In Clinical and Experimental Therapeutics

Current Issue
Previous Issues
Reprint Information
Back to The Journal of Applied Research
Click here for information on how to order reprints of this article.

Continuous-Dose Furosemide Attenuates TNF Levels in Acute Respiratory Distress Syndrome (ARDS)


Christopher Reising, MD

A. Chendrasekhar, MD

Edward Burt, PhD

P. L. Wall, DVM


Department of Surgery Education, Iowa Methodist Medical Center, 1221 Pleasant Street, Suite 550, Des Moines, Iowa 50309


Financial Support provided by Iowa Health Systems.

Presented at the 64th Annual International Assembly of the American College of Chest Physicians, Toronto, Canada, 1998.


KEY WORDS: furosemide, lung injury, oleic acid, ARDS, TNF


Acute respiratory distress syndrome (ARDS) is a clinical syndrome of progressive hypoxemia and V/Q mismatch with decreasing pulmonary compliance in the absence of congestive heart failure. Tumor necrosis factor (TNF) is a central mediator of the local and systemic inflammatory response that leads to ARDS and is significantly elevated in serum and bronchoalveolar lavage (BAL) samples in animals models of oleic acid induced lung injury. Furosemide has been shown to improve pulmonary gas exchange and intrapulmonary shunt, by a nondiuretic mechanism, in animal models of ARDS. We hypothesized that continuous-dose intravenous furosemide would attenuate the inflammatory response marked by TNF levels in a canine model of oleic acid-induced lung injury.

Methods: Eight mongrel dogs were anesthetized and given 0.1 mg/kg intravenous oleic acid to induce lung injury. Once lung injury was established (2 h), the control animals (n = 4) were continued on standard supportive therapy, and the study animals
(n = 4) were started on continuous-dose intravenous furosemide at 0.2 mg/kg per hour. Serum, BAL, and urine samples were obtained at baseline, established lung injury (2 h), and end of protocol (6 h). These specimens were analyzed with enzyme-linked immunosorbent assays for TNF at baseline, 2 hours, and 6 hours. Data were compared between groups at various stages of the model using one-way analysis of variance (ANOVA) with repeated measures.

Results: Serum TNF levels were significantly decreased at 6 hours in furosemide animals versus control animals (2.04 pg/dL + 0.45 and 1.23 pg/dL + 0.16, respectively; P <0.015). BAL TNF levels at 6 hours were significantly decreased in furosemide animals compared with control animals (2.05 pg/dL + 0.77 vs. 0.31 pg/dL + 0.36; P <0.007).

Conclusion: Continuous-dose furosemide attenuates TNF levels in serum and BAL specimens in this oleic acid model of ARDS.


Acute respiratory distress syndrome (ARDS) is a clinical syndrome with progressive hypoxemia, ventilation/perfusion mismatch, and decreasing pulmonary compliance secondary to increased capillary endothelial permeability. This capillary leak syndrome leads to high protein content interstitial and alveolar fluid accumulation, and is the result of a local and systemic inflammatory response generated by macrophages, neutrophils, and other inflammatory cells. Macrophages are responsible for the release of tumor necrosis factor (TNF), which plays a central role in driving this inflammatory process. TNF stimulates the release of other cytokines and inflammatory mediators, including IL-1, IL-6, IL-8, platelet activating factor, and leukotrienes, which escalate the inflammatory cascade and contribute to local tissue injury. Systemic injection of TNF results in a clinical picture of shock, with a capillary leak syndrome and pulmonary neutrophil sequestration.1 Elevated TNF levels have been demonstrated in the bronchoalveolar lavage (BAL) fluid and serum of patients with ARDS and in animal models of oleic acid-induced lung injury.2-7

Currently, no corrective therapy exists to reverse or attenuate the pathophysiology of inflammatory response to ARDS. Administration of antibody directed against TNF decreases mortality in animal models of septic shock if administered before endotoxin exposure, but has little effect if administered after endotoxin exposure.8 Nonsteroidal antiinflammatory drugs have been shown to decrease airway pressure and pulmonary vascular pressure by inhibiting cyclooxygenase in patients with sepsis syndrome and are currently under further investigation.9 The use of steroids has not been shown to be clinically useful in acute ARDS but could have a role in "fully established "ARDS in minimizing the fibroproliferative stage of ARDS associated with massive interstitial collagen deposition.10,11

Furosemide has been shown to improve pulmonary gas exchange and intrapulmonary shunt fraction in ARDS by a nondiuretic mechanism.12-15,19 The effect of furosemide on inflammatory cascades and cytokine response is unknown. We hypothesized that continuous-dose intravenous furosemide would attenuate the inflammatory response in this animal model of lung injury. To test our hypothesis, we designed this study to measure the effect of furosemide on the cytokine response in a canine model of oleic acid-induced lung injury.


