Continuous measurements of renal perfusion in pigs by means of intravascular Doppler

Kidney International, Vol. 59 (2001), pp. 1439–1447

Continuous measurements of renal perfusion in pigs by means of intravascular Doppler MARTIN MO¨CKEL, DIERK SCHEINERT, EVGENIJ V. POTAPOV, ERNST WELLNHOFER, VOLKER COMBE´, BORIS A. NASSERI, DIRK MAIER, SABINE MEYER, CHARLES YANKAH, ROLAND HETZER, ULRICH FREI, and KAI-UWE ECKARDT Department of Nephrology and Medical Intensive Care, Charite´, Campus Virchow-Klinikum, Humboldt-University, and Departments of Cardiothoracic Surgery and Cardiology, German Heart Center, Berlin, Germany

Continuous measurements of renal perfusion in pigs by means of intravascular Doppler. Background. Changes in renal blood flow are considered to play a significant role in the induction and maintenance of kidney failure, but are difficult to monitor with currently available techniques. The objective was to validate renal flow measurements with Doppler guidewires and to apply this technique to assess dose and time dependency of the renal vascular effects of norepinephrine (NE). Methods. In 10 anesthetized pigs, flow velocity in renal arteries (FVart) and veins (FVvein) and volumetric renal blood flow (VBF) were measured before and after intravenous bolus application of incremental doses of NE (2 to 200 ␮g). Results. FVart curves exactly reflected the changes in VBF. Beat-to-beat analysis revealed a strong linear correlation over a mean VBF range of less than 0.05 to 0.35 L/min (median correlation coefficient with FVart, r ⫽ 0.998), and significant but less close relationships were also found between VBF and FVvein. Ten seconds after the administration of 200 ␮g NE, FVart dropped from 71 to 6 cm/sec and was 90% reversible after 48 seconds. Similarly, the renal vascular resistance temporarily rose from 988 to 13711 mm Hg · min/L. In contrast, NE-induced increases in systemic vascular resistance were on average a maximum of 1.5-fold but persisted for more than 60 seconds. Conclusions. Doppler flow measurements in the renal artery provide an excellent surrogate of volumetric blood flow, which may be useful for continuous monitoring of renal hemodynamics. The renal vasculature is more sensitive when compared with the systemic vasculature, but also appears to evoke more efficient counter-regulatory mechanisms in response to NE.

A reduction in renal blood flow (RBF) commonly occurs in association with renal dysfunction and can cause kidney failure. In acute renal failure (ARF), renal hypoKey words: renal blood flow, Doppler guide wire, norepinephrine, volumetric blood flow, monitor hemodynamics. Received for publication June 20, 2000 and in revised form October 18, 2000 Accepted for publication October 23, 2000

 2001 by the International Society of Nephrology

perfusion is considered to play a key role [1]. Decreased renal perfusion may result from a reduction of mean arterial blood pressure (MABP) or decreased cardiac output or may be due to renal vasoconstriction. The combination of moderately reduced perfusion pressure and increased renal vascular resistance (RVR) is believed to promote renal failure in patients with sepsis [2] or hepatorenal syndrome [3]. The adverse effects of several “nephrotoxic” agents, including contrast media, on kidney function are also considered to be partially due to a temporary reduction in RBF [1]. Some studies have indicated that a reduction in RBF persists during the course of ARF [4, 5], while others have reported that RBF is not an important determinant of the glomerular filtration rate (GFR) depression during the maintenance phase of postischemic ARF [6, 7]. Therefore, in different clinical situations, monitoring of renal perfusion may be relevant, not only to characterize further the association between changes in renal hemodynamics and GFR, but also to identify and prevent ischemic insults to the kidney. Currently available methods for the measurement of RBF in humans have been useful to detect RBF abnormalities in various settings, but are characterized by several methodological problems that limit their accuracy and applicability. The urinary clearance of p-aminohippurate (PAH) [8] cannot be applied to oligo-anuric subjects, and in postischemic renal failure, PAH extraction is severely impaired so that the PAH clearance markedly underestimates renal plasma flow [6]. Dye dilution techniques require the simultaneous catheterization of the renal artery and vein and allow only bolus measurements [9]. For isotope techniques, the need for external counting devices limits their bedside use. Moreover, the renal uptake of commonly used radiopharmaceuticals, such as 131I-orthoiodohippurate and 99mTc-labeled mercaptoacetyltriglycerine, is not only determined by renal perfusion, but is also affected by impaired trans-tubular trans-

