Bupivacaine-induced myocardial depression and pulmonary edema: a case report

Central nervous system and cardiovascular toxicity are well-known side effects of bupivacaine. We report a case of bupivacaine-induced myocardial depression and cardiogenic pulmonary edema. A previously healthy woman developed soon after bupivacaine epidural injection of 5 mL 0.5% (25 mg) cardiogenic shock complicated with pulmonary edema. There were pronounced rales on auscultation with a butterfly sign on chest radiograph. A cardiac ultrasound showed reduced myocardial contractility, diffuse hypokinesia, left ventricular ejection fraction (LVEF) 25%, mitral and pulmonary insufficiency. Right heart catheterization showed increased pulmonary artery wedge pressure (34 mm Hg) and a pulmonary artery pressure of 48 over 33 mm Hg. These findings suggest myocardial depression owing to bupivacaine sodium channel blocking of myocardial nerve and tissue and subsequent reduction of myocardial contractility. The patient completely recovered with normalization of clinical, roentenographic, ultrasound, and hemodynamic findings and discharged 10 days later in good condition.     

Oxygen and resolution of lung injury

We investigated the effects of varying inspired oxygen concentrations on the resolution of oleic acid-induced lung injury in rabbits. Rabbits were injected intravenously with oleic acid and maintained in room air, or exposed to 60, 70, or 80% oxygen for periods of 7 or 10 days. Oleic acid caused hemorrhagic pulmonary edema with hypoxemia. Hypoxemia was more profound in the oxygen-treated animals, a difference that was significant after 7 days' exposure to 60 and 70% oxygen, and after 4 days to 80% oxygen. Mortality was increased in the animals maintained in 80% oxygen. The data suggest that environmental oxygen concentrations greater than 60% interfere with the return to normal lung function following oleic acid injury in rabbits. The hypoxemia may be due to either mismatching of ventilation and perfusion or to a diffusion block resulting from the increased septal width. There was no evidence of massive pulmonary edema as a cause of the hypoxemia. It was not possible to distinguish between injury primarily caused by oxygen and its interference with the healing process.     

FR183998 protects against the increased microvascular permeability associated with ischemia-reperfusion injury in the canine lung

FR183998 (FR) is a Na⁺/H⁺ exchange inhibitor known to protect against cardiac ischemia-reperfusion (I/R) injury. The purpose of this study is to investigate the effects of FR on pulmonary I/R injury.Eleven adult dogs were divided into two groups. FR (1mg/kg) was administered intravenously 5 min prior to ischemia and 5 min prior to reperfusion (FR group; n=5), and a vehicle was injected in the same manner (Control group; n=6). Warm ischemia was induced for 3 hr by clamping the left pulmonary artery and veins. The lung was then reperfused and animals were observed for 2 hr. Left pulmonary vascular resistance (L-PVR), cardiac output (CO), arterial oxygen pressure (PaO₂), and alveolar arterial oxygen pressure difference (A-aDO₂) were measured. The lung specimens were harvested for histological study, and polymorphonuclear neutrophils (PMNs) and the wet-to-dry lung weight ratio (W/D ratio) were quantified.^99mTc-diethyltriaminepentaacetic acid (DTPA)-human serum albumin (HSA) scintigraphy was performed.L-PVR, CO, PaO₂, A-aDO₂, and W/D ratio were maintained at significantly (p<0.05) better levels in the FR group than in the control group. Histological findings revealed alveolar damage with interstitial edema in the control group and only slight interstitial edema in the FR group. PMN infiltration was significantly (p<0.05) more severe in the control group than in the FR group. In the scintigrams,^99mTc-DTPA-HSA was more accumulated in the left lung in the control group than in the FR group.FR protects against the increased microvascular permeability resulting from I/R injury.This paper was presented in part at the Professor Albert Senn: Young Investigator Award session, 41^st Annual World Congress, International College of Angiology, Sapporo, Japan, July 1999.     

