Intensive Care Med (2007) 33:1614–1618 DOI 10.1007/s00134-007-0734-8

Christoph Langenberg Li Wan Moritoki Egi Clive N. May Rinaldo Bellomo

Received: 10 December 2006 Accepted: 6 April 2007 Published online: 16 June 2007 © Springer-Verlag 2007 This article is discussed in the editorial available at: http://dx.doi.org/10.1007/ s00134-007-0735-7.

C. Langenberg · L. Wan · M. Egi · C. N. May · R. Bellomo (u) Austin & Repatriation Medical Centre, Department of Intensive Care, Heidelberg, 3084, Victoria, Australia e-mail: [email protected] Tel.: +61-3-94965992 Fax: +61-3-94963932 C. Langenberg · L. Wan · C. N. May University of Melbourne, Howard Florey Institute, Parkville, Melbourne, Australia

BRIEF REPORT

Renal blood flow and function during recovery from experimental septic acute kidney injury

Abstract Objective: To measure renal blood flow (RBF) and renal function during recovery from experimental septic acute kidney injury (AKI). Design: Controlled experimental study. Subjects: Nine merino ewes. Setting: University physiology laboratory. Intervention: We recorded systemic and renal hemodynamics during a 96-h observation period (control) via implanted transit-time flow probes. We then compared this period with 96 h of septic AKI (48 h of Escherichia coli infusion) and subsequent recovery (48 h of observation after stopping E. coli). Measurements and results: Compared with the control period, E. coli infusion induced hyperdynamic sepsis (increased cardiac output and decreased blood pressure) and septic AKI (serum creatinine 65.4 ± 8.7 vs. 139.9 ± 33.0 µmol/l; creatinine clearance 73.8 ± 12.2 vs. 40.2 ± 17.2 ml/min; p < 0.05) with a mortality of 22%. RBF increased (278.8 ± 33.9

Introduction Acute kidney injury (AKI) affects approximately 6% of critically ill patients [1]. Severe sepsis and septic shock are the most common predisposing factors for the development of AKI in this setting [2]. Our knowledge about the pathogenesis of septic AKI is limited. An increase in renal vascular resistance resulting in renal hypoperfusion has been repeatedly proposed as central to its development [3, 4]. Because few measurements of renal blood flow (RBF) in septic patients

vs. 547.9 ± 124.8 ml/min; p < 0.05) as did renal vascular conductance (RVC). During recovery, we observed a decrease in RVC and RBF with all values returning to control levels. Indices of tubular function [fractional excretion of sodium (FENa) and urea (FEUn) and urinary sodium concentration (UNa)], which had been affected by sepsis, returned to control values after 18 h of recovery, as did serum creatinine. Conclusions: Infusion of E. coli induced a hyperdynamic circulatory state with hyperemic AKI. Recovery was associated with relative renal vasoconstriction and reduction in RBF and RVC back to control levels. Indices of tubular function normalized more rapidly than changes in RBF. Keywords Acute renal failure · Sepsis · Renal blood flow · Cardiac output · Tubular function · Recovery · Acute kidney injury

have been performed [5–7], this paradigm is derived from animal models. However, these animal studies are heterogeneous in design and have many confounders (e. g. hypodynamic circulatory state) [8, 9]. In addition, all these experimental studies have not yet shown any data on the nature and timing of renal hemodynamic and functional changes during the recovery phase of septic AKI. Accordingly, we developed a reproducible model of hyperdynamic septic AKI and assessed functional and hemodynamic parameters during recovery.