1439

1440

Mo¨ckel et al: RBF by Doppler guidewires

port in ARF [6]. Electron beam computer tomography is a noninvasive method that has a good potential to measure not only blood flow, but distribution of blood flow within the renal parenchyma [10]. However, it also does not allow more prolonged monitoring and requires contrast media. Thermodilution catheters have been used to measure blood flow in renal veins [11, 12], but complete turbulent mixing of injectate and blood in the renal vein requires injection rates of up to 100 mL/min [12]. Transcutaneous Doppler flow measurements are noninvasive and have a much better time resolution than the techniques mentioned previously in this article [5, 13, 14]. However, this method determines only relative changes in flow indices, which have not been validated against absolute flow measurements. In addition, transabdominal duplex sonography is observer dependent, and in situations in which patients are at risk of developing ARF, the conditions are far from optimal. Volumetric blood flow measurements that use a flow probe attached to the outer vessel circumference provide accurate and continuous estimates of RBF, but in humans, this technique can only rarely be applied under very selected circumstances [6]. In view of these limitations, we have decided to evaluate the potential usefulness of intravascular Doppler flow measurements, which use a guidewire with a Doppler transducer mounted at the tip, as a novel approach for continuous and prolonged monitoring of renal perfusion. The use of Doppler guidewires for blood flow velocity measurements has been established in coronary diagnostic procedures and provides an excellent surrogate measurement of volumetric coronary flow [15]. A close correlation between flow velocity determined by intravascular Doppler guidewire and volumetric blood flow has also been shown in the carotid artery of pigs [16]. Recently, the feasibility of using intravascular Doppler technology in renal arteries has been demonstrated in pigs [17] and in humans under resting conditions [18] and after angioplasty [19]. However, the relationship between flow velocity and absolute blood flow in renal vessels has not been investigated. If such a relationship is sufficiently close, flow velocity measurements might be useful to assess changes in renal perfusion under experimental and selected clinical circumstances. This preclinical study was therefore performed to evaluate the level of correlation between flow velocity measurements in renal arteries and veins of pigs with volumetric RBF as measured by extravascular Transonic flow probes. Applying this technique, we have studied timeand concentration-dependent changes in renal perfusion after the administration of norepinephrine (NE). This ␣-adrenergic catecholamine is widely used to maintain perfusion pressure in shock syndromes [20], but controversial evidence exists about its effects on renal hemodynamics.

METHODS The study protocol was approved by the state authority responsible for animal experiments (authorization code: G0353/98) and conforms to the principles of the American Physiological Society and the American Heart Association. Animals and instrumentation Ten female country pigs (weight range of 24 to 31 kg) were studied. They were anesthetized by intravenous thiopental sodium (10 mg/kg) after taking nothing by mouth for at least 12 hours. Atropine (0.05 mg/kg) and azaperon (4 to 5 mg/kg, Stresnil威) were given intramuscularly 30 minutes prior to anesthesia, and etomidate (2.5 to 3.5 mg/kg intravenously) was used for sedation. An 18-gauge intravenous line was placed in auricular veins of both ears for continuous infusion of etomidate (at 0.7 to 1.0 mg/kg/hour) and fentanyl (at 0.025 mg/kg/ hour). Repetitive doses of Pancuronium威 4 mg were given intravenously as necessary to prevent spontaneous movement. The trachea was intubated with a Hi-Lo cuffed endotracheal tube with an interior diameter of 7.0 to 8.0 mm (Mallinckrodt Medical, Athlone, Ireland). A constant-volume ventilator (Servo Ventilator 900 D, Siemens Elena, Sweden) accomplished ventilation at a tidal volume of 10 to 15 mL/kg. Enriched inspired O2 [fraction of inspired oxygen (Fio2)] 40% and NO2 were given, and arterial blood gases (ABL TM 505; Radiometer, Copenhagen, Denmark), potassium, and hemoglobin concentrations (OSM娃 Hemoximeter, Radiometer, Copenhagen, Denmark) were periodically monitored. Changes in ventilatory frequency were made as necessary to maintain an arterial Paco2 of between 38 and 42 mm Hg and an arterial pH of between 7.37 and 7.42. A standard lead II electrocardiogram was used to monitor the heart rate. Saline filled sheaths were introduced to both femoral arteries (6.5 French) and veins (8.5 French) using the Seldinger technique after surgical preparation of the vessels. To avoid blood loss from the vessel puncture sites, the sheaths were secured by normal adaptation with 3 mm Mersilene威 strings. A saline-filled 5 French polyethylene catheter was advanced into the abdominal aorta from the right femoral arterial sheath for the measurement of ABP. A saline-filled 7.5 French Swan-Ganz continuous cardiac output (CCO)/oximetry thermodilution catheter (Baxter Healthcare Corporation, Irvine, CA, USA) was inserted through the right venous sheath and advanced, flow-directed into the right atrium and right pulmonary artery. The placement of the catheter in the pulmonary artery was verified by x-ray study. The right renal artery was isolated after median laparotomy and a 4-mm ultrasonic flow probe (2.4 MHz; Transonic Systems, Ithaca, NY, USA) was placed around