Our institution's laboratory animal care and utilization committee approved this protocol. Animals were cared for in accordance with the current guidelines of the National Institutes of Health. Eight healthy mongrel dogs 19 to 29 kg were premedicated with 1 mg/kg xylazine and 0.8 mL atropine administered intramuscularly. Each animal was anesthetized with 0.8 mg/kg sodium pentothal through an intravenous bolus and 2 mg/kg per hour continuous infusion. All animals were ventilated with a standard mechanical ventilator (Nellcor Puritan-Bennett 7200, Pleasanton, CA) at tidal volumes of 10 cc/kg and respiratory rates of 20 breaths per minute adjusted to maintain pCO2 (35-45 mm Hg). Polyethylene catheters were placed in the femoral artery and right internal jugular vein by cut-down technique. A thermodilution catheter (Baxter Health Care Co, Irvine, CA) was placed in the pulmonary artery. All animals underwent laparotomies with cystostomy catheter placement for urine collection.

After baseline cardiovascular, pulmonary, and blood gas measurements were taken, 0.1 mg/kg oleic acid was injected intravenously. The animals were randomly divided into 2 groups, control (n = 4) and furosemide (n = 4). Serial measurements of cardiac index (CI), stroke volume index (SVI), pulmonary capillary wedge pressure (PCWP), and arterial blood gas analysis were performed at baseline, 2 hours, and 6 hours. Serial urine, BAL, and serum specimens were obtained at baseline, 2 hours, and 6 hours. Specimens were stored at -70˚C.

After establishment of lung injury (2 h), defined as a P02/FI02 of less than 200 with a PCWP of less than 18 mm Hg, continuous-dose furosemide (0.2 mg/kg per h) was initiated in the experimental group. Volume resuscitation from lung injury was achieved with isotonic saline and guided by parameters of maintaining PCWP (8-12 mm Hg) and SVI within 10% of baseline. Ventilators were adjusted to maintain oxygen saturation 90% or greater by adjusting FI02 up to 80% and increasing positive end-expiratory pressure (5-20 cm H2O). After completion of the experimental protocol, the animals were euthanized.

TNF concentrations in serum, BAL, and urine specimens were measured in duplicate using a commercially available enzyme-linked immunosorbent assay (ELISA) kit (Cytokit Red TNF assay, College Park, MD). This ELISA system measures total, bound and unbound, TNF in bodily fluids using precoated goat antirabbit TNF antibody microtiter wells, streptavidin-conjugated alkaline phosphatase, and an amplified color generation system. Microtiter wells and reagents were provided in the kit. Data was collected using an automated plate reader and plotted using computer software.

Statistical analysis was performed on ELISA data using averaged values for each stage of the lung injury model. Each stage values were compared with baseline values using analysis of variance with repeated measures. As part of the analysis of variance, the Student-Newman-Keuls test was used for the multiple comparisons between groups. Statistical significance threshold was P <0.05.


All animals survived the protocol. There was no statistical difference in CI, SVI, or PCWP between furosemide-treated animals and control animals at baseline, 2 hours, or 6 hours. All animals treated with oleic acid had significant elevations in serum and BAL TNF levels at 2 hours. No statistical difference in TNF levels was identified in serum specimens between furosemide-treated animals and control animals at baseline or 2 hours (Table 1). There was a significant decrease in serum TNF levels at 6 hours in the furosemide-treated animals compared with control animals (2.04 ng/dL + 0.45 vs. 1.23 ng/dL + 0.16, respectively; P <0.015; Table 1). Similarly, there was no difference in BAL TNF levels at baseline and 2 hours between the 2 groups; however, there was a significant decrease in BAL TNF levels in furosemide-treated animals compared with control animals (2.04 ng/dL + 0.77 vs. 0.31 ng/dL + 0.36; P <0.007; Table 2). Urine TNF levels (concentration) were not significantly different at 6 hours between furosemide-treated animals and control animals (0.52 ng/dL + 0.35 vs. 0.53 ng/dL + 0.65; P = 0.98).


