UvA-DARE is a service provided by the library of the University of Amsterdam (http://dare.uva.nl)

UvA-DARE (Digital Academic Repository)

Experimental strategies in the treatment of acute renal failure in sepsis

Johannes, T.

Link to publication

Citation for published version (APA):Johannes, T. (2011). Experimental strategies in the treatment of acute renal failure in sepsis.

General rightsIt is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s),other than for strictly personal, individual use, unless the work is under an open content license (like Creative Commons).

Disclaimer/Complaints regulationsIf you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, statingyour reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Askthe Library: https://uba.uva.nl/en/contact, or a letter to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam,The Netherlands. You will be contacted as soon as possible.

Download date: 17 Jun 2020

Experimental strategies in the treatment of acute renal failure in sepsis

Experimental strategies in the treatment of acute renal failure in sepsis Thesis, University of Amsterdam, The Netherlands, with summary in Dutch

ISBN: 978-94-610-8137-7

Printed by Gildeprint Drukkerijen, Enschede, The Netherlands

Copyright © 2011 by Tanja Johannes All studies described in this thesis were conducted at

Department of Translational Physiology, Academic Medical Center, University of Amsterdam, The Netherlands Performance of the studies presented in this thesis was financially supported by grants of: - ƒortüne Program to Tanja Johannes (No. 1168-0-0, Medical Faculty, University of Tübingen, Germany) - Deutsche Forschungsgemeinschaft to Tanja Johannes (JO 577/1-1) - Eli Lilly and Co. to Can Ince

Experimental strategies in the treatment of acute renal failure in sepsis

ACADEMISCH PROEFSCHRIFT

ter verkrijging van de graad van doctor

aan de Universiteit van Amsterdam

op gezag van de Rector Magnificus prof. dr. D.C. van den Boom

ten overstaan van een door het college voor promoties

ingestelde commissie

in het openbaar te verdedigen

in de Agnietenkapel

op dinsdag 1 maart 2011, te 14:00 uur

door

Tanja Johannes geboren te Waiblingen, Duitsland

Promotiecommissie

Promotor Prof. dr. ir. C. Ince

Overige leden Prof. dr. J.H. Ravesloot

Prof. dr. S. Florquin

Prof. dr. R.T. Krediet

Prof. dr. R.J. Stolker

Prof. dr. T.M. van Gulik

Prof. dr. K.E. Unertl

Faculteit der Geneeskunde

“Science is a wonderful thing if one does not have to earn one's living at it.” Albert Einstein

for Bert & Maita

Table of contents Chapter 1 Introduction

9

Chapter 2

Influence of fluid resuscitation on renal microvascular PO2 in a normo-tensive rat model of endotoxemia. Critical Care, 10: R88, 2006

13

Chapter 3

Nonresuscitated endotoxemia induces microcirculatory hypoxic areas in the renal cortex in the rat. Shock, 31(1): 97-103, 2009

33

Chapter 4 Low-dose dexamethasone-supplemented fluid resuscitation reverses endotoxin-induced acute renal failure and prevents cortical microvascular hypoxia. Shock, 31(5): 521-528, 2009

47

Chapter 5 Iloprost preserves renal oxygenation and restores kidney function in endotoxemia-related acute renal failure in the rat. Critical Care Medicine, 37(4): 1423-32, 2009

65

Chapter 6

Activated Protein C restores kidney function in a dose-dependent manner in endotoxin-induced acute renal failure in the rat. Submitted for publication

83

Chapter 7 Summary and conclusion

97

Addendum Samenvatting en conclusie

104

List of abbreviations 108

Acknowledgements 109

List of publications 110

Curriculum vitae 112

CHAPTER INTRODUCTION

Chapter 1

10

Sepsis and septic shock are the dominant causes of acute renal failure (ARF) in the intensive care setting, with an associated increase in the likelihood of death, prolonged hospital stay, and increased costs of care (1, 2). The incidence of ARF proportionally rises with the severity of sepsis, occuring in 19% of patients with sepsis, 23% patients with severe sepsis and 51% patients with septic shock (3-5).

To date, no specific renal protective or treatment modalities are available and this often results in the need for supportive treatment (6). Interventions to protect the kidney against acute kidney injury (AKI) include preliminary optimization of renal perfusion by volume therapy supplemented by administration of vasopressors (7). If these strategies fail costly renal-replacement therapy might be necessary.

Unfortunately the pathogenesis of septic ARF remains only partially understood (8, 9). From animal studies we know that alterations in renal blood flow, glomerular and peritubular microcirculation, tubular cell function and structure as well as derangements in cellular bioenergetics are playing a role (8, 10, 11). Furthermore kidney tissue hypoxia may be a contributing factor in the progression of acute kidney failure to AKI (11). Under hypoxic conditions, cells may be injured by the lack of oxygen or by formation of reactive oxygen species. Data with regard to renal hypoxia in the clinical situation are lacking and current concepts regarding renal oxygenation during ARF are presumptive and largely derive from experimental studies.

Within the kidney there is a high heterogeneity of oxygen tensions (12). Although the kidney receives 25% of cardiac output, the presence of oxygen shunting between arterial and venous vessels keeps renal tissue oxygen tensions surprisingly low. The main blood supply is directed to the renal cortex, with oxygen tensions of 30 to 50 mmHg. Due to countercurrent oxygen exchange the oxygen tension in the renal medulla does not rise above 10 to 25 mmHg (13). The fact that not all regions within the kidney are equally well provided with oxygen makes the organ rather sensitive to hypoxic injury. The few experimental studies that have investigated changes in renal tissue oxygenation under septic conditions present contrasting results (14-17). Till now, the relationship between renal oxygen delivery, consumption and tissue oxygenation is poorly understood, especially with regard to biological response and functional consequences.

As mentioned before the mainstay of current conservative therapeutic interventions is aimed at improving kidney perfusion by i.e. fluid expansive therapy (18). However, measures that enhance renal perfusion may have the potential to exacerbate kidney tissue hypoxia by increasing oxygen consumption. Therefore, these clinical interventions may contribute to the development of AKI instead of preventing its progression.

Due to the clinical importance it is of utmost interest to gain further insight into the pathophysiology of acute renal failure in sepsis and to find more specific therapeutic interventions aimed at preventing the development of AKI.

Aim of this thesis The presented thesis is based on the assumption that hypoxia and microcirculatory dysfunction play a central role in the pathogenesis of acute renal failure in sepsis. We hypothesized that strategies aimed on preventing hypoxia and microcirculatory dysfunction result in maintained kidney function. This hypothesis was tested in a clinically relevant model of septic renal failure and resuscitation in the rat.

Introduction

11

Outline of this thesis

Chapter 1 Fluid resuscitation is an early therapeutic strategy in the treatment of septic shock, with the aim to restore blood flow and oxygen delivery to vital organs. Enhancing kidney perfusion by fluid expansive therapy however may have the potential to exacerbate kidney hypoxia due to increased oxygen consumption. Furthermore the decision of which resuscitation solution should be used remains controversial. There is an ongoing debate about the potential of hydroxyethyl starches to impair renal function. In this chapter we tested the hypothesis that renal microvascular PO2 and oxygen consumption are impaired during endotoxemia; that this effect is associated with a diminished renal function; that fluid resuscitation with either colloids or crystalloids improves an impaired µPO2 and oxygen consumption and restores kidney function; and that colloids are better at resuscitating than crystalloids in this context.

Chapter 2 In several animal models of acute renal failure in sepsis, no profound tissue hypoxia or decrease in microcirculatory PO2 can be seen. In this chapter we hypothesized that heterogeneity of microcirculatory oxygen supply to demand in the kidney is obscured when looking at the average PO2 during endotoxemia. Therefore in this study we applied a new technique for studying microvascular oxygen histograms instead of average PO2 readings during endotoxemia. We investigated whether the observed changes were due to changes in renal blood flow or due to endotoxemia. Furthermore we looked at the effects of fluid resuscitation on the slope of the microvascular oxygen histograms.

Chapter 3 An important molecule known to act on the renal vascular tone and therefore being involved in the regulation of the intrarenal oxygen supply is nitric oxide (NO). The main production of NO under septic conditions derives from the enzyme iNOS. This enzyme can be blocked by the glucocorticosteroid dexamethasone. In this chapter we hypothesized that inhibition of iNOS by low-dose dexamethasone would improve an impaired intrarenal oxygenation and kidney function.

Chapter 4 In the kidney, prostaglandins uphold the balance between vasodilatation and vasoconstriction and hereby maintain homeostasis and physiologic renal function. Prostacyclin is a potent vasodilator and involved in the regulation of renal and glomerular hemodynamics, in the secretion of renin and in tubular transport processes. Due to these effects prostacyclin must be considered to play a potential renal protective role in septic acute renal failure. In this chapter we tested the hypothesis that exogenous prostacyclin would counterbalance an endotoxemia-induced intrarenal vasoconstriction and would therefore have beneficial effects on kidney function.

Chapter 5 Activated protein C has been demonstrated to have beneficial effects on inflammation and coagulation during sepsis. Both factors impair the renal microcirculation and therefore disturb oxygen transport to the tissue. In this chapter we hypothesized that treatment with recombinant activated protein C in two different concentrations improves renal microvascular oxygenation and kidney function in endotoxin-induced acute renal failure in the rat.

Chapter 1

12

References 1. Klenzak J, Himmelfarb J. Sepsis and the kidney. Crit Care Clin 2005; 21: 211-22. 2. Oppert M, Engel C, Brunkhorst FM et al. Acute renal failure in patients with severe sepsis and septic shock - a

significant independent risk factor for mortality. Nephrol Dial Transplant 2008; 23: 904-9. 3. Rangel-Frausto MS, Pittet D, Costigan M et al. The natural history of the systemic inflammatory response syndrome

(SIRS). JAMA 1995; 273:117-23. 4. Lopes JA, Jorge S, Resina C et al. Acute kidney injury in patients with sepsis: a contemporary analysis. Int J Infect

Dis 2009; 13: 176-81. 5. Uchino S, Kellum JA, Bellomo R et al. Acute renal failure in critically ill patients: a multinational, multicenter study.

JAMA 2005; 294: 813-8. 6. Palevsky PM, Zhang JH, O'Connor TZ et al. Intensity of renal support in critically ill patients with acute kidney injury.

N Engl J Med 2008; 359: 7-20. 7. Rivers E, Nguyen B, Havstad S et al. Early goal-directed therapy in the treatment of severe sepsis and septic shock.

N Engl J Med 2001; 345: 1368-77. 8. Wan L, Bagshaw SM, Langenberg C et al. Pathophysiology of septic acute kidney injury: what do we really know?

Crit Care Med 2008; 36: 198-203. 9. Ishikawa K, May CN, Gobe G et al. Pathophysiology of septic acute kidney injury: a different view of tubular injury.

Contrib Nephrol 2010; 165: 18-27. 10. Doi K, Leelahavanichkul A, Yuen PS, Star RA. Animal models of sepsis and sepsis-induced kidney injury. J Clin

Invest 2009; 119: 2868-78. 11. Schrier RW, Wang W. Acute renal failure and sepsis. N Engl J Med 2004; 351: 159-69. 12. Baumgartl H, Zimelka W, Lubbers DW. Evaluation of PO(2) profiles to describe the oxygen pressure field within the

tissue. Comp Biochem Physiol A Mol Integr Physiol 2002; 132: 75-85. 13. Brezis M, Rosen S. Hypoxia of the renal medulla - its implications for disease. N Engl J Med 1995; 332: 647-55. 14. Nitescu N, Grimberg E, Guron G. Low-dose candesartan improves renal blood flow and kidney oxygen tension in

rats with endotoxin-induced acute kidney dysfunction. Shock 2008; 30: 166-72. 15. Gullichsen E, Nelimarkka O, Halkola L, Niinikoski J. Renal oxygenation in endotoxin shock in dogs. Crit Care Med

1989; 17: 547-50. 16. James PE, Bacic G, Grinberg OY et al. Endotoxin-induced changes in intrarenal pO2, measured by in vivo electron

paramagnetic resonance oximetry and magnetic resonance imaging. Free Radic Biol Med 1996; 21: 25-34. 17. Linder MM, Hartel W, Alken P, Muschaweck R. Renal tissue oxygen tension during the early phase of canine

endotoxin shock. Surg Gynecol Obstet 1974; 138: 171-3. 18. De Vriese AS. Prevention and treatment of acute renal failure in sepsis. J Am Soc Nephrol 2003; 14: 792-805

13

CHAPTER INFLUENCE OF FLUID RESUSCITATION ON RENAL MICROVASCULAR PO2 IN A NORMOTENSIVE RAT MODEL OF ENDOTOXEMIA

Tanja Johannes, Egbert G Mik, Boris Nohé, Nicolaas JH Raat, Klaus E Unertl, and Can Ince Critical Care, 10: R88, 2006

Chapter 2

14

Abstract

Introduction Septic renal failure is often seen in the intensive care unit but its pathogenesis is only partly understood. This study, performed in a normotensive rat model of endotoxemia, tests the hypotheses that endotoxemia impairs renal microvascular PO2 (µPO2) and oxygen consumption (VO2,ren), that endotoxemia is associated with a diminished kidney function, that fluid resuscitation can restore µPO2, VO2,ren and kidney function, and that colloids are more effective than crystalloids. Methods Male Wistar rats received a one-hour intravenous infusion of lipopolysaccharide, followed by resuscitation with HES130/0.4 (Voluven®), HES200/0.5 (HES-STERIL® 6%) or Ringer's lactate. The renal µPO2 in the cortex and medulla and the renal venous PO2 were measured by a recently published phosphorescence lifetime technique. Results Endotoxemia induced a reduction in renal blood flow and anuria, while the renal µPO2 and VO2,ren remained relatively unchanged. Resuscitation restored renal blood flow, renal oxygen delivery and kidney function to baseline values, and was associated with oxygen redistribution showing different patterns for the different compounds used. HES200/0.5 and Ringer's lactate increased the VO2,ren, in contrast to HES130/0.4. Conclusion The loss of kidney function during endotoxemia could not be explained by an oxygen deficiency. Renal oxygen redistribution could for the first time be demonstrated during fluid resuscitation. HES130/0.4 had no influence on the VO2,ren and restored renal function with the least increase in the amount of renal work.

Fluid Resuscitation

15

Introduction

The kidney is one of the most commonly injured organs in critically ill patients. Acute renal failure is a complication in sepsis, with a prevalence ranging from 25% in severe sepsis to 50% in septic shock (1). Sepsis seems to have an additional impact on outcome, as mortality can be up to 75% among patients with acute septic renal failure (2, 3). The pathogenesis of sepsis-induced renal failure is multifactorial and is characterized by a reduction in the glomerular filtration rate that may occur despite a maintained renal blood flow (RBF) and normal systemic hemodynamics (4).

The morphology of the kidney can range from normal appearing tissue to endothelial damage, medullary blockade with tubular necrosis and disseminated fibrin thrombi (5). Theories on the patho-genesis suggest an uncontrolled and inappropriate release of various inflammatory mediators leading to direct cytotoxic effects or an impairment of the microvascular autoregulation (6). The latter might cause a maldistribution of renal microcirculatory blood flow and oxygen supply. Regarding renal tissue oxygenation, there is a high heterogeneity of oxygen tensions within the organ due to the anatomy of the renal microvasculature (7, 8). The fact that not all regions within the kidney are equally well provided with oxygen makes the organ rather sensitive to hypoxic injury (9). The few studies that have investigated changes in renal tissue oxygenation during endotoxemia present contrasting results (10-12). The relationship between renal oxygen delivery, consumption and tissue oxygenation, especially with regard to biological response and functional consequences, is still poorly understood and the role of oxygen in septic renal failure remains controversial (10, 13, 14).

Fluid resuscitation is an early therapeutic strategy in the treatment of septic shock, with the aim of restoring blood flow and oxygen delivery to vital organs (15). The decision of which solution should be used during resuscitation remains controversial, especially with regard to the kidney. There is an ongoing discussion about the potential of hydroxyethyl starches to impair renal function (16-18). In well-hydrated patients without preexisting renal dysfunction, however, application of starches seems to be safe (19, 20). Fluid resuscitation not only has an influence on systemic hemodynamics but also dilutes the blood, resulting in beneficial effects on the microvasculature (21, 22).

A recently published study from our group demonstrates that resuscitation with HES200/0.5 (HES-STERIL® 6%) could successfully restore a decreased mucosal microvascular PO2 (µPO2) of the pig's intestine after lipopolysaccharide (LPS) infusion (23). In contrast to the mucosal µPO2, the serosal µPO2 remained decreased. The gut mucosa and serosa can be regarded as two differently behaving anatomical compartments, and the same accounts for the kidney cortex and the kidney medulla. The renal tissue PO2 is regionally different, with values around 50 Torr (6.7 kPa) in the cortex and 20 Torr (2.7 kPa) in the medulla (9). As the tissue PO2 reflects the balance between oxygen delivery and consumption of oxygen in viable cells and tissues (24), its observation in a model of septic renal failure can give important information, particularly because renal hypoxia seems to play an important role in the pathogenesis of the disease (9, 25).

The primary objective of the present study is to test the hypothesis that treatment of endotoxemia by fluid resuscitation with either colloids or crystalloids improves an impaired µPO2, resulting in restoration of oxygen consumption and kidney function. Secondary to the primary objective our study involves a detailed description of changes in oxygenation during endotoxemia and a comparison of different resuscitation fluids. Four distinct hypotheses can be identified: that renal µPO2 and oxygen consumption are impaired during endotoxemia; that this effect is associated with a diminished renal function; that fluid resuscitation with either colloids or crystalloids improves an impaired µPO2 and

Chapter 2

16

oxygen consumption and restores kidney function; and that colloids are better at resuscitating than crystalloids in this context.

In the present study we applied a new technique recently developed and validated by our group (26) to a normotensive rat model of endotoxemia. This phosphorescence quenching technique allows the noninvasive quantitative measurement of cortical microvascular PO2 (cµPO2) and medullary microvascular PO2 (mµPO2) and the detection of the renal venous PO2 (PrvO2). A continuous noninvasive measurement of renal oxygen consumption has been made possible with this unique possibility. Furthermore, we determined the glomerular filtration rate and tubular sodium reabsorption, the major energyconsuming and therefore oxygen-consuming process in the kidney.

Materials and methods

Animals

All experiments in this study were approved and reviewed by the Animal Research Committee of the Academic Medical Center at the University of Amsterdam. Care and handling of the animals were in accordance with the guidelines for Institutional and Animal Care and Use Committees. Experiments were performed on 37 Wistar male rats (Charles River, Maastricht, The Netherlands) with a mean ± standard deviation body weight of 282 ± 16 g. Surgical preparation Rats were anesthetized with an intraperitoneal injection of a mixture of 90 mg/kg ketamine (Nimatek®; Eurovet, Bladel, The Netherlands), 0.5 mg/kg medetomidine (Domitor®; Pfizer, New York, NY, USA) and 0.05 mg/kg atropine-sulfate (Centrafarm, Etten-Leur, The Netherlands). After tracheotomy the animals were mechanically ventilated with a FiO2 of 0.4. For drug and fluid administration, four vessels were cannulated with polyethylene catheters (outer diameter, 0.9 mm; Braun, Melsungen, Germany).

A catheter in the right carotid artery was connected to a pressure transducer to monitor the arterial blood pressure and the heart rate. The right jugular vein was cannulated and the catheter tip inserted to a depth close to the right atrium, allowing continuous central venous pressure measurement. Catheters of the same size were placed in the right femoral artery and vein and were used for withdrawal of blood and continuous infusion of Ringer's lactate at a rate of 15 ml/kg/hour (Baxter, Utrecht, The Netherlands). The body temperature of the rat was maintained at 37 ± 0.5°C during the entire experiment. The ventilator settings were adjusted to maintain an arterial PCO2 between 35 and 40 Torr (4.7–5.3 kPa). All preceding steps were described in detail in a previous study (27).

The kidney was exposed, decapsulated and immobilized in a Lucite kidney cup (K. Effenberger, Pfaffingen, Germany) via a 4 cm incision of the left flank. The renal vessels were carefully separated from each other under preservation of the nerves. A 0.5 × 1.0 cm2 piece of aluminum foil was placed on the dorsal site of the renal vein to prevent contribution of underlying tissue to the phosphorescence signal in the venous PO2 measurement. A perivascular ultrasonic transient time flow probe (type 0.7 RB; Transonic Systems Inc., Ithaca, NY, USA) was placed around the left renal artery and connected to a flow meter (T206; Transonic Systems Inc.) to allow continuous measurement of RBF (28). The left ureter was isolated, ligated and cannulated with a polyethylene catheter for urine collection. The operation field was covered with plastic foil throughout the entire

Fluid Resuscitation

17

experiment, to prevent evaporation of body fluids. The experiment was ended by infusion of 1 ml of 3 M potassium chloride inducing a sudden cardiac arrest. Finally, the kidney was removed and weighed, and correct placement of the catheters was checked post mortem.

Hemodynamic and blood gas measurements The mean arterial pressure (MAP) was continuously measured in the carotid artery, calculated as: MAP (mmHg) = diastolic pressure + (systolic pressure – diastolic pressure)/3. Furthermore the blood flow of the renal artery (ml/minute) was measured and recorded continuously. An arterial blood sample (0.2 ml) was taken from the femoral artery at three different time points: first time point, 0 minutes = baseline (t0); second time point, 50 minutes = endotoxemia (t1); and third time point, ~70 minutes = resuscitation (t2). The blood samples were replaced by the same volume of HES130/0.4 (Voluven®, 6% HES 130/0.4; Fresenius Kabi Nederland B.V., Schelle, Belgium). The samples were used for determination of blood gas values (ABL505 blood gas analyzer; Radiometer, Copenhagen, Denmark), as well as for determination of the hematocrit concentration, hemoglobin concentration, hemoglobin oxygen saturation, and sodium and potassium concentrations (OSM 3; Radiometer).

Measurement of renal microvascular oxygenation and renal venous PO2 Oxygen-dependent quenching of phosphorescence was used to detect changes in µPO2 and to measure the PO2 in the renal vein (PrvO2). In brief, after infusion a water-soluble phosphorescent dye (Oxyphor G2; Oxygen Enterprises, Ltd. Philadelphia, PA, USA) binds to albumin. This phosphor-albumin complex is confined to the circulation and emits phosphorescence with a wavelength around 800 nm, if excited by a flash of light (29). The phosphorescence intensity decreases at a rate dependent on the surrounding oxygen concentration. The relationship between the measured decay

time and the PO2 is given by the Stern-Volmer relation: 1/τ = (1/τ0) + kq [O2], where τ is the measured decay time, τ0 is the decay time at an oxygen concentration of zero and kq is the quenching constant.

For oxygenation measurements within the rat renal cortex and the outer medulla, a dual-wavelength phosphorimeter was used. This new method was recently described and validated elsewhere (26). Oxyphor G2 (a two-layer glutamate dendrimer of tetra-(4-carboxy-phenyl) benzoporphyrin) gets excited with light of 440 nm and 632 nm, respectively, which allows a continuous and simultaneous measurement in two different depths, the kidney cortex and the outer medulla. On the basis of a high tissue penetration and the fact of the low light absorbance of blood within the near-infrared spectrum, Oxyphor G2 is also well suited for oxygen measurements in full blood. Using a frequency-domain phosphorimeter and a very thin reflection probe, the technique of oxygen-dependent quenching of phosphorescence was applied for noninvasive detection of the PrvO2.

Calculation of renal oxygen delivery, renal oxygen consumption, renal oxygen extraction and vascular resistance

Renal oxygen delivery was calculated as DO2ren (ml/minute) = RBF × arterial oxygen content (1.31 × hemoglobin × SaO2) + (0.003 × PaO2), where SaO2 is arterial oxygen saturation and PaO2 is arterial partial pressure of oxygen. Renal oxygen consumption was calculated as VO2ren (ml/minute/g) = RBF × (arterial – renal venous oxygen content difference). Renal venous oxygen content was calculated as (1.31 × hemoglobin × SrvO2) + (0.003 × PrvO2). The SrvO2 was calculated using Hill's equation with p50 = 37 Torr (4.9 kPa) and Hill coefficient = 2.7 (30). The renal oxygen extraction ratio was

Chapter 2

18

calculated as O2ERren (%) = VO2ren/DO2ren. Since values of renal venous pressure were not available, an estimation of the vascular resistance of the renal artery flow region was made: MAP – RBF ratio (U) = (MAP/RBF) × 100 (31).

Assessment of kidney function Creatinine clearance (Clearcrea) was assessed as an index of the glomerular filtration rate according to the standard procedure to measure the function of the investigated kidney (13, 32). Calculations of the clearance were made with the standard formula: clearance (ml/minute) = (U × V)/P, where U is the urine concentration of creatinine, V is the urine volume per unit time and P is the plasma concentration of creatinine. The specific elimination capacity for creatinine of the left kidney was normalized to the organ weight. Urine samples from the left ureter were collected at 10-minute intervals for analysis of urine volume and creatinine concentration. Plasma samples for analysis of creatinine were obtained at the midpoint of each 10-minute urine collection period. The concentrations of creatinine in urine and plasma were determined by colorimetric methods.

Furthermore, all urine samples were analyzed for the sodium concentration. The urine sodium concentration (UNa+; mmol/l) was multiplied by the urine volume per unit time to obtain sodium excretion (UNa+ × V). The cost of sodium transport (VO2/TNa+) is the ratio of the total amount of VO2ren over the total amount of sodium reabsorbed (TNa+, mmol/minute), which was determined according to: TNa+ = (Clearcrea × PNa+) - UNa+ × V, where PNa+ is the plasma sodium concentration.

Experimental protocol After an operating time of 60 minutes, two optical fibers for phosphorescence measurements were placed both 1 mm above the decapsulated kidney surface and 1 mm above the renal vein. Oxyphor G2 (1.2 ml/kg; Oxygen Enterprises, Ltd) was subsequently infused intravenously for 15 minutes. After 40 minutes µPO2 and PrvO2 were continuously measured during the entire experiment, and 10 minutes later the baseline blood sample (0.2 ml) was taken via the femoral artery catheter. At this time point the rats were randomized between the nonresuscitation group (n = 8), the resuscitation with HES130/0.4 group (n = 8), the resuscitation with HES200/ 0.5 group (n = 8), the resuscitation with Ringer's lactate group (n = 8), and the control group (n = 5).

In total 32 animals were assigned to receive a one-hour infusion of LPS (10 mg/kg, serotype 0127:B8; Sigma-Aldrich, Zwijndrecht, the Netherlands) to induce endotoxemia. Five animals served as time controls. A second blood sample was taken 50 minutes after the start of LPS infusion and was analyzed as already described. Directly after cessation of LPS infusion, one group of animals received fluid resuscitation with Voluven® (6% HES 130/0.4; Fresenius Kabi) at a rate of 20 ml/hour. For all resuscitation groups the resuscitation target was defined as a five-minute steady plateau in RBF. A second group of rats received fluid resuscitation with HES200/0.5 (HES-STERIL®, 6% HES 200/0.5; Fresenius Kabi) at a rate of 20 ml/hour. A third group of animals received Ringer's lactate (Baxter) as the resuscitation fluid; to ensure the same volume effect, the infusion rate was 60 ml/hour. A fourth group served as controls and did not receive fluid resuscitation after LPS infusion.

The experiment was ended 10 minutes after cessation of fluid resuscitation or at a corresponding time point for the control groups by intravenous bolus injection of 3 M KCl.

Statistical analysis Values are reported as the mean ± standard deviation, unless indicated otherwise. The decay curves of phosphorescence were analyzed using Labview 6.1 software (National Instruments,

Fluid Resuscitation

19

Austin, TX, USA). Statistics were performed using GraphPad Prism version 4.0 for Windows (GraphPad Software, San Diego, CA, USA). Differences within groups were first tested with the one-way analysis of variance for repeated measurements. When appropriate, post-hoc analyses were performed with the Student-Newman-Keuls post test. Intergroup differences were analyzed using the unpaired t test. P < 0.05 was considered significant.

Results

Systemic variables Systemic hemodynamic changes for the time points of baseline (t0), endotoxemia (t1) and resuscitation (t2) are presented in Table 1. Baseline values in the experimental and control groups were no different. LPS infusion induced a slight decrease in the MAP compared with the control group. Resuscitation with HES200/0.5 (HES-STERIL® 6%) and Ringer's lactate restored the MAP to baseline values, whereas after resuscitation with HES130/0.4 (Voluven®) the MAP remained at 96 ± 26 mmHg. Although the MAP significantly increased in the time control group to 129 ± 6 mmHg at t2, all groups showed normotensive values during the entire experiment. After LPS infusion the heart rate increased significantly from 263 ± 21 beats/minute at t0 to 294 ± 30 beats/minute at t2 in the nonresuscitation group. The heart rate increased in all groups receiving fluid resuscitation (versus baseline and control values, P < 0.05). Fluid resuscitation also increased the central venous pressure significantly regardless of the type of fluid. The RBF decreased dramatically during LPS infusion to 50% of baseline values and did not recover in the nonresuscitation group. Both resuscitation with colloids and crystalloid restored the RBF to baseline values. After a sudden decrease in RBF from 5.4 ± 1.0 to 2.0 ± 0.7 ml/minute with LPS infusion, HES200/0.5 restored the RBF most effectively to 6.7 ± 1.7 ml/minute (versus baseline and control values, P < 0.05).

The calculated renal vascular resistance showed a 50% increase from 26 ± 6 dyne/s/cm5 at baseline to 51 ± 22 dyne/s/cm5 at t2 in the nonresuscitation group. This increase in renal vascular resistance was present in all groups receiving LPS and could be normalized to baseline values with fluid resuscitation.

The pH remained 7.4 at all time points in the control group. The pH decreased in all groups receiving LPS from 7.4 at baseline, to 7.3 at t1 and to 7.2 at t2. Fluid resuscitation could not preserve this drop in pH. The negative base excess decreased from -2.1 ± 2.2 mmol/l for all experimental groups at t0 to -8.4 ± 2.4 mmol/l at t1. A further drop to 12.2 ± 6.0 mmol/l in the nonresuscitation group could be prevented by fluid resuscitation.

The resuscitation target for all groups receiving fluid resuscitation was defined as a five-minute steady plateau in RBF. On completion of the experiment, animals resuscitated with HES130/0.4 and HES200/0.5 received an average amount of 5.8 ± 1.3 and 8.0 ± 1.1 ml fluids, respectively, until a plateau in RBF was reached. To reach the same resuscitation target and volume effect, 23.0 ± 4.5 ml Ringer's lactate were administered. Hematocrit values did not change in the control groups. With fluid resuscitation the hematocrit decreased about 21%, 23% and 16% for HES130/0.4, HES200/0.5 and RL, respectively, compared with baseline values.