Presence and quantification of neuropeptide Y in pulmonary edema fluids in rats

The present study was undertaken to determine the contribution of neuropeptide Y to edema occurrence in neurogenic and hydrostatic pulmonary edema. To induce neurogenic pulmonary edema, fibrinogen and thrombin were injected into the cisterna magna; and to evoke hydrostatic pulmonary edema, saline was infused intravenously. Concentrations of neuropeptide Y in serum and edema fluid were measured using enzyme-linked immunosorbent assay (ELISA), which showed a mean value of 158 nM (95% confidence limit 124-202 nM) in the neurogenic edema fluid, significantly higher than that in the hydrostatic one. Using immunohistochemistry, fluorescent signals reactive to neuropeptide Y were found in the alveolar macrophages and edema fluid in case of fibrin-induced pulmonary edema, but were almost absent in hydrostatic edema and absent in normal lungs. Mean ratio of protein concentrations in edema fluid to that in serum was 74.9 +/- 0.9% in fibrin-induced pulmonary edema, being higher than that in hydrostatic one. From these results, we conclude that neuropeptide Y has a relationship to the high protein concentration ratio, i.e., to increased pulmonary vascular permeability, and consequently may contribute to the development of neurogenic pulmonary edema in rats.     

人源性脐带间充质干细胞治疗实验性肺动脉高压所致右心衰竭的疗效观察

目的 拟通过建立MCT诱导的PAH大鼠模型,观察人脐带间充质干细胞（UC-MSCs）干预对大鼠PAH和右心衰竭的治疗效应。方法 实验动物分为3组（空白对照组、PAH组和UC-MSCs组）,在MCT腹腔注射1周后进行干预,UC-MSCs组舌下静脉注射UC-MSCs悬液,空白对照组和PAH组舌下静脉注射等量生理盐水。第4周大鼠行右心超声心动图、右心导管测压、右心肥厚指数以及肺组织病理等检测。结果 与对照组相比,PAH组大鼠右心室游离壁厚度(RVWT)和右心室内径(RVID)显著增大,肺动脉血流加速时间与射血时间比值（PAT/PET）显著下降,右心收缩压（RVSP）和右心肥厚指数(RVHI)显著增高,肺小动脉血管壁厚度（WT）明显增厚。与PAH组相比,UC-MSCs组RVWT和RVID显著减小,PAT/PET明显升高,RVSP和RVHI明显降低,WT明显变薄。结论 利用MCT腹腔注射成功制备PAH大鼠模型,经舌下静脉注射UC-MSCs可以显著降低肺动脉压力,改善右心功能,逆转肺血管重构。     

Afterload reduction with phentolamine in patients with acute pulmonary edema

Phentolamine in amounts of 10 to 40 μg/kg/min was infused intravenously for the emergency treatment of acute pulmonary edema due to left ventricular failure. Fourteen patients with arteriosclerotic heart disease, ranging in age from 52 to 87 years, had clinical and roentgenographic signs of pulmonary edema. The pulmonary artery wedge pressure was increased to an average of 24 mm Hg and the cardiac index was decreased to 1.9 liters/min/m² or less prior to the administration of phentolamine. A reduction in the pulmonary artery wedge pressure to 14 mm Hg and an increase in the cardiac index to 2.5 liters/min/m² was observed in response to this alpha adrenergic blocking agent. Reduction in peripheral resistance with phentolamine was associated with reversal of pulmonary edema.     