Acute respiratory distress syndrome continues to be a significant challenge for clinicians. Although we are better at supporting these patients through their acute and chronic phases of lung injury and shock, there is no corrective antidote in our armamentarium to attenuate this inflammatory response or its sequelae. Ali et al. pioneered the first major work on the pulmonary vasoactivity and the nondiuretic capacity of furosemide in elaborate canine models of oleic acid-induced lung injury.12-15 Previously, we demonstrated significant improvements in lung injury scores, oxygenation, and intrapulmonary shunt fraction, independent of cardiac filling pressures, in this canine model of oleic acid lung injury.19 Several mechanisms have been proposed to explain how furosemide effects these improvements in pulmonary performance. Most notably, furosemide might not block hypoxic pulmonary vasoconstriction and preferentially dilate vessels in nonflooded alveolar units, thereby improving ventilation perfusion mismatch.14 To our knowledge, no previous studies have examined the possibility that furosemide might affect the inflammatory cascades in this model of lung injury. This study was designed to examine the cytokine response of oleic acid-induced lung injury and to document any attenuation of this cytokine response by continuous-dose furosemide.

In this model, oleic acid causes increased pulmonary vascular permeability followed by an escalating inflammatory process, which amplifies this injury and leads to noncardiogenic pulmonary edema simulating ARDS. Olanof et al. documented the inflammatory component of this model by demonstrating increased levels of prostaglandin, thromboxane, and TNF locally and systemically.6,17,18 We were able to demonstrate a similar rise in TNF levels both locally (BAL specimens) and systemically (serum levels) after lung injury. We were also able to demonstrate that continuous furosemide infusion after the induction of lung injury attenuates TNF levels, both locally and systemically. However, the mechanism by which furosemide affects this attenuation is unclear. At this point, we can only speculate that furosemide might inhibit the production of inflammatory cytokines like TNF at the level of their synthesis in the macrophage and neutrophil. Additionally, whether this attenuated response of TNF levels contributes to the previously documented improvements in pulmonary performance cannot be established from this study.

Although the urine volumes in the furosemide-treated animals were significantly higher than control animals, urine TNF concentrations were statistically similar between the 2 groups. This indicates that the total amount of TNF eliminated in urine by the furosemide-treated animals was significantly higher than control animals. We know that furosemide does not affect the glomerular filtration rate of the kidney; therefore, we speculate that furosemide might stimulate the nephron to actively or passively secrete TNF. Further studies in this area are necessary to explain these findings.

One limitation of this study is that we did not collect lung histology data to determine if there was a concomitant decrease in the neutrophil sequestration in the pulmonary interstitium along with the attenuated TNF levels in the furosemide-treated animals. Additional research to investigate this issue is necessary. Future studies could also look at the TNF activation status of these cells in the pulmonary interstitium with PCR techniques.

A second limitation of this study is the narrow scope of our assessment of the local and systemic inflammatory response generated by this oleic acid model. We focused on TNF, exclusive of other cytokines, prostaglandins, leukotrienes, complement, or adhesion molecules. We chose to look at TNF for our cytokine quantification for several reasons. First, TNF is the central cytokine mediator of the inflammatory response. Secondly, commercially available cytokine assays for canines are limited and, as stated previously, the sequence homology of TNF between human and canines is well preserved.

Our study was also limited by size (n = 8); however, we were able to demonstrate statistically significant decreases in serum and BAL TNF concentrations in our furosemide-treated animals compared with control animals. This model of lung injury might not reflect a similar pathophysiology of ARDS as that seen clinically with sepsis, trauma, or aspiration. Therefore, extrapolation of our results to clinical patient management strategies is premature. In conclusion, continuous-dose furosemide attenuated TNF levels, both in the bronchoalveolar lavage and serum, of animals with oleic acid-induced pulmonary injury.