An example of an experiment is shown in Figure 1. The MAP and RBF started to decrease with the onset of LPS infusion. While the MAP dropped only slightly and began to recover to baseline values after 20 minutes, the RBF remained decreased at 50% of baseline. Resuscitation with

Chapter 2

20

HES130/ 0.4 restored RBF to values ~20% above baseline. The cµPO2 and the mµPO2 only slightly changed during the one-hour LPS infusion. With the onset of fluid resuscitation there was a redistribution of cortical oxygenation towards the medulla.

Baseline (t0) Endotoxemia (t1) Resuscitation (t2)

MAP (mm Hg) NR HES130/0.4 HES200/0.5 RL C

117 ± 6 118 ± 6 119 ± 9

113 ± 10 113 ± 4

105 ± 13 † 102 ± 21 † 105 ± 11 † 102 ± 18 † 127 ± 3 *

96 ± 19 *† 96 ± 26 *† 114 ± 14 123 ± 26

129 ± 6 *‡

HR (bpm) NR HES130/0.4 HES200/0.5 RL C

263 ± 21 268 ± 25 252 ± 16 247 ± 14 261 ± 9

278 ± 27 † 277 ± 22 † 269 ± 27 * 264 ± 26 * 248 ± 9 *

294 ± 30 *† 299 ± 24 *† 295 ± 16 *† 280 ± 19 *† 256 ± 7 ‡

CVP (mm Hg) NR HES130/0.4 HES200/0.5 RL C

3.8 ± 1.3 4.0 ± 0.9 4.2 ± 0.6 3.8 ± 0.9 4.2 ± 0.9

3.9 ± 0.6 4.0 ± 0.9 3.9 ± 0.6 4.0 ± 1.4 4.5 ± 1.8

3.7 ± 0.8 6.3 ± 1.2 *†‡ 6.0 ± 1.5 *‡ 6.7 ± 1.7 *†‡

4.6 ± 1.1

RBF (mL*min-1) NR HES130/0.4 HES200/0.5 RL C

4.9 ± 0.9 4.8 ± 1.0 5.4 ± 1.0 4.9 ± 0.9 5.6 ± 0.9

2.5 ± 1.1 *† 2.1 ± 0.9 *† 2.0 ± 0.7 *† 3.0 ± 1.2 *†

5.1 ± 1.0

2.1 ± 1.3 *† 4.9 ± 1.5 ‡

6.7 ± 1.1 *†‡ 5.7 ± 1.3 *‡ 5.1 ± 1.0 ‡

RVR (dyne*sec/cm5) NR HES130/0.4 HES200/0.5 RL C

26 ± 6 25 ± 6 23 ± 5 24 ± 7 21 ± 3

50 ± 22 *† 58 ± 32 *

57 ± 22 *† 38 ± 9 *† 26 ± 5 *

51 ± 22 *† 21 ± 8 ‡

17 ± 1 †‡ 21 ± 2 †‡ 26 ± 5 *‡

TABLE 1. Systemic Hemodynamics

Values represent mean ± SD. * P < 0.05 vs baseline. † P < 0.05 vs control. ‡ P < 0.05 vs NR. NR = non-resuscitation; HES130/0.4 = Voluven®; HES200/0.5 = HES-STERIL® 6%; RL = Ringer‘s lactate; C = control group; MAP = mean arterial blood pressure; HR = heart rate; CVP = central venous pressure; RBF = renal blood flow; RVR = renal vascular resistance.

FIGURE 1. Example experiment. Lipopolysaccharide (LPS) infusion resulted in a slight initial decline in the mean arterial pressure (MAP) and a marked decrease in renal blood flow (RBF). Whereas the MAP recovered after 20 minutes, the RBF remained unchanged. Fluid resuscitation with 6 ml HES130/0.4 restored RBF to 20% above baseline values. Cortical (cµPO2) and medullary (mµPO2) microvascular PO2 did not change during LPS infusion. Upon fluid resuscitation cµPO2 markedly decreased.

Fluid Resuscitation

21

Renal oxygenation Data of the oxygenation parameters of the kidney are shown in Figures 2, 3, 4. Baseline values in the experimental and control groups were not significantly different. The cµPO2 and mµPO2 decreased significantly during the experiment in all groups: from 71 ± 8 Torr (9.5 ± 1.1 kPa) at t0 to 53 ± 9 Torr (7.1 ± 1.2 kPa) at t2 for the cµPO2, and from 54 ± 5 Torr (7.2 ± 0.7 kPa) at t0 to 43 ± 10 Torr (5.7 ± 1.3 kPa) at t2 for the mµPO2. LPS infusion had no effect on microvascular oxygenation. The medullary PO2 could be significantly restored in animals receiving resuscitation with HES200/0.5 (versus baseline and control values, P < 0.05). The PrvO2 was significantly lower at t2 than at baseline in all groups except the HES130/0.4 group. In the group receiving HES130/0.4, the PrvO2 increased in 50% of the animals and the PrvO2 was unchanged or decreased in the other 50%, explaining a rather high standard deviation. Although no major changes in renal µPO2 occurred, fluid resuscitation regardless of the type of fluid was accompanied by redistribution between the cortical and medullary PO2. This redistribution is demonstrated as changes in µPO2 (shown in Figure 4), defined as the difference in cµPO2 and mµPO2.

CµPO2 decreased whereas mµPO2 was unchanged in animals receiving HES130/0.4, resulting in a µPO2 of 9 ± 5 Torr (1.2 ± 0.7 kPa), which was significantly lower compared with baseline and with the control group (µPO2 12 ± 5 Torr (1.6 ± 0.7 kPa)). In the HES200/0.5 and Ringer's lactate groups, the µPO2 was 7 ± 3 Torr (0.9 ± 0.4 kPa) respectively. When resuscitated with HES200/0.5 both the cµPO2 and mµPO2 increased (mµPO2 > cµPO2), whereas when receiving Ringer's lactate the cµPO2 decreased and the mµPO2 stayed almost unchanged.

In the experimental groups DO2ren decreased immediately during LPS infusion. In the nonresuscitation group DO2ren decreased from 1.15 ± 0.25 ml/minute at baseline to 0.58 ± 0.23 ml/minute at t1, and reached its lowest reading with 0.45 ± 0.27 ml/minute at t2. Fluid resuscitation restored DO2ren to ~0.93 ml/minute, which was slightly but significantly lower than baseline values. VO2ren significantly increased over time from 0.10 ± 0.02 at baseline to 0.18 ± 0.05 ml/minute/g at t2 in the control group. This increase was not present in animals receiving LPS, in whom VO2ren remained around 0.10 ml/minute/g. Fluid resuscitation with HES200/0.5 and Ringer's lactate significantly increased VO2ren to 0.18 ± 0.06 ml/minute/g and 0.29 ± 0.22 ml/minute/g, respectively (versus nonresuscitation, P < 0.05). Resuscitation with HES130/0.4 had no effect on the renal oxygen consumption. Resuscitation with HES200/0.5 and Ringer's lactate let to a marked increase in O2ERren compared with the nonresuscitation group – in contrast to HES130/0.4, which showed no statistical difference.

Chapter 2

22

FIGURE 2. Measured renal oxygenation parameters. (a) Cortical microvascular PO2 (µPO2), (b) medullary µPO2 and (c) renal venous PO2 at baseline (t0), endotoxemia (t1) and resuscitation (t2) in the control (C) group (n = 5), nonresuscitation (NR) group (n = 8), HES130/0.4 resuscitation group (n = 8), HES200/0.5 resuscitation group (n = 8) and Ringer's lactate (RL) resuscitation group (n = 8). *P < 0.05 versus baseline, #P < 0.05 versus control group, •P < 0.05 versus NR group. Rats are individually presented and connected by lines.

Fluid Resuscitation

23

FIGURE 3. Calculated renal oxygenation parameters. (a) Renal oxygen delivery (DO2ren), (b) renal oxygen consumption (VO2ren) and (c) renal oxygen extraction (O2ERren) at baseline (t0), endotoxemia (t1) and resuscitation (t2) in the control (C) group (n = 5), nonresuscitation (NR) group (n = 8), HES130/ 0.4 resuscitation group (n = 8), HES200/0.5 resuscitation group (n = 8) and Ringer's lactate (RL) resuscitation group (n = 8). *P < 0.05 versus baseline, #P < 0.05 versus control group, •P < 0.05 versus NR group. Rats are individually presented and connected by lines.

Chapter 2

24

FIGURE 4. µPO2 between cortical and medullary microvascular PO2 calculated as a measure of oxygen redistribution. Cortical microvascular PO2 (cµPO2) and medullary microvascular PO2 (mµPO2) are shown for baseline (t0), endotoxemia (t1) and resuscitation (t2) in (a) control (C) group (n = 5), (b) nonresuscitation (NR) group (n = 8), (c) HES130/0.4 resuscitation group (n = 8), (d) HES200/0.5 resuscitation group (n = 8) and (e) Ringer's lactate (RL) resuscitation group (n = 8). *P < 0.05 versus baseline, #P < 0.05 versus control group, •P < 0.05 versus NR group. Data presented as mean ± standard deviation. µPO2, microvascular PO2. µPO2, the difference in cµPO2 and mµPO2.

Fluid Resuscitation

25

Renal function The Clearcrea of the left kidney did not change over the time in control rats. In the experimental

groups the averaged Clearcrea was 0.78 ± 0.31 ml/minute/g left kidney weight. The averaged weight of the left kidney was 1.26 ± 0.10 g. At time point t1 all animals receiving LPS were anuric. Fluid resuscitation by all tested fluids restored Clearcrea to baseline values, as presented in Figure 5.

The baseline values for reabsorptive metabolic costs (VO2/TNa+) were similar in all experimental groups. In the control group the VO2/TNa+ quotient was slightly lower then in the other groups and increased nonsignificantly over the time. With resuscitation VO2/TNa+ trended upwards in all groups. In animals receiving Ringer's lactate the VO2/TNa+ increased from 1.21 ± 0.42 at t0 to 2.34 ± 0.87 at t2, which was statistically significant (data shown in Figure 6).

FIGURE 5. Creatinine clearance as an index of the glomerular filtration rate. Creatinine clearance measured at baseline (t0), endotoxemia (t1) and resuscitation (t2) in the control (C) group (n = 5), nonresuscitation (NR) group (n = 8), HES130/0.4 resuscitation group (n = 8), HES200/0.5 resuscitation group (n = 8) and Ringer's lactate (RL) resuscitation group (n = 6). As lipopolysaccharide infusion was regularly associated with anuria, no clearance could be calculated for these animals. Data were normalized per gram of left kidney weight. Data presented as mean ± standard error of the mean.

Chapter 2

26

FIGURE 6. Kidney oxygen consumption per sodium reabsorbed as an index of metabolic cost. Oxygen consumption per sodium reabsorbed (VO2/ TNa+) measured at baseline (t0) and resuscitation (t2) in the control (C) group (n = 5), nonresuscitation (NR) group (n = 8), HES130/0.4 resuscitation group (n = 8), HES200/0.5 resuscitation group (n = 8) and Ringer's lactate (RL) resuscitation group (n = 6). *P < 0.05 versus baseline. Testing was performed using the Student paired t test. Data are presented as mean ± standard error of the mean.

Discussion

The main findings in our study can be summarized as follows. Endotoxemia severely diminished renal function despite having only a minimal effect on the renal cµPO2 and mµPO2 and on renal oxygen consumption. Fluid resuscitation restored renal blood flow and re-established kidney function accompanied this by redistribution of µPO2. Finally, HES130/0.4 was the only resuscitation fluid tested that did not significantly increase VO2ren.

In a normotensive model of endotoxemia we tested four hypotheses: that renal µPO2 and renal oxygen consumption are impaired during endotoxemia; that this effect is associated with a diminished kidney function; that treatment of endotoxemia by fluid resuscitation with either colloids or crystalloids can improve an impaired µPO2 and oxygen consumption and restore kidney function; and that colloids are more beneficial than crystalloids in this context.

These hypotheses must be partly rejected. As regards the first two hypotheses, in contrast to previous investigation in our model, the renal µPO2 and renal oxygen consumption were only minimally affected during endotoxemia, whereas the kidney function was totally diminished. The loss of kidney function therefore cannot be explained by an oxygen deficiency. As hypothesized in the

Fluid Resuscitation

27

third hypothesis, all resuscitation fluids restored the RBF and kidney function to baseline values. Regardless of which resuscitation fluid was used, oxygen redistribution between the cortex and medulla of the kidney was observed. HES200/0.5 and Ringer's lactate significantly increased the renal oxygen consumption, in contrast to HES130/0.4. Regarding the final hypothesis, as both colloids and crystalloids restored kidney function to baseline values, it might be difficult to choose one in favor of the other. Regarding the renal oxygen consumption, however, renal resuscitation with HES130/0.4 might cause the least amount of renal work.

Administration of LPS was characterized by an increased heart rate, a slight reduction in MAP, a marked decline in RBF, an increase in renal vascular resistance, and a reduction in the glomerular filtration rate resulting in anuria. As the initially slightly reduced MAP recovered to baseline values within 20 minutes, we define our model as normotensive endotoxemia.

This study has some limitations. First, we used anesthetized animals, which could affect the hemodynamic response and renal vascular response to LPS and fluid resuscitation. The changes in the MAP, heart rate and RBF, however, were qualitatively similar to previously published data (13, 33, 34). Second, we did not measure lactate or cytokine levels to verify theseptic status. We did, however, observe a continuous decrease in blood pH, accompanied by an increase in negative base excess regarded as indicative of sepsis. Third, we used a continuous infusion of LPS to induce endotoxemia. This approach is not the same as human septic shock because our animal model is acute and neglects the effects of disease progression on organ dysfunction over many hours, a problem encountered in many animal models of sepsis (35).

Only a few studies have been performed investigating the behavior of renal tissue oxygenation during endotoxemia. Gullichsen and colleagues (33) observed in dogs that, after endotoxin infusion, the cortical PO2 was markedly decreased despite a transient increase in RBF. During the next hours of the experiment the cortical PO2 and MAP gradually increased even though the RBF remained relatively depressed. These latter results are in agreement with our observations. Investigations by James and colleagues (11) in mice using electron paramagnetic resonance oximetry showed an initial decrease in cortical PO2 and a slight increase in medullary PO2 during LPS infusion, which recovered to the control values after 40 minutes. This recovery of PO2 in both regions was attributed to a direct toxic effect of LPS on cellular mitochondrial function, resulting in decreased oxygen utilization.

Different findings exist about the influence of endotoxemia on the renal oxygen extraction and oxygen consumption (13, 14, 33), and the role of oxygen supply herein is controversial. In dogs (33) and sheep (14), for example, an impaired oxygen extraction and decreased oxygen consumption was found. In rats, however, oxygen consumption was not impaired by endotoxemia since the kidney was able to increase the oxygen extraction in the presence of a decreased renal oxygen delivery (13). An interesting finding in the latter study was that oxygen consumption did not decrease while the effective tubular sodium reabsorption was reduced, indicating that another oxygen-consuming mechanism was induced in their septic model.

Overall, our findings are in agreement with these findings by Heemskerk and colleagues (13). Oxygen consumption did not change or was only slightly impaired during endotoxemia and increased during resuscitation. The latter increase was accompanied by an increase in the VO2/TNa+ ratio. The causes of that increase in VO2/TNa+ are unknown, but mitochondrial dysfunction (uncoupling) cannot be excluded. In our model µPO2 was well preserved during endotoxemia, indicating that although DO2ren decreased about 50% this was still sufficient to maintain adequate

Chapter 2

28

oxygenation. In our model, therefore, the role of oxygen supply on changes in O2ERren and VO2ren is likely to be limited. Although no significant difference in baseline oxygenation parameters existed between the groups, it is clear from Figures 2 and 3 that a fairly large variation existed between individual animals in all parameters except the mµPO2. The study of Heemskerk and colleagues (13) reported individual values in the same manner and showed similar variations in renal DO2 and VO2 values.

Several studies indicate an intrarenal redistribution in blood flow from the cortex toward the medulla during endotoxemia (25, 36, 37). This suggests that the renal cortex and medulla have different responses to endotoxin in animal experiments. As the renal medulla is especially vulnerable to hypoxia and ischemia (9), sepsis-induced medullary ischemia has been suspected as a major mechanism of such renal injury. In our model the values of cµPO2 and mµPO2 only declined slightly. The loss in renal function can therefore probably not be attributed to hypoxia. Furthermore, there was no significant evidence of intrarenal redistribution of oxygen during endotoxemia as could be expected with redistribution in blood flow. However, a significant redistribution of oxygen during resuscitation could be observed.

Acute renal failure in sepsis is considered, at least in part, the result of renal ischemia. That is why restoration of adequate RBF by fluid resuscitation should therefore be the primary means of renal protection in critically ill patients. There are few data on the influence of fluid resuscitation on tissue oxygenation (38). Fluid resuscitation with HES200/0.5 (HES-STERIL® 6%) was successful in correcting intestinal mucosal µPO2, but not in normalizing serosal µPO2 in a normodynamic, lowdose endotoxic pig model (23). In our model, fluid resuscitation was associated with oxygen redistribution. This redistribution occurred regardless of the type of used fluid and was smallest, although significant, for HES130/0.4. In the HES200/0.5 group both the cµPO2 and mµPO2 increased, and this may most probably be explained by an increase in cortical and medullary blood flow. When Ringer's lactate was used for resuscitation, the cµPO2 decreased whereas the mµPO2 was rather unaffected. A possible explanation of the decrease in PO2 in the cortex might be due to an increase in oxygen-consuming reabsorption of sodium in the cortical collecting tubules following an increase in the glomerular filtration rate.

In our study all resuscitation fluids were able to normalize the RBF and oxygen delivery. In the only comparable study the gelatin- based resuscitation fluid Haemaccel® was able to restore reduced oxygen delivery during endotoxemia; the oxygen consumption was unaffected (39). We observed the same effects for HES130/0.4. The application of HES200/0.5 and Ringer's lactate, however, was accompanied by a significant increase in oxygen consumption. Based on the current data we can only speculate on the exact reason(s) why both these fluids significantly increased the renal oxygen consumption compared with baseline values in contrast to HES130/0.4. To reach our chosen resuscitation endpoint (a plateau in RBF) higher volumes of HES200/0.5 and Ringer's lactate where needed, resulting in both a higher volume load and a higher total sodium load. Moreover, systemic hemodynamic parameters were better restored with HES200/0.5 and Ringer's lactate, accompanied by a significant increase in RBF. These factors are all known to influence renal oxygen consumption (40-42). The increase in renal oxygen consumption was accompanied by an increase in oxygen extraction, and therefore a decrease in renal venous PO2, for HES200/0.5 and Ringer's lactate. A point of interest from Figure 6 is that the VO2/TNa+ ratio has the tendency to increase more in the case of HES200/0.5 and Ringer's lactate as compared with HES130/0.4. This indicates that at

Fluid Resuscitation

29

least part of the oxygen consumption is used by processes other than sodium reabsorption, or alternatively that sodium reabsorption is energetically less effective (higher oxygen cost).

The glomerular filtration rate as an index of kidney function could be restored to baseline values regardless of the type of resuscitation fluid. In two other studies the application of a colloid solution could not restore an endotoxemia-induced reduction in the glomerular filtration rate (32, 39), although in our model the fluid resuscitation with the aim of restoring RBF was more aggressive using much higher volumes. The controversy cannot be explained by the fact that we used creatinine clearance as a marker for the glomerular filtration rate/acute renal failure since Heemskerk and colleagues (39) also used creatinine while Heneka and colleagues (32) used inulin clearance. In this respect it is interesting to note that markers for acute renal failure that show a faster response, such as serum cystatin C (43), recently became available and these markers could be the better choice in acute models of kidney failure. It is difficult to arrive at a definitive conclusion regarding the safety of hydroxyethyl starches in respect to renal function. As all fluids restored kidney function during endotoxemia to the same extent, and both crystalloid and colloid solutions increased oxygen consumption during resuscitation, no conclusion can be made in a decision between crystalloids and colloids. HES130/0.4 was able, however, to restore renal function with the least increase in the amount of renal work, as indicated by unaffected renal oxygen consumption. Since under conditions of critical oxygen supply to the tissue one strives to minimize oxygen consumption (44-46), the latter could prove clinically beneficial in more severe shock states. Conclusion

The present study for the first time monitored simultaneously and continuously renal cortical and medullary µPO2 and renal oxygen consumption in a rat model of endotoxemia and resuscitation. Our model was associated with an impaired kidney function during endotoxemia. As only minor changes in renal µPO2 and oxygen consumption could be observed during endotoxemia, the loss of kidney function cannot be explained by an oxygen deficiency. Regardless of the type of resuscitation fluid, the renal blood flow and kidney function could be restored to baseline values. Fluid resuscitation was associated by a redistribution of oxygen, which showed different patterns for the different compounds used. HES200/0.5 and Ringer's lactate increased the renal oxygen consumption, in contrast to HES130/0.4. It can therefore be concluded that fluid resuscitation using HES130/0.4 resuscitates the kidney without increasing the renal work load. This could prove clinically significant in conditions where one strives for oxygen consumption as low as possible.

Key messages

• Endotoxemia severely diminished renal function. • Only minor changes in renal µPO 2 and VO2ren could be observed during endotoxemia, and

therefore the loss of kidney function cannot be explained by an oxygen deficiency. • Fluid resuscitation restored RBF and re-established kidney function at the expense of a high

VO2ren. • HES130/0.4 was the only resuscitation fluid tested that did not significantly increase VO2ren. • Fluid resuscitation was accompanied by redistribution between cortical and medullary PO2.

Chapter 2

30

Competing interests

The authors declare that they have no competing interests.

Authors' contributions

TJ made a substantial contribution to the conception and design of the study, performed the animal experiments, was involved in the acquisition, analysis and interpretation of data, and wrote the manuscript. EGM designed and built the technical devices (phosphorimeters), was involved in the interpretation of data and helped to draft the manuscript. BN contributed to the conception and design of study and consulted for clinically relevant importance. NJHR revised the manuscript critically for important intellectual content. KEU gave final approval of the version to be published. CI conceived of the study, participated in coordination and gave final approval of the version to be published. All authors read and approved the final manuscript.

Acknowledgements

Supported in part by a grant of the fortuene-programme to Tanja Johannes (No. 1168-0-0, Medical Faculty, University of Tuebingen, Germany). The authors kindly thank Prof. Dr Hartmut Oßwald, Department of Pharmacology and Toxicology, University of Tuebingen, Germany for his helpful support. Furthermore, the authors would like to thank Anneke Koeman, Department of Physiology of the Academic Medical Center, Amsterdam, The Netherlands, for her help with the animal experiments.

References

1. Uchino S, Kellum JA, Bellomo R, Doig GS, Morimatsu H, Morgera S, Schetz M, Tan I, Bouman C, Macedo E, et al.: Acute renal failure in critically ill patients: a multinational, multicenter study. JAMA 2005, 294:813-818.

2. Brivet FG, Kleinknecht DJ, Loirat P, Landais PJ: Acute renal failure in intensive care units – causes, outcome, and prognostic factors of hospital mortality; a prospective, multicenter study. French Study Group on Acute Renal Failure. Crit Care Med 1996, 24:192-198.

3. Neveu H, Kleinknecht D, Brivet F, Loirat P, Landais P: Prognostic factors in acute renal failure due to sepsis. Results of a prospective multicentre study. The French Study Group on Acute Renal Failure. Nephrol Dial Transplant 1996, 11:293-299.

4. Thijs A, Thijs LG: Pathogenesis of renal failure in sepsis. Kidney Int Suppl 1998, 66:S34-S37. 5. Klenzak J, Himmelfarb J: Sepsis and the kidney. Crit Care Clin 2005, 21:211-222. 6. Wan L, Bellomo R, Di Giantomasso D, Ronco C: The pathogenesis of septic acute renal failure. Curr Opin Crit Care

2003, 9:496-502. 7. Baumgartl H, Leichtweiss HP, Lubbers DW, Weiss C, Huland H: The oxygen supply of the dog kidney:

measurements of intrarenal pO2. Microvasc Res 1972, 4:247-257. 8. Schurek HJ, Jost U, Baumgartl H, Bertram H, Heckmann U: Evidence for a preglomerular oxygen diffusion shunt in

rat renal cortex. Am J Physiol 1990, 259:F910-F915. 9. Brezis M, Rosen S: Hypoxia of the renal medulla – its implications for disease. N Engl J Med 1995, 332:647-655. 10. Gullichsen E: Renal perfusion and metabolism in experimental endotoxin shock. Acta Chir Scand Suppl 1991,

560:7-31. 11. James PE, Bacic G, Grinberg OY, Goda F, Dunn JF, Jackson SK, Swartz HM: Endotoxin-induced changes in

intrarenal pO2, measured by in vivo electron paramagnetic resonance oximetry and magnetic resonance imaging. Free Radic Biol Med 1996, 21:25-34.

12. Linder MM, Hartel W, Alken P, Muschaweck R: Renal tissue oxygen tension during the early phase of canine endotoxin shock. Surg Gynecol Obstet 1974, 138:171-173.

13. Heemskerk AE, Huisman E, van Lambalgen AA, van den Bos GC, Hennekes M, Thijs LG, Tangelder GJ: Renal function and oxygen consumption during bacteraemia and endotoxaemia in rats. Nephrol Dial Transplant 1997, 12:1586-1594.

14. Weber A, Schwieger IM, Poinsot O, Klohn M, Gaumann DM, Morel DR: Sequential changes in renal oxygen consumption and sodium transport during hyperdynamic sepsis in sheep. Am J Physiol 1992, 262:F965-F971.

Fluid Resuscitation

31

15. Dellinger RP, Carlet JM, Masur H, Gerlach H, Calandra T, Cohen J, Gea-Banacloche J, Keh D, Marshall JC, Parker MM, et al.: Surviving Sepsis Campaign guidelines for management of severe sepsis and septic shock. Crit Care Med 2004, 32:858-873.

16. Deman A, Peeters P, Sennesael J: Hydroxyethyl starch does not impair immediate renal function in kidney transplant recipients: a retrospective, multicentre analysis. Nephrol Dial Transplant 1999, 14:1517-1520.

17. Schortgen F, Lacherade JC, Bruneel F, Cattaneo I, Hemery F, Lemaire F, Brochard L: Effects of hydroxyethylstarch and gelatin on renal function in severe sepsis: a multicentre randomized study. Lancet 2001, 357:911-916.

18. Van der Sande FM, Luik AJ, Kooman JP, Verstappen V, Leunissen KM: Effect of intravenous fluids on blood pressure course during hemodialysis in hypotensive-prone patients. J Am Soc Nephrol 2000, 11:550-555.

19. Dehne MG, Muhling J, Sablotzki A, Dehne K, Sucke N, Hempelmann G: Hydroxyethyl starch (HES) does not directly affect renal function in patients with no prior renal impairment. J Clin Anesth 2001, 13:103-111.

20. Suttner S, Boldt J: Volume replacement with hydroxyethyl starch: is there an influence on kidney function? Anasthesiol Intensivmed Notfallmed Schmerzther 2004, 39:71-77.

21. Schramm S, Wettstein R, Wessendorf R, Jakob SM, Banic A, Erni D: Acute normovolemic hemodilution improves oxygenation in ischemic flap tissue. Anesthesiology 2002, 96:1478-1484.

22. van Bommel J, Trouwborst A, Schwarte L, Siegemund M, Ince C, Henny C: Intestinal and cerebral oxygenation during severe isovolemic hemodilution and subsequent hyperoxic ventilation in a pig model. Anesthesiology 2002, 97:660-670.

23. Siegemund M, van Bommel J, Schwarte LA, Studer W, Girard T, Marsch S, Radermacher P, Ince C: Inducible nitric oxide synthase inhibition improves intestinal microcirculatory oxygenation and CO2 balance during endotoxemia in pigs. Intensive Care Med 2005, 31:985-992.

24. Zuurbier CJ, van Iterson M, Ince C: Functional heterogeneity of oxygen supply-consumption ratio in the heart. Cardiovasc Res 1999, 44:488-497.

25. Di Giantomasso D, Morimatsu H, May CN, Bellomo R: Intrarenal blood flow distribution in hyperdynamic septic shock: effect of norepinephrine. Crit Care Med 2003, 31:2509-2513.

26. Johannes T, Mik EG, Ince C: Dual-wavelength phosphorimetry for determination of cortical and subcortical microvascular oxygenation in rat kidney. J Appl Physiol 2006, 100:1301-1310.

27. Raat NJ, Verhoeven AJ, Mik EG, Gouwerok CW, Verhaar R, Goedhart PT, de Korte D, Ince C: The effect of storage time of human red cells on intestinal microcirculatory oxygenation in a rat isovolemic exchange model. Crit Care Med 2005, 33:39-45.

28. Welch WJ, Deng X, Snellen H, Wilcox CS: Validation of miniature ultrasonic transit-time flow probes for measurement of renal blood flow in rats. Am J Physiol 1995, 268:F175-F178.

29. Dunphy I, Vinogradov SA, Wilson DF: Oxyphor R2 and G2: phosphors for measuring oxygen by oxygen-dependent quenching of phosphorescence. Anal Biochem 2002, 310:191-198.

30. Ellis CG, Bateman RM, Sharpe MD, Sibbald WJ, Gill R: Effect of a maldistribution of microvascular blood flow on capillary O(2) extraction in sepsis. Am J Physiol Heart Circ Physiol 2002, 282:H156-H164.

31. Inoue RY, Gontijo JA, Franchini KG: Hemodilution mediates hemodynamic changes during acute expansion in unanesthetized rats. Am J Physiol Regul Integr Comp Physiol 2000, 279:R2243-R2251.

32. Heneka MT, Loschmann PA, Osswald H: Polymerized hemoglobin restores cardiovascular and kidney function in endotoxin- induced shock in the rat. J Clin Invest 1997, 99:47-54.

33. Gullichsen E, Nelimarkka O, Halkola L, Niinikoski J: Renal oxygenation in endotoxin shock in dogs. Crit Care Med 1989, 17:547-550.

34. Levy B, Mansart A, Bollaert PE, Franck P, Mallie JP: Effects of epinephrine and norepinephrine on hemodynamics, oxidative metabolism, and organ energetics in endotoxemic rats. Intensive Care Med 2003, 29:292-300.

35. Esmon CT: Why do animal models (sometimes) fail to mimic human sepsis? Crit Care Med 2004, 32:S219-S222. 36. Cronenwett JL, Lindenauer SM: Distribution of intrarenal blood flow during bacterial sepsis. J Surg Res 1978,

24:132-141. 37. Ravikant T, Lucas CE: Renal blood flow distribution in septic hyperdynamic pigs. J Surg Res 1977, 22:294-298. 38. Lang K, Boldt J, Suttner S, Haisch G: Colloids versus crystalloids and tissue oxygen tension in patients undergoing

major abdominal surgery. Anesth Analg 2001, 93:405-409. 39. Heemskerk AE, Huisman E, van Lambalgen AA, van den Bos GC, Hennekes MW, Thijs LG, Tangelder GJ:

Influence of fluid resuscitation on renal function in bacteremic and endotoxemic rats. J Crit Care 1997, 12:120-131. 40. Bauer C, Kurtz A: Oxygen sensing in the kidney and its relation to erythropoietin production. Annu Rev Physiol

1989, 51:845-856. 41. Epstein FH, Agmon Y, Brezis M: Physiology of renal hypoxia. Ann NY Acad Sci 1994, 718:72-81. 42. Kinter WB, Pappenheimer JR: Role of red blood corpuscles in regulation of renal blood flow and glomerular filtration

rate. Am J Physiol 1956, 185:399-406. 43. Herget-Rosenthal S, Marggraf G, Husing J, Goring F, Pietruck F, Janssen O, Philipp T, Kribben A: Early detection

of acute renal failure by serum cystatin C. Kidney Int 2004, 66:1115-1122.