Acute pulmonary edema

Fluid movement from the pulmonary capillaries into the interstitial space occurs continuously and is drained by the lymphatics. With increased leakage or decreased clearance, excessive extravascular lung water accumulates, initially as interstitial edema and subsequently as alveolar edema. The most common cause of pulmonary edema is an increase in microvascular hydrostatic pressure. An increased permeability of the capillaries is the other mechanism of production of pulmonary edema. An acute, critical reduction in colloid osmotic pressure may play a contributory role in pulmonary edema even at normal hydrostatic pressures. Dyspnea, diaphoresis, and anxiety characterize the clinical picture. A history of heart disease and congestive heart failure may be present in CPE, whereas evidence of an inciting event or disease process suggests NCPE. Hypoxia, decreased lung compliance, and increased shunt fraction are seen in both types of pulmonary edema, but the duration of pulmonary edema tends to be more severe and prolonged in NCPE. Evidence of increased permeability in NCPE distinguishes it from CPE. Clinically, this is assumed when pulmonary edema is demonstrated at normal PCWP and when edema fluid protein concentration and COP are close to those of plasma. The management of pulmonary edema consists of the improvement of gas exchange by methods that range from supplemental oxygen administration to mechanical ventilatory support with PEEP, depending on the severity of the disturbance in lung function. Improvement in myocardial function and a decrease in pulmonary congestion are accomplished with diuretics and morphine; in those patients who do not respond to this therapy, manipulation of preload, afterload, and myocardial contractility by vasodilators and inotropic agents may be required. In acute pulmonary edema, intravenously administered agents with a short half-life and rapid onset of action are preferred. The role of colloids in the treatment of pulmonary edema is controversial. The indications for the use of corticosteroids in ARDS are controversial, and an optimum dose has not been determined. Many clinicians tend to choose steroids to treat these patients, but the value of these agents in this setting awaits the results of controlled trials now under way.     

Neutrophil-derived epoxide, 9,10-epoxy-12-octadecenoate, induces pulmonary edema

We have observed that neutrophils biosynthesize linoleate epoxide, 9,10-epoxy-12-octadecenoate, and have named it leukotoxin because of its cytotoxic effect. In this experiment, the effect of leukotoxin on the lung was investigated. Acute effect of leukotoxin: Using Wistar rats, leukotoxin (100 mumol/kg) was injected intravenously for the leukotoxin group, and linoleate (100 mumol/kg) for the linoleate group. Physiological saline was injected as the control. Ten min after injection, rats were divided into 3 groups: (1) lungs were isolated, and lung wet weight, and dry weight were measured; (2) lung lavages were performed, and albumin concentration and activity of angiotensin converting enzyme (ACE) were measured; (3) morphological changes were studied by light and electron microscope. After administration of leukotoxin, lung wet weight/body weight ratios and dry weight/wet weight ratios were increased. Albumin concentration and ACE activity in lung lavages were also increased. Pulmonary edema was also confirmed by light microscopic findings. Alveolar epithelial cell damage and endothelium damage were also observed. Linoleate had no significant effect on these biochemical parameters and morphological findings. Subacute effect of leukotoxin: Twelve hr after administration of leukotoxin (50 mumol/kg) or linoleate (50 mumol/kg), the same studies were performed as in the acute experiments. Immediately after administration of leukotoxin, no significant effect was observed. However, 12 hr later similar changes were observed as in the acute experiments. Linoleate did not show any significant effect 12 hr after injection. These results indicate that leukotoxin biosynthesized by neutrophils might be closely related to the genesis of inflammatory edema.     

Hyperdynamic sepsis in baboons: I. Aspects of hemodynamics

The baboon has a number of advantages as a shock model, as its physiological as well as its biochemical behaviour is similar to man. Therefore we have tried to set up a model to mimic the early hyperdynamic phase of clinical sepsis. Seven baboons, 21-25 kg body weight were kept under EEG servocontrolled anesthesia for 8 hr. During this time live E. coli (ATC #33985) 2 x 10(10) BW/8 hr were continuously infused intravenously. Adequate fluid supply with Ringer's solution (up to 40 ml/kg/hr) was given to keep the pulmonary artery wedge pressure at baseline levels; this procedure resulted in a hyperdynamic response with a cardiac output (CO) 20-35% above baseline and a decrease (20-39%) in mean arterial pressure (MAP), leading to a 50% decrease in peripheral resistance. The pulmonary vascular changes were reflected in an increase of the mean pulmonary pressure (PA) to 42% above baseline and a marked rise in pulmonary vascular resistance (PVR) to 50% above baseline with no additional changes in pulmonary gas exchange. After 8 hr both CO and MAP were inversely correlated (r = 0.9-1) with dramatically increased catecholamine plasma levels (15 times above baseline). With continuous infusion of live E. coli (blood levels 10(5)-10(6) CFU/ml) and massive fluid supply we have successfully mimicked hyperdynamic sepsis with severe organ failure after an 8-hr observation period.     