Mo¨ckel et al: RBF by Doppler guidewires

it to measure volumetric renal blood flow (VBF). The flow probe was fixed in place and sutured at the retroperitoneum for stability, and the connecting cable was led out through a small lateral cut in the abdominal skin. Ultrasonic gel was placed around the flow probe to minimize movement artifacts and to improve the acoustic signal. The abdomen was closed with interrupted sutures. All animals received a 0.9% saline infusion at a flow rate of 10 mL/kg/hour. Additional volume (saline 0.9 or 6% hydroxy-ethyl-starch solution) was infused to maintain the central venous pressure at between 5 and 10 mm Hg. A 4 French multipurpose guiding catheter (Tempo娃 Cobra 1; Cordis, Waterloo, Belgium) was directed from the left femoral arterial sheath into the right renal artery. This catheter nearly occluded the vessel. A 0.018 inch (0.46 mm) Doppler guidewire (FloWire威; Cardiometrics Inc., Mountain View, CA, USA) was introduced through the guiding catheter into the renal artery and placed under x-ray control precisely at the position of the flow probe. This catheter carried a 12 MHz pulsed wave Doppler probe with a sample volume of 5.2 mm to measure renal artery flow velocity (FVart). The Doppler catheter was connected with a FloMap威 ultrasound machine (Cardiometrics, Inc.). After the Doppler guidewire placement, the guiding catheter was drawn back into the aorta. This procedure was completed in less than 60 seconds to avoid ischemic kidney damage caused by the decrease of blood flow induced by the guiding catheter. A 5 French multipurpose catheter (Tempo娃 Cobra 1; Cordis) was directed from the left femoral venous sheath into the right renal vein under x-ray control. Another 0.018 inch Doppler guidewire was introduced in the renal vein to measure renal venous flow velocity (FVvein). As the renal vein lumen in pigs is much larger than the renal artery lumen, the renal vein catheter was left in place to improve the Doppler guidewire placement in the middle of the vessel and to minimize wall and movement artifacts. The vascular catheters were connected to low-displacement transducers (disposable transducer DT-XX; Becton Dickinson, Heidelberg, Germany). The reference point for all vascular pressures was the midcardiac plane. Heart rate, blood, central venous and pulmonary artery pressures were displayed on a standard Hewlett Packard patient monitor (M104AIB; HP, McMinnville, OR, USA). Continuous cardiac output and mixed venous oxygen saturation (Svo2) data and blood temperatures were displayed on a Vigilance Monitor (Baxter Healthcare Corporation). Online registration software development and postprocessing data analysis To enable online waveform comparison and postprocessing beat-to-beat comparison of flow and velocity curves, special software for multiple signal source inte-

1441

gration was developed. This software runs on a standard laptop computer (TECRA 8000; Toshiba Europe, Neuss, Germany; Pentium威 II, 300 MHz processor) and allows sampling at a rate of up to 1 kHz. The data acquisition software MODA 2.0 were compiled by the use of the Borland C⫹⫹ Builder V4 Professional (Inprise Corporation, Scotts Valley, CA, USA), which enables the recording and display of up to eight analog channels simultaneously. Interfacing (connection) of the different analog devices was realized by BNC cable to an input connector box, which was connected to a PCMCIA analog/digital board (PCM-DAS16/16娃; Computer Boards, Inc., Middleboro, MA, USA) inside the laptop. Cardiac output, body temperature, and central venous saturation were transmitted via a serial port cable from the Vigilance Monitor. For the purpose of postprocessing analysis of the data MATLAB V5.3 (Scientific Computers, Aachen, Germany) scripts were employed. These scripts were used to calculate the time integrals of flow, flow velocity, and pressure for each heart beat. The difference between average peak velocity (APV) provided by the FloMap and the calculated flow-velocity integral is only related to the integration time, which is two heart beats or two seconds in the Cardiometrics device. Further scripts were employed for creating diagrams showing flow, flow velocity, and pressure curves for individual as well as all 10 pigs. The resulting values were saved in ASCII format and further calculations were performed using Excel 97 (Microsoft) software (discussed later in this article). Experimental protocol The animals were prepared as described previously in this article. Baseline values and calibration signals were recorded first. NE (Hoechst, Frankfurt/Main, Germany) was administered intravenously as a 2 to 5 mL bolus in saline via the femoral venous sheath. The NE dose was increased by the following steps: 2 ␮g, 10 ␮g, 5 times then 20, 50, 100, and 200 ␮g. Data that included the renal artery flow velocity, and in parallel VBF, were recorded after injection of the bolus until renal flow had returned to baseline values. The NE application was repeated to assess the renal venous flow velocity. At the conclusion of the experiment, the animals were killed by intravenous injection of KCl. Statistical analysis and calculations Data are displayed as medians and 25th/75th percentiles. Linear correlation coefficients (Pearson) were calculated for the comparison of VBF and renal flow velocity (mean velocity calculated over one cardiac cycle; discussed previously in this article). The Kruskal–Wallis test was used to calculate significance in a series of measurements (Table 1). RVR was calculated using this formula:

1442

Mo¨ckel et al: RBF by Doppler guidewires

Table 1. Effect of five repetitive applications of norepinephrine 20 ␮g on renal arterial flow velocity (intra-arterial flow wire) in eight pigs Application no.

1

2

3

4

5

P value

FVartmin cm/sec

18.1 6.8/34.4 10.0 9.3/12.3 29.5 26.8/39.0

16.5 7.0/31.2 10.5 9.0/12.8 27.5 25.5/35.3

19.3 5.7/29.1 10.0 8.3/13.3 32.0 27.5/37.5

19.5 9.2/28.2 10.5 9.0/13.8 32.0 24.3/36.5

23.1 5.2/31.7 10.5 9.0/13.0 30.5 28.3/35.8

0.985

tmin sec t90% sec

0.991 0.860

Values are medians and 25%/75% percentiles. Abbreviations are: FVartmin, minimum renal arterial flow velocity after the application of norepinephrine (NE) 20 ␮g intravenously; tmin, time interval until minimum FVart was reached after application of NE; t90%, time at which renal blood flow reduction returned to 90% of baseline. P values are from the Kruskal-Wallis test.

RVR [mm Hg · min/L] ⫽

MABP ⫺ CVP VBF

where VBF is the mean volumetric renal blood flow (L/min), MABP is the mean aortic blood pressure (mm Hg), and CVP is the central venous pressure (mm Hg). Systemic vascular resistance (SVR) was calculated using this formula: SVR[mm Hg · min/L] ⫽

MABP ⫺ CVP CO

where CO is the cardiac output (L/min). Diagrams were drawn by the use of Excel 97威 (Microsoft Corporation, Redmond, WA, USA) and Matlab威 V5.3 (Scientific Computers, Aachen, Germany) software. In Figure 2, the mean values of one cardiac cycle every five seconds of each animal were used to calculate median and interquartile range of values. RESULTS The placement of extravascular flow probes and intravascular Doppler flow guidewires in the renal arteries and veins was easily established in all animals investigated. Under baseline conditions median (25th/75th percentiles) VBF was 142 mL/min (116/152, 5.3 mL/kg/minute, 4.1/6.0, N ⫽ 10). These values corresponded to median FVart of 81 cm/sec (50/101, N ⫽ 10). FVvein was approximately 10 times lower, and the median value was 8.3 cm/sec (7.0/11.8, N ⫽ 8). Under all conditions investigated, changes in VBF were paralleled by changes in FV. As demonstrated in initial recording (Fig. 1A), simultaneously registered curves of VBF and FVart were superimposed and virtually indistinguishable from each other, unless the time scale was stretched (Fig. 1B). The slight phase displacement of the trace of ABP of approximately 0.3 of a second, which became apparent with a change of scale of the time axis in Figure 1B, was due to the travel time of the pulse wave in the fluid filled systems that were used for blood pressure recordings. Cyclic alterations of