Chapter 2

32

44. Kress JP, O'Connor MF, Pohlman AS, Olson D, Lavoie A, Toledano A, Hall JB: Sedation of critically ill patients during mechanical ventilation. A comparison of propofol and midazolam. Am J Respir Crit Care Med 1996, 153:1012-1018.

45. Manthous CA, Hall JB, Kushner R, Schmidt GA, Russo G, Wood LD: The effect of mechanical ventilation on oxygen consumption in critically ill patients. Am J Respir Crit Care Med 1995, 151:210-214.

46. Rivers EP, McIntyre L, Morro DC, Rivers KK: Early and innovative interventions for severe sepsis and septic shock: taking advantage of a window of opportunity. CMAJ 2005, 173:1054-1065.

33

CHAPTER NONRESUSCITATED ENDOTOXEMIA INDUCES MICROCIRCULATORY HYPOXIC AREAS IN THE RENAL CORTEX IN THE RAT

Tanja Johannes, Egbert G Mik, and Can Ince Shock, 31(1): 97-103, 2009

Chapter 3

34

Abstract

The pathophysiology of acute renal failure (ARF) in sepsis is only partly understood. In several animal models of septic ARF, no profound tissue hypoxia or decrease in microcirculatory PO2 (µPO2) can be seen. We hypothesized that heterogeneity of microcirculatory oxygen supply to demand in the kidney is obscured when looking at the average µPO2 during endotoxemia. In 20 anesthetized and ventilated rats, MAP, renal blood flow (RBF), and creatinine clearance (CLcrea) were recorded. Renal µPO2 was measured by phosphorescence quenching, allowing measurement of µPO2 distributions. Five animals received a 1-h LPS infusion (10 mg kg-1 h-1). In 5 rats, RBF was mechanically reduced to 2.1 ± 0.2 mL min-1. Five animals served as time control. LPS infusion significantly reduced RBF to 2.1 ± 0.2 mL min-1 and induced anuria. Average cortical µPO2 decreased from 68 ± 4 to 52 ± 6 mmHg, with a significant left shift in the cortical oxygen histogram toward hypoxia. This shift could not be observed in animals receiving mechanical RBF reduction. In these animals, CLcrea was reduced to 50%. An additional group of rats (n = 5) received fluid resuscitation. In these animals, RBF was restored to baseline, CLcrea increased approximately 50%, and the cortical microcirculatory hypoxic areas disappeared after resuscitation. In conclusion, endotoxemia was associated with the occurrence of cortical microcirculatory hypoxic areas that are not detected in the average PO2 measurement, proving the hypothesis of our study. These observations suggest the involvement of hypoxia in the pathogenesis of endotoxemia-induced ARF.

Microcirculatory Hypoxic Areas

35

Introduction

Acute renal failure (ARF) is an often-seen complication in critically ill patients. In sepsis, ARF has a prevalence ranging from 25% in severe sepsis to 50% in septic shock (1) and a mortality up to 75% in patients with acute septic renal failure (2, 3). The pathogenesis of sepsis-induced renal failure is multifactorial. An inappropriate release of various inflammatory mediators and an imbalance between vasoactive substances causing direct cytotoxic effects and impairment of the microvasculature seem to be key factors (4). Septic ARF is characterized by intrarenal vasoconstriction leading to renal hypoperfusion with a redistribution of the cortical blood flow toward the medulla (5-7). However, the relationship between renal oxygen delivery, consumption, and tissue oxygenation, especially with regard to biological response and functional consequences, is still poorly understood. Therefore, the role of microcirculatory dysfunction and subsequent oxygen deficiency in septic renal failure remains controversial (8-10). In a recent study, we only detected a minor influence of endotoxemia on mean microvascular PO2 readings in both cortex and outer medulla (11). Because of the far-reaching consequences, it is necessary to have a more detailed look on intrarenal oxygenation during endotoxemia before drawing the conclusion that endotoxemia-induced ARF is not associated with local hypoxia.

In this study, we hypothesized that heterogeneity of microcirculatory oxygen supply to demand, leading to local microvascular hypoxia, in the kidney is obscured when looking at the average microvascular PO2 during endotoxemia. To investigate this hypothesis, we performed rat experiments in which endotoxemia was induced by intravenous infusion of LPS. Heterogeneity of intrarenal oxygenation was studied by a phosphorescence quenching technique, allowing recovery of microvascular PO2 histograms from cortex and outer medulla (12). The analysis of renal oxygen distribution and the appearance of cortical microcirculatory hypoxic areas during endotoxemia led to a second series of experiments in which a mechanical reduction in renal blood flow (RBF) to values comparable to the flow reduction seen during endotoxemia was induced. This second series of experiments should answer the question if the cortical microcirculatory hypoxic areas were specific for endotoxemia and not simply a nonspecific phenomenon due to reduction of RBF. A third group of fluidresuscitated animals was included to study the effect of fluid resuscitation on the reversibility of microcirculatory hypoxic areas and the possible influence of (relative) hypovolemia.

Material and methods

Animals All experiments in this study were reviewed and approved by the Animal Research Committee of the Academic Medical Center at the University of Amsterdam. Handling and care of the animals were performed in accordance with the guidelines for Institutional and Animal Care and Use Committees. For the experiments, 20 male Wistar rats (Charles River, Maastricht, The Netherlands) with a body weight of 286 ± 18 g were used (mean ± SD).

Surgical preparation Rats were anesthetized by injection of ketamine (90 mg kg-1, i.p.; Nimatek; Eurovet, Bladel, The Netherlands), medetomidine (0.5 mg kg-1, i.p.; Domitor; Pfizer, New York, NY), and atropine sulfate (0.05 mg kg-1, i.p.; Centrafarm, Etten-Leur, The Netherlands). Mechanical ventilation (FIO2, 0.4) was performed via tracheotomy. Four vessels were cannulated with polyethylene catheters (Braun,

Chapter 3

36

Melsungen, Germany; outer diameter, 0.9 mm) for drug and fluid administration. Catheterization of the right carotid artery allowed monitoring of arterial blood pressure and heart rate. The right jugular vein and the right femoral artery and vein were cannulated and used for withdrawal of blood and continuous infusion of Ringer lactate (15 mL kg-1 h-1; Baxter, Uden, The Netherlands). Body temperature was kept at 37 ± 0.5°C, and arterial PCO2 was maintained between 35 and 40 mmHg by adjustment of ventilator settings.

The left kidney was exposed via a flank incision, decapsulated, and immobilized in a Lucite kidney cup (K. Effenberger, Pfaffingen, Germany). Under preservation of the nerves, the renal vessels were carefully separated from each other. A small piece of aluminum foil was placed on the dorsal side of the renal vein to prevent contribution of underlying tissue to the phosphorescence signal (venous PO2 measurement). For continuous measurement of RBF, a perivascular ultrasonic transient time flow probe (type 0.7 RB; Transonic Systems, Inc., Ithaca, NY) was placed around the left renal artery and connected to a flow meter (T206; Transonic Systems Inc.). The left ureter was isolated, ligated, and cannulated with a polyethylene catheter to allow urine collection. The temperature of the kidney surface was measured and kept at approximately 37°C. The experiment was ended by infusion of 1 mL of 3 M potassium chloride inducing sudden cardiac arrest.

Hemodynamics, blood gas measurements, and kidney function MAP (in millimeters of mercury) was continuously measured in the carotid artery and calculated as diastolic pressure + (systolic pressure - diastolic pressure) / 3. The RBF (in milliliters per minute) was measured and recorded continuously. At two different time points, an arterial blood sample was taken from the femoral artery: first time point, 0 min = baseline (t0); second time point, 60 min = endotoxemia or time equivalent (t1). Samples were analyzed for blood gas values (ABL505 blood gas analyzer; Radiometer, Bronshoj, Denmark) for determination of hemoglobin, hematocrit, hemoglobin oxygen saturation, sodium, and potassium concentration (OSM 3; Radiometer). The clearance of creatinine (CLcrea) was assessed as an index for glomerular filtration rate. For analysis of urine volume and creatinine concentration, urine samples were collected at 10-min intervals. Plasma samples were obtained at the midpoint of each 10-min urine collection period and analyzed for creatinine levels. The concentrations of creatinine in urine and plasma were determined by colorimetric methods.

Measurement of renal microvascular oxygenation and renal venous PO2 The applied method of oxygen-dependent quenching of phosphorescence for detection of changes in microvascular PO2 (µPO2) and measurement of renal venous PO2 (PrvO2) is described in detail elsewhere (12, 13). In brief, after intravenous infusion, a water-soluble phosphorescent dye (Oxyphor G2; Oxygen Enterprises, Ltd., Philadelphia, Pa) binds to albumin, forming a stable complex (14-16). This complex is well confined to the circulation and if excited by a flash of light emits phosphorescence with a wavelength approximately 800 nm (17). The phosphorescence decay depends on the surrounding oxygen concentration. The relationship between the measured decay time and the PO2 is given by the Stern-Volmer relation:

+=0

11!! qk [O2]

where τ is the measured decay time, τ0 is the decay time at an oxygen concentration of zero, and kq is the quenching constant. For oxygenation measurements within the rat renal cortex and outer medulla, a dual-wavelength phosphorimeter was used. This allowed the continuous and

Microcirculatory Hypoxic Areas

37

simultaneous measurement of two different depths. The heterogeneity in oxygen pressure was analyzed by fitting a sum of small rectangular distributions to the distributions of quencher concentration in the phosphorescence data, an approach published by Golub et al. (18). Calculation of renal oxygen delivery and oxygen consumption per unit time The renal oxygen delivery was calculated as DO2ren (in milliliters per minute) = RBF x arterial oxygen content (1.31 x Hb x SaO2) + (0.003 x PaO2). The renal oxygen consumption per unit time was calculated as VO2ren (in milliliters per minute per gram) = RBF x arterial - renal venous oxygen content difference. The renal venous oxygen content was calculated as (1.31 x Hb x SrvO2) + (0.003 x PrvO2) (13).

Experimental protocol After 60 min of surgery, two optical fibers for oxygenation measurements were placed both 1 mm above the decapsulated kidney surface and 1 mm above the renal vein. Oxyphor G2 (1.2 mg kg-1 in 15 min; Oxygen Enterprises) was infused intravenously. The measurement of µPO2 and PrvO2 was started 40 min later. The baseline blood sample was then taken. At this time point, the rats were randomized between the LPS (LPS; n = 5) and time control groups (TC; n = 5).

In total, 5 animals received a 1-h infusion of LPS (10 mg kg-1; serotype 0127:B8; Sigma, Zwijndrecht, The Netherlands) to induce endotoxemia. Five rats served as time controls. A second blood gas was taken 60 min after start of LPS infusion and analyzed as described before. In a second series of experiments, a third group of rats was prepared for a mechanical reduction in RBF (FR; n = 5). In detail, in these animals, a thin silastic catheter was placed loosely around the renal artery at its junction with the abdominal aorta. The movement of a small plastic ring surrounding the catheter loop allowed for controlled partial occlusion of the vessel and an RBF reduction to 2 mL min-

1. The experiments were ended 60 min after start of LPS infusion or a corresponding time point by intravenous bolus injection of 3 M KCl.

An additional group of animals (n = 5) received fluid resuscitation 1 h after LPS infusion (10 mg kg-1). For resuscitation, HES (130 kd; Voluven; 6% HES 130/0.4; Fresenius Kabi, Utrecht, The Netherlands) was infused at a rate of 20 mL h-1 until a 5-min steady plateau in RBF was reached. In this group, three measurement points were defined: first time point, 0 min = baseline (t0); second time point, 60 min = endotoxemia (t1); and 90 min = after 30 min of fluid resuscitation (t2).

Statistical analysis Values are reported as mean ± SD unless otherwise indicated. Analysis software for the monoexponential fit procedures of the phosphorescence curves was written in Labview 6.1 software (National Instruments, Austin, Tex). Statistics and the PO2 histogram recovery were performed using GraphPad Prism version 4.0 for Windows (GraphPad Software, San Diego, Calif). For testing differences within groups and intergroup differences, two-way ANOVA for repeated measurements with Bonferroni post test was performed. The oxygen histograms were analyzed using two-way ANOVA as previously described. Every bin of the histogram was separately tested in time and versus control. P values less than 0.01 were considered significant.

Chapter 3

38

Results

Systemic hemodynamics No differences in systemic hemodynamics existed between the three groups at baseline (Table 1). Endotoxemia induced a significant decrease in MAP from 116 ± 7 mmHg at baseline (t0) to 97 ± 17 mmHg at t1. Furthermore, RBF dropped significantly from 5.8 ± 0.8 to 2.1 ± 0.2 mL min-1 at t1 (1 h after start of LPS infusion; P < 0.01 vs. baseline). A sample experiment showing the response of MAP and RBF during endotoxemia over time is shown in Figure 1A.

In the mechanical blood flow reduction group, RBF was reduced from 6.1 ± 0.7 to 2.1 ± 0.2 mL min-1 without affecting MAP (example shown in Fig. 3A). In the control group, MAP and RBF were stable over time.

MAP

(mmHg)

RBF

(ml/min)

c!PO2

(mmHg)

m!PO2

(mmHg)

DO2ren

(ml/min)

VO2ren

(ml/min/g)

CLcrea

(!l/min/g)

TC (n = 5)

t0 114 ± 3 6.5 ± 0.7 67 ± 3 52 ± 3 1.5 ± 0.2 0.12 ± 0.03 984 ± 199

t1 121 ± 3 5.8 ± 0.7 61 ± 9 48 ± 3 1.3 ± 0.2 0.17 ± 0.05 1169 ± 346

LPS (n = 5)

t0 116 ± 7 5.8 ± 0.8 68 ± 4 55 ± 3 1.4 ± 0.2 0.10 ± 0.02 891 ± 119

t1 97 ± 17† 2.1 ± 0.2*† 52 ± 6*" 45 ± 2* 0.5 ± 0.1*† 0.07 ± 0.05† anuric

FR (n = 5)

t0 111 ± 5 6.1 ± 0.7 67 ± 7 49 ± 7 1.4 ± 0.2 0.20 ± 0.08 755 ± 134

t1 106 ± 17 2.1 ± 0.2*† 63 ± 4 46 ± 3 0.4 ± 0.1*† 0.07 ± 0.04* 427 ± 206†

Values represent mean ± SD. * P < 0.01 vs baseline; † P < 0.01 vs control; " P < 0.01 vs flow reduction. TC = control group; LPS = group receiving LPS; FR = flow reduction group. t0 = baseline; t1= 1h after LPS-infusion or equivalent time. MAP = mean arterial pressure; RBF = renal blood flow; c!PO2 = cortical microvascular PO2; m!PO2 = medullary microvascular PO2; DO2ren = renal O2-delivery; VO2ren = renal O2-consumption; CLcrea = creatinine clearance.

TABLE 1. Systemic and regional variables

Renal oxygenation parameters Data of the oxygenation parameters of the kidney are shown in Table 1. The baseline values in the experimental groups and the control group were not different. Renal oxygen delivery, VO2ren, and µPO2 did not change in the control group. At the end of the 60-min infusion of LPS, there was a mild, but significant, reduction in cortical and medullary µPO2 of 16 and 10 mmHg, respectively. An example of the behavior of µPO2 during endotoxemia is shown in Figure 1B. Despite the only mild reduction of 16 mmHg in the mean cortical microvascular PO2, 1 h after LPS infusion, the cortical oxygen histogram was significantly (P < 0.01) shifted to the left compared with baseline and control (Fig. 2). This indicates the presence of microvascular hypoxic areas in which the detected PO2 values were less than 10 mmHg. The occurrence of microvascular hypoxic areas is accompanied by a significant reduction in microvascular regions with high PO2, denoted by bins 7 and 8 of the oxygen histogram (97.5 and 112.5 mmHg, respectively). In contrast, the medullary oxygen histogram did not significantly change during endotoxemia. The distributions in the control group were stable over time.

Microcirculatory Hypoxic Areas

39

Renal oxygen delivery decreased from 1.4 ± 0.2 at baseline to 0.5 ± 0.1 at t1 (P < 0.01 vs. baseline). VO2ren did not change compared with baseline but significantly decreased compared with the control group.

In the mechanical flow reduction group, the renal cortical and medullary µPO2 were not affected and remained high with 63 ± 4 and 46 ± 3 mmHg, respectively (example shown in Fig. 3B). The oxygen distributions before and after RBF reduction are demonstrated in Figure 4. After flow reduction to 2 mL min-1, there was no significant change in the cortical or medullary µPO2 or in the oxygen distribution of the two investigated kidney regions. Renal oxygen delivery significantly decreased from 1.4 ± 0.2 at t0 to 0.4 ± 0.1 upon flow reduction (P < 0.01 vs. baseline). Furthermore, there was a significant drop in VO2ren from 0.20 ± 0.08 at t0 to 0.07 ± 0.04 mL min-1 g-1 at t1 compared with baseline.

FIGURE 1. Sample experiment for endotoxemia. MAP and RBF progressively dropped 10 min after start of LPS infusion. Whereas MAP slowly recovered after 10 min, RBF remained low approximately 2 mL min-1. The renal cortical (cµPO2) and outer medullary (mµPO2) microvascular PO2 only slightly dropped during LPS infusion. t0 = baseline; t1 = 1 h after start of LPS infusion.

Chapter 3

40

FIGURE 2. Histograms showing oxygen distributions in the cortex (A) and the outer medulla (B) of the rat kidney for time control and for endotoxemia. t0 = baseline; t1 = 1 h after start of LPS infusion. *P < 0.01 versus baseline; †P < 0.01 vs. control; ‡P < 0.01 vs flow reduction.

FIGURE 3. Sample experiment for RBF reduction. MAP transiently increased after reducing the RBF to values approximately 2 mL min-1. The intervention had no influence on renal cortical (cµPO2) and outer medullary (mµPO2) microvascular PO2. t0 indicates baseline; t1, 1 h after start of LPS infusion.

Microcirculatory Hypoxic Areas

41

FIGURE 4. Histograms showing oxygen distributions in the cortex (A) and the outer medulla (B) of the rat kidney during reduction in RBF (FR) by ligation of the renal artery. t0 indicates baseline; t1, 1 h after start of blood flow reduction. *P < 0.01 vs. control. Kidney function Creatinine clearance was stable during the experimental period in the control group. In the endotoxemia group, CLcrea was 891 ± 119 µL min-1 g-1 at baseline and became zero due to the occurrence of anuria at t1 (Table 1).

In the flow reduction group, CLcrea was 50% reduced compared with baseline (P < 0.01 vs. baseline).

Fluid resuscitation In an additional group of endotoxemic rats, fluid resuscitation was given to prove the reversibility of microcirculatory hypoxic areas and to exclude hypovolemia as a cause of this phenomenon (Fig. 5). Endotoxemia was accompanied by a slight reduction in MAP and a significant drop in RBF of 62% at t1 (P < 0.01 vs. baseline). Fluid resuscitation had no effect on MAP, but restored RBF to baseline, and animals started to urinate after anuria. Creatinine clearance increased approximately 50%, which was significantly lower than at baseline. Cortical and medullary µPO2 dropped 12 and 9 mmHg, respectively, at t1 and did not significantly change upon fluid resuscitation.

Chapter 3

42

FIGURE 5. Fluid resuscitation. A, Influence of fluid resuscitation on systemic and regional variables. t0 indicates baseline; t1, 1 h after start of LPS infusion; t2, after 30 min of fluid resuscitation. Histograms showing oxygen distributions in the cortex (B) and the outer medulla (C) of the rat kidney during fluid resuscitation. *P < 0.01 vs. baseline (ANOVA for repeated measurements with Newman-Keuls posttest).

The cortical oxygen histogram, however, was significantly (P < 0.01) shifted to the left compared with baseline (Fig. 5B) at t1. This left shift was reversed by fluid resuscitation. In contrast, the medullary oxygen histogram did not significantly change during endotoxemia and resuscitation. Renal oxygen delivery significantly decreased approximately 62% at t1 and increased 50% upon fluid resuscitation (P < 0.01 vs. baseline). VO2ren was unaffected by endotoxemia and resuscitation. Discussion

In a rat model, we studied the effect of endotoxemia on regional microvascular oxygenation of the kidney. Local µPO2 is depending on both VO2 and DO2, and changes in µPO2 therefore denote changes in their balance. Classically, VO2 becomes only dependent on available oxygen at very low PO2 levels (several millimeters of mercury) (19, 20). Previous studies on renal oxygenation in sepsis reported PO2 values of several tens of millimeters of mercury (6, 21), arguing against tissue hypoxia to a level that affects cellular metabolism and induces oxidative stress. However, insight in the heterogeneity of PO2 is lacking, and the occurrence of profound local hypoxia cannot be excluded. A more detailed look into the heterogeneity of microvascular/tissue PO2 would be appropriate before

Microcirculatory Hypoxic Areas

43

rejecting hypoxia as a possible contributing factor in the pathogenesis of sepsis-induced ARF. We tested the hypothesis that the occurrence of local microvascular hypoxia due to heterogeneity in microcirculatory oxygen supply to demand in the kidney is obscured when looking at the average microvascular PO2.

In this study, endotoxemia induced a significant reduction in both the average cortical and outer medullary µPO 2, obtained by monoexponential analysis of the phosphorescencesignal. In accordance with our previous report (11), both average renal cortical and medullary PO2 during endotoxemia are well more than 40 mmHg, not at all suggestive of hypoxia-induced metabolic impairment. A more sophisticated analysis of the phosphorescence signals, allowing the recovery of heterogeneity in µPO2 values, shows the occurrence of regions with profound hypoxia in the cortex (approximately 10% of the signal originated from regions with µPO2 values less than 15 mmHg). In contrast, the µPO2 distribution in the outer medulla did not show a significant left shift during endotoxemia.

LPS infusion significantly decreased RBF and DO2ren, reduced MAP by approximately 9 mmHg from baseline, and was accompanied by anuria. In contrast, mechanical RBF reduction, with a reduction in DO2ren comparable to the effect of LPS, did not induce hypoxic areas in neither cortex nor outer medulla and was not associated with anuria (CLcrea did decrease compared with baseline, but not significantly). Mechanical flow reduction is known to significantly alter renal perfusion pressure (22, 23), and in view of the modest effect of 50% reduction of RBF, a small reduction in MAP is not likely to hamper kidney function. These findings argue against prerenal causes being the predominant factor in the appearance of microcirculatory hypoxic areas and the impairment in kidney function (anuria). However, because prerenal acute kidney injury is always associated with changes in intrarenal hemodynamics, the strict separation between prerenal and/or intrarenal causes from the hemodynamic point of view is difficult. Concerning intrarenal changes, a distorted relationship between oxygen supply and demand on a local level due to endotoxin, probably due to increased intrarenal vasoconstriction, can explain our findings. In this view, the lack of large PO2 changes in the mechanical flow reduction group can be explained by concomitant falls in DO2 and VO2 (reduced filtration and resorption) in the presence of a good functioning microcirculation.

Although macrohemodynamic alterations do not seem to be the primary cause of anuria and the occurrence of hypoxic areas during endotoxemia, hypovolemia might be a contributing factor. For this reason, and to investigate whether the hypoxic areas were reversible, we studied the effects of fluid resuscitation. In comparison, and in reference to our previous studies, RBF increased upon resuscitation to baseline value (11). Fluid resuscitation restored kidney function to 50% of baseline and was accompanied by reduction of hypoxic areas in the cortex. These findings suggest a correlation between the occurrence of hypoxic areas and functional impairment of the kidney. Although fluid resuscitation had no effect on MAP, arguing against severe hypovolemia, the partial recovery of renal function does suggest some prerenal, that is, hemodynamically mediated component.

Although reports on heterogeneity in microvascular/tissue PO2 in sepsis are lacking, more data exist on heterogeneity in microvascular blood flow due to the availability of several measurement techniques (24, 25). In various pathologies, a maldistribution in microvascular blood flow was demonstrated in tissues other than kidney (26, 27). Several animal studies reported an intrarenal redistribution of blood flow during endotoxemia (28, 29). The distribution of blood flow from cortex toward the medulla is suggesting that endotoxemia causes an increase in arteriovenous shunting,

Chapter 3

44

resulting in a fall in tissue perfusion. Those effects add to preglomerular arteriovenous diffusive oxygen shunting, described by Schurek et al. (30) under physiological conditions. Shunting in combination with locally reduced blood flow (21) (theoretically) makes the cortex prone to hypoxia during endotoxemia.

When interpreting our findings, the reader should keep in mind some limitations of our animal model. Short-term models of rodent endotoxemia, like in our study, are accompanied by hypodynamic state and marked reduction in RBF with extensive intrarenal vasoconstriction (31). In contrast, studies in large animal models show a hyperdynamic response to endotoxemia without reduction in RBF (32, 33). For this reason, the results of animal studies have to be interpreted carefully, and still, a direct clinical relevance remains equivocal. Furthermore, due to the lack of a steady state of creatinine balance, changes in CLcrea as an index of glomerular filtration should only be regarded as a gross indication.

Overall, we demonstrate that endotoxemia is associated with the occurrence of microcirculatory hypoxic areas that are not detected using techniques that provide an average PO2 value, proving the hypothesis of our study. Our data are suggesting a role for regional hypoxia in the development of endotoxemia-induced renal dysfunction. However, a causal relation between local hypoxia and loss of kidney function remains to be proven. Nevertheless, our findings are significant because previous studies (5, 6, 11) failed to demonstrate profound tissue hypoxia to levels that might impair mitochondrial respiration. Because the used phosphorescence technique allows detection of local hypoxia but is not able to identify the anatomical location, we can only speculate regarding which part of the cortex might be affected. The peritubular capillaries are identified as being prone to damage and dysfunction during endotoxemia (34-36) and seem to be a good candidate for further research. Furthermore, the question remains if 20% of the PO2 values less than 22.5 mmHg and 10% of the values less than 7.5 mmHg are sufficient to induce anuria. A possible explanation can be the recently described phenomenon of “oxygen conformance of cellular metabolism” (37-39). If already mild hypoxia reduces cellular oxygen consumption (as opposed to the classical view [19, 20]), then our PO2 histograms show indeed that a large portion of the renal cortex might be in an oxygen-dependent state.

We conclude that hypoxia should not be discarded as a contributing cause in the pathogenesis of endotoxemia induced ARF, and further studies into the cause of hypoxia, its localization, and its role in the development of sepsis-induced ARF are indicated.

Microcirculatory Hypoxic Areas

45

References 1. Uchino S, Kellum JA, Bellomo R, Doig GS, Morimatsu H, Morgera S, Schetz M, Tan I, Bouman C, Macedo E, et al.:

Acute renal failure in critically ill patients: a multinational, multicenter study. JAMA 294:813-818, 2005. 2. Neveu H, Kleinknecht D, Brivet F, Loirat P, Landais P: Prognostic factors in acute renal failure due to sepsis.

Results of a prospective multicentre study. The French Study Group on Acute Renal Failure. Nephrol Dial Transplant 11:293-299, 1996.

3. Brivet FG, Kleinknecht DJ, Loirat P, Landais PJ: Acute renal failure in intensive care unitsVcauses, outcome, and prognostic factors of hospital mortality; a prospective, multicenter study. French Study Group on Acute Renal Failure. Crit Care Med 24:192-198, 1996.

4. Wan L, Bellomo R, Di Giantomasso D, Ronco C: The pathogenesis of septic acute renal failure. Curr Opin Crit Care 9:496-502, 2003.

5. Gullichsen E, Nelimarkka O, Halkola L, Niinikoski J: Renal oxygenation in endotoxin shock in dogs. Crit Care Med 17:547-550, 1989.

6. James PE, Bacic G, Grinberg OY, Goda F, Dunn JF, Jackson SK, Swartz HM: Endotoxin-induced changes in intrarenal PO2, measured by in vivo electron paramagnetic resonance oximetry and magnetic resonance imaging. Free Radic Biol Med 21:25-34, 1996.

7. Linder MM, Hartel W, Alken P, Muschaweck R: Renal tissue oxygen tension during the early phase of canine endotoxin shock. Surg Gynecol Obstet 138:171-173, 1974.

8. Gullichsen E: Renal perfusion and metabolism in experimental endotoxin shock. Acta Chir Scand Suppl 560:7-31, 1991.

9. Heemskerk AE, Huisman E, van Lambalgen AA, van den Bos GC, Hennekes M, Thijs LG, Tangelder GJ: Renal function and oxygen consumption during bacteraemia and endotoxaemia in rats. Nephrol Dial Transplant 12:1586-1594, 1997.

10. Weber A, Schwieger IM, Poinsot O, Klohn M, Gaumann DM, Morel DR: Sequential changes in renal oxygen consumption and sodium transport during hyperdynamic sepsis in sheep. Am J Physiol 262:F965-F971, 1992.

11. Johannes T, Mik EG, Nohe B, Raat NJ, Unertl KE, Ince C: Influence of fluid resuscitation on renal microvascular PO2 in a normotensive rat model of endotoxemia. Crit Care 10:R88, 2006.

12. Johannes T, Mik EG, Ince C: Dual-wavelength phosphorimetry for determination of cortical and subcortical microvascular oxygenation in rat kidney. J Appl Physiol 100:1301-1310, 2006.

13. Mik EG, Johannes T, Ince C: Monitoring of renal venous PO2 and kidney oxygen consumption in rats by a near-infrared phosphorescence lifetime technique. Am J Physiol Renal Physiol 294:F676-F681, 2008.

14. Rietveld IB, Kim E, Vinogradov SA: Dendrimers with tetrabenzoporphyrin cores: near infrared phosphors for in vivo oxygen imaging. Tetrahedron 59: 3821-3831, 2003.

15. Rozhkov V, Wilson D, Vinogradov S: Phosphorescent Pd porphyrin-dendrimers: tuning core accessibility by varying the hydrophobicity of the dendritic matrix. Macromolecules 35:1991-1993, 2002.

16. Vinogradov SA, Lo LW, Wilson DF: Dendritic polyglutamic porphyrins: probing porphyrin protection by oxygen-dependent quenching of phosphorescence. Chem Eur J 5:1338-1347, 1999.