Indomethacin and ketanserin block ethchlorvynol-induced hypoxemia in dogs

Intravenous injection of ethchlorvynol (ECV) leads to hypoxemia and a permeability pulmonary edema. Whether the hypoxemia is directly attributable to the pulmonary edema or caused by release of mediators has not been explored. Three groups of dogs were studied: (1) ECV, (2) indomethacin--ECV, and (3) ketanserin--ECV. In group 1, 25 to 30 mg/kg of ECV caused a significant fall in PaO2 at 4 min (92 +/- 12.6 to 77 +/- 21 mm Hg, p less than 0.05), which persisted throughout the experiment. The P(A-a)O2 gradient widened significantly at 3 min (22 +/- 11 to 31 +/- 16.8 mm Hg, p less than 0.05) and remained abnormal for the remainder of the experiment. There was no significant fall in PaO2 in groups 2 and 3. Lung tissue water to dry weight ratio increased significantly in all groups at 60 min. Lung tissue water to dry weight ratios were normal at 10 min after ECV injection in additional groups. It was concluded that ECV causes hypoxemia, which is mediated by cyclooxygenase products and 5 hydroxytryptamine. This hypoxemia can be prevented by the administration of drugs that block these products.     

The pulmonary lesion of smoke inhalation in an ovine model

Inhalation injury was induced in chronically instrumented sheep (n = 9) by insufflating them with smoke from burning cotton cloth. Sham animals (n = 9) were insufflated with air. There were no temporal changes in any measured parameter of the sham animals. Smoke induced a depression in PaO2. There was a threefold elevation in protein-rich pulmonary lymph which was sustained for over 48 hours. The lymph-to-plasma oncotic pressure ratio was increased. The cardiac index, left atrial pressure, and pulmonary arterial pressure remained unchanged in both groups. After smoke inhalation, the interstitial levels of neutrophils increased while interstitial antiprotease activity was depressed. The lung lymph concentration of 6-keto prostaglandin F1a, the major metabolite of prostacyclin, was increased. These data suggest that the pulmonary injury following smoke inhalation is the result of an increase in lung microvascular permeability to protein with resultant pulmonary edema. The mechanisms responsible for these changes appear to be related to direct injury to the tracheobronchial tree by cytotoxic agents in the smoke; polymorphonuclear leukocytes; and, possibly, eicosanoids.     

Effects of therapeutic paracentesis on systemic and hepatic hemodynamics and on renal and hormonal function

Thirteen patients with cirrhosis and tense ascites (six with and seven without peripheral edema) underwent 4- to 15-liter paracentesis without intravenous "colloid" replacement. Cardiac output increased from 6.6 +/- 0.7 liters per min at baseline to 8.2 +/- 0.7 liters per min (p less than 0.003) 1 hr after large-volume paracentesis completion and fell to 7.5 +/- 0.69 liters per min (p less than 0.05 vs. baseline, p less than 0.02 vs. 1 hr) 24 hr after large-volume paracentesis completion. There was no change in mean arterial pressure or mean pulmonary artery pressure. Central venous pressure fell from 9.1 +/- 0.8 mm Hg at baseline to 8.6 +/- 1.4 mm Hg 1 hr post-large-volume paracentesis to 6.8 +/- 1.0 mm Hg (p less than 0.005 vs. baseline, p less than 0.02 vs. 1 hr value) at 24 hr, and pulmonary capillary wedge pressure fell from 13.1 +/- 0.9 to 11.1 +/- 1.3 mm Hg 1 hr after large-volume paracentesis and to 9.89 +/- 1.2 (p less than 0.01 vs. baseline, p less than 0.03 vs. 1 hr after large-volume paracentesis) at 24 hr. Heart rate fell from 90 +/- 3.0 to 85 +/- 2.9 beats per min (p less than 0.01) 1 hr after large-volume paracentesis completion, but increased to 89 +/- 2.5 beats per min (p less than 0.02 vs. 1 hr after large-volume paracentesis) at 24 hr.(ABSTRACT TRUNCATED AT 250 WORDS)     

Documentation of pulmonary capillary permeability in the adult respiratory distress syndrome accompanying human sepsis.