ABP with a frequency of approximately 14 per minute were respiration associated and led to parallel cyclic changes in renal perfusion (Fig. 1A). Figure 1A depicts the parallel recording of ABP, VBF, and FVart following the bolus application of NE 20 ␮g at zero time. It can be seen that this resulted in an increase in heart rate and a rise in blood pressure after approximately five seconds, and a rapid decline in renal perfusion after approximately eight seconds. The reduction in renal perfusion resulted in a brief cessation of diastolic flow, but this was completely reversible approximately 30 seconds after the injection. In contrast, the elevation in ABP persisted for approximately two minutes. This effect of NE was highly reproducible in the cohort. The time course of changes in VBF, FVart, RVR, systemic vascular resistance (SVR), and MABP in all 10 animals are summarized in Figure 2. The effect of NE on the systemic vasculature lasted longer when compared with the renal vessels (Fig. 2, D and E). When the effect of incremental doses of NE was tested, a temporary reduction of renal perfusion was observed over the entire dose range that was investigated (2 to 200 ␮g). With an increasing dose of NE, both the extent and duration of reduction in RBF increased. This is shown in Figure 3A, which illustrates median renal arterial flow velocity in 10 pigs 10 and 25 seconds after bolus applications of NE. RVR 10 seconds after bolus application, as calculated from VBF and mean arterial pressure, was exponentially related to NE dose and increased by up to 14-fold (Fig. 3C). After the maximum dose of NE (200 ␮g), RVR increased from baseline (988 mm Hg · min/L) to a maximum value of 13711 mm Hg · min/L 10 seconds after intravenous NE application (Fig. 3C). To test for possible changes in renal vascular responsiveness to NE, repetitive bolus applications of NE 20 ␮g were given to each animal after time intervals of two to five minutes. As shown in Table 1, NE effects were highly reproducible and did not change with time. When comparing FV measurements in the renal artery and vein, parallel changes were observed under all conditions investigated and both arterial and FVvein correlated with VBF. As an example, the upper panel of Figure 4 shows

Mo¨ckel et al: RBF by Doppler guidewires

1443

Fig. 1. Original recording of arterial blood pressure (ABP), renal arterial flow velocity (FVart), and volumetric renal blood flow (VBF) following the application of norepinephrine (NE) 20 ␮g at zero time. As the flow and the velocity curve are superimposed (A), in (B) the time scale was increased to show the beat-to-beat comparison of the variables at higher resolution.

Fig. 2. Time course of volumetric renal blood flow (VBF), renal arterial flow velocity (FVart), renal vascular resistance (RVR), systemic vascular resistance (SVR), and mean arterial blood pressure (MABP) following the application of NE 20 ␮g in 10 animals (median and interquartile range). RVR was calculated from the pressure difference (MABPcentral venous pressure) and VBF. SVR was calculated from the pressure difference (MABP-central venous pressure) and cardiac output, as measured with a Swan-Ganz continuous cardiac output thermodilution catheter.

1444

Mo¨ckel et al: RBF by Doppler guidewires

investigated (Table 2). When comparing the relationship between arterial and venous FV measurements and VBF (Fig. 4), it became obvious that flow velocity in the vein is in the range of values in which the correlation is less optimal (that is, below 15 cm/sec). Thus, although FVvein and (arterial) VBF were still significantly correlated, median correlation coefficients between flow and velocity were always lower than in the renal artery, and the variation between animals was greater (Table 2).

Fig. 3. Dose dependency of the effect of norepinephrine (NE) on renal arterial flow velocity (FVart, A), mean arterial blood pressure (MABP, B), and renal vascular resistance (RVR, C). RVR was calculated from renal artery flow (extravascular flow probe) and the pressure difference (MABP-central venous pressure). Symbols are: (䊉) values 10 seconds after the application of NE; (䊊) values 25 seconds after the application of NE. Note that the reduction in FVart was partially reversible after 25 seconds, although the MABP tended to be higher than after 10 seconds.

the correlation between blood flow and FVart in one animal using values that were recorded during the application of different doses of NE. While the relationship was continuously linear above a RBF of 0.05 L/min, it tended to flatten with lower blood flow rates. Similar relationships were found in all of the animals. The regression curves were not exactly superimposed due to interindividual variability of the renal artery diameter. In all animals, however, the correlation coefficients between FVart and VBF were close to 1.0 under all conditions

DISCUSSION This study shows that the insertion of Doppler flow guidewires into the main renal vessels enables continuous monitoring of renal perfusion and provides an easy way to accurately detect and quantitate changes in RBF. Although a significant correlation between VBF and FV could be demonstrated in the renal vein, the much closer relationship between VBF and FVart favors the use of Doppler guidewires primarily in the arterial system. Importantly, FVart was linearly related to VBF over a broad range of values, even to as low as one seventh of the baseline flow rate. It appeared to become somewhat less accurate only with flow rates below 50 mL/min. Since the relationship between total RBF and FV depends on the size of the renal artery, these results may not be directly extrapolated to humans. However, when related to body weight, renal hemodynamics in pigs is similar to those of humans [21]. Previous studies have found flow velocities of up to 60 cm/sec (peak systolic velocity, which corresponds to an APV of ⬃30 cm/sec) in nonstenotic human renal arteries [22]. These values in human renal arteries fit well within the linear range established in our experiments. In contrast to transabdominal ultrasound, the technique is invasive and if applied to renal arteries may theoretically create vasospasm and damage to the endothelium. However, given the small dimensions of the guide wire and its flexibility, we believe that the risks are small. Under selected circumstances, the technique therefore warrants further careful evaluation. Particularly in patients in whom other hemodynamic parameters are already invasively monitored and who are at a predictable risk for developing renal dysfunction, such as patients undergoing cardiac surgery [23], this technique may reveal information about the changes in renal hemodynamics that cannot be obtained with any other method previously used to measure RBF in humans. In addition, the efficacy of certain drugs that have been shown to ameliorate renal perfusion in animal experiments, such as endothelin antagonists [24], could be directly assessed in humans. In brain-dead organ donors, the monitoring of renal perfusion might be helpful to guide supportive therapy toward an optimal preservation of renal function. For such indications, the fact that only relative