17. Dunphy I, Vinogradov SA, Wilson DF: Oxyphor R2 and G2: phosphors for measuring oxygen by oxygen-dependent quenching of phosphorescence. Anal Biochem 310:191-198, 2002.

18. Golub AS, Popel AS, Zheng L, Pittman RN: Analysis of phosphorescence in heterogeneous systems using distributions of quencher concentration. Biophys J 73:452-465, 1997.

19. Jones DP, Mason HS: Gradients of O2 concentration in hepatocytes. J Biol Chem 253:4874-4880, 1978. 20. Wilson DF, Erecinska M, Drown C, Silver IA: The oxygen dependence of cellular energy metabolism. Arch Biochem

Biophys 195:485-493, 1979. 21. Nitescu N, Grimberg E, Guron G: Low-dose candesartan improves renal blood flow and kidney oxygen tension in

rats with endotoxin-induced acute kidney dysfunction. Shock 30:166-172, 2008. 22. Moosavi SM, Johns EJ: Effect of renal perfusion pressure on renal function, renin release and renin and

angiotensinogen gene expression in rats. J Physiol 520(Pt 1):261-269, 1999. 23. Blaine EH, Zimmerman MB: Renin secretion during dynamic changes in renal perfusion pressure. Am J Physiol

236:F546-F551, 1979. 24. Millar CG, Thiemermann C: Intrarenal haemodynamics and renal dysfunction in endotoxaemia: effects of nitric oxide

synthase inhibition. Br J Pharmacol 121:1824-1830, 1997. 25. Di Giantomasso D, Morimatsu H, May CN, Bellomo R: Intrarenal blood flow distribution in hyperdynamic septic

shock: effect of norepinephrine. Crit Care Med 31:2509-2513, 2003. 26. Lamm WJ, Bernard SL, Wagner WW Jr, Glenny RW: Intravital microscopic observations of 15-microm microspheres

lodging in the pulmonary microcirculation. J Appl Physiol 98:2242-2248, 2005. 27. Ellis CG, Bateman RM, Sharpe MD, Sibbald WJ, Gill R: Effect of a maldistribution of microvascular blood flow on

capillary O(2) extraction in sepsis. Am J Physiol Heart Circ Physiol 282:H156-H164, 2002.

Chapter 3

46

28. Millar CG, Thiemermann C: Carboxy-PTIO, a scavenger of nitric oxide, selectively inhibits the increase in medullary perfusion and improves renal function in endotoxemia. Shock 18:64-68, 2002.

29. Cronenwett JL, Lindenauer SM: Distribution of intrarenal blood flow during bacterial sepsis. J Surg Res 24:132-141, 1978.

30. Schurek HJ, Jost U, Baumgartl H, Bertram H, Heckmann U: Evidence for a preglomerular oxygen diffusion shunt in rat renal cortex. Am J Physiol 259: F910-F915, 1990.

31. Heyman SN, Darmon D, Goldfarb M, Bitz H, Shina A, Rosen S, Brezis M: Endotoxin-induced renal failure. I. A role for altered renal microcirculation. Exp Nephrol 8:266-274, 2000.

32. Langenberg C, Wan L, Egi M, May CN, Bellomo R: Renal blood flow and function during recovery from experimental septic acute kidney injury. Intensive Care Med 33:1614-1618, 2007.

33. Offner PJ, Robertson FM, Pruitt BA Jr: Effects of nitric oxide synthase inhibition on regional blood flow in a porcine model of endotoxic shock. J Trauma 39:338-343, 1995.

34. Richman AV, Gerber LI, Balis JU: Peritubular capillaries. A major target siteof endotoxin-induced vascular injury in the primate kidney. Lab Invest 43:327-332, 1980.

35. Achparaki A, Kotzampassi K, Eleftheriadis E, Foroglou C: Ultrastructural changes of the renal cortex after septic shock in rats. Histol Histopathol 3: 133-146, 1988.

36. Wu L, Tiwari MM, Messer KJ, Holthoff JH, Gokden N, Brock RW, Mayeux PR: Peritubular capillary dysfunction and renal tubular epithelial cell stress following lipopolysaccharide administration in mice. Am J Physiol Renal Physiol 292:F261-F268, 2007.

37. Chandel NS, Budinger GR, Choe SH, Schumacker PT: Cellular respiration during hypoxia. Role of cytochrome oxidase as the oxygen sensor in hepatocytes. J Biol Chem 272:18808-18816, 1997.

38. Schumacker PT, Chandel N, Agusti AG: Oxygen conformance of cellular respiration in hepatocytes. Am J Physiol 265:L395-L402, 1993.

39. Budinger GR, Chandel N, Shao ZH, Li CQ, Melmed A, Becker LB, Schumacker PT: Cellular energy utilization and supply during hypoxia in embryonic cardiac myocytes. Am J Physiol 270:L44-L53, 1996.

47

CHAPTER LOW-DOSE DEXAMETHASONE-SUPPLEMENTED FLUID RESUSCITATION REVERSES ENDOTOXIN-INDUCED ACUTE RENAL FAILURE AND PREVENTS CORTICAL MICROVASCULAR HYPOXIA

Tanja Johannes, Egbert G Mik, Karin Klingel, Hans-Jürgen Dieterich, Klaus E Unertl, and Can Ince Shock, 31(5): 521-528, 2009

Chapter 4

48

Abstract

There is growing evidence that impairment in intrarenal oxygenation and hypoxic injury might contribute to the pathogenesis of septic renal failure. An important molecule known to act on the renal microvascular tone and therefore consequently being involved in the regulation of intrarenal oxygen supply is NO. The main production of NO under septic conditions derives from iNOS, an enzyme that can be blocked by dexamethasone (DEX). In an animal model of endotoxininduced renal failure, we tested the hypothesis that inhibition of iNOS by low-dose DEX would improve an impaired intrarenal oxygenation and kidney function. Twenty-two male Wistar rats received a 30-min intravenous infusion of LPS (2.5 mg/kg) and consecutively developed endotoxemic shock. Two hours later, in 12 animals, fluid resuscitation was initiated. Six rats did not receive resuscitation; four animals served as time control. In addition to the fluid, six animals received a bolus of low-dose DEX (0.1 mg/kg). In these animals, the renal iNOS mRNA expression was significantly suppressed 3 h later. Dexamethasone prevented the appearance of cortical microcirculatory hypoxic areas, improved renal oxygen delivery, and significantly restored oxygen consumption. Besides a significant increase in MAP and renal blood flow, DEX restored kidney function and tubular sodium reabsorption to baseline values. In conclusion, treatment with low-dose DEX in addition to fluid resuscitation reversed endotoxin-induced renal failure associated by an improvement in intrarenal microvascular oxygenation. Therefore, low-dose DEX might have potential application in the prevention of septic acute renal failure.

Dexamethasone

49

Introduction

Sepsis is an important risk factor for the development of acute renal failure (ARF) in critically ill patients (1). The mortality rate in patients with septic shock and severe ARF remains high (2). Besides the fact that the pathogenesis of septic ARF is still not completely understood, there is furthermore a lack in specific therapies. There is growing evidence that the renal microcirculation is markedly affected under septic condition (3). For this reason, not only global renal blood flow (RBF) is impaired (4). An inappropriate release of vasoactive mediators leads to intrarenal maldistribution of blood flow, resulting in an inadequate oxygen supply to renal tissue (5, 6). Particularly, the role of hypoxic injury in the development of ARF is gaining more and more interest based on the findings that intrarenal shunt diffusion of oxygen and heterogeneous consumption can cause marked oxygen gradients even under physiological conditions (7).

An important molecule in the regulation of renal microvascular tone is NO (8). Its physiological production via induction of NO-synthase is disturbed under septic conditions, and an inappropriate high release of NO is the consequence (9). This overproduction of NO is mainly derived from iNOS, leading to peroxynitrite-related tubular injury, downregulation of renal endothelial NO-synthase (10), and massive systemic vasodilation, a symptom difficult to counteract by vasoconstrictors (11).

Glucocorticosteroids can enhance the effects of vasopressors in sepsis. This effect is due to an improved vascular reactivity to catecholamines (12). Glucocorticosteroids have been advocated for decades as an approach in the therapy of septic shock (13), in particular with the idea to correct for a sepsis-related relative adrenal insufficiency (14). However, multicenter clinical trials using high-dose treatment with glucocorticosteroids have failed to demonstrate beneficial effects (13, 15). There is evidence that glucocorticosteroid therapy in “physiologic dose” (low-dose) in sepsis improved survival (16, 17). Dexamethasone (DEX) has been shown not only to sensitize blood vessels to catecholamines (12), but also to inhibit iNOS production by repressing gene expression (18). There are a few studies that could demonstrate beneficial effects of DEX on circulatory and renal dysfunction during endotoxemia (19, 20). In particular, Tsao et al. (19) could show that low-dose DEX attenuated renal dysfunction in conscious rats, when administered simultaneously with LPS.

However, these studies used different doses of DEX, and their relevance to clinical practice was uncertain because essential information concerning the state of shock, adequate resuscitation procedures, direct measurement of kidney function, and comparison of early versus delayed treatment has been lacking. There is evidence that the improvement of renal dysfunction by low-dose DEX was related to iNOS inhibition (19). This conclusion was supported by the findings that selective iNOS inhibition was associated with improved renal function in septic models (21, 22). However, the mechanism by which DEX-induced iNOS inhibition is improving renal function is still unknown.

The objectives of the present study were to investigate in a rat model of endotoxin-induced shock associated with ARF, first, the effects of standard fluid resuscitation plus additional application of low-dose DEX on systemic and renal hemodynamics; second, to prove whether low-dose DEX had beneficial effects on renal oxygenation; third, to demonstrate that low-dose DEX would improve renal function; and finally, to evaluate, if DEX-induced suppressive effects on renal iNOS expression are related to changes in intrarenal oxygenation.

Chapter 4

50

Material and methods

Animals The Animal Research Committee of the Academic Medical Center at the University of Amsterdam approved the experiments of this study. In accordance with the guidelines for Institutional and Animal Care and Use Committees, care and handling of the animals were performed. In total, 22 Wistar male rats (Harlan, Horst, the Netherlands), with a mean body weight of 267 g (SD, 54 g), were used for the experiments.

Surgical preparation Rats were anesthetized by i.p. injection of a mixture of 90 mg/kg ketamine, 0.5 mg/kg medetomidine, and 0.05 mg/kg atropine-sulfate. Animals were mechanically ventilated (fraction of inspired oxygen, 0.4), and ventilation was adjusted to keep the arterial PCO2 between 35 and 40 mmHg. For drug and fluid administration, four vessels were cannulated with polyethylene catheters (PE 50). A catheter in the right carotid artery was used for monitoring of arterial blood pressure and heart rate (HR) and connected to a pressure transducer. To allow continuous central venous pressure (CVP) measurement, the right jugular vein was cannulated, and the catheter tip was inserted to the level of the right atrium. Two other catheters were placed in the right femoral artery and vein and used for withdrawal of blood and continuous infusion of 15 mL kg-1 h-1 Ringer’s lactate (Baxter, Utrecht, The Netherlands) and infusion of 50 mg kg-1 h-1 ketamine (Eurovet, Bladel, The Netherlands). The rat’s body temperature was maintained around 37°C ± 0.5°C. The left kidney was exposed via flank incision, decapsulated, and immobilized in a Lucite kidney cup (K. Effenberger, Pfeffingen, Germany). A perivascular ultrasonic transient time flow probe (type 0.7 RB; Transonic Systems Inc, Ithaca, NY) was placed around the left renal artery and connected to a flowmeter (T206; Transonic Systems Inc). For urine collection, the left ureter was isolated, ligated, and cannulated. The operation field was covered with Saran wrap throughout the entire experiment to prevent evaporation of body fluids. At the end of the experiment, the left kidney was removed, weighed, and fixed in paraformaldehyde. The experiment was ended by 1-mL infusion of 3 M potassium chloride, inducing sudden cardiac arrest in the animal. Finally, the correct placement of the catheters was checked postmortem.

Hemodynamic and blood gas measurements MAP, HR, CVP, and RBF were continuously measured and recorded. An arterial blood sample (0.4 mL) was taken from the femoral artery at three different time points: first time point: 0 min = baseline (t0), second time point: 2 h = endotoxemia (t1), and third time point: ~5 h = resuscitation (t2). The same volume of hydroxyethyl-starch (Voluven, 6% HES 130/0.4; Fresenius Kabi, Utrecht, the Netherlands) replaced the sampling volume of blood. The samples were analyzed for blood gas values, blood pH, ionized sodium, and potassium using an automatic blood gas analyzer (ABL505 blood gas analyzer; Radiometer, Copenhagen, Denmark). Hematocrit, hemoglobin (Hb) concentration, and Hb oxygen saturation were measured using a hemoxymeter (OSM3; Radiometer). Furthermore, samples were analyzed for plasma creatinine concentration and nitrate/nitrite/S-nitrosothiols (NOx) plasma levels.

Dexamethasone

51

Measurement of renal microvascular oxygenation and renal venous PO2 Measurements of microvascular PO2 (µPO2) within the kidney cortex and outer medulla were performed by oxygen-dependent quenching of phosphorescence using a dual-wavelength phosphorimeter (23). The renal venous PO2 (rvPO2) was detected by the same method (24). Briefly described, a water-soluble phosphorescent dye (Oxyphor G2; Oxygen Enterprises Ltd, Philadelphia, Pa), intravenously infused, binds to albumin (25). The phosphoralbumin complex emits phosphorescence with a wavelength around 800 nm if excited by a flash of light. Dependent on the oxygen concentration, the phosphorescence intensity decreases, and by Stern-Volmer relation, the relationship between the measured decay time and the PO2 can be calculated (26). The intrarenal heterogeneity in µPO2 was analyzed by fitting a sum of small rectangular distributions to the distributions of quencher concentration in the phosphorescence data (27).

Calculation of renal oxygen delivery, consumption and extraction, and renal vascular resistance Renal oxygen delivery (DO2ren) was calculated as: RBF x arterial O2 content (1.31 x Hb x SaO2) + (0.003 x PaO2). Renal oxygen consumption (VO2ren) was calculated as: RBF x (arterial - renal venous O2 - content difference). Renal venous O2 content was calculated as (1.31 x Hb x SrvO2) + (0.003 x rvPO2) (24). An estimation of the renal vascular resistance was made by: RVR = MAP / RBF.

Assessment of kidney function The clearance of creatinine (CrCl) was assessed as an index of glomerular filtration rate. Calculations of the clearance were done with standard formula: clearance [in milliliter per minute] = (U x V) / P, where U is the concentration of creatinine in urine, V is the urine volume per unit time, and P is the concentration of creatinine in plasma. The specific elimination capacity for creatinine of the left kidney was normalized to the organ weight. Urine samples from the left ureter were collected at 10-min intervals for analysis of urine volume, urine sodium, and creatinine concentration. Urine sodium concentration was determined by flame photometry, and creatinine concentration (plasma and urine) was measured by the enzymic-Jaffe method. At the midpoint of each 10-min interval, plasma samples for analysis of creatinine were obtained. To calculate the amount of sodium reabsorbed (TNa+; in millimoles per minute), the CrCl was multiplied by the plasma sodium concentration and subtracted by the sodium excretion per unit time.

Plasma NOx measurement For measurement of NOx, 50-µL plasma samples were centrifuged at 20,200g for 10 min and immediately frozen at -80°C. NOx was measured using an NO analyzer (GE Analytical Instruments Inc, Boulder, Colo) (28). Briefly described, NOx is measured using vanadium (III) chloride (Sigma-Aldrich, Zwijndrecht, the Netherlands) at 90°C. Vanadium (III) chloride reduces nitrate/nitrite to NO gas, which is measured as the luminescence of the ozone NO reaction in a vacuum chamber. A calibration curve of sodium nitrate was constructed using the mean of 10 measured values of each calibration point. The raw data were analyzed using integration method for estimating the area under the curve.

Chapter 4

52

iNOS mRNA expression At the end of the experiment, the kidney was removed, weighed, and immediately fixed in 4% paraformaldehyde overnight. In n = 3 kidneys per group, iNOS mRNA expression was determined by in situ hybridization (ISH). 35S-labeled antisense RNA probes for detection of iNOS were synthesized by in vitro transcription from the dual-promoter plasmid pSPT18 containing a 725-base pair murine iNOS cDNA fragment. Control mRNA probes were obtained from the vector pSPT18. Pretreatment, hybridization, and washing conditions of dewaxed 5-µm paraffin renal tissue sections were performed as described previously (29). Slide preparations were subjected to autoradiography, exposed for 3 weeks at 4°C, and counterstained with hematoxylin and eosin. Slides were digitalized using phase-contrast microscopy (20x objective, DMIRB; Leica, Bensheim, Germany) and a ProgRes C10 digital camera (JenaOptik, Jena, Germany). Quantitative analysis of autoradiographic signals obtained by ISH representing iNOS-expressing renal tissue was done by self-written image analysis software. Areas of hybridization-positive renal tissue were referred to the total area of 0.24 mm2 renal tissue and expressed in percent iNOS mRNA-positive (% iNOS+) pixels.

Experimental protocol After an operating time of 60 min, two optical fibers for phosphorescence measurements were placed both 1 mm above the kidney surface and 1 mm above the renal vein. Then, a 15-min intravenous infusion of Oxyphor G2 (10 mg/kg; Oxygen Enterprises Ltd) was started. After 40 min, µPO2 and rvPO2 were continuously measured. Ten minutes later, a baseline blood sample (0.4 mL) was taken. At this time point, the rats were randomized between the control (n = 4), the nonresuscitation (n = 6), and the fluid resuscitation group (n = 6). All animals receiving DEX (n = 6) also received standard fluid resuscitation.

After recording the baseline values, a 30-min infusion of LPS (2.5 mg/kg, serotype 0127:B8; Sigma-Aldrich) to induce endotoxemia was started. Ninety minutes after stopping LPS infusion, a second blood sample was taken and analyzed. At t1 (2 h), two groups of animals received a bolus of HES130 (5 mL/kg) followed by continuous infusion (5 mL kg-1 h-1). In addition to the fluid resuscitation, one group of rats received a bolus of DEX (0.1 mg/kg). A second group (nonresuscitation) did not receive fluid resuscitation after LPS infusion. All animals received the same fluid volume except the nonresuscitation group. The experiment was ended 10 min after having stopped the treatment or a corresponding time point for the control groups.

Statistical analysis Values are presented as mean ± SD, unless otherwise indicated. To develop a software environment to allow data acquisition and analysis of decay and monoexponential fit procedures of the phosphorescence curves, Labview 6.1 software (National Instruments, Austin, Tex) was used. Statistics and recovery of the PO2 histogram were performed using GraphPad Prism version 4.0 for Windows (GraphPad Software, San Diego, Calif). For data analysis within each group and intergroup differences, two-way ANOVA for repeated measurements with Bonferroni posttest was performed. The oxygen histograms were analyzed using two-way ANOVA as described above. Every bin of the histogram was separately tested in time and versus control. P < 0.05 was considered significant.

Dexamethasone

53

Results

Mortality during study In the present study consisting of a total of 22 rats, none of the animals died during the experimental protocol.

Systemic and renal hemodynamics are restored by low-dose DEX In a rat model, we investigated the effects of endotoxin on systemic hemodynamic parameters (Fig. 1).

FIGURE 1. Cardiovascular and renal hemodynamic parameters during endotoxemia and resuscitation. Values represent mean ± SD. *P < 0.05 vs. baseline, †P < 0.05 vs. control, ‡P < 0.05 vs. fluid resuscitation. t0 = baseline; t1 = endotoxemia; t2 = resuscitation. NR indicates nonresuscitation; FR, fluid resuscitation.

Chapter 4

54

Two hours after LPS infusion (t1), MAP decreased significantly by 45%, and the animals developed shock. The HR and CVP were not affected by endotoxemia. Regarding the RBF, there was a significant decrease of ~60% at t1. Endotoxemia was additionally accompanied by an increase in renal vascular resistance (RVR). In all animals receiving LPS, the pH dropped from 7.4 at baseline to 7.2 at t1. In animals in the nonresuscitation group, the pH further decreased to 7.1, indicating severe metabolic acidosis. The negative base excess dropped from -1.6 ± 0.3 at baseline to -14.8 ± 1.8 at t1. In animals in the nonresuscitation group, MAP and RBF decreased to values of 60% and 82% of baseline, respectively. Fluid resuscitation alone could not restore MAP or RBF, but prevented a further drop in pH and restored the negative base excess from -13.7 ± 3.2 at t1 to -11.0 ± 3.2 at 5 h after LPS infusion (t2). The RVR was normalized to baseline values by fluid application. Treatment with low-dose DEX in addition to the standard fluid resuscitation significantly reversed the fall in MAP and RBF. Low-dose DEX prevented a decline in pH and improved the negative base excess from -16.8 ± 2.5 at t1 to -12.9 ± 3.1 at t2. In all animals receiving LPS, the CVP at the end of the experiment significantly increased compared with baseline. This is probably indicative of better cardiac preload in animals receiving fluid resuscitation and probably a decrease in cardiac contractility in the nonresuscitation group. However, to better interpret these observations, the measurement of cardiac output would have been necessary. A typical example of an experiment of the DEX group is shown in Figure 2.

FIGURE 2. Example experiment. MAP and RBF progressively dropped during LPS infusion. At 120 min, MAP and RBF are about 50% decreased compared with baseline. After an initial increase with start of fluid resuscitation plus DEX, MAP and RBF stabilized. With 290 min, MAP and RBF started to recover toward baseline values. Following anuria during septic shock, the animal started to urinate under resuscitation. Creatinine clearance was totally restored at the end of the experiment.

Dexamethasone

55

Renal microvascular hypoxia and oxygen availability improved by low-dose DEX By using the technique of oxygen-depending quenching of phosphorescence, we evaluated the renal microvascular oxygenation (µPO2) and measured the rvPO2 (Table 1). Neither the renal cortical µPO2 nor the outer medullary µPO2 were influenced by LPS infusion. However, endotoxemia was accompanied by a significant left shift in the cortical and outer medullary microvascular oxygen histogram and the appearance of areas with an average PO2 lower than 10 mmHg at t2 in the nonresuscitation group (Fig. 3).

Baseline (t0) Endotoxemia (t1) Resuscitation (t2)

Cortical !PO2 (mmHg) NR FR

DEX C

62 ± 5 68 ± 7 66 ± 9 68 ± 5

57 ± 14 69 ± 12 60 ± 20 62 ± 6

22 ± 7 *† 44 ± 7 * 43 ± 9 * 58 ± 7

Medullary !PO2 (mmHg) NR FR

DEX C

46 ± 5 50 ± 5 50 ± 7 52 ± 2

45 ± 13 53 ± 11 42 ± 16 51 ± 5

19 ± 7 *† 38 ± 6 * 37 ± 7 * 45 ± 6

rvPO2 (mmHg) NR FR

DEX C

59 ± 7 55 ± 6 53 ± 9 58 ± 9

38 ± 10 *† 39 ± 17 *†

42 ± 9 * 59 ± 5

18 ± 4 *† 32 ± 11 *† 27 ± 14 *†

55 ± 4

DO2ren (ml/min) NR FR DEX C

1.18 ± 0.21 1.37 ± 0.29 1.35 ± 0.22 1.28 ± 0.07

0.43 ± 0.18 *† 0.43 ± 0.19 *† 0.53 ± 0.27 *†

1.17 ± 0.05

0.19 ± 0.11 *† 0.38 ± 0.23 *† 0.67 ± 0.15 *†

1.15 ± 0.26

VO2ren (ml/min/g) NR FR

DEX C

0.19 ± 0.09 0.25 ± 0.08 0.25 ± 0.04 0.25 ± 0.10

0.14 ± 0.08 0.13 ± 0.10 * 0.14 ± 0.07 * 0.18 ± 0.05

0.10 ± 0.05 *† 0.15 ± 0.11 * 0.27 ± 0.09 ‡ 0.23 ± 0.05

O2ERren (%) NR FR

DEX C

23 ± 7 27 ± 5 30 ± 8 26 ± 9

48 ± 15 *† 49 ± 25 *† 42 ± 11 †

21 ± 4

82 ± 8 *† 58 ± 19 *† 62 ± 19 *†

28 ± 5

Values represent mean ± SD. * P < 0.05 vs baseline. † P < 0.05 vs control. ‡ P < 0.05 vs fluid resuscitation. NR = non-resuscitation; FR = fluid resuscitation; DEX = dexamethasone; C = control. !PO2 = microvascular partial pressure of oxygen; rvPO2 = renal venous PO2; DO2ren = renal O2-delivery; VO2ren = renal O2-consumption; O2ERren = renal O2-extraction.

TABLE 1. Renal Oxygenation Parameters

Chapter 4

56

FIGURE 3. Histograms showing the distribution of renal microvascular oxygenation. A shows the µPO2 distribution for the cortex. B is shows the µPO2 distribution for the outer medulla. Values represent mean ± SD. *P < 0.01 vs. baseline, †P < 0.01 vs. control, ‡P < 0.01 vs. fluid resuscitation. t0 = baseline; t1 = endotoxemia; t2 = resuscitation. NR indicates nonresuscitation; FR, fluid resuscitation.

Dexamethasone

57

Furthermore the DO2ren significantly decreased during endotoxemia because of the drop in RBF and accompanied by a significant reduction in renal oxygen consumption (VO2ren). The renal oxygen extraction, however, increased around 30% from baseline and was linked to a decrease in rvPO2. In all animals, the average µPO2 reading was reduced at t2. In both groups receiving fluid resuscitation, µPO2 was 50% higher than in the nonresuscitation group at t2. Dexamethasone significantly prevented the cortical microcirculatory hypoxic areas compared with the fluid resuscitation group. Furthermore DO2ren was slightly improved in rats of the DEX group. In those rats, the renal oxygen consumption was significantly higher than in the fluid resuscitation group, suggesting improved oxygen availability.

Renal kidney function and sodium reabsorption restored by low-dose DEX In all animals, the CrCl was analyzed as index of glomerular filtration rate (Fig. 4). LPS infusion was constantly accompanied by anuria. In the fluid resuscitation group, CrCl was 70% lower than baseline at t2. When additionally to the fluid resuscitation DEX was given, the CrCl was significantly restored to baseline values. Low-dose DEX treatment furthermore normalized the renal sodium reabsorption.

FIGURE 4. Kidney function and tubular sodium reabsorption during endotoxemia and resuscitation. Values represent mean ± SD. *P <0.05 vs. baseline, †P < 0.05 vs. control, ‡P < 0.05 vs. fluid resuscitation. t0 = baseline; t2 = resuscitation. 1 indicates animal is anuric. TNa+ indicates tubular sodium reabsorption; NR, nonresuscitation; FR, fluid resuscitation.

Chapter 4

58

Plasma NO concentration during endotoxemia and resuscitation

The concentration of NOx in plasma was used as indicator for NO production. As concentrations of plasma NOx consist of 20 to 40 µmol/L nitrates, 0.2 to 1 µmol/L nitrites and 0.002 to 0.005 µmol/L S-nitrosothiols, the value obtained by the vanadium method was mostly nitrate. Therefore, a calibration curve using sodium nitrate was constructed. In animals of the time control group, the NOx concentration was stable around 20 µmol/L (Fig. 5). During endotoxemia, the main production of NO derives from iNOS. This time-dependent process needed to be determined. For this reason, in n = 4 rats, the time-dependent increase in plasma NOx production was investigated. There was a significant increase in NOx levels 5 h after the start of LPS infusion. Therefore, this time point was chosen as final measurement point of the experimental protocol. Interestingly, the increase in plasma NOx concentration was not suppressed by DEX application and was in all experimental groups 12-fold elevated compared with baseline at t2.

FIGURE 5. Plasma NOx levels during endotoxemia and resuscitation. Values represent mean ± SD. *P < 0.05 vs. baseline, †P < 0.05 vs. control. A, 1-5 h measurement of NOx post LPS (n = 4), NR. B, FR. NR indicates nonresuscitation; FR, fluid resuscitation.

Renal iNOS mRNA expression suppressed by low-dose DEX To investigate the effects of endotoxemia on the renal iNOS mRNA expression, the kidney was removed at the end of the experiment and analyzed for the typical cluster formation of silver grains of iNOS by ISH. Three different regions of the kidney were investigated: the cortex and the outer and the inner medulla (Fig. 6). The iNOS mRNA expression in percentage of positive pixels was low for all regions in the control group. In kidneys from the nonresuscitation group, the amount of iNOS mRNA expression increased significantly by a factor of 2.3 compared with control. In animals of the fluid resuscitation group, the iNOS mRNA expression was also significantly increased in the cortex and inner medulla. The region of the outer medulla, although showing elevated expression, did not reach statistical significance because of a relatively high SD. In the kidneys of the DEX group, the amount of cortical iNOS-specific pixels was significantly lower than in the fluid resuscitation group. Also, the outer medullary iNOS mRNA expression was suppressed by low-dose DEX. For direct

Dexamethasone

59

comparison of iNOS mRNA expression in the fluid resuscitation and DEX group, different hematoxylin-eosin stained kidney slices are shown in Figure 7.

FIGURE 6. Calculated iNOS mRNA expression of renal tissue. The iNOS mRNA expression in renal cortex and outer and inner medulla was determined using ISH. Results are expressed in percentage of iNOS mRNA positive pixels per 0.24 mm2 renal tissue. Values represent mean ± SD. *P < 0.05 vs. control, †P < 0.05 vs. fluid resuscitation. NR indicates nonresuscitation; FR, fluid resuscitation.

FIGURE 7. iNOS mRNA expression as typical cluster formation. Comparison of iNOS mRNA expression in hematoxylin-eosin-stained renal tissue of the fluid resuscitation and the DEX group. Arrows indicate cluster formation of grains that can be attributed to single iNOS mRNA-expressing cells. FR indicates fluid resuscitation. Original magnification x20.

Chapter 4

60

Discussion

The main findings of the present study were that low-dose DEX significantly improved systemic and renal hemodynamic and oxygenation parameters in comparison to standard fluid resuscitation. Although the average µPO2 was unaffected by the DEX treatment, the appearance of microcirculatory hypoxic areas in the cortical oxygen histogram was reversed in this group. Dexamethasone had hereby beneficial effects on the renal oxygen availability. The improvement in hemodynamics and in intrarenal oxygenation was accompanied by a restoration of CrCl and normalization of tubular sodium reabsorption. Dexamethasone significantly reduced renal iNOS expression but had no suppressive effects on the plasma NOx concentration.

Animal models can never be directly translated to the human pathophysiology. Besides this fact, many protocols do not mimic a clinical setting and therefore lack a direct clinical relevance. The aim of our study was to investigate the effects of glucocorticoid treatment during endotoxemia in a model in which treatment is started in the presence of severe hemodynamic alterations like those seen in clinical practice. Early goal-directed therapy using fluid resuscitation to prevent hypoperfusion of vital organs is the standard therapy in sepsis (30). Therefore, fluid resuscitation was the primary therapeutic approach in our model after profound septic shock had developed. In addition to standard fluid resuscitation, a bolus of low-dose DEX was given. In most previous studies, DEX had been given together with endotoxin or even before (11, 19). The reason for the choice of this strategy is based on the fact that there is a reduced receptor-binding capacity of glucocorticoids under septic conditions (31). It is postulated that nitrosation of the glucocorticoid receptor by NO is the underlying mechanism in this context (11).