1615

Fig. 1 Systemic hemodynamic parameters of seven sheep (n = 6 for CVP control) during a 96-h control period and during a 48-h period of sepsis induced by infusion of E. coli followed by a subsequent 48-h recovery period (mean ± SD). The line at 48 h indicates the

start of recovery after administration of a single bolus of gentamicin and cessation of E. coli infusion. *p < 0.05. HR, Heart rate; CO, cardiac output; TPC, total peripheral conductance; MAP, mean arterial pressure; CVP, central venous pressure

1616

Materials and methods

Analogue signals of mean arterial pressure (MAP), central venous pressure (CVP), cardiac output (CO), and RBF were collected as previously described [10]. Animal preparation Total peripheral conductance (TPC; CO/MAP) and renal vascular conductance (RVC; RBF/MAP) were calculated The institutional Animal Ethics Committee approved (conductance is the reciprocal value of resistance). this study. We procured nine female Merino ewes (34.2–47.3 kg) for chronic instrumentation. The animals underwent two separate operative procedures to implant Protocol and measurements transit-time flow probes around the pulmonary and renal arteries [10]. They were allowed to recover for at least Initially, all animals were studied during a 96-h control 2 weeks. Cannulae were then inserted intra-arterially and period with fluid administration only (normal saline 1.0 ml/kg/h). After a week, sepsis and recovery period intravenously [10].

Fig. 2 Renal hemodynamic and functional parameters of seven sheep during a 96-h control period and during a 48-h period of sepsis induced by infusion of E. coli followed by a subsequent 48-h recovery period (mean ± SD). The line at 48-h indicates the start of recovery after administration of a single bolus of gentamicin and

cessation of E. coli infusion. *p < 0.05. RVC, Renal vascular conductance; RBF, renal blood flow; UNa, urine sodium concentration; FENa, fractional excretion of sodium; FEUn, fractional excretion of urea nitrogen

1617

were studied. Sepsis was induced by the administration of a bolus of live E. coli [3.9 × 109 colony-forming units (cfu) in 15 ml saline] followed by a continuous infusion of 3.0 × 108 cfu/h for 48 h. During sepsis, normal saline was administered at the same rate as in control to prevent marked hypovolemia. After 48 h, a bolus of 120 mg gentamicin was administered intravenously and E. coli infusion was stopped. The sheep were then monitored for 48 h (recovery) with fluid administration continued as described above. Urinary output was measured and urine sampled every 90 min. Arterial blood samples were obtained every 12 h, as previously described [10].

after induction of sepsis before steadily decreasing to 0.66 ± 0.25 vs. 1.72 ± 0.77 ml/h/kg in control (p < 0.05; Fig. 2). After a brief polyuric phase in recovery, it returned to control levels. Tubular functional changes