1445

Mo¨ckel et al: RBF by Doppler guidewires

Fig. 4. Correlation of volumetric renal blood flow (VBF) and flow velocity in one experimental animal (using data obtained for all NE doses). Correlation with (A) renal arterial flow velocity (FVart) and (B) renal venous flow velocity (FVvein). Because of the larger diameter of the renal vein, the renal venous flow velocity is approximately 10 times lower. Correlation coefficients for all animals investigated are listed in Table 2.

Table 2. Correlation coefficients between simultaneous measurements of volumetric renal blood flow (extravascular flow probe) and arterial or venous blood flow velocity (intravascular Doppler guide wire) following the application of increasing doses of norepinephrine in 10 pigs

Arterial flow velocity Median (Min–max) Venous flow velocity Median (Min–max)

2 ␮g NE

10 ␮g NE

20 ␮g NE

50 ␮g NE

100 ␮g NE

200 ␮g NE

0.98 (0.55–1.00)

0.98 (0.88–1.00)

1.00 (0.95–1.00)

0.99 (0.91–1.00)

0.99 (0.95–1.00)

0.99 (0.73–1.00)

0.51 (⫺0.09–0.92)

0.66 (⫺0.32–0.90)

0.81 (⫺0.08–0.89)

0.86 (0.35–0.88)

0.89 (0.37–0.92)

0.80 (0.33–0.88)

Data are Pearson correlation coefficients. Abbreviations are: Min, minimum; max, maximum; NE, norepinephrine intravenously.

changes in perfusion can be detected by flow velocity measurements is not a significant limitation. Moreover, if information about “absolute” flow rates is required, this can be calculated by combining flow velocity recordings with a technique that gives an approximate estimate of the cross-sectional area of the vessel. This is possible with intravascular ultrasound (IVUS), which is frequently used in coronary arteries, in combination with Doppler ultrasound and by angiography [17, 18]. Such calculated “absolute” flow rates may, however, overestimate the RBF because of the laminar flow profile. It should be noted that close correlations between flow and velocity found in the present study during vasoconstriction of small renal arteries may not be true for flow

reductions related to renal artery stenosis because of the existence of turbulent movement. Some of the advantages of the flow wire technique, combined with other hemodynamic measurements, have been shown in this study by assessing the effects of NE on RBF. NE is an ␣-adrenergic catecholamine that has previously been shown to reduce RBF after intra-arterial injection [25, 26]. NE-induced renal ischemia has been used as a model for ARF in rats [27] and dogs [28]. However, the effects of NE on the renal vasculature after systemic (intravenous) application are less clear. Several studies reported that RBF may be enhanced by NE in healthy animals [29, 30], and in models of endotoxin shock, the RBF was reported to be enhanced after