With the concept to correct for a sepsis-related relative adrenal insufficiency, glucocorticosteroids have been used for years in the treatment of septic shock. However, a recently published multicenter clinical trial, the CORTICUS study, failed to demonstrate beneficial effects in sepsis (32). One reason why low-dose glucocorticosteroid therapy did not improve survival might be that the inclusion time window was subsequently increased to 72 h. We postulate that lowdose glucocorticosteroid therapy has to be started as part of hemodynamic stabilization to demonstrate beneficial effects. Endotoxemia was associated with a significant drop in MAP and RBF in our model. These severe hemodynamic alterations could be significantly restored by treatment with low-dose DEX compared with fluid resuscitation alone. Explanations of the underlying mechanisms for these observations are made throughout the following paragraphs and focus mainly on the measured alterations in iNOS mRNA expression and NOx plasma levels.

There are not much data on the influence of endotoxemia on tissue oxygenation. In previous investigations of our own group, infusion of endotoxin was accompanied by only mild changes in µPO2 (6). In different studies, the impact of endotoxemia on tissue oxygenation was slightly varying; however, endotoxemia was never accompanied by profound tissue hypoxia (33). The reasons for this observation are diverse and ranging from mitochondrial dysfunction to a decrease in oxygen consumption or might be simply explained by the fact that the DO2ren, although reduced, was still high enough to maintain adequate tissue oxygenation. Recently, we demonstrated that endotoxemia was accompanied by the appearance of microcirculatory hypoxic areas in the cortical oxygen histogram (27). Such areas were present in our current model after 3 h of endotoxemia in the nonresuscitation group. Treatment with low-dose DEX prevented the occurrence of the cortical microcirculatory hypoxic areas. The attempt to verify the link between these hypoxic areas and the

Dexamethasone

61

renal anatomy is still missing. However, the speculation that the peritubular capillary region of the cortex might be affected in this context is legitimate. Because DEX application improved the DO2ren, one can argue that this increase in oxygen availability might have prevented peritubular capillary dysfunction, which is an early event in endotoxemia that contributes to tubular stress and renal injury (34).

Low-dose DEX significantly improved kidney function and normalized tubular sodium reabsorption. If the restoration in renal function was due to the improvement in macrohemodynamics and/or due to changes in the local microcirculation needs to be proven in future investigations. Inulin clearance was considered as an accurate measure for kidney function in our model. However, due to reservations regarding spectrofluorometric interference in the presence of Oxyphor G2, changes in CrCl were used as an index of glomerular filtration. Due to the lack of a steady state of creatinine balance, these changes should be regarded only as a gross indication.

Dexamethasone is known to inhibit the gene expression of iNOS (18), which we were able to confirm in our model. Beneficial effects on kidney function and macrohemodynamics have been linked to the inhibition in NO production (19). This makes sense in the context that, in sepsis, NO is responsible for peripheral vasodilation, and this activates the sympathetic and renin-angiotensin systems, which maintain blood pressure at the expense of renal vasoconstriction (35). Although given 2 h after endotoxin, DEX significantly restored MAP and RBF in our model. Furthermore, it significantly improved kidney function and normalized the tubular sodium reabsorption. Dexamethasone is known for its absence of mineralocorticoid action, and the normalization of tubular sodium reabsorption is therefore most likely attributed to the attenuation of the deleterious effects of LPS. Alternatively, this action of DEX on sodium transport might be due to stimulation of the Na-K-ATPase activity of the medullary thick ascending limb (36).

Beneficial effects of low-dose DEX on renal function can be attributed to iNOS inhibition. Selective iNOS inhibition has been demonstrated to improve kidney function in several studies of endotoxemia (22, 21). The application of nonselective NO synthase inhibitors inhibiting also endothelial NO synthase (eNOS) showed negative effects on kidney function (37). The production of NO by eNOS is important for autoregulation of renal microvascular tone, and its inhibition therefore is thought to be detrimental for the kidney (8). There is evidence that low-dose DEX can also downregulate eNOS in the kidney as demonstrated in nonendotoxemic mice (38). That DEX in our model did not exhibit its beneficial effects solely because of iNOS inhibition is supported by the finding of high plasma NOx concentration. In contrast to a reduction in renal iNOS expression, the plasma NOx content was 10 times higher in the DEX group than in the control after 5 h of endotoxemia. Other studies (11, 19) showed a reduction in plasma NOx levels upon DEX treatment at an earlier time point in relation to LPS administration. Because NOx concentration is a systemic parameter, it is necessary to consider that the measured NOx derives also from other tissue than the kidney. We demonstrated that renal iNOS expression was suppressed by low-dose DEX. However, it is known that up to 1.4 times more iNOS is expressed in lung tissue (39), which could be a remaining source for the elevated NOx.

MAP and renal blood low were restored in the DEX group. As NO is postulated to be mainly responsible for septic vasodilation, how can the effects of DEX be explained? Dexamethasone has been shown to improve endotoxin-related vascular hyporeactivity by modulating ATP-sensitive potassium channels (12). This interaction can explain the increased vascular reactivity to catecholamines by glucocorticoids. Translated to our model, it can mean that DEX might have

Chapter 4

62

improved the vascular reactivity to endogenous catecholamines known to be elevated during endotoxemia (40). The increase in MAP and therefore an increase in renal perfusion pressure could at least partly explain the improvement in kidney function, accompanied by the fact that DEX normalized the RVR.

Low-dose DEX clearly improved macrohemodynamics and renal function during endotoxemia. These findings cannot be solely related to iNOS inhibition because plasma levels of NOx were not suppressed. Our data suggest that the improvement in renal function was directly related to prevention of microcirculatory hypoxic areas in the DEX-treated animals. The underlying mechanisms still need to be uncovered. However, due to the impressive effects of low-dose DEX, it is worthwhile to consider its application in the treatment of septic ARF, keeping in mind the possibility of immunosuppression.

Acknowledgements

The authors thank Prof Dr Hartmut Oßwald, Department of Pharmacology and Toxicology, University of Tuebingen, Germany, for his kind support. The authors also thank Christof Zanke, Department of Anesthesiology and Critical Care, University of Tuebingen, Germany, for writing the software to analyze the iNOS mRNA expression.

References

1. Thijs A, Thijs LG: Pathogenesis of renal failure in sepsis. Kidney Int Suppl 66:S34-S37, 1998. 2. Brivet FG, Kleinknecht DJ, Loirat P, Landais PJ: Acute renal failure in intensive care units-causes, outcome, and

prognostic factors of hospital mortality; a prospective, multicenter study. French Study Group on acute renal failure. Crit Care Med 24:192-198, 1996.

3. Heyman SN, Darmon D, Goldfarb M, Bitz H, Shina A, Rosen S, Brezis M: Endotoxin-induced renal failure. I. A role for altered renal microcirculation. Exp Nephrol 8:266-274, 2000.

4. Nitescu N, Grimberg E, Ricksten SE, Herlitz H, Guron G: Endothelin B receptors preserve renal blood flow in a normotensive model of endotoxininduced acute kidney dysfunction. Shock 29:402-409, 2008.

5. Millar CG, Thiemermann C: Carboxy-PTIO, a scavenger of nitric oxide, selectively inhibits the increase in medullary perfusion and improves renal function in endotoxemia. Shock 18:64-68, 2002.

6. Johannes T, Mik EG, Nohe B, Raat NJ, Unertl KE, Ince C: Influence of fluid resuscitation on renal microvascular PO2 in a normotensive rat model of endotoxemia. Crit Care 10:R88, 2006.

7. Schurek HJ, Jost U, Baumgartl H, Bertram H, Heckmann U: Evidence for a preglomerular oxygen diffusion shunt in rat renal cortex. Am J Physiol 259:F910-F915, 1990.

8. Ito S, Carretero OA, Abe K: Role of nitric oxide in the control of glomerular microcirculation. Clin Exp Pharmacol Physiol 24:578-581, 1997.

9. Millar CG, Thiemermann C: Intrarenal haemodynamics and renal dysfunction in endotoxaemia: effects of nitric oxide synthase inhibition. Br J Pharmacol 121:1824-1830, 1997.

10. Schrier RW, Wang W: Acute renal failure and sepsis. N Engl J Med 351: 159-169, 2004. 11. Duma D, Silva-Santos JE, Assreuy J: Inhibition of glucocorticoid receptor binding by nitric oxide in endotoxemic

rats. Crit Care Med 32:2304-2310, 2004. 12. d’Emmanuele V, Lippolis L, Autore G, Popolo A, Marzocco S, Sorrentino L, Pinto A, Sorrentino R: Dexamethasone

improves vascular hyporeactivity induced by LPS in vivo by modulating ATP-sensitive potassium channels activity. Br J Pharmacol 140:91-96, 2003.

13. Annane D, Bellissant E, Bollaert PE, Briegel J, Keh D, Kupfer Y: Corticosteroids for severe sepsis and septic shock: a systematic review and metaanalysis. BMJ 329:480, 2004.

14. Rothwell PM, Udwadia ZF, Lawler PG: Cortisol response to corticotrophin and survival in septic shock. Lancet 337:582-583, 1991.

15. Mansart A, Bollaert PE, Seguin C, Levy B, Longrois D, Mallie JP: Hemodynamic effects of early versus late glucocorticosteroid administration in experimental septic shock. Shock 19:38-44, 2003.

Dexamethasone

63

16. Annane D, Sebille V, Charpentier C, Bollaert PE, Francois B, Korach JM, Capellier G, Cohen Y, Azoulay E, Troche G, et al.: Effect of treatment with low doses of hydrocortisone and fludrocortisone on mortality in patients with septic shock. JAMA 288:862-871, 2002.

17. Bollaert PE, Charpentier C, Levy B, Debouverie M, Audibert G, Larcan A: Reversal of late septic shock with supraphysiologic doses of hydrocortisone. Crit Care Med 26:645-650, 1998.

18. Radomski MW, Palmer RM, Moncada S: Glucocorticoids inhibit the expression of an inducible, but not the constitutive, nitric oxide synthase in vascular endothelial cells. Proc Natl Acad Sci USA 87:10043-10047, 1990.

19. Tsao CM, Ho ST, Chen A, Wang JJ, Li CY, Tsai SK, Wu CC: Low-dose dexamethasone ameliorates circulatory failure and renal dysfunction in conscious rats with endotoxemia. Shock 21:484-491, 2004.

20. Leach M, Hamilton LC, Olbrich A, Wray GM, Thiemermann C: Effects of inhibitors of the activity of cyclo-oxygenase-2 on the hypotension and multiple organ dysfunction caused by endotoxin: a comparison with dexamethasone. Br J Pharmacol 124:586-592, 1998.

21. Matejovic M, Krouzecky A, Martinkova V, Rokyta R Jr, Kralova H, Treska V, Radermacher P, Novak I: Selective inducible nitric oxide synthase inhibition during long-term hyperdynamic porcine bacteremia. Shock 21:458-465, 2004.

22. Rosselet A, Feihl F, Markert M, Gnaegi A, Perret C, Liaudet L: Selective INOS inhibition is superior to norepinephrine in the treatment of rat endotoxic shock. Am J Respir Crit Care Med 157:162-170, 1998.

23. Johannes T, Mik EG, Ince C: Dual-wavelength phosphorimetry for determination of cortical and subcortical microvascular oxygenation in rat kidney. J Appl Physiol 100:1301-1310, 2006.

24. Mik EG, Johannes T, Ince C: Monitoring of renal venous PO2 and kidney oxygen consumption in rats by a near-infrared phosphorescence lifetime technique. Am J Physiol Renal Physiol 294:F676-F681, 2008.

25. Dunphy I, Vinogradov SA, Wilson DF: Oxyphor R2 and G2: phosphors for measuring oxygen by oxygen-dependent quenching of phosphorescence. Anal Biochem 310:191-198, 2002.

26. Vinogradov SA, Lo LW, Wilson DF: Dendritic polyglutamic porphyrins: probing porphyrin protection by oxygen-dependent quenching of phosphorescence. Chem Eur J 5:1338-1347, 1999.

27. Johannes T, Mik EG, Ince C: Nonresuscitated endotoxemia induces microcirculatory hypoxic areas in the renal cortex in the rat. Shock 31(1):97-103, 2009

28. Yang BK, Vivas EX, Reiter CD, Gladwin MT: Methodologies for the sensitive and specific measurement of S-Nitrosothiols, iron-nitrosyls, and nitrite in biological samples. Free Radic Res 37:1-10, 2003.

29. Szalay G, Sauter M, Hald J, Weinzierl A, Kandolf R, Klingel K: Sustained nitric oxide synthesis contributes to immunopathology in ongoing myocarditis attributable to interleukin-10 disorders. Am J Pathol 169: 2085-2093, 2006.

30. Rivers E, Nguyen B, Havstad S, Ressler J, Muzzin A, Knoblich B, Peterson E, Tomlanovich M: Early goal-directed therapy in the treatment of severe sepsis and septic shock. N Engl J Med 345:1368-1377, 2001.

31. Stith RD, McCallum RE: Down regulation of hepatic glucocorticoid receptors after endotoxin treatment. Infect Immun 40:613-621, 1983.

32. Sprung CL, Annane D, Keh D, Moreno R, Singer M, Freivogel K, Weiss YG, Benbenishty J, Kalenka A, Forst H, et al.: Hydrocortisone therapy for patients with septic shock. N Engl J Med 358:111-124, 2008.

33. Nitescu N, Grimberg E, Guron G: Low-dose candesartan improves renal blood flow and kidney oxygen tension in rats with endotoxin-induced acute kidney dysfunction. Shock 30:166-172, 2008

34. Wu L, Tiwari MM, Messer KJ, Holthoff JH, Gokden N, Brock RW, Mayeux PR: Peritubular capillary dysfunction and renal tubular epithelial cell stress following lipopolysaccharide administration in mice. Am J Physiol Renal Physiol 292:F261-F268, 2007.

35. Singh P, Deng A, Weir MR, Blantz RC: The balance of angiotensin II and nitric oxide in kidney diseases. Curr Opin Nephrol Hypertens 17: 51-56, 2008.

36. Doucet A, Hus-Citharel A, Morel F: In vitro stimulation of Na-K-ATPase in rat thick ascending limb by dexamethasone. Am J Physiol 251:F851-F857, 1986.

37. Schwartz D, Mendonca M, Schwartz I, Xia Y, Satriano J, Wilson CB, Blantz RC: Inhibition of constitutive nitric oxide synthase (NOS) by nitric oxide generated by inducible NOS after lipopolysaccharide administration provokes renal dysfunction in rats. J Clin Invest 100:439-448, 1997.

38. Wallerath T, Godecke A, Molojavyi A, Li H, Schrader J, Forstermann U: Dexamethasone lacks effect on blood pressure in mice with a disrupted endothelial NO synthase gene. Nitric Oxide 10:36-41, 2004.

39. Kan W, Zhao KS, Jiang Y, Yan W, Huang Q, Wang J, Qin Q, Huang X, Wang S: Lung, spleen, and kidney are the major places for inducible nitric oxide synthase expression in endotoxic shock: role of P38 mitogen-activated protein kinase in signal transduction of inducible nitric oxide synthase expression. Shock 21:281-287, 2004.

40. Jones SB, Kotsonis P, Majewski H: Endotoxin enhances norepinephrine release in the rat by peripheral mechanisms. Shock 2:370Y375, 1994.

65

CHAPTER ILOPROST PRESERVES RENAL OXYGENATION AND RESTORES KIDNEY FUNCTION IN ENDOTOXEMIA-INDUCED ACUTE RENAL FAILURE IN THE RAT

Tanja Johannes, Can Ince, Karin Klingel, Klaus E Unertl, and Egbert G Mik Critical Care Medicine, 37(4): 1423-32, 2009

Chapter 5

66

Abstract Objective: To investigate that exogenous prostacyclin would counterbalance an endotoxemia-induced intrarenal vasoconstriction and would therefore have beneficial effects on kidney function. Design: Prospective, randomized, controlled study. Setting: University medical center research laboratory. Subjects: Eighteen male Wistar rats. Interventions: In anesthetized and ventilated animals, arterial blood pressure (mean arterial blood pressure [MAP]) and renal blood flow (RBF) were recorded. Renal microvascular PO2 (µPO2) and renal venous PO2 were continuously measured by phosphorescence lifetime technique. All animals received a 30-minute infusion of lipopolysaccharide (LPS) (2.5 mg/kg) to induce endotoxemia. One group of rats was not resuscitated. A second group received fluid resuscitation 90 minutes after stop of LPS infusion. In a third group of rats, the prostacyclin analogue iloprost (100 ng/kg/min) was continuously infused in addition to fluid resuscitation. Furthermore, in all the animals, plasma NOx levels, renal inducible nitric-oxide synthase (iNOS) messenger RNA (mRNA) expression, and creatinine clearance were determined. Measurements and Main Results: During LPS infusion, MAP and RBF progressively dropped to 50% of baseline at 120 minutes. After an initial increase in MAP and RBF, start of fluid resuscitation with iloprost resulted in the stabilization of both parameters. All animals became anuric during endotoxemia. Only in animals receiving iloprost was creatinine clearance totally restored at the end of the experiment. Iloprost had no significant effects on average µPO2, but prevented the occurrence of cortical microcirculatory hypoxic areas. NOx levels and iNOS mRNA expression were significantly increased in all animals receiving LPS after 5 hours. There was no difference in NOx concentration between the different groups. In animals receiving iloprost, iNOS mRNA expression was significantly suppressed in the inner medulla. Conclusions: Iloprost significantly restored kidney function of endotoxemic rats to baseline values. This beneficial effect of iloprost on renal function might be addressed to an improvement in intrarenal oxygenation.

Iloprost

67

Introduction

Acute renal failure (ARF) is a serious condition in patients with septic shock with 70% mortality (1, 2). Unfortunately, the pathogenesis is only partly understood, and there is need to further identify the mechanisms responsible for the development of ARF in sepsis and define strategies to prevent the development of ARF due to septic conditions. So far, it is known that sepsis-associated kidney dysfunction is characterized by a reduction in renal blood flow (RBF) and glomerular filtration rate (3). Morphology might range from endothelial damage in glomerular and interstitial vessels, disseminated fibrin thrombi, and aggregation of leukocytes and platelets to acute tubular necrosis (4). In sepsis, there is a release of a vast array of inflammatory cytokines, vasoactive mediators (nitric oxide [NO] adenosine, etc.), thrombogenic substances, and arachidonic acid metabolites (5). Overall, the predominant pathogenetic factor in the development of ARF seems to be renal hypoperfusion due to an imbalance between renal vasoconstriction and vasodilation (6). Whether this results in hypoxia as a key pathogenic factor remains controversial (7); but recently, we demonstrated that endotoxemia is associated with the occurrence of local hypoxia in the renal cortex in rats (8).

In the kidney, prostaglandins uphold the balance between vasodilator and vasoconstrictor to maintain homeostasis and physiologic kidney function (9). Because of its potent vasodilatory effects, prostacyclin (PGI2) must be considered to play a potential renal protective role in septic ARF. The synthesis of PGI2 along the nephron is quite significant and highest in the glomeruli and the inner medulla. The documented role of PGI2 in the kidney is the regulation of renal and glomerular hemodynamics, renin secretion, and tubular transport processes (6). In animal experimental studies, the stable PGI2 analogue iloprost preserved kidney function against anoxia in rabbits (10), and had beneficial effects in ischemia/ reperfusion-induced renal injury in a rat model (11). Furthermore, in a rat study of endotoxemia, iloprost inhibited an increase in plasma NO2-/NO3- and lung tissue inducible nitric-oxide synthase (iNOS) expression (12). In a clinical study, iloprost was successfully used to prevent contrast-mediated nephropathy (13). On the basis of the physiologic actions of PGI2 and the described renalprotective effects of iloprost, we hypothesized that this PGI2 analogue might have a potential role in a renal protective strategy in sepsis.

This study in a rat model of endotoxemia was undertaken to examine whether 1) application of exogenous PGI2 (iloprost) would counterbalance a sepsis-induced intrarenal vasoconstriction and would therefore have beneficial effects on kidney function; 2) this effect would be related to an intrarenal shift in microvascular oxygenation; and 3) iloprost-mediated effects were associated with changes in renal iNOS messenger RNA (mRNA) expression and total body NO production.

Material and methods

Experiments were conducted in accordance with the guidelines for Institutional and Animal Care and Use Committees’ care and handling of the animals and approved by the Animal Research Committee of the Academic Medical Center at the University of Amsterdam.

Animal Preparation and Monitoring. Male Wistar rats (259 ± 42 g; Harlan, Horst, The Netherlands) were anesthetized with a mixture of ketamine (90 mg/kg), medetomidine (0.5 mg/kg), and atropine-sulfate (0.05 mg/kg) intraperitonally. A tracheostomy was performed, and the animals were mechanically ventilated (FIO2 0.4). The right

Chapter 5

68

carotid artery was cannulated and used for monitoring of arterial blood pressure and heart rate. To allow continuous central venous pressure measurement, a catheter was inserted into the right jugular vein to a level of the right atrium. A catheter in the right femoral artery was used for withdrawal of blood via the right femoral vein for continuous infusion of Ringer’s lactate (15 mL/kg/h; Baxter, The Netherlands) and ketamine (50 mg/kg/h; Nimatek; Eurovet, The Netherlands). Rectal temperature was maintained at 37°C. Through flank incision, the left kidney was exposed, decapsulated, and immobilized. Left RBF was measured with a perivascular flow probe connected to a flow meter (T206; Transonic Systems, Ithaca, NY). The left ureter was isolated, ligated, and cannulated for urine collection. At the end of the experiment, the left kidney was removed and weighed. All preceding steps are described in detail in a previous study (14).

Experimental Protocol. At the start of the experiment, rats were randomized between control (n = 4), nonresuscitation (n = 6), fluid resuscitation (n = 6), and iloprost (n = 6) groups. All animals receiving iloprost received standard fluid resuscitation with hydroxyethyl starch (130 kDa). At the end of surgery (60 minutes), two optical fibers for phosphorescence measurements were placed both 1 mm above the kidney surface and 1 mm above the renal vein and an intravenous infusion of Oxyphor G2 (10 mg/kg in 15 minutes; Oxygen Enterprises, USA) was started. Forty minutes later, µPO2 and rvPO2 were continuously recorded. Then, a baseline blood sample (0.4 mL) was taken. In 18 rats, a 30-minute infusion of lipopolysaccharide (LPS, 2.5 mg/ kg; serotype 0127:B8, Sigma, The Netherlands) was given to induce septic shock. Ninety minutes after stop of LPS infusion, two groups of animals received fluid resuscitation (5 mL/kg followed by 5 mL/kg/hr; Voluven, 6% hydroxyethyl starch 130/0.4; Fresenius Kabi, The Netherlands). Additional to the fluid resuscitation, in one group of animals iloprost (100 ng/kg/min; Ilomedine, Schering, The Netherlands) was continuously infused. Continuous infusion was chosen because of the pharmacokinetic properties of the drug (a short plasma half-life of 20–30 minutes) and continued throughout the end of the protocol. Another group (nonresuscitation) did not receive fluid resuscitation after LPS infusion. All animals received the same fluid volume except the nonresuscitation group. The experiment was ended 10 minutes after stop of treatment or a corresponding time point for the control groups. One animal of the nonresuscitation group died during the experiment and was replaced.

Measurement of Renal Microvascular Oxygenation and Renal Venous PO2. The renal microvascular PO2 (µPO2) within the kidney cortex and outer medulla was measured by oxygen-dependent quenching of phosphorescence using a dual-wavelength-phosphorimeter (15). The renal venous PO2 (rvPO2) was detected by the same method (16). Briefly described, Oxyphor G2 (Oxygen Enterprises, Philadelphia, PA), intravenously infused, binds to albumin (17). If excited by a flash of light (wavelength ~440 or 630 nm), the Oxyphor G2-albumin complex emits phosphorescence (wavelength ~800 nm). Dependent on the oxygen concentration, the phosphorescence intensity decreases and the relationship between the measured decay-time and the PO2 can be estimated using the Stern-Volmer relation (18 –20). Heterogeneity in oxygen pressure was analyzed by fitting a sum of small rectangular distributions to the distributions of quencher concentration in the phosphorescence data (21).

Blood Gas Measurements. An arterial blood sample (0.4 mL) was taken via the femoral artery at three different time points: first time point, 0 minutes = baseline (t0); second time point, 120 minutes = endotoxemia (t1); and third

Iloprost

69

time point, ~300 minutes = resuscitation (t2). The same volume of hydroxyethyl starch replaced the sampling volume of blood. Samples were analyzed for the determination of blood gas values (ABL505 blood gas analyzer; Radiometer, Denmark), hematocrit, Hb concentration, and HbSO2 (OSM3; Radiometer, Denmark). Furthermore, in each sample sodium and potassium concentration, plasma creatinine concentration, and NOx plasma levels were measured. Calculations of Renal Oxygenation. Renal oxygen delivery (DO2ren) was calculated as RBF x arterial oxygen content: RBF x (1.31 x Hb x SaO2) + (0.003 x PaO2). Renal oxygen consumption (VO2ren) was calculated as RBF x (arterial - renal venous oxygen content difference). Renal venous oxygen content is: (1.31 x Hb x SrvO2) + (0.003 x rvPO2) (16). An estimation of the renal vascular resistance (RVR) of the renal artery flow region was made: RVR = MAP/RBF (22).

Measurement of Renal Function. Creatinine clearance (Clcrea) was assessed as an index of glomerular filtration rate. Calculations of the clearance were performed with standard formula: Clcrea (mL/min) = (U x V)/P, where U is the urine creatinine concentration, V is the urine volume per unit time, and P is the plasma creatinine concentration. The specific elimination capacity for creatinine of the left kidney was normalized to the organ weight. For analysis of urine volume and creatinine concentration, urine samples from the left ureter were collected at 10-minute intervals. At the midpoint of each 10-minute interval, plasma creatinine concentration was analyzed. Analysis of samples was performed using Jaffe method. In all, the urine samples’ sodium concentration was determined. To calculate sodium excretion (UNa+ x V), sodium concentration in urine (UNa+; mmol/L) was multiplied by the urine flow (V). The cost of sodium transport (VO2/TNa+) is the relation of the total amount of VO2ren over the total amount of sodium reabsorbed (TNa+). TNa+ (mmol x min-1) was calculated according to: (Clcrea x PNa+) – UNa+ x V, where PNa+ is the plasma concentration of sodium.

Plasma NOx Measurement. NOx (nitrate/nitrite/S-nitrosothiols) in plasma was determined using a NO analyzer (Sievers Instruments, Boulder, CO). For measurement of NOx plasma, samples (50 µL) were centrifuged at 20.200g for 10 minutes and immediately frozen at -80°C. A detailed description of this method is published elsewhere (23). Briefly described, NOx is determined using vanadium (III) chloride (Sigma-Aldrich, Zwijndrecht, The Netherlands). Vanadium (III) chloride reduces nitrate/nitrite to NO gas, which is released in a closed system apparatus of an ozone-based chemiluminescent assay at a temperature of 90°C. A calibration curve of sodium nitrate was constructed using the mean of ten measured values for each calibration point. As concentrations of plasma nitrate are 20–40 µM/L, of nitrite 0.2–1 µM/L and of S-nitrosothiols 0.002–0.005 µM/L, the value obtained by the vanadium assay is mostly nitrate. The raw data were analyzed using the integration method for estimating the area under the curve.

iNOS mRNA Expression. After termination of the experiment, the kidney was immediately fixed in 4% paraformaldehyde overnight and embedded in paraffin. iNOS mRNA expression was determined by in situ hybridization. 35S-labeled antisense RNA probes for the detection of iNOS mRNA were synthesized by in vitro transcription from the dual promoter plasmid pSPT18 containing a 725-bp murine iNOS cDNA fragment. Control mRNA probes were obtained from the vector pSPT18. Pretreatment, hybridization, and washing conditions of dewaxed 5-µm paraffin renal tissue sections were

Chapter 5

70

performed as described previously (24). Slide preparations were subjected to autoradiography, exposed for 3 weeks at 4°C, and counterstained with hematoxylin and eosin. Image capturing was performed using phase-contrast microscopy (20x objective, DMIRB; Leica, Bensheim, Germany) and a digital camera (ProgRes C10; JenaOptik, Jena, Germany). Quantitative analysis of autoradiographic signals obtained by in situ hybridization representing iNOS mRNA expressing renal tissue was done by self-written image analysis software. Areas of hybridization-positive renal tissue were referred to the total area of tissue sections and expressed in percentage of iNOS mRNA positive (% iNOS+) pixels per 0.24 mm2 renal tissue (n = three kidneys for each group).

Data Presentation and Statistics. Values are presented as mean ± SD, unless otherwise indicated. Analysis of the mono-exponential fit procedures of the phosphorescence curves was performed using Labview 6.1 software (National Instruments, Austin, TX). For PO2 histogram recovery and statistics, GraphPad Prism version 4.0 for Windows (GraphPad Software, San Diego, CA) was used. For data analysis within each group and intergroup differences, two-way analysis of variance for repeated measurements with Bonferroni posttest was performed. p < 0.05 were considered significant.

Results

Systemic and Regional Hemodynamics. The baseline values in the experimental groups and the control group were not different (Table 1). In the nonresuscitation group, MAP decreased 43% at t1 and dropped further to 60% at t2 compared with baseline (p < 0.05 vs. control). The heart rate was significantly decreased compared with control 5 hours after start of LPS infusion. The central venous pressure was significantly increased at t1 and t2. After LPS infusion, RBF dropped 62% (t1) and reached its lowest reading at t2 with 80% compared with baseline (p < 0.05 vs. control). The RVR increased by 61% at t2. Fluid resuscitation could not restore MAP. There was a significant increase in central venous pressure at t1 and t2 compared with baseline. After fluid resuscitation, RBF was still 63% lower than at baseline. After a significant increase in RVR at t1 (p < 0.05 vs. control), RVR was normalized by fluid resuscitation.

In the iloprost group, MAP decreased 39% after LPS infusion compared with baseline. A further drop in MAP was prevented in animals receiving iloprost as a supplement to standard fluid resuscitation. In this group, MAP was 38% lower than at baseline at t2 (p < 0.05 vs. control). The heart rate remained unchanged during the experimental period. There was a significant increase in central venous pressure at t1 and t2 compared with baseline. In the iloprost group, RBF was restored after resuscitation, slightly but not significantly, and 48% lower than at baseline. At t2, RVR was normalized to baseline values.