The urine sodium concentration decreased during sepsis (196.9 ± 76.7 vs. 59.1 ± 54.4 mmol/l, p < 0.05) but returned to control levels by 66 h (Fig. 2). The fractional excretion of sodium (FENa) showed a similar pattern and, in sepsis, it decreased to a minimum of 0.34 ± 0.44% compared with 1.40 ± 0.59% in control. The fractional excretion of urea nitrogen (FEUn) deStatistical analysis creased from 55.3 ± 4.9% in control to 28.4 ± 16.6% in sepsis (p < 0.05, Fig. 2) and its recovery pattern was Data are presented as means ± standard deviation. The 12- similar to that of FENa. h mean values for each variable for the control and recovery periods were compared with the sepsis period using the Wilcoxon ranked sign test. A p-value < 0.05 was consid- Discussion ered statistically significant. We developed a large animal model of septic AKI with a hyperdynamic circulatory state similar to that seen in ICU patients with septic ARF [11] and Results which contrasts with most previous animal studTwo of nine sheep died from severe sepsis (mortality of ies, where hypodynamic sepsis was induced [8, 9]. 22%) before entering the recovery phase and could not be We also maintained fluid administration throughout the whole experiment to avoid marked hypoincluded in the data analysis. Infusion of E. coli induced a hyperdynamic volemia. In this model, AKI was associated with renal vasodicirculatory state with an increased CO (4.0 ± 0.7 vs. 8.0 ± 2.3 l/min; Fig. 1) and a decreased MAP latation and renal hyperemia. Conversely, recovery was (95.2 ± 12.2 to 78.5 ± 14.7 mmHg, p < 0.05; Fig. 1) associated with relative vasoconstriction and a decrease in with peripheral vasodilatation (TPC 43.3 ± 9.5 vs. RBF to control levels. These observations challenge the 110.3 ± 52.1 ml/min/mmHg, p < 0.05). The CVP re- widely held paradigm of renal ischemia as the underlying mained steady. All variables returned to normal during cause for the development of septic AKI [3]. In addition, our study shows that recovery of renal function can recovery (Fig. 1). proceed in parallel with renal vascular vasoconstriction and decreased RBF. This vasoconstrictive renal recovery Renal hemodynamics and functional parameters pattern provides evidence of an association between changes in renal hemodynamics and changes in glomeruRenal vasodilatation with an increase in RVC (3.0 ± 0.3 lar filtration rate (GFR). In our study, GFR was estimated vs. 7.5 ± 2.9 ml/min/mmHg, p < 0.05; Fig. 2) oc- by serum creatinine and creatinine clearance. Although curred and was accompanied by an increase in RBF creatinine clearance was measured in a non-steady-state (547.9 ± 124.8 vs. 278.8 ± 33.9 ml/min in control, period and overestimates GFR due to tubular creatinine p < 0.05). During recovery, RVC and RBF returned to secretion, the changes reported were consistent with the control levels. However, at the end of the recovery period, development of oliguria and a rise in serum creatinine. We RBF was less than in controls (Fig. 2). administered the potentially nephrotoxic aminoglycoside Serum creatinine increased significantly (139.9 ± 33.0 gentamicin. However, its nephrotoxicity occurs only after vs. 65.4 ± 8.7 µmol/l; Fig. 2) and creatinine clearance prolonged administration and tubular accumulation [12] decreased (40.2 ± 17.2 vs. 73.8 ± 12.2 ml/min in control, and cannot reasonably account for the changes in renal p < 0.05). During recovery, serum creatinine rapidly blood flow during recovery. returned to control levels. Creatinine clearance also inRecovery of tubular function occurred within 48 h, creased above control levels in the baseline period at 66 h making it unlikely that severe tubular necrosis would (109.0 ± 37.2 vs. 74.7 ± 24.4 ml/min; Fig. 2) and then have accompanied the development of septic AKI in this returned to control levels. Urinary output increased briefly model. Furthermore, UNa decreased in the sepsis period,

1618

as is seen in “pre-renal” states with preserved tubular function [13]. This “pre-renal ARF pattern” of tubular function occurred despite the presence of marked renal hyperemia and despite the continuous infusion of fluids. The FENa was also < 1% in the sepsis period, which is similar to the finding seen in some septic humans [14]. FENa is a widely used parameter to differentiate between “pre-renal” (< 1%) and “intra-renal” (> 1%) causes of ARF [15]. Our findings, however, call into question the ability of this marker to distinguish between the presence or absence of hypoperfusion. The FEUn, another marker, which suggests preserved tubular function when its value is < 35% [16], was also decreased in our septic model. The above tubular findings are all internally consistent and suggest preserved tubular function. As about 90% of renal flow is delivered to the glomeruli, we speculate that changes in tone (vasodilatation) of afferent and efferent arterioles may initially be responsible for loss of GFR and that later, during recovery, similar but opposite changes (vasoconstriction) may be responsible for the return of GFR to normal. Loss of GFR in the setting of increased RBF might theoretically occur if afferent arteriolar vasodilatation were coupled with even greater efferent arteriolar vasodilatation and consequent loss of intraglomerular filtration pressure (the driving physical force for GFR). It is possible that recovery of GFR in this setting is due to the return of normal or near

normal afferent and efferent vessel tone. We note that all this remains theoretical speculation and that further research is needed to establish whether these are indeed the mechanisms at work in this model of septic AKI and associated recovery.