1446

Mo¨ckel et al: RBF by Doppler guidewires

NE infusion [31]. In critically ill patients, the efficacy [32–34] and safety [35, 36] of NE in treating refractory hypotension in septic shock has been demonstrated, but knowledge about its visceral vasoconstriction properties has raised concerns about the advisability of its sustained infusion in hemodynamically unstable patients. Part of the discrepancy of the effects of intrarenal arterial and systemic application of NE on RBF may theoretically result from the elevation of arterial perfusion pressure (after systemic application) that counterbalances increases in RVR. This balance would be particularly relevant if the dose dependency of the effects of NE on systemic and RVR were different. When testing this hypothesis by applying incremental doses of NE, we were unable to find a dose that increased renal perfusion through a preferential effect on perfusion pressure (Fig. 3). In contrast, temporary renal vasoconstriction was found over the whole dose range investigated. In addition, by using repetitive bolus applications, we were unable to obtain any evidence of a reduction in the renal vasoconstrictive effect of NE (Table 1), although this finding does not exclude that adaptive mechanisms may occur with more prolonged and continuous application of the drug. It is of note, however, that by using bolus applications, the systemic and renal hemodynamic effects can be separated in so far as the effect of NE on ABP slightly preceded the reduction in RBF (Fig. 1A). In addition and more importantly, it persisted considerably beyond the recovery of renal perfusion (Figs. 1 and 2). This suggests that NE might be cleared from its receptors in the kidney more rapidly than in other vascular beds that determine SVR or that counter-regulatory mechanisms are more effective in the kidney than in other parts of the vascular system. Previous studies have shown that nitric oxide (NO) attenuates the renal vasoconstrictive effect of NE [25, 26, 37]. It is tempting to speculate that the overproduction of NO in septic shock explains that, unlike the effects under baseline conditions, NE infusion during endotoxic shock can actually increase RBF and that this effect is not the result of an increase in perfusion pressure alone, as recently demonstrated in dogs [31]. ACKNOWLEDGMENTS The study was supported by University Research Funds. We thank Dr. M. Meißler and his team at the Institute for Animal Experiments, Charite´/Virchow-Klinikum (Berlin, Germany) for handling the animals prior to the experiments and for his expert help in oro-tracheal intubation of the pigs. Preliminary results were presented at the annual meeting of the German Society of Nephrology (Freiburg, Germany) in 1999 and were published in abstract form (Kidney Blood Press Res 22:323, 1999; J Am Soc Nephrol 10:384A, 1999). This article contains data from the doctoral thesis of D. Maier. Ms. T. Derwent (German Heart Center, Berlin, Germany) kindly performed a final language editing of the manuscript. Reprint requests to Martin Mo¨ckel, M.D., and Kai-Uwe Eckardt, M.D., Charite´/Campus Virchow-Klinikum, Department of Nephrology and Medical Intensive Care, Augustenburger Platz 1, 13353, Berlin, Germany. E-mail: [email protected] and [email protected]

APPENDIX Abbreviations used in this article are: ABP, arterial blood pressure; APV, average peak velocity; ARF, acute renal failure; CCO, continuous cardiac output; FIO2, fraction of inspired oxygen; FVart, flow velocity in renal arteries; FVvein, flow velocity in renal veins; GFR, glomerular filtration rate; IVUS, intravascular ultrasound; NE, norepinephrine; NO, nitric oxide; PAH, para-aminohippurate; RBF, renal blood flow; RVR, renal vascular resistance; SvO2, oximetry; SVR, systemic vascular resistance; VBF, volumetric blood flow.

REFERENCES 1. Thadhani R, Pascual M, Bonventre JV: Acute renal failure. N Engl J Med 334:1448–1460, 1996 2. Khan RZ, Badr KF: Endotoxin and renal function: Perspectives to the understanding of septic acute renal failure and toxic shock. Nephrol Dial Transplant 14:814–818, 1999 3. Eckardt KU: Renal failure in liver disease. Intensive Care Med 25:5–14, 1999 4. Hollenberg NK, Adams DF, Oken DE, et al: Acute renal failure due to nephrotoxins. N Engl J Med 282:1329–1334, 1970 5. Stevens PE, Gwyther SJ, Hanson ME, et al: Noninvasive monitoring of renal blood flow characteristics during acute renal failure in man. Intensive Care Med 16:153–158, 1990 6. Corrigan G, Ramaswamy D, Kwon O, et al: PAH extraction and estimation of plasma flow in human postischemic acute renal failure. Am J Physiol 277:F312–F318, 1999 7. Reubi FC, Gossweiler N, Gu¨rtler R: Renal circulation in man studied by means of a dye-dilution method. Circulation 33:426–442, 1966 8. Battilana C, Zhang HP, Olshen RA, et al: PAH extraction and estimation of plasma flow in diseased human kidneys. Am J Physiol 261:F726–F733, 1991 9. Reubi FC, Gu¨rtler R, Gossweiler N: A dye dilution method of measuring renal blood flow in man, with special reference to the anuric subject. Proc Soc Exp Biol Med 111:760–764, 1962 10. Lerman LO, Taler SJ, Textor SC, et al: Computed tomographyderived intrarenal blood flow in renovascular and essential hypertension. Kidney Int 49:846–854, 1996 11. Brenner M, Schaer GL, Mallory DL, et al: 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, 1990 12. Haywood GA, Stewart JT, Counihan PJ, et al: Validation of bedside measurements of absolute human renal blood flow by a continuous thermodilution technique. Crit Care Med 20:659–664, 1992 13. Greene ER, Venters MD, Avasthi PS, et al: Noninvasive characterization of renal artery blood flow. Kidney Int 20:523–529, 1981 14. Stevens PE, Bolsin S, Gwyther SJ, et al: Practical use of duplex Doppler analysis of the renal vasculature in critically ill patients. Lancet 1:240–242, 1989 15. Kern MJ, Bach RG, Mechem CJ, et al: Variations in normal coronary vasodilatory reserve stratified by artery, gender, heart transplantation and coronary artery disease. J Am Coll Cardiol 28:1154–1160, 1996 16. Chaloupka JC, Vinuela F, Kimme SC, et al: Use of a Doppler guide wire for intravascular blood flow measurements: A validation study for potential neurologic endovascular applications. Am J Neuroradiol 15:509–517, 1994 17. Beregi JP, Lahoche A, Willoteaux S, et al: Renal artery vasomotion: In vivo assessment in the pig with intravascular Doppler. Fundam Clin Pharmacol 12:613–618, 1998 18. Savader SJ, Lund GB, Osterman FAJ: Volumetric evaluation of blood flow in normal renal arteries with a Doppler flow wire: A feasibility study. J Vasc Interv Radiol 8:209–214, 1997 19. Savader SJ, Lund GB, Venbrux AC: Doppler flow wire evaluation of renal artery blood flow before and after PTA: Initial results. J Vasc Interv Radiol 9:451–460, 1998 20. Levy B, Bollaert PE, Charpentier C, et al: Comparison of norepinephrine and dobutamine to epinephrine for hemodynamics, lactate metabolism, and gastric tonometric variables in septic shock:

Mo¨ckel et al: RBF by Doppler guidewires

21. 22. 23. 24.

25.

26.

27.

28.

29.

A prospective, randomized study. Intensive Care Med 23:282–287, 1997 Link L, Weidmann P, Probst P, et al: Renal handling of norepinephrine and epinephrine in the pig. Pflu¨gers Arch 405:66–69, 1985 van Der Hulst V, Van Baalen J, Kool LS, et al: Renal artery stenosis: Endovascular flow wire study for validation of Doppler US. Radiology 200:165–168, 1996 Llopart T, Lombardi R, Forselledo M, et al: Acute renal failure in open heart surgery. Ren Fail 19:319–323, 1997 Roux S, Qiu C, Sprecher U, et al: Protective effects of endothelin receptor antagonists in dogs with aortic cross-clamping. J Cardiovasc Pharmacol 34:199–205, 1999 Fitzgerald RD, Dechtyar I, Templ E, et al: Cardiovascular and catecholamine response to surgery in brain-dead organ donors. Anaesthesia 50:388–392, 1995 Matsumura Y, Egi Y, Maekawa H, et al: Enhancement of norepinephrine and angiotensin II-induced renal effects by NG-nitro-larginine, a nitric oxide synthase inhibitor. Biol Pharm Bull 18:496– 500, 1995 Conger JD, Robinette JB, Guggenheim SJ: Effect of acetylcholine on the early phase of reversible norepinephrine-induced acute renal failure. Kidney Int 19:399–409, 1981 Burke TJ, Cronin RE, Duchin KL, et al: Ischemia and tubule obstruction during acute renal failure in dogs: Mannitol in protection. Am J Physiol 238:F305–F314, 1980 Visscher CA, De Zeeuw D, De Jong PE, et al: Drug-induced

30.

31. 32. 33. 34. 35. 36. 37.

1447

changes in renal hippurate clearance as a measure of renal blood flow. Kidney Int 48:1617–1623, 1995 Schaer GL, Fink MP, Parrillo JE: Norepinephrine alone versus norepinephrine plus low-dose dopamine: Enhanced renal blood flow with combination pressor therapy. Crit Care Med 13:492–496, 1985 Bellomo R, Kellum JA, Wisniewski SR, et al: Effects of norepinephrine on the renal vasculature in normal and endotoxemic dogs. Am J Respir Crit Care Med 159:1186–1192, 1999 Meadows D, Edwards JD, Wilkins RG, et al: Reversal of intractable septic shock with norepinephrine therapy. Crit Care Med 16:663–666, 1988 Desjars P, Pinaud M, Potel G, et al: A reappraisal of norepinephrine therapy in human septic shock. Crit Care Med 15:134–137, 1987 Redl WE, Armbruster C, Edelmann G, et al: The effects of norepinephrine on hemodynamics and renal function in severe septic shock states. Intensive Care Med 19:151–154, 1993 Desjars P, Pinaud M, Bugnon D, et al: Norepinephrine therapy has no deleterious renal effects in human septic shock. Crit Care Med 17:426–429, 1989 Martin C, Saux P, Auffray JP, et al: Thermodilution cardiac output measurements by injection in pulmonary artery vs. CVP catheter. Crit Care Med 11:460–461, 1983 Parekh N, Dobrowolski L, Zou AP, et al: Nitric oxide modulates angiotensin II- and norepinephrine-dependent vasoconstriction in rat kidney. Am J Physiol 270:R630–R635, 1996