Iloprost

71

Baseline (t0) Endotoxemia (t1) Resuscitation (t2)

MAP (mmHg) non-resuscitation fluid resuscitation

iloprost time control

114 ± 5 108 ± 7 107 ± 3 109 ± 4

65 ± 5 *† 64 ± 16 *† 65 ± 8 *† 107 ± 7

46 ± 7 *† 58 ± 12 *† 67 ± 4 *† 103 ± 7

HR (bpm) non-resuscitation fluid resuscitation

iloprost time control

262 ± 21 261 ± 20 258 ± 30 271 ± 28

279 ± 32 263 ± 21 271 ± 27 274 ± 35

250 ± 62 † 277± 50 290 ± 24 321 ± 28

CVP (mmHg) non-resuscitation fluid resuscitation

iloprost time control

5.7 ± 2.4 6.8 ± 0.9 5.9 ± 2.0 5.5 ± 1.1

7.2 ± 1.4 * 8.1 ± 0.7 * 7.0 ± 2.0 * 6.5 ± 1.4

7.9 ± 1.9 * 8.0 ± 2.0 * 7.4 ± 2.2 * 6.8 ± 1.4 *

RBF (mL!min-1) non-resuscitation fluid resuscitation

iloprost time control

5.5 ± 0.3 5.8 ± 0.9 6.0 ± 0.6 6.3 ± 0.8

2.1 ± 1.0 *† 1.6 ± 0.9 *† 2.2 ± 1.0 *†

5.6 ± 0.3

1.1 ± 0.6 *† 2.2 ± 1.4 *† 3.1 ± 1.4 *†

6.1 ± 0.9

RVR (dyne!sec/cm5) non-resuscitation fluid resuscitation

iloprost time control

21 ± 2 19 ± 3 18 ± 2 18 ± 2

38 ± 19 * 51 ± 28 *† 41 ± 32 *

19 ± 2

54 ± 27 *† 36 ± 18 26 ± 12 17 ± 3

TABLE 1. Systemic hemodynamic parameters

Values represent mean ± SD. * P < 0.05 vs baseline. † P < 0.05 vs control. MAP = mean arterial blood pressure; HR = heart rate; CVP = central venous pressure; RBF = renal blood flow; RVR = renal vascular resistance.

Example Experiment. A typical example of an experiment is shown in Figure 1. During LPS infusion, MAP and RBF progressively dropped. At 120 minutes, MAP and RBF are about 50% decreased compared with baseline values. With the start of fluid resuscitation and continuous infusion of iloprost there was an initial increase in MAP and RBF followed by stabilization of both parameters. In four of six animals, there was an undulation in MAP and RBF during resuscitation with iloprost. After a period of anuria during septic shock the animal started to urinate under resuscitation. Clcrea was totally restored at the end of the experiment. After an initial increase in microvascular PO2 (µPO2) during LPS infusion, cortical (cµPO2) and outer medullary (mµPO2) microvascular PO2 slightly dropped during resuscitation. Furthermore, the renal venous PO2 (rvPO2) progressively decreased during the experimental period.

Chapter 5

72

FIGURE 1. Example experiment. (A) During lipopolysaccharide (LPS) infusion mean arterial pressure (MAP) and renal blood flow (RBF) progressively drop. Compared with baseline values MAP and RBF are about 50% decreased at 120 minutes. After an initial increase in MAP and RBF with start of fluid resuscitation + iloprost both parameters are undulating and stabilizing. Following anuria during septic shock the animal starts to urinate under resuscitation. Creatinine clearance (Clcrea) is totally restored at the end of the experiment. (B) After an initial increase in microvascular PO2 (µPO2) during LPS infusion cortical (C) and outer medullary (m) µPO2 slightly drop during resuscitation. Renal venous PO2 (rvPO2) progressively decreases with start of LPS application. Renal Oxygenation Parameters. There was a significant decrease in cµPO2 and mµPO2 at t2 compared with baseline in the control group (Fig. 2). The renal oxygen delivery, consumption, and extraction (DO2ren, VO2ren, and renal oxygen extraction) remained unchanged over time.

In the nonresuscitation group, µPO2 was stable around baseline values at t1. However, 5 hours post-LPS, cµPO2 and mµPO2 were 22 ± 8 and 18 ± 8 mm Hg, respectively (p < 0.05 vs. control). This significant reduction in the average µPO2 was accompanied by a significant left shift in both the cortical and outer medullary oxygen histogram at t1 and t2 (Fig. 3). DO2ren decreased from 1.20 ± 0.23 at baseline to 0.42 ± 0.20 at t1 and to 0.20 ± 0.12 mL/min at t2 (p < 0.05 vs. control) in the nonresuscitation group. In the same group, VO2ren was significantly decreased compared with baseline at t2, whereas renal oxygen extraction in- creased from 50 ± 15 at t1 to 82% ± 9% at t2 (p < 0.05 vs. control).

After fluid resuscitation, cµPO2 and mµPO2 were 41 ± 13 and 35 ±11 mmHg, respectively, which was only slightly but not significantly lower than in the control group. DO2ren was significantly decreased at t1 and t2 (p < 0.05 vs. control). Furthermore, there was a significant drop in VO2ren with 0.13 ± 0.10 at t1 and 0.16 ± 0.12 mL/min/g at t2 compared with baseline in animals receiving fluid resuscitation. Renal oxygen extraction was 65% ± 15% at t2 (p < 0.05 vs. control).

Iloprost

73

In the iloprost group, cµPO2 was slightly but not significantly reduced at t2. The medullary PO2 was at t2 42 ± 7 mm Hg and, therefore, as high in the control group. In this group, there was a significant prevention in the appearance of cortical microcirculatory hypoxic areas compared with the fluid resuscitation group. A drop in DO2ren could not be prevented by iloprost (p < 0.05 vs. control). However, iloprost restored VO2ren from 0.12 ± 0.07 mL/min/g at t1 to baseline values (0.26 ± 0.10 mL/min/g). The renal oxygen extraction was with 87% ± 12% at t2, the highest of all groups (p < 0.05 vs. control, fluid resuscitation).

FIGURE 2. Renal oxygenation during septic shock and resuscitation. Values represent mean ± SD *p < 0.05 vs. baseline. xp < 0.05 vs. control. +p < 0.05 vs. fluid resuscitation. T0, baseline; t1, endotoxemia; t2, resuscitation. CµPO2, cortical microvascular PO2; mµPO2, outer medullary microvascular PO2; DO2ren, renal oxygen delivery; VO2ren, renal oxygen consumption; O2ERren, renal oxygen extraction.

Chapter 5

74

FIGURE 3. Renal microvascular oxygen histograms during septic shock and resuscitation. (A) Cortical microvascular oxygen histograms, (B) Outer medullary microvascular oxygen histograms. Values represent mean ± SD *p < 0.05 vs. baseline. xp < 0.05 vs. control. �p < 0.05 vs. fluid resuscitation. T0, baseline; t1, endotoxemia; t2, resuscitation.

Iloprost

75

FIGURE 3 B. (Continued).

Chapter 5

76

Kidney Function, Tubular Sodium Resorption, and Metabolic Cost. There was no change in Clcrea, TNa+ and VO2/TNa+ relation over time in the control group (Fig. 4). At t1 all LPS-treated animals were anuric. This condition did not improve in the nonresuscitation group.

After fluid resuscitation, Clcrea was one third of baseline (p < 0.05 vs. control). The tubular sodium resorption was significantly reduced at t2 (p < 0.05 vs. control) and related to a significant increase in metabolic cost (VO2/TNa+).

Resuscitation with iloprost restored Clcrea to baseline values (0.77 ± 0.24 mL/ min/g). Furthermore, the tubular sodium resorption was normalized with no increase in VO2/TNa+ relation.

FIGURE 4. Kidney function, tubular sodium resorption, and metabolic cost during septic shock and resuscitation. Values represent mean ± SD. Changes in % between t0 and t2. *p < 0.05 vs. baseline. xp < 0.05 vs. control. �p < 0.05 vs. fluid resuscitation (1). Animal is anuric. Clcrea, creatinine clearance; TNa+, tubular sodium resorption; VO2/TNa+, oxygen consumption per sodium reabsorbed (metabolic cost).

Plasma NOx Concentration. A calibration curve for sodium nitrate was constructed to determine the NOx concentration (Fig. 5). In n = 4 animals, plasma NOx was measured hourly for 5 hours following LPS infusion. After 4 hours, NOx was significantly higher than at baseline. At t2 (5 hours) plasma NOx levels were significantly increased compared with time control. There was no change in plasma NOx concentration in the control group.

In the nonresuscitation group, NOx increased significantly from 17 ± 4 at baseline to 214 ± 67 µM/L at t2 (p < 0.05 vs. control).

Fluid resuscitation had no influence on plasma NOx levels (223 ± 61 µM/L at t2). In animals receiving resuscitation with iloprost, NOx was with 255 ± 20 µM/L at t2 highest of all experimental groups, but not significantly so.

Iloprost

77

FIGURE 5. Plasma NOx (nitrate/nitrite) levels during septic shock and resuscitation. Values represent mean ± SD *p < 0.05 vs. baseline. xp < 0.05 vs. control. T0, baseline; t1, endotoxemia; t2, resuscitation. (A) Measurement of plasma NOx 1–5 hours postlipopolysaccharide (n = 4). Insert showing calibration curve of sodium nitrate. (B) Plasma NOx in control and experimental group for three different time points. AUC, area under curve. iNOS mRNA Expression. The expression of renal iNOS mRNA was increased in all animals receiving LPS (Fig. 6). In the nonresuscitation group, this increase was significant in all regions of the kidney. In the fluid resuscitation and iloprost group, iNOS mRNA expression did not reach significance in the inner medulla (p < 0.05 vs. control). For direct comparison of iNOS mRNA expression in the three experimental and the control groups, hematoxylin and eosin-stained kidney slices are shown for cortex, outer, and inner medulla (Figs. 6B–D).

Chapter 5

78

FIGURE 6. Inducible nitric-oxide synthase (iNOS) messenger RNA (mRNA) expression of renal tissue determined by in situ hybridization. (A) iNOS mRNA expression in cortex, outer, and inner medulla. Results are expressed in percent of iNOS mRNA positive (% iNOS+) cells per 0.24 mm2 renal tissue. Values represent mean ± SD *p < 0.05 vs. control. (B–D) iNOS mRNA expression as typical cluster formation of silver grains seen in the nonresucitation, fluid resuscitation, and iloprost group in hematoxylin and eosin-stained renal tissue.

Iloprost

79

Discussion

In this study, we demonstrated that the prostaglandin analogue iloprost restored kidney function in a rat model of endotoxemia. The application of iloprost had beneficial effects on renal vascular resistance, normalized renal oxygen consumption, and preserved cortical and outer medullary microvascular oxygenation by preventing the occurrence of hypoxic regions.

Renal hypoperfusion because of intrarenal vasoconstriction is considered to be one of the most important pathogenetic factors in the development of ARF in sepsis (25). Therefore, the idea to use a vasodilator, like PGI2, for the treatment of sepsis-related renal failure is understandable. In a recently published study of endotoxemia, the effects of decreased renal PGI2 by cyclo-oxygenase inhibition and increased renal PGI2 with transgenic mice were investigated (26). This study clearly demonstrated that endogenous PGI2 contributed to renal protection in endotoxemia-induced acute kidney injury because of the observation that glomerular filtration rate significantly decreased when a cyclooxygenase inhibitor was given. In transgenic animals, however, the protective vasodilating effect was probably overridden by excessive activation of the renin-angiotensin system because of an increase in renal cyclic AMP and renin. These negative effects are possibly related to an excess in PGI2. Unfortunately, there are no data about plasma concentrations of PGI2.

On the basis of the findings and the postulated underlying mechanisms of this study, we wanted to investigate if exogenously given prostacylin in the form of iloprost can be used as a therapeutic strategy in the treatment of endotoxemia-related ARF.

In our model, rats received after 2 hours of LPS standard fluid resuscitation supplemented by a continuous infusion of 100 ng/kg/min iloprost. After 5 hours, the clearance of creatinine restored to baseline values in the iloprost group, whereas in animals receiving fluid resuscitation alone the creatinine clearance was only 30% of baseline. Similar results were shown in a study of endotoxin-shocked rabbits. Here, glomerular filtration rate was substantially maintained in animals receiving the PGI2 analogue taprostene (27). In our investigations, RBF only slightly increased when iloprost was infused, besides normalization in renal vascular resistance. We could recently demonstrate the appearance of cortical microcirculartory hypoxic areas in endotoxin-induced renal failure in the rat (8). The shift in the cortical oxygen histogram toward anoxia could also be shown in this study in nonresuscitated animals 5 hours after LPS infusion, and here also the outer medullary region was affected. This shift was attenuated in rats receiving iloprost resulting in no occurrence of hypoxic areas in both the cortical and outer medullary regions. This phenomenon might contribute to the improvement in kidney function. We could demonstrate that the tubular sodium resorption normalized in the iloprost group with no increase in renal oxygen consumption. Theses observations presuppose an adequate oxygenation in that part of the kidney where most of the oxygen is needed for the absorption of solute (28). The finding that the renal oxygen extraction increased in the iloprost group can be easily explained by the fact that the renal oxygen consumption significantly increased despite an only slight increase in RBF.

Septic ARF is characterized by severe intrarenal vasoconstriction in face of a profound vasodilatation in the systemic circulation (29). This vasodilation is mediated by the release of large amounts of NO after the induction of iNOS (30). As nitrate and nitrite are the primary oxidation products of NO, plasma NOx (nitrate/nitrite) levels can be used as an indicator of NO formation. Under physiologic conditions, constitutive NOS is involved in the regulation of the intrarenal blood flow and tubular resorption (31). During sepsis, renal expression of iNOS is increasing extensively

Chapter 5

80

followed by dysregulation of intrarenal vascular tone (2). However, many approaches of simply blocking NOS failed and had detrimental effects on kidney function (32, 33), whereas selective inhibition of iNOS seems to be beneficial (34, 35). In our model, there was a significant increase in plasma NOx levels 5 hours after LPS administration that did not differ in the different groups. With a more specific look at the kidney, we analyzed iNOS mRNA expression by in situ hybridization for cortex, outer, and inner medulla. iNOS expression was up-regulated after 5 hours of endotoxemia in all investigated zones of the kidney (36). In the iloprost group, iNOS expression was suppressed in the outer medulla compared with the nonresuscitation group. Whether in this region eNOS could have exhibited its inhibitory effects on the vasoconstrictive system (32) and therefore have normalized renal sodium resorption remains speculative. Potential interaction between iNOS and eNOS is postulated and supported by the finding that increased iNOS activity inhibits eNOS activity (37).

The renin–angiotensin–aldosterone system plays an important role in maintaining blood pressure, in electrolyte and fluid homeostasis of organisms, and depends on the concentration of the protease renin. Hyperactivation of systemic renin–angiotensin–aldosterone system during sepsis is well documented. However, LPS was demonstrated to significantly down-regulate intrarenal renin–angiotensin–aldosterone system through inhibition of renin activity (38). This observation could be related to the development of ARF in the context of endotoxemia. Endothelial autacoids such as PGI2 are known to stimulate renin secretion (26) and might therefore be beneficial in endotoxemia-releated ARF.

A potential application of iloprost in septic patients is given as iloprost is already in clinical use, for example, in the treatment of primary pulmonary hypertension (39). However, there are contradictory results in clinical studies where iloprost was administered in sepsis (40–43). To profit from the potential renal protective vasodilatory effect of iloprost, it is necessary to make sure that the patient receives an adequate fluid resuscitation to avoid hypoperfusion of vital organs. Inadequate filling might be a possible explanation why different clinical studies did not show uniformly positive effects of iloprost administration in sepsis. It is known that high plasma concentrations of prostacylin are associated with headache, nausea, and emesis (44). However, most of the patients in septic renal failure are sedated and mechanically ventilated, and, therefore, these side effects are negligible.

Although our study solely focused on the kidney, for potential clinical application, it is mandatory to consider the effects of iloprost to other organs. However, at least for the liver and lungs positive effects of PGI2 analogues have been reported in the literature. In a porcine model of endotoxemia, iloprost ameliorated hepatic metabolic disturbances, and thereby, hepatic energy balance (45). Furthermore, low-dose PGI2 had beneficial in LPS-induced inflammation in the rat, by reducing albumin leakage in the lung and improving blood oxygenation (46). Another limitation of our study is that it does not investigate long-term survival and outcome.

Our data demonstrate that the continuous infusion of the stable PGI2 analogue iloprost substantially restored kidney function in a rat model of endotoxemia. These observations were associated with preserved cortical and outer medullary microvascular oxygenation and a significant protection against the occurrence in microcirculatory hypoxic areas. On the basis of these results, one could hypothesize that iloprost infusion as adjuvant to standard fluid resuscitation might be useful as a renal protective strategy to preserve the renal oxygenation and kidneyfunction in the early stage of sepsis. Adequate fluid resuscitation should then avoid potential iloprost-induced hypotension and an early start of continuous iloprost infusion (because of the short plasma half-life

Iloprost

81

of 20 to 30 minutes) could restore kidney function and oxygenation before the onset of structural kidney damage like acute tubular necrosis. Of course, the true clinical value of iloprost in this setting and the prolonged beneficial effects on kidney function remain to be evaluated.

Acknowlegement

We thank Christof Zanke, Department of Anesthesiology and Critical Care, University of Tuebingen, Germany for writing the software to analyze the iNOS RNA expression.

References 1. Neveu H, Kleinknecht D, Brivet F, et al: Prognostic factors in acute renal failure due to sepsis. Results of a

prospective multicentre study. The French Study Group on Acute Renal Failure Nephrol Dial Transplant. Nephrol Dial Transplant 1996; 11:293–299

2. Schrier RW, Wang W: Acute renal failure and sepsis. N Engl J Med 2004; 351:159–169 3. Heyman SN, Darmon D, Goldfarb M, et al: Endotoxin-induced renal failure. I. A role for altered renal

microcirculation. Exp Nephrol 2000; 8:266–274 4. Wan L, Bellomo R, Di Giantomasso D, et al: The pathogenesis of septic acute renal failure. Curr Opin Crit Care

2003; 9:496–502 5. Badr KF: Sepsis-associated renal vasoconstriction: Potential targets for future therapy. Am J Kidney Dis 1992;

20:207–213 6. Nasrallah R, Hebert RL: Prostacyclin signaling in the kidney: Implications for health and disease. Am J Physiol

Renal Physiol 2005; 289:F235–F246 7. Wan L, Bagshaw SM, Langenberg C, et al: Pathophysiology of septic acute kidney injury: What do we really know?

Crit Care Med 2008; 36(Suppl 4):S198–S203 8. Johannes T, Mik EG, Ince C: Non-resuscitated endotoxemia induces microcirculatory hypoxic areas in the renal

cortex in the rat. Shock 2009; 31:97–103 9. Nasrallah R, Nusing RM, Hebert RL: Localization of IP in rabbit kidney and functional role of the PGI(2)/IP system in

cortical collecting duct. Am J Physiol Renal Physiol 2002; 283:F689–F698 10. Turker RK, Demirel E, Ercan ZS: Iloprost preserves kidney function against anoxia. Prostaglandins Leukot Essent

Fatty Acids 1988; 31:45–52 11. Mizutani A, Okajima K, Uchiba M, et al: Antithrombin reduces ischemia/reperfusioninduced renal injury in rats by

inhibiting leukocyte activation through promotion of prostacyclin production. Blood 2003; 101: 3029–3036 12. Isobe H, Okajima K, Uchiba M, et al: Antithrombin preventsendotoxin-induced hypotension by inhibiting the

induction of nitric oxide synthase in rats. Blood 2002; 99: 1638–1645 13. Spargias K, Adreanides E, Giamouzis G, et al: Iloprost for prevention of contrast-mediated nephropathy in high-risk

patients undergoing a coronary procedure. Results of a randomized pilot study. Eur J Clin Pharmacol 2006; 62:589–595

14. Johannes T, Mik EG, Nohe B, et al: Influence of fluid resuscitation on renal microvascular PO2 in a normotensive rat model of endotoxemia. Crit Care 2006; 10:R88

15. Johannes T, Mik EG, Ince C: Dual-wavelength phosphorimetry for determination of cortical and subcortical microvascular oxygenation in rat kidney. J Appl Physiol 2006; 100:1301–1310

16. Mik EG, Johannes T, Ince C: Monitoring of renal venous PO2 and kidney oxygen consumption in rats by a near-infrared phosphorescence lifetime technique. Am J Physiol Renal Physiol 2008; 294:F676–F681

17. Dunphy I, Vinogradov SA, Wilson DF: Oxyphor R2 and G2: Phosphors for measuring oxygen by oxygen-dependent quenching of phosphorescence. Anal Biochem 2002; 310: 191–198

18. Rozhkov V, Wilson D, Vinogradov S: Phosphorescent Pd porphyrin-dendrimers: Tuning core accessibility by varying the hydrophobicity of the dendritic matrix. Macromolecules 2002; 35:1991–1993

19. Vinogradov SA, Lo LW, Wilson DF: Dendritic polyglutamic porphyrins: Probing porphyrin protection by oxygen-dependent quenching of phosphorescence. Chem Eur J 1999; 5:1338–1347

20. Mik EG, Donkersloot C, Raat NJ, et al: Excitation pulse deconvolution in luminescence lifetime analysis for oxygen measurements in vivo. Photochem Photobiol 2002; 76:12–21

21. Golub AS, Popel AS, Zheng L, et al: Analysis of phosphorescence in heterogeneous systems using distributions of quencher concentration. Biophys J 1997; 73:452–465

22. Inoue RY, Gontijo JA, Franchini KG: Hemodilution mediates hemodynamic changes during acute expansion in unanesthetized rats. Am J Physiol Regul Integr Comp Physiol 2000; 279:R2243–R2251

Chapter 5

82

23. Yang BK, Vivas EX, Reiter CD, et al: Methodologies for the sensitive and specific measurement of S-nitrosothiols, iron-nitrosyls, and nitrite in biological samples. Free Radic Res 2003; 37:1–10

24. Szalay G, Sauter M, Hald J, et al: Sustained nitric oxide synthesis contributes to immunopathology in ongoing myocarditis attributable to interleukin-10 disorders. Am J Pathol 2006; 169:2085–2093

25. Koch T, Geiger S, Ragaller MJ: Monitoring of organ dysfunction in sepsis/systemic inflammatory response syndrome: Novel strategies. J Am Soc Nephrol 2001(Suppl 12) 17:S53–S59

26. Wang W, Zolty E, Falk S, et al: Prostacyclin in endotoxemia-induced acute kidney injury: Cyclooxygenase inhibition and renal prostacyclin synthase transgenic mice. Am J Physiol Renal Physiol 2007; 293: F1131–F1136

27. Schneider J: Beneficial effects of the prostacyclin analogue taprostene on cardiovascular, pulmonary and renal disturbances in endotoxin-shocked rabbits. Eicosanoids 1991; 4:99–105

28. Epstein FH, Agmon Y, Brezis M: Physiology of renal hypoxia. Ann NY Acad Sci 1994; 718:72–81 29. Thijs A, Thijs LG: Pathogenesis of renal failure in sepsis. Kidney Int Suppl 1998; 66: S34–S37 30. Tsao CM, Ho ST, Chen A, et al: Low-dose dexamethasone ameliorates circulatory failure and renal dysfunction in

conscious rats with endotoxemia. Shock 2004; 21:484–491 31. Mount PF, Power DA: Nitric oxide in the kidney: Functions and regulation of synthesis. Acta Physiol (Oxf) 2006;

187:433–446 32. Schwartz D, Mendonca M, Schwartz I, et al: Inhibition of constitutive nitric oxide synthase (NOS) by nitric oxide

generated by inducible NOS after lipopolysaccharide administration provokes renal dysfunction in rats. J Clin Invest 1997; 100:439–448

33. Schramm L, Weierich T, Heldbreder E, et al: Endotoxin-induced acute renal failure in rats: Effects of L-arginine and nitric oxide synthase inhibition on renal function. J Nephrol 2005; 18:374–381

34. Liaudet L, Fishman D, Markert M, et al: L-canavanine improves organ function and tissue adenosine triphosphate levels in rodent endotoxemia. Am J Respir Crit Care Med 1997; 155:1643–1648

35. Tiwari MM, Brock RW, Megyesi JK, et al: Disruption of renal peritubular blood flow in lipopolysaccharide-induced renal failure: Role of nitric oxide and caspases. Am J Physiol Renal Physiol 2005; 289: F1324 –F1332

36. Holmqvist B, Olsson CF, Svensson ML, et al: Expression of nitric oxide synthase isoforms in the mouse kidney: Cellular localization and influence by lipopolysaccharide and Tolllike receptor 4. J Mol Histol 2005; 36: 499–516

37. Rengasamy A, Johns RA: Regulation of nitric oxide synthase by nitric oxide. Mol Pharmacol 1993; 44:124–128 38. Almeida WS, Maciel TT, Di Marco GS, et al: Escherichia coli lipopolysaccharide inhibits renin activity in human

mesangial cells. Kidney Int 2006; 69:974–980 39. Hoeper MM, Schwarze M, Ehlerding S, et al: Long-term treatment of primary pulmonary hypertension with

aerosolized iloprost, a prostacyclin analogue. N Engl J Med 2000; 342:1866–1870 40. Kiefer P, Tugtekin I, Wiedeck H, et al: Hepato-splanchnic metabolic effects of the stable prostacyclin analogue

iloprost in patients with septic shock. Intensive Care Med 2001; 27:1179–1186 41. Lehmann C, Taymoorian K, Wauer H, et al: Effects of the stable prostacyclin analogue iloprost on the plasma

disappearance rate of indocyanine green in human septic shock. Intensive Care Med 2000; 26:1557–1560 42. Radermacher P, Buhl R, Santak B, et al: The effects of prostacyclin on gastric intramucosal pH in patients with

septic shock. Intensive Care Med 1995; 21:414–421 43. Hannemann L, Reinhart K, Meier-Hellmann A, et al: Prostacyclin in septic shock. Chest 1994; 105:1504–1510 44. Stiebellehner L, Petkov V, Vonbank K, et al: Long-term treatment with oral sildenafil in addition to continuous IV

epoprostenol in patients with pulmonary arterial hypertension. Chest 2003; 123:1293–1295 45. Trager K, Matejovic M, Zulke C, et al: Hepatic O2 exchange and liver energy metabolism in hyperdynamic porcine

endotoxemia: Effects of iloprost. Intensive Care Med 2000; 26: 1531–1539 46. Dubniks M, Grande PO: The effects of activated protein C and prostacyclin on arterial oxygenation and protein

leakage in the lung and the gut under endotoxaemia in the rat. Acta Anaesthesiol Scand 2008; 52:381–387

83

CHAPTER ACTIVATED PROTEIN C RESTORES KIDNEY FUNCTION IN A DOSE-DEPENDENT MANNER IN ENDOTOXIN-INDUCED ACUTE RENAL FAILURE IN THE RAT

Tanja Johannes, Emre Almac, Egbert G Mik, Matthieu Legrand, Klaus E Unertl, and Can Ince Submitted for publication

Chapter 6

84

Abstract

Introduction Activated protein C (APC) has been shown to have beneficial effects on the inflammatory process and coagulation during sepsis. Inflammation and coagulopathy impair the microvasculature and therefore disturb oxygen transport to the tissue. The hypothesis of our study was that APC-treatment improves renal microvascular oxygenation and kidney function in endotoxin-induced acute renal failure in the rat. Methods In 21 anesthetized and ventilated (FiO2 0.5) male Wistar rats arterial blood pressure and renal blood flow were recorded. The renal microvascular PO2 was continuously measured by phosphorescence lifetime technique. All animals received a LPS-bolus (10mg/kg) to induce endotoxemic shock. All rats received fluid resuscitation (HES 130kD) 1h after LPS-application. In one group of animals APC (Drotrecogin Alpha, Xigris®, Eli-Lilly) was continuously infused in a concentration of 10µg/kg/h. Another group received a continuous infusion of 100µg/kg/h APC. Results After LPS-bolus MAP and RBF progressively dropped to 40% and 60% of baseline at 1h respectively. Treatment with APC 100 in addition to fluid resuscitation prevented a further decline of both parameters. APC, independently of concentration, had no significant effects on average µPO2, but prevented the occurrence of cortical microcirculatory hypoxic areas. All animals had a significant decrease in CLcrea 1h after LPS. Only in animals receiving APC 100, CLcrea was significantly restored at the end of the experiment (3h). Conclusion APC 100 significantly restored kidney function compared to standard fluid resuscitation during endotoxemia. This was accompanied by protection against the occurrence of cortical microcirculatory hypoxic areas. Furthermore, this application best improved MAP.

Activated Protein C

85

Introduction

The prevalence of acute renal failure (ARF) in sepsis is high (approximately 40%) and the mortality reaches up to 75% in ICU patients with septic shock (1, 2). The pathogenesis of septic renal failure remains unclear, however, and current strategies to overcome renal dysfunction are mainly supportive rather then curative (3).

There is growing evidence that microcirculatory dysfunction accompanied by tissue dysoxia might play a key role in the development of septic ARF (4, 5). An inappropriate release of pro-inflammatory mediators is as well involved in the pathogenesis of sepsis as disturbances in the coagulation system, both leading to microcirculatory dysfunction with consecutive organ failure (6, 7).

It has been shown that the reduction in plasma levels of protein C is associated with an increased risk of death in patients with sepsis (8, 9). Activated protein C (APC) is an important endogenous protein that modulates coagulation and inflammation by promoting fibrinolysis and inhibiting thrombosis and inflammation (10, 11). Different experimental and clinical studies could demonstrate that APC improved outcome of severe sepsis (8, 12-15). Recently two studies showed beneficial effects of APC in acute kidney injury in a rat model of LPS-induced (16), and CLP-related renal failure (17). Furthermore activated protein C reduced ischemia/reperfusion-induced renal injury in rats (18).

Based on the physiological actions of activated protein C and the recently demonstrated beneficial effects of APC on renal endothelial dysfunction we hypothesized that recombinant human activated protein C might have a potential role in a renal protective strategy in sepsis. We therefore investigated in a rat model of endotoxin-induced acute renal failure 1) the influence of APC treatment on systemic and regional hemodynamics 2) the effects of APC on the renal microvascular oxygenation and 3) the influence of APC on renal function. Furthermore our study should unveil a possible relation between renal dysfunction and microvascular hypoxia. Material and methods

Experiments were approved by the Animal Research Committee of the Academic Medical Center at the University of Amsterdam and conducted in accordance with the guidelines for Institutional and Animal Care and Use Committees (IACUC) care and handling of the animals. The experiments were performed in 23 Wistar male rats (Harlan, Horst, The Netherlands) with a body weight of 318 ± 37 g.