Conclusions We studied renal recovery after AKI in a model of hyperdynamic sepsis. We found that recovery was characterized by the rapid reversal of all vasodilatory processes induced by sepsis with return of renal vasomotor tone to normal and return of RBF to control levels. The association of recovery of GFR with a decrease in RBF suggests (but does not prove) that hemodynamic factors may be of importance in the pathophysiology of both loss and recovery of GFR in this model. The rapid changes in tubular function also suggest that extensive acute tubular necrosis is unlikely in our model. Further studies of this model incorporating interventions and hormonal or cytokines measurements may further elucidate the mechanisms responsible for the development of and recovery from AKI. Acknowledgements. The study was supported by a grant from the Australian and New Zealand College of Anaesthetists. Christoph Langenberg was funded by the Else Kröner–Fresenius Foundation (Germany).

References 1. Uchino S, Kellum JA, Bellomo R, Doig GS, Morimatsu H, Morgera S, Schetz M, Tan I, Bouman C, Macedo E, Gibney N, Tolwani A, Ronco C (2005) Acute renal failure in critically ill patients: a multinational, multicenter study. Jama 294:813–818 2. Uchino S, Doig GS, Bellomo R, Morimatsu H, Morgera S, Schetz M, Tan I, Bouman C, Nacedo E, Gibney N, Tolwani A, Ronco C, Kellum JA (2004) Diuretics and mortality in acute renal failure. Crit Care Med 32:1669–1677 3. Schrier RW, Wang W (2004) Acute renal failure and sepsis. N Engl J Med 351:159–169 4. Lameire N, Van Biesen W, Vanholder R (2005) Acute renal failure. Lancet 365:417–430 5. Brenner M, Schaer GL, Mallory DL, Suffredini AF, Parrillo JE (1990) Detection of renal blood flow abnormalities in septic and critically ill patients using a newly designed indwelling thermodilution renal vein catheter. Chest 98:170–179

6. Lucas CE, Rector FE, Werner M, Rosenberg IK (1973) Altered renal homeostasis with acute sepsis. Clinical significance. Arch Surg 106:444–449 7. Rector F, Goyal S, Rosenberg IK, Lucas CE (1973) Sepsis: a mechanism for vasodilatation in the kidney. Ann Surg 178:222–226 8. Langenberg C, Bellomo R, May C, Wan L, Egi M, Morgera S (2005) Renal blood flow in sepsis. Crit Care 9:R363–R374 9. Langenberg C, Bellomo R, May CN, Egi M, Wan L, Morgera S (2006) Renal vascular resistance in sepsis. Nephron Physiol 104:1–11 10. Langenberg C, Wan L, Egi M, May CN, Bellomo R (2006) Renal blood flow in experimental septic acute renal failure. Kidney Int 69:1996–2002 11. Parker MM, Shelhamer JH, Natanson C, Alling DW, Parrillo JE (1987) Serial cardiovascular variables in survivors and nonsurvivors of human septic shock: heart rate as an early predictor of prognosis. Crit Care Med 15:923–929

12. Nagai J, Takano M (2004) Molecular aspects of renal handling of aminoglycosides and strategies for preventing the nephrotoxicity. Drug Metab Pharmacokinet 19:159–170 13. Miller TR, Anderson RJ, Linas SL, Henrich WL, Berns AS, Gabow PA, Schrier RW (1978) Urinary diagnostic indices in acute renal failure: a prospective study. Ann Intern Med 89:47–50 14. Vaz AJ (1983) Low fractional excretion of urine sodium in acute renal failure due to sepsis. Arch Intern Med 143:738–739 15. Espinel CH (1976) The FENa test. Use in the differential diagnosis of acute renal failure. Jama 236:579–581 16. Carvounis CP, Nisar S, GuroRazuman S (2002) Significance of the fractional excretion of urea in the differential diagnosis of acute renal failure. Kidney Int 62:2223–2229