1. Tracey KJ, Beytler B, Lowry SF, et al: Shock and tissue injury induced by recombinant cachectin. Science 234:470, 1986.

2.         Millar AB, Singer M, Meager A: Tumor necrosis factor in bronchopulmonary secretions of patients with adult respiratory distress syndrome. Lancet 2:712, 1989.

3.         Meduri GU, Kohler G, Headley S, et al: Inflammatory cytokines in BAL of patients with ARDS. Chest 108:1303-1314, 1995.

4.         Armstrong L, Millar AB: Relative production of tumor necrosis factor alpha and interleukin 10 in adult respiratory distress syndrome. Thorax 52:442-446, 1997.

5.         Goodman RB, Streiter RM, Martin DP, et al: Inflammatory cytokines in patients with persistence of acute respiratory distress syndrome. Am J Respir Crit Care Med 154:602-611, 1996.

6.         Syrbu S, Thrall RS, Smilowitz HM: Sequential appearance of inflammatory mediators in rat bronchoalveolar lavage fluid after oleic acid-induced lung injury. Exp Lung Res 22:33-49, 1996.

7.         Chendrasekhar A, Prabhakar G, Timberlake GA: Alveolar cytokine responses to different modes of mechanical ventilation in acute lung injury. FASEB J 10:A108, 1996.

8.         Demling RH: Role of mediators in human ARDS. J Crit Care 3:56, 1988.

9.         Bernhard GR, Reines HD, Halushka PV, et al: Prostacyclin and thromboxane A2 formation is increased in human sepsis syndrome: Effects of cyclooxygenase inhibition. Am Rev Respir Dis 144:1095, 1991.

10.         Bernhard GR, Luce JM, Sprung CL, et al: High-dose corticosteroids in patients with ARDS. N Engl J Med 317:1565, 1987.

11.         Meduri GU, Chinn AJ, Leeper KV, et al: Corticosteroid rescue treatment of progressive fibroproliferation in late ARDS. Chest 105:1516, 1994.

12.         Ali J, Chernicki W, Wood LDH: Effect of furosemide in canine low pressure pulmonary edema. J Clin Invest 64:1494, 1979.

13.         Ali J, Fischer CJ, Duke K, et al: How does furosemide reduce shunt in low pressure pulmonary edema? [Abstract] Physiologist 22:2, 1979.

14.         Ali J, Unhurt H, Skoog C, et al: Effect of lung edema on pulmonary vasoactivity of furosemide. J Surg Res 35:383, 1983.

15.         Ali J, Wood LDH: Pulmonary vascular effects on gas exchange in pulmonary edema. J Appl Physiol 57:160, 1984.

16.         Schuster DP: ARDS: Clinical lessons from the oleic acid model of acute lung injury. Am J Respir Crit Care Med 149:245, 1994.

17.         Hageman JR, McCulloch K, Hunt CE, et al: Oleic acid induced lung injury increases plasma prostaglandin levels. Prostaglandins Leukotrienes Essential Fatty Acids 35:157-164, 1989.

18.         Olanoff LS, Reines HD, Spicer KM, et al: Effects of oleic acid on pulmonary capillary leak and thromboxanes. J Surg Res 36:597-605, 1984.

19.         Reising CA, Chendrasekhar A, et al: Continuous dose furosemide as a therapeutic approach to acute respiratory distress syndrome. J Surg Res 82:56-60, 1999.


Table 1. Serum TNF Levels

Serum                      Time 0                     2 hours                      6 hours

Control            1.09 ng/dL + 0.90     1.40 ng/dL + 0.63     2.04 ng/dL + 0.45

Furosemide    1.65 ng/dL + 1.12     2.12 ng/dL + 0.47     1.23 ng/dL + 0.16

P value                       N/D                           N/D                        p<0.015


Table 2. BAL TNF Levels

BAL                        Time 0                     2 hours                      6 hours

Control            0.31 ng/dL + 0.61     1.48 ng/dL + 1.39     2.05 ng/dL + 0.77

Furosemide    1.25 ng/dL + 1.30     1.22 ng/dL + 0.21     0.31 ng/dL + 0.36

P value                       N/D                           N/D                         <0.007


©2000-2013. All Rights Reserved. Veterinary Solutions LLC is an authorized retailer for The Journal of Applied Research