Animal preparation and monitoring. Rats were anesthetized by intraperitonal injection of a mixture of 90mg/kg ketamine, 0.5mg/kg medetomidine and 0.05mg/kg atropine-sulphate. Animals were mechanically ventilated with a FiO2 of 0.5 via tracheostomy. The right carotid artery was cannulated and used for monitoring of arterial blood pressure and heart rate. A catheter in the right femoral artery was used for blood gas sampling. The right femoral vein was cannulated for continuous infusion of Ringer’s lactate (15mL/kg/h; Baxter, Utrecht, The Netherlands) and ketamine (50mg/kg/h; Nimatek®; Eurovet, Bladel, The Netherlands). The animal’s temperature was maintained at 37°C. The left kidney was exposed via flank incision, decapsulated and immobilized in a Lucite kidney cup. A perivascular flow probe connected to a flow meter (T206; Transonic Systems Inc., Ithaca, NY, USA) was used to measure renal blood flow (RBF). The left ureter was isolated, ligated and cannulated for urine collection. At the end of the experiment the left kidney was removed and weighed.

Chapter 6

86

Experimental protocol. The experimental schema is demonstrated in Figure 1.

surgical procedure and Oxyphor G2 infusion (=2h)

LPS

bolu

s (10

mg/

kg)

fluid resuscitation 5 ml/kg bolus ! 5ml/kg/h infusion± APC infusion (10 !g/kg/h) or (100 !g/kg/h)

2h1h

t0 t1 t2

FIGURE 1. Experimental schema. At start of the experiment rats were randomized between control (n=5), fluid resuscitation (n=6), APC 10 (n=6) and APC 100 (n=6) groups. All animals receiving APC received standard fluid resuscitation with hydroxyethl starch (HES 130 kD). At the end of surgery (1h) an intravenous infusion of Oxyphor G2 (10 mg/kg in 15 min; Oxygen Enterprises Ltd., Philadelphia, PA, USA), necessary for the oxygen measurements, was started. Forty minutes later µPO2 and rvPO2 were continuously recorded via two optical fibers placed both 1mm above the kidney surface and 1mm above the renal vein. Then, the baseline blood sample (0.4mL) was taken. In 18 rats a bolus of lipopolysaccharide (LPS, 10 mg/kg; serotype 0127:B8, Sigma-Aldrich, Zwijndrecht, The Netherlands) was given to induce septic shock. One hour after the LPS-bolus three groups of animals received fluid resuscitation (5 mL/kg followed by 5 mL/kg/h; Voluven®, 6 % HES 130/0.4; Fresenius Kabi, Schelle, Belgium) for two hours. In addition to the fluid resuscitation, in one group of animals APC (10 µg/kg/h; recombinant human activated protein C; Drotrecogin Alpha, Xigris®, Eli-Lilly Indianapolis, IN, USA) was continuously infused. Another group received APC in a concentration of 100 µg/kg/h. All animals received the same fluid volume. The experiment was ended 10 min after stop of treatment or a corresponding time point for the control groups.

Measurement of renal microvascular oxygenation and renal venous PO2. The renal microvascular PO2 (µPO2) within the kidney cortex and outer medulla were measured by oxygen-dependent quenching of phosphorescence using a dual-wavelength-phosphorimeter (19). The renal venous PO2 (rvPO2) was detected by the same method (20). Briefly described, Oxyphor G2 (Oxygen Enterprises Ltd., Philadelphia, PA, USA), intravenously infused, binds to albumin (21). If excited by a flash of light (wavelength ~ 440 or 630 nm) the Oxyphor G2-albumin complex emits phosphorescence (wavelength ~800nm). Dependent on the oxygen concentration the phosphorescence intensity decreases and the relationship between the measured decay-time and the PO2 can be estimated using the Stern-Volmer relation (22-24). Heterogeneity in oxygen pressure was analyzed by fitting a sum of small rectangular distributions to the distributions of quencher concentration in the phosphorescence data (5, 25, 26).

Activated Protein C

87

Blood gas measurements. An arterial blood sample (0.4 mL) was taken via the femoral artery at three different time points. First time point: 0 minutes = baseline (t0), second time point: 1h after LPS-bolus (t1) and third time point: 3h after LPS-bolus or equivalent time (t2). The same volume of hydroxyethyl starch replaced the sampling volume of blood. Samples were analyzed for determination of blood gas values (ABL505 blood gas analyzer; Radiometer, Copenhagen, Denmark), hematocrit, Hb-concentration, and HbSO2 (OSM3; Radiometer, Copenhagen, Denmark). Furthermore in each sample the plasma creatinine concentration was measured.

Calculations of renal oxygenation. Renal oxygen delivery (DO2ren) was calculated as RBF x arterial O2-content: RBF x (1.31 x Hb x SaO2) + (0.003 x PaO2). Renal oxygen consumption (VO2ren) was calculated as RBF x (arterial – renal venous O2-content difference). Renal venous O2-content is: (1.31 x Hb x SrvO2) + (0.003 x rvPO2) (20). An estimation of the vascular resistance (RVR) of the renal artery flow region was made: RVR = MAP /RBF.

Measurement of renal function. Creatinine clearance (CLcrea) was assessed as an index of glomerular filtration rate (GFR). Calculations of the clearance were done with standard formula: CLcrea (mL/min) = (U x V) /P, where U is the urine creatinine concentration, V is the urine volume per unit time and P is the plasma creatinine concentration. The specific elimination capacity for creatinine of the left kidney was normalized to the organ weight. For analysis of urine volume and creatinine concentration urine samples from the left ureter were collected at 10-min intervals. At the midpoint of each 10-min-interval plasma creatinine concentration was analyzed. Analysis of samples was performed using Jaffé method. Data presentation and statistics. Values are presented as mean ± SEM, unless otherwise indicated. Analysis of the monoexponential fit procedures of the phosphorescence curves was performed using Labview 6.1 software (National Instruments, Austin, TX, USA). For PO2 histogram recovery and statistics GraphPad Prism version 4.0 for Windows (GraphPad Software, San Diego, CA, USA) was used. For data analysis within each group and intergroup differences two-way ANOVA for repeated measurements with Bonferroni post test was performed. P values < 0.05 were considered significant.

Results

Systemic and regional variables. Table 1 is showing the data for the measured systemic and regional variables. There was no statistically significant difference in baseline readings between the different groups. One hour (t1) after LPS-bolus mean arterial pressure (MAP) was in all group significantly reduced compared to baseline and control. There was no change in heart rate (HR) at t1. After LPS-bolus renal blood flow (RBF) progressively decreased and was at t1 in all experimental groups, reduced by more than 60%. At t1 renal vascular resistance (RVR) was significantly increased in all experimental groups compared to baseline and/or control. Between t1 and t2 MAP was not significantly different in the experimental groups although the animals receiving APC 100 showed a slight recovery in MAP at t2 compared to t1. When compared to baseline HR was at t2 significantly higher in the experimental

Chapter 6

88

groups. At t2 RBF was in all experimental groups significantly reduced compared to baseline and control. RVR was normalized after receiving resuscitation (t2).

MAP (mmHg)

HR (bpm)

RBF (ml/min)

RVR

(dyne/sec/cm5) rvPO2

(mmHg)

TC (n = 5)

t0 109 ± 1 254 ± 12 6.3 ± 0.1 17 ± 1 74 ± 4

t1 112 ± 7 261 ± 11 5.8 ± 0.3 19 ± 1 73 ± 2

t2 112 ± 5 267 ± 16 6.0 ± 0.3 19 ± 1 64 ± 5

LPS + FR (n = 6)

t0 102 ± 5 247 ± 16 7.2 ± 0.2 14 ± 1 69 ± 6

t1 71 ± 10*† 275 ± 15 2.9 ± 0.8*† 35 ± 12* 52 ± 6*†

t2 67 ± 11*† 296 ± 9* 3.9 ± 0.7*† 19 ± 4 42 ± 3*†

APC 10 (n = 6)

t0 101 ± 4 257 13 6.8 ± 0.7 16 ± 1 67 ± 4

t1 62 ± 8*† 272 ± 18 3.3 ± 0.8*† 27 ± 6* 58 ± 4

t2 58 ± 7*† 296 ± 15* 3.6 ± 0.5*† 19 ± 4 43 ± 3*†

APC 100 (n = 6)

t0 102 ± 1 262 ± 7 6.3 ± 0.3 16 ± 1 65 ± 7

t1 73 ± 7*† 291 ± 22 2.7 ± 0.6*† 35 ± 7*† 47 ± 7*†

t2 80 ± 7*† 320 ± 17*† 3.9 ± 0.3*† 21 ± 1 41 ± 4*†

Values represent mean ± SEM. * P < 0.05 vs baseline; † P < 0.05 vs control. TC = time control; LPS + FR = group receiving lipopolysaccharide + fluid resuscitation; APC 10 = treatment with 10 !g/kg/h activated protein C; APC 100 = treatment with 100 !g/kg/h activated protein C. Animals receiving APC also received LPS + FR. t0 = baseline; t1 = 1h after LPS-bolus; t2 = 3 h after LPS-bolus or equivalent time. MAP = mean arterial pressure; HR = heart rate; RBF = renal blood flow; RVR = renal vascular resistance; rvPO2 = renal venous PO2.

TABLE. 1 Systemic and regional variables

Renal oxygenation. Table 1 and Figure 2 are showing the measured renal oxygenation parameters. There was no statistically significant difference in baseline readings between the different groups. Average cortical µPO2 did significantly drop at t1. In the fluid resuscitation and the APC 100 group the rvPO2 was significantly reduced compared to baseline and control after LPS at t1 (Table 1).

At t2 cµPO2 and mµPO2 were in all groups significantly lower than at baseline and in the APC 10 group significant for the cµPO2 compared to control. The reduction in the average µPO2 after LPS-bolus was accompanied by a significant left shift in the cortical oxygen histogram of the experimental groups (Fig. 3A) at t2. This left shift was not present in the outer medullary region (Fig. 3B). In

Activated Protein C

89

animals receiving APC 10 or 100 there was a significant reduction in the appearance of cortical microcirculatory hypoxic areas compared to the fluid resuscitation group.

The renal venous PO2 dropped in all experimental groups significantly compared to baseline and control at t2. The renal oxygen delivery (DO2ren) was at t1 and t2 in all experimental groups significantly reduced compared to baseline and control (Fig. 4A). In contrast, the renal oxygen consumption (VO2ren) did not significantly change during endotoxemia and resuscitation (Fig. 4B). At the end of the experiment (t2) there was a significant increase in renal oxygen extraction (O2ERren) in all groups compared to baseline and control (Fig. 4C).

FIGURE 2. Measured renal microvascular oxygenation. Values represent mean ± SEM. * P < 0.05 vs baseline. † P < 0.05 vs control. LPS + FR = group receiving lipopolysaccharide + fluid resuscitation; APC 10 = treatment with 10 µg/kg/h activated protein C, APC 100 = treatment with 100 µg/kg/h activated protein C. Animals receiving APC also received LPS + FR. t0 baseline; t1 1h after LPS-bolus; t2 3h after LPS-bolus or equivalent time.

Chapter 6

90

FIGURE 3. Renal microvascular oxygen histograms. Panel A cortical oxygen histogram. Values represent mean ± SEM. * P < 0.05 vs baseline. † P < 0.05 vs control. # P < 0.05 vs FR. LPS + FR = group receiving lipopolysaccharide + fluid resuscitation; APC 10 = treatment with 10 µg/kg/h activated protein C, APC 100 = treatment with 100 µg/kg/h activated protein C. Animals receiving APC also received LPS + FR. t0 baseline; t1 1h after LPS-bolus; t2 3h after LPS-bolus or equivalent time. µPO2 = microvascular PO2.

A

Activated Protein C

91

FIGURE 3. Renal microvascular oxygen histograms. Panel B outer medullary oxygen histogram.

B

Chapter 6

92

FIGURE 4. Calculated renal oxygenation parameters. Values represent mean ± SEM. * P < 0.05 vs baseline. † P < 0.05 vs control. LPS + FR = group receiving lipopolysaccharide + fluid resuscitation; APC 10 = treatment with 10 µg/kg/h activated protein C, APC 100 = treatment with 100 µg/kg/h activated protein C. Animals receiving APC also received LPS + FR. t0 baseline; t1 1h after LPS-bolus; t2 3h after LPS-bolus or equivalent time. DO2ren = renal oxygen delivery; VO2ren = renal oxygen consumption; O2ERren = renal oxygen extraction. Kidney function. There was no significant change in creatinine clearance (CLcrea) in the time control. All animals of the experimental groups had a significant reduction in CLcrea and decrease in urine output after LPS-bolus (t1). All treated animals started to increase urine volume after start of resuscitation. Only in the APC 100 group CLcrea was totally restored to baseline compared to animals receiving fluid resuscitation alone (Fig. 5).

FIGURE 5. Kidney function. Values represent mean ± SEM. * P < 0.05 vs baseline. † P < 0.05 vs control. # P < 0.05 vs FR. LPS + FR = group receiving lipopolysaccharide + fluid resuscitation; APC 10 = treatment with 10 µg/kg/h activated protein C, APC 100 = treatment with 100 µg/kg/h activated protein C. Animals receiving APC also received LPS + FR. t0 baseline; t1 1h after LPS-bolus; t2 3h after LPS-bolus or equivalent time. CLcrea = creatinine clearance.

Activated Protein C

93

Discussion

In the present study we demonstrated that recombinant human activated protein C in a concentration of 100µg/kg/h restored kidney function in a rat model of endotoxemia. The application of APC 100 added to standard fluid resuscitation maintained mean arterial pressure and preserved cortical microvascular oxygenation by preventing the occurrence of hypoxic areas.

Renal hypoperfusion is considered to be one of the most important pathogenetic factors in the development of ARF in sepsis (27). Disturbances in the autoregulation of microvessels, endothelial damage, tubular necrosis or disseminated fibrin thrombi leading to blockade of the intrarenal microvasculature (28, 29). The serine protease protein C plays an important role in controlling thrombosis and inflammation. Furthermore it could be

demonstrate that protein C exhibits cyto-protective effects (11, 30). There is evidence that reduction in plasma levels of protein C is prognostic for outcome in sepsis (31). In a recent study of LPS-induced kidney injury the ability of activated protein C to restore impaired renal hemodynamics could be shown. It is postulated that this restoration was due to suppression of local NO and the angiotensin system, as well as inhibition of leukocyte-endothelial cell interactions (16). Another study of the same group could demonstrate a clear link between an acquired protein C deficiency and renal dysfunction in a rat model of polymicrobial sepsis (17).

Based on the findings of these previous studies and the known acting mechanisms of protein C we wanted to investigate, if recombinant human activated protein C in addition to standard fluid resuscitation can be used as a therapeutic strategy in the treatment of endotoxin-related ARF. In the present study, rats received 1h after LPS-bolus standard fluid resuscitation supplemented by continuous infusion of 10 or 100µg/kg/h APC. After 3h the clearance of creatinine restored to baseline values in the APC 100 group. In animals receiving fluid resuscitation alone the creatinine clearance was only 50% of baseline. In our investigations neither APC 10 nor APC 100 could significantly restore mean arterial pressure or renal blood flow. Similar findings for mean arterial pressure were demonstrated for endotoxin-induced shock in rabbits; here APC was infused in a concentration of 160µg/kg/h (32). Recently, we showed the occurrence of cortical microcirculartory hypoxic areas in endotoxin-induced renal failure in the rat (5). In that study a shift in the cortical oxygen histogram toward anoxia could be demonstrated in non-resuscitated animals 5h after LPS-infusion. Such a shift in the cortical oxygen histogram was also present in the fluid resuscitation group of our current study and was attenuated in rats receiving APC 10 or 100.

There is not much published data about APC in sepsis-induced ARF. We did not investigate the direct acting mechanisms of APC in our present study. However, the results of our study and the previous mentioned study of Gupta et al. (16) are two fitting pieces in the puzzle of explaining the changes in the microvascular oxygen histograms and in understanding more about the pathophysiology of septic ARF. Septic ARF is characterized by severe intrarenal vasoconstriction in face of a profound vasodilatation in the systemic circulation (28). These mechanisms can be explained by activation of the renin-angiotensin-system (RAS) in counteraction to peripheral vasodilation mediated by NO. In the septic kidney RAS activation is leading to an aggravation of renal hypoperfusion via vasoconstriction by angiotensin II (ANG II) (33). Gupta et al. could demonstrate that APC treatment blocked the LPS-induced activation of the renal RAS and suppressed the induction of renal iNOS. Using two-photon intravital microscopy they showed that this was accompanied by improved intrarenal blood flow and reduced leukocyte adhesion. For our

Chapter 6

94

observations this could mean that counteracting the vasoconstrictive action of ANG II with APC might explain the protection against the occurrence of microcirculatory hypoxic areas in the renal cortex. This is in agreement with our previously published speculation that the peritubular capillary region of the cortex can be the anatomical correlate to the observed hypoxic areas (5). In their work Gupta et al. could furthermore demonstrate that treatment with APC reduced peritubular dysfunction. Peritubular capillary dysfunction is an early event in endotoxemia that contributes to tubular stress and renal injury (34). The interpretation of the above-demonstrated results emphasizes once more the possible important role of hypoxia in the development of acute renal failure in sepsis.

When interpreting our findings, the reader should keep in mind some limitations of our animal model. Such models can never be directly translated to the human pathophysiology and the results of animal studies have to be interpreted carefully. Another limitation of our study is that it does not investigate long-term survival and outcome. However, the aim of our study was to investigate the acute effects of APC treatment in a model of endotoxemic shock in which treatment is started in the presence of severe hemodynamic alterations like seen in clinical practice. Early goal-directed therapy using fluid resuscitation to prevent hypoperfusion of vital organs is the standard therapy in sepsis (35). Therefore, fluid resuscitation was the primary therapeutic approach in our model after profound septic shock had developed. So treatment with activated protein C supplemented standard fluid resuscitation. In contrast to a recently published protocol of our group (36), we gave 10 mg/kg LPS as a bolus resulting in a much more severe model of endotoxemic shock. As the lethality of non-resuscitated animals reached more than 50% such group was not included in the present study. Due to the lack of a steady state of creatinine balance, changes in CLcrea as an index of glomerular filtration should only be regarded as a gross indication.

Our data demonstrate that the continuous infusion of the activated protein C in a concentration of 100µg/kg/h substantially restored kidney function in a rat model of endotoxemia. This observation was associated by preserved cortical microvascular oxygenation and a significant protection against the occurrence of microcirculatory hypoxic areas. Based on these results one could hypothesize that APC 100 infusion as adjuvant to standard fluid resuscitation might be useful as a renal protective strategy to preserve renal oxygenation and kidney function in the early stage of sepsis.

Acknowledgement

We acknowledge the support by Eli Lilly Co. in form of an educational grant to Can Ince.

References 1. Neveu H, Kleinknecht D, Brivet F, Loirat P, Landais P: Prognostic factors in acute renal failure due to sepsis.

Results of a prospective multicentre study. The French Study Group on Acute Renal Failure. Nephrol. Dial. Transplant. 1996; 11(2):293-9.

2. Oppert M, Engel C, Brunkhorst FM et al.: Acute renal failure in patients with severe sepsis and septic shock--a significant independent risk factor for mortality: results from the German Prevalence Study. Nephrol. Dial. Transplant. 2008; 23(3):904-9.

3. Palevsky PM, Zhang JH, O'Connor TZ et al.: Intensity of renal support in critically ill patients with acute kidney injury. N. Engl. J. Med. 2008; 359(1):7-20.

4. Molitoris BA, Sutton TA: Endothelial injury and dysfunction: role in the extension phase of acute renal failure. Kidney Int. 2004; 66(2):496-9.

5. Johannes T, Mik EG, Ince C: Non-resuscitated endotoxemia induces mircocirculatory hypoxic areas in the renal cortex in the rat. Shock 2009; 31(1):97-103.

Activated Protein C

95

6. Linas SL, Whittenburg D, Repine JE: Role of neutrophil derived oxidants and elastase in lipopolysaccharide-mediated renal injury. Kidney Int. 1991; 39(4):618-23.

7. Schrier RW, Wang W: Acute renal failure and sepsis. N. Engl. J. Med. 2004; 351(2):159-69. 8. Bernard GR, Vincent JL, Laterre PF et al.: Efficacy and safety of recombinant human activated protein C for severe

sepsis. N. Engl. J. Med. 2001; 344(10):699-709. 9. Fourrier F: Recombinant human activated protein C in the treatment of severe sepsis: an evidence-based review.

Crit Care Med 2004; 32(11 Suppl):S534-S541. 10. Bernard GR: Drotrecogin alfa (activated) (recombinant human activated protein C) for the treatment of severe

sepsis. Crit. Care Med. 2003; 31(1 Suppl):S85-S93. 11. Esmon CT: Protein C anticoagulant pathway and its role in controlling microvascular thrombosis and inflammation.

Crit. Care Med. 2001; 29(7 Suppl):S48-S51. 12. Favory R, Lancel S, Marechal X, Tissier S, Neviere R: Cardiovascular protective role for activated protein C during

endotoxemia in rats. Intensive Care Med 2006; 32(6):899-905. 13. Grinnell BW, Joyce D: Recombinant human activated protein C: a system modulator of vascular function for

treatment of severe sepsis. Crit. Care Med. 2001; 29(7 Suppl):S53-S60. 14. Hoffmann JN, Vollmar B, Laschke MW et al.: Microhemodynamic and cellular mechanisms of activated protein C

action during endotoxemia. Crit. Care Med. 2004; 32(4):1011-7. 15. Taylor FB, Jr., Chang A, Esmon CT, D'Angelo A, Vigano-D'Angelo S, Blick KE: Protein C prevents the

coagulopathic and lethal effects of Escherichia coli infusion in the baboon. J. Clin. Invest. 1987; 79(3):918-25. 16. Gupta A, Rhodes GJ, Berg DT, Gerlitz B, Molitoris BA, Grinnell BW: Activated protein C ameliorates LPS-induced

acute kidney injury and downregulates renal INOS and angiotensin 2. Am. J. Physiol. Renal Physiol. 2007; 293(1):F245-F254.

17. Gupta A, Berg DT, Gerlitz B et al.: Role of protein C in renal dysfunction after polymicrobial sepsis. J. Am. Soc. Nephrol. 2007; 18(3):860-7.

18. Mizutani A, Okajima K, Uchiba M, Noguchi T: Activated protein C reduces ischemia/reperfusion-induced renal injury in rats by inhibiting leukocyte activation. Blood 2000; 95(12):3781-7.

19. Johannes T, Mik EG, Ince C: Dual-wavelength phosphorimetry for determination of cortical and subcortical microvascular oxygenation in rat kidney. J. Appl. Physiol. 2006; 100(4):1301-10.

20. Mik EG, Johannes T, Ince C: Monitoring of renal venous PO2 and kidney oxygen consumption in rats by a near-infrared phosphorescence lifetime technique. Am. J. Physiol. Renal Physiol. 2008; 294(3):F676-F681.

21. Dunphy I, Vinogradov SA, Wilson DF: Oxyphor R2 and G2: phosphors for measuring oxygen by oxygen-dependent quenching of phosphorescence. Anal. Biochem. 2002; 310(2):191-8.

22. Rietveld IB, Kim E, Vinogradov SA: Dendrimers with tetrabenzoporphyrin cores: near infrared phosphors for in vivo oxygen imaging. Tetrahedron 2003; 59:3821-31.

23. Rozhkov V, Wilson D, Vinogradov S: Phosphorescent Pd porphyrin-dendrimers: Tuning core accessibility by varying the hydrophobicity of the dendritic matrix. Macromolecules 2002; 35:1991-3.

24. Vinogradov SA, Lo LW, Wilson DF: Dendritic polyglutamic porphyrins: Probing porphyrin protection by oxygen-dependent quenching of phosphorescence. Chem.–Eur. J. 1999; 5:1338-47.

25. Golub AS, Popel AS, Zheng L, Pittman RN: Analysis of phosphorescence in heterogeneous systems using distributions of quencher concentration. Biophys. J. 1997; 73(1):452-65.

26. Mik EG, Johannes T, Zuurbier CJ et al.: In vivo mitochondrial oxygen tension measured by a delayed fluorescence lifetime technique. Biophys. J. 2008; 95(8):3977-90.

27. Koch T, Geiger S, Ragaller MJ: Monitoring of organ dysfunction in sepsis/systemic inflammatory response syndrome: novel strategies. J Am Soc Nephrol 2001; 12 Suppl 17:S53-S59.

28. Thijs A, Thijs LG: Pathogenesis of renal failure in sepsis. Kidney Int. Suppl. 1998; 66:S34-S37. 29. Wan L, Bagshaw SM, Langenberg C, Saotome T, May C, Bellomo R: Pathophysiology of septic acute kidney injury:

what do we really know? Crit. Care Med. 2008; 36(4 Suppl):S198-S203. 30. Joyce DE, Gelbert L, Ciaccia A, DeHoff B, Grinnell BW: Gene expression profile of antithrombotic protein c defines

new mechanisms modulating inflammation and apoptosis. J. Biol. Chem. 2001; 276(14):11199-203. 31. Fisher CJ, Jr., Yan SB: Protein C levels as a prognostic indicator of outcome in sepsis and related diseases. Crit.

Care Med. 2000; 28(9 Suppl):S49-S56. 32. Roback MG, Stack AM, Thompson C, Brugnara C, Schwarz HP, Saladino RA: Activated protein C concentrate for

the treatment of meningococcal endotoxin shock in rabbits. Shock 1998; 9(2):138-42. 33. Burns KD, Homma T, Harris RC: The intrarenal renin-angiotensin system. Semin. Nephrol. 1993; 13(1):13-30. 34. Wu L, Tiwari MM, Messer KJ et al.: Peritubular capillary dysfunction and renal tubular epithelial cell stress following

lipopolysaccharide administration in mice. Am. J. Physiol. Renal Physiol. 2007; 292(1):F261-F268. 35. Rivers E, Nguyen B, Havstad S et al.: Early goal-directed therapy in the treatment of severe sepsis and septic

shock. N. Engl. J. Med. 2001; 345(19):1368-77. 36. Johannes T, Mik EG, Nohe B, Raat NJ, Unertl KE, Ince C: Influence of fluid resuscitation on renal microvascular

PO2 in a normotensive rat model of endotoxemia. Crit. Care 2006; 10(3):R88.

97

CHAPTER SUMMARY AND CONCLUSION

Chapter 7

98

Introduction

The development of acute renal failure (ARF) in sepsis is frequent, with a prevalence up to 40%. However, its pathogenesis remains only partially understood. To date, no specific renal protective or treatment modalities are available. Therefore the treatment of intensive care unit patients with severe acute kidney injury (AKI) often results in supportive therapy like costly renal-replacement. Renal tissue hypoxia may be a contributing factor in the progression of kidney failure. The mainstay of current conservative therapeutic interventions is aimed at increasing kidney perfusion and maintaining renal function by, for example, fluid expansive therapy. However, measures that enhance kidney perfusion have the potential to exacerbate kidney hypoxia by increasing oxygen consumption. Therefore, these clinical interventions may contribute to the development of AKI instead of preventing its progression. Because of this paradox, it is of utmost clinical importance to gain further insight into the patho-physiology of ARF and to find more specific therapeutic interventions aimed at reversing ARF and preventing the development of AKI.

Thesis

In this thesis we assumed that hypoxia and microcirculatory dysfunction play a role in the pathogenesis of septic renal failure and hypothesized that strategies aimed on preventing hypoxia and microcirculatory dysfunction result in maintained kidney function. To this end we established a rat model of endotoxin-induced ARF with signs of severe hemodynamic alterations like those seen in clinical practice. Early goal-directed therapy using fluid resuscitation to prevent hypoperfusion of vital organs is the standard therapy in sepsis. Therefore, fluid resuscitation was the primary therapeutic approach in all our studies. In addition to standard fluid resuscitation different treatment strategies were performed. All substances we tested are in clinical use and act on regulating vascular tone and/or secondarily influencing microcirculatory blood flow. Besides evaluating renal function by means of creatinine clearance we established a method that allowed comprehensive measurements of renal oxygenation. Using oxygen-dependent quenching of phosphorescence we were able to simultaneously measure cortical and outer-medullary microvascular PO2 (µPO2) and renal venous PO2. Recovery of microvascular PO2 histograms from cortex and outer medulla allowed us to study heterogeneity of intrarenal oxygenation.

Proof of thesis There is evidence that tissue hypoxia and microcirculatory dysfunction are contributors in the development of septic ARF. However, only a few studies looked at renal tissue oxygenation and kidney function in different animal models under septic conditions. As mentioned before, fluid resuscitation is an early therapeutic strategy in the treatment of septic shock, with the aim of restoring blood flow and oxygen delivery to vital organs. In Chapter 1 we tested the hypothesis that renal µPO2 and oxygen consumption (VO2,ren) are impaired during endotoxemia; that this effect is associated with a diminished renal function; that fluid resuscitation with either colloids or crystalloids improves an impaired µPO2 and oxygen consumption and restores kidney function; and that colloids are better at resuscitating than crystalloids in this context. In our model endotoxemia was accompanied by a reduction in renal blood flow and anuria, while the renal µPO2 and VO2,ren

remained relatively unchanged. Fluid resuscitation restored renal blood flow, renal oxygen delivery and kidney function to baseline values, and was associated with a redistribution of oxygen showing different patterns for the different compounds used. In contrast to HES130/0.4, HES200/0.5 and

Summary and conclusion

99

Ringer's lactate increased the renal oxygen consumption. For the first time we demonstrated the presence of renal oxygen redistribution during fluid resuscitation, but from our results we concluded that the loss of kidney function during endotoxemia could not be explained by an oxygen deficiency. HES130/0.4 had no influence on the VO2,ren and restored renal function with the least increase in the amount of renal work.

Based on the finding that no profound tissue hypoxia or decrease in microcirculatory PO2 could be seen during endotoxemia we performed a second study. In Chapter 2 we hypothesized that heterogeneity of microcirculatory oxygen supply to demand in the kidney is obscured when looking at the average PO2 during endotoxemia. Like demonstrated before, the average PO2 remained relatively unchanged during endotoxemia. Only after analysis of renal oxygen distribution we were able to demonstrate the appearance of microcirculatory hypoxic areas in the rat renal cortex. This finding led to a second series of experiments in which renal blood flow was mechanically reduced to values comparable to the flow reduction seen during endotoxemia. This second series of experiments should answer the question whether the cortical microcirculatory hypoxic areas were specific for endotoxemia and not simply a nonspecific phenomenon due to reduction of renal blood flow. The significant left shift in the cortical oxygen histogram toward hypoxia like seen during endotoxemia could not be observed in animals receiving mechanical reduction in renal blood flow. In a third group of animals we studied the effect of fluid resuscitation on the reversibility of microcirculatory hypoxic areas. In these animals the cortical microcirculatory hypoxic areas disappeared after resuscitation and renal function restored to 50% of baseline. In conclusion, endotoxemia was associated with the occurrence of cortical microcirculatory hypoxic areas that are not detected in the average PO2 measurement, proving the hypothesis of our study. These observations suggest the involvement of hypoxia in the pathogenesis of endotoxemia-induced ARF.