Current evidence suggests that pulmonary edema accompanying human sepsis may result either from changes in the serum oncotic and hydrostatic pressures or an increase in the permeability of the pulmonary microvasculature. In this study, we compared the "clearance" of injected 131I-labeled human serum albumin from blood to bronchoalveolar secretions in intubated patients with pulmonary edema secondary to sepsis or myocardial infarction. A significantly increased mean +/- SE clearance of the radionuclide was seen in patients with sepsis (0.34 +/- 0.03 ml per hour) compared to those with myocardial infarction (0.043 +/- 0.008 ml per hour) (P less than 0.001), although both groups had similar degrees of edema on chest radiographs. Because the patients with sepsis had no severe decrease in serum oncotic pressure (18.4 +/- 5.0 mm Hg) or evidence of left heart failure, as determined by the pulmonary wedge pressure (11.0 +/- 6.8 mm Hg), we concluded that the genesis of the pulmonary edema in sepsis was due to an increase in pulmonary microvascular permeability, as measured by the increased clearance of 131I-labeled human serum albumin.     

Role of fibronectin in pancreatitis-associated lung injury

Acute pancreatitis (AP) in humans can lead to increased vascular permeability in the lungs and respiratory failure. Fibronectin plays an important role in maintaining the structural integrity of the pulmonary epithelium and endothelium. However, its importance in pancreatitis-associated lung injury has not been defined. AP was produced by infusing caerulein (5 ug/kg/hr) in rats for 8 or 24 hr. Lung injury was assessed histologically and by determining lung microvascular permeability by bronchoalveolar lavage (BAL) analysis. Organ distribution of a target particle given intravenously was determined by the vascular clearance of magnetic iron oxide particles. Plasma fibronectin was measured by the enzyme-linked immunosorbent assay technique. After 8 hr of cerulein infusion, serum amylase increased 8-fold. Pancreatitis correlated with lung injury. BAL at 8 hr showed a 90% increase (P < 0.05) in albumin levels. Histological analysis at 8 hr revealed an increased number of leukocytes within the lungs. By 8 hr, plasma fibronectin significantly decreased 25% (P < 0.05) and the pulmonary uptake of iron oxide increased 111% (P < 0.05). By 24 hr, these effects had nearly resolved. These results indicate that decreases in serum fibronectin and increases in pulmonary leukocyte margination during acute pancreatitis may compromise the integrity of the air–blood barrier and also increase the pulmonary uptake of circulating pathogenic materials, thus making lung injury more likely.     

Pulmonary transvascular fluid dynamics in sheep during hemorrhage

The effects of hemorrhage on pulmonary hemodynamics and lung transvascular fluid dynamics were studied in sheep. We found that 2 hr of hemorrhage caused a fall in lung lymph flow (p less than 0.05) and no significant change in lymph protein concentration. The fall in lymph flow was not due to decreased vascular surface area since the regional distribution of pulmonary perfusion was not altered during hemorrhage; however, the decrease in lymph flow was associated with decrease (p less than 0.05) in the calculated pulmonary microvascular pressure. The extravascular lung water lung content per g bloodless dry lung was increased (p less than 0.05) in the hemorrhaged sheep from the control values. Pulmonary edema was not due to increased lung vascular endothelial permeability since the net transvascular protein flux was not increased. The finding that pulmonary edema occurred despite the consistent decreases in lymph flow suggests that edema may be due to hemorrhage-induced lymphatic "failure" or that edema fluid is sequestered in spaces (e.g., endothelial cells) where if cannot be drained by the lymphatics.