After demonstrating the occurrence of cortical microcirculatory hypoxic areas during endotoxemia we were searching for treatment strategies aimed on preventing microvascular hypoxia and maintaining renal function. Nitric oxide (NO) is an important molecule known to act on the renal microvascular tone. Therefore NO is consequently being involved in the regulation of intrarenal oxygen supply. Under septic conditions the main production of NO derives from iNOS, an enzyme that can be blocked by the clinically used glucocorticosteroid dexamethasone (DEX). In Chapter 3 of this thesis we tested the hypothesis that inhibition of iNOS by low-dose DEX would improve an impaired intrarenal oxygenation and kidney function. Two hours after development of endotoxemic shock one group of rats received a bolus of low-dose DEX (0.1 mg/kg) in addition to standard fluid resuscitation. In these animals, the renal iNOS mRNA expression was significantly suppressed 3 hours later. Furthermore low-dose dexamethasone prevented the appearance of cortical microcirculatory hypoxic areas, improved renal oxygen delivery, and significantly restored oxygen consumption. Besides a significant increase in systemic hemodynamic, DEX restored renal function and tubular sodium reabsorption to baseline values. In conclusion, the treatment of rats with low-dose dexamenthasone in addition to fluid resuscitation reversed endotoxin-induced renal failure associated by an improvement in intrarenal microvascular oxygenation. Therefore, low-dose DEX might have potential application in the prevention of septic acute renal failure.

Another class of substances known to modulate vascular tone are prostaglandins. In the kidney, prostaglandins uphold the balance between vasodilator and vasoconstrictor to maintain homeostasis and physiologic kidney function. Prostacyclin (PGI2) must be considered to play a potential renal protective role in septic ARF due to its potent vasodilatory effects. In Chapter 4 we test the

Chapter 7

100

hypothesis that exogenous prostacyclin would counterbalance an endotoxemia-induced intrarenal vasoconstriction and would therefore have beneficial effects on kidney function. Continuous infusion of the prostacyclin analogue iloprost (100 ng/kg/min) 2 hours after start of LPS-infusion in addition to fluid resuscitation resulted in the stabilization of hemodynamic parameters. All animals became anuric during endotoxemia. Only in animals receiving iloprost was creatinine clearance totally restored at the end of the experiment. Iloprost had no significant effects on average µPO2, but prevented the occurrence of cortical microcirculatory hypoxic areas. The renal expression of iNOS mRNA was significantly increased in all animals receiving LPS after 5 hours. Only in animals receiving iloprost, iNOS mRNA expression was significantly suppressed in the inner medulla. In conclusion the prostacyclin analogue iloprost significantly restored kidney function of endotoxemic rats to baseline values. This beneficial effect of iloprost on renal function might be addressed to an improvement in intrarenal oxygenation.

In the last experimental chapter of this thesis we looked at the effects of a substance with no direct effects on the vascular tone of the renal microcirculation. Activated protein C (APC) has been shown to have beneficial effects on the inflammatory process and coagulation during sepsis. Inflammation and coagulopathy impair the microvasculature and therefore disturb oxygen transport to the tissue. The hypothesis of our study presented in Chapter 5 was that APC-treatment improves renal microvascular oxygenation and kidney function in endotoxin-induced acute renal failure in the rat. Two hours after LPS-bolus rats received in addition to fluid resuscitation continuous infusion of either 10 or 100 µg/kg/h of APC. Treatment with APC 100 prevented a further decline of mean arterial blood pressure and renal blood flow. APC, independently of concentration, had no significant effects on average µPO2, but prevented the occurrence of cortical microcirculatory hypoxic areas. All animals receiving LPS had a significant decrease in creatinine clearance. Only in animals receiving APC 100, kidney function was significantly restored at the end of the experiment. In conclusion, APC 100 significantly restored renal function compared to standard fluid resuscitation in a rat model of endotoxemia. This was accompanied by protection against the occurrence of cortical microcirculatory hypoxic areas. Furthermore, this application best improved mean arterial blood pressure. Based on these results one could hypothesize that APC 100 infusion as adjuvant to standard fluid resuscitation might be useful as a renal protective strategy to preserve renal oxygenation and kidney function in the early stage of sepsis.

Conclusion In this thesis we present the result of different experimental therapeutic strategies to preserve kidney function in a rat model of endotoxemia. Therefore we established a model showing severe hemodynamic alterations and loss in kidney function a few hours after development of septic shock. Aim of our investigations was to create a setting representing a pathology like seen in ICU patients in septic shock. Our resuscitation strategies were also based on clinical standard, where fluid resuscitation to stabilize systemic hemodynamic parameters was the primary step. All different treatment strategies were therefore an addition to fluid resuscitation.

In our first study we could demonstrate a loss in kidney function during endotoxemia, which could be reverse by fluid resuscitation. Though the average µPO2 was only slightly affected during endotoxemia we could show the appearance of cortical microcirculatory hypoxic areas as left shift in the cortical oxygen histogram. Furthermore we demonstrated a correlation between impaired microvascular oxygenation and kidney function proving the hypothesis of this thesis. Strategies

Summary and conclusion

101

aiming on reversing microvascular hypoxia had beneficial effects on renal function. However, the proof of a causal correlation between renal hypoxia and functional consequences is still lacking. Further research is necessary to unveil a direct link between renal tissue hypoxia, critical oxygen delivery and renal function in the pathogenesis of acute renal failure in sepsis.

Clinical implication Before trying to translate our findings to human septic renal failure, it is important to comment on our specific animal model. Although our model is akin to clinical conditions animal models can never be directly translated to human pathophysiology and the clinic setting. Rodent models of endotoxemia are frequently accompanied by a hypodynamic state, intrarenal vasoconstriction and decreased renal blood flow in contrast with the hyperdynamic response seen in large animal models or in human sepsis. For this reason, the results of animal studies have to be interpreted carefully in terms of clinical relevance.

Prior to test our renal protective treatment strategies in humans it is mandatory to check the reproducibility of our result in a large animal model for example in a porcine model of endotoxin-induced renal failure. If the outcome of such a study could demonstrate to have beneficial protective effect than a big step towards a randomized clinical trial would be made. Especially in terms of the fact that all the substances we tested are registered drugs in clinical use.

Due to clinical importance of finding renal protective strategies to prevent or treat septic acute renal failure the performance of such clinical trials would be of utmost interest.

103

ADDENDUM SAMENVATTING EN CONCLUSIE LIST OF ABBREVIATIONS ACKNOWLEDGEMENTS LIST OF PUBLICATIONS CURRICULUM VITAE

104

Samenvatting en conclusie Introductie

De ontwikkeling van acuut nierfalen (acute renal failure, ARF) in sepsis komt vaak voor en heeft een prevalentie tot zo’n 40%. Echter, de pathogenese is nog steeds niet volledig duidelijk. Tot op heden zijn er geen specifieke protectieve maatregelen dan wel behandelingsmodaliteiten voorhanden. De behandeling van intensive care patiënten met ernstige AKI (acute kidney injury, acute nierschade) mondt daardoor vaak uit in ondersteunende maatregelen zoals dure nierfunctievervangende therapie. Renale weefselhypoxie is een mogelijke factor in de ontwikkeling van AKI. De meeste gangbare conservatieve therapeutische interventies richten zich op het vergroten van de perfusie van de nier en het handhaven van de nierfunctie. Een voorbeeld hiervan is volume expansieve therapie. Echter, interventies die de geforceerd de nierfunctie proberen te verbeteren zouden potentieel toename van de hypoxie door verhoging van de zuurstofconsumptie kunnen veroorzaken. Het is daardoor mogelijk dat deze klinische interventies uiteindelijk bijdragen aan de ontwikkeling van AKI in plaats van het voorkomen ervan. Vanwege deze paradox is het van wezenlijk klinisch belang meer inzicht te krijgen in de pathofysiologie van ARF en om specifieke behandelingsopties te vinden die gericht zijn op het verbeteren van ARF en het voorkomen van AKI.

These

In dit proefschrift zijn we uitgegaan van de veronderstelling dat hypoxie en microcirculatoire dysfunctie een rol spelen in de pathogenese van septisch nierfalen. De hypothese was derhalve dat strategieën gericht op het voorkomen van hypoxie en microcirculatoire dysfunctie resulteren in het behoud van nierfunctie. Om deze hypothese te toetsen hebben we een rattenmodel ontwikkeld van endotoxine-geïnduceerde ARF met de tekenen van ernstige hemodynamische veranderingen zoals die gezien worden in de klinische praktijk. “Early goal-directed therapy” met behulp van vloeistof- resuscitatie is de standaard therapie voor de behandeling van sepsis. Resuscitatie d.m.v. infusietherapie was daarom de primaire therapeutische behandeling in al onze studies. De experimentele behandelingsstrategieën werden getest bovenop de standaard vloeistofresuscitatie. Alle door ons geteste farmaca worden klinisch gebruikt en hebben een directe werking op de renale vasculaire tonus en/of beïnvloeden secundair de microvasculaire doorbloeding. Naast evaluatie van de nierfunctie d.m.v. meting van de kreatinine klaring hebben we een methode toegepast om nauwgezet de renale oxygenatie te bestuderen. Gebruik makend van de methode van zuurstof-afhankelijke uitdoving van fosforescentie zijn we in staat geweest om simultaan corticale en buitenste medullaire microvasculaire PO2 (µPO2) te meten in combinatie met de PO2 in de vena renalis. Door meting van microvasculaire PO2 histogrammen in de cortex en buitenste medulla zijn we in staat geweest om de heterogeniteit in intrarenale oxygenatie te bestuderen.

Bewijs van de these Er zijn aanwijzingen dat weefselhypoxie en microcirculatoire dysfunctie bijdragen aan de ontwikkeling van septisch ARF. Het aantal studies dat gekeken heeft naar weefseloxygenatie en nierfunctie in verschillende septische diermodellen is echter beperkt. Zoals hierboven opgemerkt is resuscitatie met infuusvloeistoffen een vroege therapeutische strategie in de behandeling van septische shock, gericht op herstel van de zuurstof toevoer naar vitale organen. In hoofdstuk 1

105

hebben we de hypothese getest dat de renale µPO2 en zuurstofconsumptie (VO2,ren) zijn verslechterd bij endotoxemie; dat dit gepaard gaat met verminderde nierfunctie; dat vloeistofresuscitatie met colloïdale of kristalloïde vloeistoffen de µPO2, VO2,ren en nierfunctie herstelt; en dat colloïdale vloeistoffen in dit opzicht beter zijn dan kristalloïde vloeistoffen. In ons model was endotoxemie geassocieerd met een reductie in renale doorbloeding en anurie terwijl de renale µPO2 en VO2,ren relatief onveranderd bleven. Vloeistofresuscitatie herstelde de renale doorbloeding, het renale zuurstofaanbod en de nierfunctie naar baseline waarden en was geassocieerd met een distributie van zuurstof die afhing van de resuscitatievloeistof. In tegenstelling tot HES130/0.4, veroorzaakten HES200/0.5 en Ringer's lactaat een toename in renale zuurstofconsumptie. Voor de eerste keer hebben we het bestaan van renale zuurstofredistributie tijdens vloeistofresuscitatie aangetoond, maar uit deze resultaten hebben we geconcludeerd dat het verlies van nierfunctie tijdens endotoxemie niet verklaard kon worden door een tekort aan zuurstof. HES130/0.4 had geen invloed op de VO2,ren en herstelde de nierfunctie met de geringste toename in renale arbeid.

Naar aanleiding van de vinding dat geen uitgebreide weefselhypoxie en afname van microcirculatoire PO2 werd gezien tijdens endotoxemie hebben we een tweede studie uitgevoerd. In hoofdstuk 2 hebben we de hypothese getoetst dat heterogeniteit in de verhouding van zuurstofaanbod en zuurstofvraag op microcirculatoir niveau tijdens endotoxemie verborgen blijft als alleen de gemiddelde PO2 waarde wordt gemeten. Zoals eerder aangetoond bleef de gemiddelde PO2 vrijwel onveranderd tijdens endotoxemie. Alleen d.m.v. analyse van de distributie van renale zuurstofspanning waren we in staat om het ontstaan van hypoxische microcirculatoire gebieden aan te tonen in de renale cortex van de rat. Deze vinding heeft geleid tot een tweede serie experimenten waarbij de renale bloedflow mechanisch werd gereduceerd tot het niveau zoals die tijdens endotoxemie werd waargenomen. Deze tweede serie experimenten moest antwoord geven op de vraag of de hypoxische microcirculatoire gebieden in de cortex specifiek waren voor endotoxemie en niet alleen een niet-specifiek fenomeen waren t.g.v. reductie van de bloedflow. De significante linksverschuiving in het corticale PO2 histogram in de richting van hypoxie, zoals geobserveerd tijdens endotoxemie, werd niet gezien in dieren waarbij een mechanische reductie van de bloedflow werd bewerkstelligd. In een derde groep dieren bestudeerden we de effecten van vloeistof-resuscitatie op de reversibiliteit van de microcirculatoire hypoxische gebieden. In deze dieren verdwenen de corticale microcirculatoire hypoxische gebieden na resuscitatie en herstelde de nierfunctie zich tot 50% van de baselinewaarde. Concluderend was endotoxemie geassocieerd met het ontstaan van corticale microcirculatoire hypoxische gebieden die niet werden gedetecteerd door gemiddelde PO2 metingen, hetgeen de hypothese van onze studie bewijst. Deze observaties suggereren een rol voor hypoxie in de pathogenese van endotoxemie-geïnduceerde ARF.

Na het aantonen van het ontstaan van microcirculatoire hypoxische gebieden tijdens endotoxemie zijn we op zoek gegaan naar therapeutische strategieën gericht op het tegengaan van microvasculaire hypoxie en behoud van nierfunctie. Stikstofoxide (NO) is een belangrijk molecuul waarvan bekend is dat het de renale microvasculaire tonus beïnvloed. Hierdoor speelt NO een rol in de regulatie van het renale zuurstofaanbod. Onder septische condities is de belangrijkste bron van NO de productie door iNOS, een enzym waarvan de werking kan worden onderdrukt door de klinisch gebruikte glucocorticosteroid dexamethason (DEX). In hoofdstuk 3 van dit proefschrift hebben we de hypothese getoetst dat inhibitie van iNOS d.m.v. lage dosis DEX leidt tot verbetering van een verminderde intrarenale oxygenatie en nierfunctie. Twee uur na de ontwikkeling van endotoxemische shock kreeg een groep ratten een bolus lage dosering DEX (0.1 mg/kg) toegediend

106

in aanvulling op standaard vloeistofresuscitatie. In deze dieren was drie uur later de renale iNOS mRNA expressie significant onderdrukt. Daarnaast voorkwam dexamethason het ontstaan van corticale microcirculatoire hypoxische gebieden, verbeterde het renale zuurstofaanbod en herstelde de zuurstofconsumptie van de nier significant. Naast een significantie verbetering van de systemische hemodynamiek herstelde DEX de nierfunctie en de tubulaire natrium terugresorptie tot beginwaarden. De conclusie van dit hoofdstuk is dat de behandeling van ratten met lage dosis dexamethason in aanvulling op vloeistofresuscitatie het endotoxine-geïnduceerde nierfalen herstelde en dat dit herstel geassocieerd was met een verbetering van de intrarenale microvasculaire oxygenatie. Lage dosis DEX zou potentieel een rol kunnen spelen als preventieve maatregel tegen het ontstaan van acuut septisch nierfalen.

Een andere klasse van stoffen die de vasculaire tonus moduleren zijn de prostaglandines. In de nier houden prostaglandines de balans tussen vasodilatatie en vasoconstrictie in stand en spelen zo een rol bij homeostase en nierfunctie. Prostacycline (PGI2) moet geacht worden een potentiele beschermende rol voor de nier te kunnen spelen in septische ARF, vanwege zijn potente vasodilatoire effecten. In hoofdstuk 4 hebben we de hypothese getoetst dat exogeen toegediende prostacycline een endotoxemie-geïnduceerde intrarenale vasoconstrictie tegengaat en zo positieve effecten heeft of de nierfunctie. Continue infusie van de prostacycline analoog iloprost (100 ng/kg/min), gestart 2 uur na start van LPS-infusie in aanvulling op vloeistofresuscitatie, resulteerde in een stabilisatie van hemodynamische parameters. Alle dieren werden anuur tijdens endotoxemie. Alleen bij de dieren die iloprost kregen was de kreatinine klaring aan het einde van het experiment geheel hersteld. Iloprost had geen significant effect op gemiddelde µPO2, maar ging het ontstaan van corticale microcirculatoire hypoxische gebieden tegen. Na 5 uur was de expressie van iNOS mRNA significant toegenomen in de nieren van alle dieren die LPS kregen toegediend. Alleen in de dieren die iloprost kregen was de iNOS mRNA expressie significant onderdrukt in de binnenste medulla. Concluderend, de prostacycline analoog iloprost herstelde de nierfunctie van endotoxemische ratten terug naar aanvangswaarden. Dit positieve effect van iloprost op de nierfunctie zou mogelijk kunnen worden toegeschreven aan een verbeterde intrarenale oxygenatie.

In het laatste experimentele hoofdstuk van dit proefschrift hebben we gekeken naar de effecten van een substantie waarvan geen directe effecten op vasculaire tonus of renale microcirculatie bekend waren. Het is aangetoond dat Activated Protein C (APC) positieve effecten heeft op het inflammatoire proces en de coagulatie tijdens sepsis. Inflammatie en coagulopathie hebben een negatieve invloed op microcirculatoire functie en hinderen het zuurstoftransport naar het weefsel. De hypothese van onze studie in hoofdstuk 5 was dat behandeling met APC de renale microvasculaire oxygenatie en nierfunctie verbetert in endotoxine-geïnduceerd acuut nierfalen in de rat. Twee uur na toediening van een LPS-bolus kregen ratten in toevoeging op vloeistofresuscitatie een continue infusie van 10 of 100 µg/kg/uur APC. Behandeling met APC 100 voorkwam een verdere achteruitgang van de gemiddelde arteriële bloeddruk en de renale bloedflow. APC, onafhankelijk van concentratie, had geen significant effect of de gemiddelde µPO2, maar voorkwam het ontstaan van corticale microcirculatoir hypoxische gebieden. Alle dieren die LPS kregen toegediend hadden een significante afname in kreatinineklaring. Alleen in de dieren uit de APC 100 groep was de nierfunctie aan het einde van het experiment significant hersteld. Concluderend, APC 100 leidde tot een significant herstel van nierfunctie in vergelijking tot standaard vloeistofresuscitatie in een rattenmodel van endotoxemie. Dit effect hing samen met een bescherming tegen het ontstaan van corticale microvasculaire hypoxische gebieden. Daarnaast had APC 100 het beste effect op de gemiddelde

107

arteriële bloeddruk. Deze resultaten suggereren dat APC 100 infusie als adjudant bij standaard vloeistofresuscitatie bruikbaar zou kunnen zijn als strategie ter behoud van renale oxygenatie en nierfunctie in de vroege fase van sepsis.

Conclusie In dit proefschrift presenteren we de resultaten van verschillende experimentele therapeutische strategieën gericht op behoud van nierfunctie in een rattenmodel van endotoxemie. We hebben een model ontwikkeld dat ernstige hemodynamische veranderingen en, enige uren na ontwikkeling van septische shock, verlies van nierfunctie laat zien. Het doel van onze studies was om enigszins de pathologie zoals die gezien word bij patiënten op de intensive care na te bootsen. Onze resuscitatie strategieën waren daarbij tevens gebaseerd op de hedendaagse klinische standaard waarbij vloeistofresuscitatie ter stabilisatie van hemodynamische parameters de primaire stap was. Alle experimentele behandelingen werden bovenop deze standaard vloeistoftherapie uitgevoerd.

In onze eerste studie waren we in staat om een verlies aan nierfunctie te demonstreren tijdens endotoxemie die reversibel was door vloeistofresuscitatie. Hoewel de gemiddelde µPO2 slechts weinig was gedaald tijdens endotoxemie konden we het ontstaan van corticale microcirculatoire hypoxische gebieden aantonen door een linksverschuiving in het corticale PO2 histogram. Daarnaast hebben we een correlatie laten zien tussen verminderde microvasculaire oxygenatie en nierfunctie, wat de these bewijst. Strategieën gericht op voorkomen of herstellen van de microvasculaire hypoxie hebben positieve effecten op de nierfunctie. Echter, het bewijs van een causale relatie tussen renale hypoxie en functionele consequenties ontbreekt nog. Meer onderzoek is nodig om een eventueel direct verband tussen renale weefselhypoxie en nierfunctiestoornissen in de pathogenese van acuut nierfalen bij sepsis aan te tonen.

Klinische implicatie Voordat pogingen worden ondernomen om onze vindingen te vertalen naar septisch nierfalen bij patiënten is het belangrijk enige nadelen van ons specifieke proefdiermodel in acht te nemen. Hoewel zo goed mogelijk klinische condities werden nagebootst kan geen enkel proefdiermodel direct vertaald worden naar humane pathofysiologie in de kliniek. In kleine knaagdieren gaat endotoxemie vaak gepaard met een hypodynamische status, intrarenale vasoconstrictie en een afname in renale bloedflow. Dit is in tegenstelling tot de hyperdynamische respons die vaker gezien wordt in grote proefdieren en bij mensen. Derhalve moeten de resultaten van dierstudies met enige voorzichtigheid geïnterpreteerd worden voor wat betreft hun klinische relevantie.

Voordat de door ons voorgestelde therapieën ter bescherming van de nieren in mensen wordt getest lijkt het meer opportuun om de reproduceerbaarheid van onze resultaten te onderzoeken in bijvoorbeeld een varkensmodel van sepsis. Indien de uitkomst van zo’n studie ook positieve beschermende effecten zou aantonen dan zou daarmee een grote stap in de richting van een gerandomiseerde klinische trial gedaan zijn. Met name doordat alle door ons geteste farmaca reeds in klinisch gebruik zijn.

Vanwege het grote klinische belang van het vinden van nierfunctie beschermende of nierfunctie verbeterende therapieën zou de uitvoering van zulke klinische studies zeer interessant zijn.

108

List of abbreviations AKI acute kidney injury

APC activated protein C

ARF acute renal failure

Clearcrea creatinine clearance CVP central venous pressure

DEX dexamethasone

DO2ren renal oxygen delivery

FR fluid resuscitation

GFR glomerular filtration rate

HR heart rate

Ilo iloprost

iNOS inducible nitric oxide synthase

LPS lipopolysaccharide

MAP mean arterial pressure

µPO2 microvascular PO2

NO nitric oxide

NOx concentration of nitrate/nitrite/S-nitrosothiols

NR non-resuscitation

O2ERren renal oxygen extraction

rvPO2 renal venous PO2

RBF renal blood flow

RVR renal vascular resistance

TC time control

TNa+ tubular sodium reabsorption

VO2ren renal oxygen consumption

109

Acknowledgements This manuscript would not be complete without acknowledging the people who made this work

possible. I want to thank Professor Can Ince who gave me the opportunity to work in his research

department in Amsterdam. His enthusiasm and the collection of interesting technology within his

department made me travel to Amsterdam and spent there several years of my life. Looking back at

this research period and seeing now the ultimate result, this thesis, I definitely conclude that this time

was well spent. Also, I like to express my gratitude to all co-workers within the department of

Translational Physiology who supported me with my work and made my stay unforgettable.

Secondly, I want to thank my husband Bert Mik for all the years of patience and mentorship that

guided me through my research work. He built the technical devices we needed for the oxygen

measurements and taught me how to use them. Without his positive thoughts and his personal

investment in me, I probably would never have finished this manuscript. My thanks also go to

Professor Klaus Unertl who supported my research and my training in anesthesiology from the first

moment we met. Furthermore, I have to thank Dr. Boris Nohé. He confirmed my interest in basic

science and guided me through the first steps of my academic work. My thanks also go to Professor

Stolker who gave me a job in his department and who is fully supporting my current research.

Furthermore, I want to thank the members of the promotion committee for their assessment of my

thesis and their presence during the PhD ceremony.

Above all, I want to thank Hanni and Rolf who made it possible for me to have the career of my

choice. I thank them for their guidance and efforts to make this career come true. Finally I want to

thank my late great-grand nanny who always believed in me – I will never forget you.

110

List of publications

Mik EG, Johannes T, Fries M Clinical microvascular monitoring: a bright future without a future? Crit. Care Med. 2009; 37(11): 2980-1. Johannes T, Mik EG, Klingel K, Goedhart PT, Zanke C, Nohe B, Dieterich HJ, Unertl KE, Ince C Effects of 1400W and/or nitroglycerin on renal oxygenation and kidney function during endotoxaemia in anaesthetised rats. CEPP. 2009; 36 (9): 870–879. Legrand M, Almac E, Mik EG, Johannes T, Kandil A, Bezemer R, Payen D, Ince C L-NIL prevents renal microvascular hypoxia and increase of renal oxygen consumption after ischemia-reperfusion in rats. Am. J. Physiol. Renal Physiol. 2009; 296(5): F1109-17. Mik EG, Ince C, Eerbeek O, Heinen A, Stap J, Hooibrink B, Schumacher CA, Balestra GM, Johannes T, Beek JF, Nieuwenhuis AF, van Horssen P, Spaan JA, Zuurbier CJ Mitochondrial oxygen tension within the heart. J. Mol. Cell. Cardiol. 2009; 46(6): 943-51. Raat NJ, Hilarius PM, Johannes T, de Korte D, Ince C, Verhoeven AJ Rejuvenation of stored human red cells reverses the renal microvascular oxygenation deficit in an isovolemic transfusion model in rats. Transfusion. 2009; 49(3): 427-4. Johannes T, Ince C, Klingel K, Unertl KE, Mik EG Iloprost preserves renal oxygenation and restores kidney function in endotoxemia-related acute renal failure in the rat. Crit. Care Med. 37(4): 1423-32, 2009 Johannes T, Mik EG, Klingel K, Dieterich HJ, Unertl KE, Ince C Low-dose dexamethasone supplemented fluid resuscitation reverses endotoxin-induced acute renal failure and prevents cortical microvascular hypoxia. Shock. 2009; 31(5): 521-528. Johannes T, Mik EG, Ince C Non-resuscitated endotoxemia induces microcirculatory hypoxic areas in the renal cortex in the rat. Shock. 2009; 31(1): 97-103. Ploppa A, Ayers DM, Johannes T, Unertl KE, Durieux ME The Inhibition of Human Neutrophil Phagocytosis and Oxidative Burst by Tricyclic Antidepressants. Anesth. Analg. 2008; 107(4): 1229-35. Mik EG, Johannes T, Zuurbier C, Heinen A, Houben-Weerts JHPM, Balestra GM, Stap J, Beek, JF, Ince C In vivo mitochondrial oxygen tension measured by a delayed fluorescence lifetime technique. Biophys. J. 2008; 95(8): 3977-90. Legrand M, Mik EG, Johannes T, Payen D, Ince C Renal hypoxia and dysoxia after reperfusion of the ischemic kidney. Mol. Med. 2008: 14(7-8): 502-16. Mik EG, Johannes T, Ince C Monitoring of renal venous PO2 and kidney oxygen consumption in rats by a near-infrared phosphorescence lifetime technique. Am. J. Physiol. Renal Physiol. 2008; 294(3): F676-81.

111

Johannes T, Mik EG, Nohe B, Unertl KE, Ince C Acute Decrease in Renal Microvascular PO2 during Acute Normovolemic Hemodilution. Am. J. Physiol. Renal Physiol. 2007; 292(2): F796-803. Johannes T, Mik EG, Nohe B, Raat NJH, Unertl KE, Ince C Influence of fluid resuscitation on renal microvascular PO2 in a normotensive rat model of endotoxemia. Crit. Care. 2006; 10(3): R88. Johannes T, Mik EG, Ince C Dual-wavelength phosphorimetry for determination of cortical and sub-cortical microvascular oxygenation in rat kidney. J. Appl. Physiol. 2006; 100(4): 1301-10. Nohe B, Johannes T, Reutershan J, Rothmund A, Haeberle HA, Ploppa A, Schroeder TH, Dieterich HJ Synthetic Colloids Attenuate Leukocyte-Endothelial Interactions by Inhibition of Integrin Function. Anesthesiology. 2005; 103 (4): 759-767. Nohe B, Schmidt V, Johannes T, Reutershan J, Schroeder TH, Dieterich HJ, Schroeder S Maldistribution of activated leukocytes during systemic inflammation. Perfusion. 2005; 18: 4-16. Nohe B, Johannes T, Schmidt V, Schroeder TH, Kiefer RT, Unertl K, Dieterich HJ Effects of reduced shear stress on inflammatory reactions in vitro Effects of pathological flow conditions on leukocyte-endothelial interactions and monocyte tissue factor expression in human cell cultures. Anaesthesist. 2005; 54 (8): 773-780. Nohe B, Johannes T, Dieterich HJ Antiinflammatory effects of omega-3 fatty acids vary at different stages of inflammation. Am. J. Physiol. Heart Circ. Physiol. 2003; 285 (5): 2248-2249. Nohe B, Ruoff H, Johannes T, Zanke C, Unertl K, Dieterich HJ A fish oil emulsion used for parenteral nutrition attenuates monocyte-endothelial interactions under flow. Shock. 2002; 18 (3): 217-222. Nohe B, Zanke C, Johannes T, Kiefer T, Dieterich HJ Effects of magnetic cell separation on monocyte adhesion to endothelial cells under flow. APMIS. 2002; 110 (4): 299-308.

112

Curriculum vitae

Tanja Johannes was born on December 22, 1974 in Waiblingen, Germany. She attended pre-university education at Maria-Merian Gymnasium in Waiblingen. In 1996, she started her medical studies at Eberhard-Karls-University of Tübingen. During this time she developed interest in basic science and started to work on a research project at the Laboratory of Anesthesiology at the University of Tübingen. In 2002, this project was successfully rounded up by a doctoral thesis with the title “Einfluss einer parenteralen Fischölemulsion auf die Zelladhäsion und Tissue Factor Expression von Monozyten und Endothelzellen in vitro“ and graded with "summa cum laude“. In the same year she finished her medical studies and started training in Anesthesiology and Intensive Care Medicine at the Department of Anesthesiology and Intensive Care Medicine at the University of Tübingen. Besides clinical work she was always involved in research projects and from 2004 to 2006 she worked on the basis of a research grant in the Department of Translational Physiology at the Academic Medical Center, University Amsterdam, The Netherlands. Till today her research focus lies on pathophysiological changes in renal oxygenation in sepsis and hemodilution in various animal models. Her research was supported by different project grants and is published in numerous international journals. In 2009 she was registered as anesthesiologist. In the same year she took up her current position as medical staff member at the Department of Anesthesiology at Erasmus Medical Center, University of Rotterdam, The Netherlands. Tanja is married to Bert Mik with whom she has a little daughter named Maita.