Patients who develop acute renal failure (ARF) before or after liver transplantation (LTx) have a six- to eightfold more frequent mortality compared with those who undergo LTx and do not develop this complication (1). The need for postoperative renal replacement therapy (continuous venous-venous hemofiltration, and/or dialysis) is associated with a mortality up to 60% (2). Preoperative poor renal function (serum creatinine [CRE] >1.4 mg/dL) is the strongest significant predictor for postoperative ARF (3). Many drugs have been used for renal protection during major surgery, most commonly, dopamine, furosemide, and mannitol (4,5). Fenoldopam is a parenteral direct-acting systemic and renal vasodilator that was initially used for treatment of hypertension, but is now under evaluation at smaller doses as a renal protectant. Fenoldopam selectively binds to DA1 receptors without any interaction with DA2 and β1 receptors. At larger concentrations than those required to activate DA1 receptors, fenoldopam is an α receptor antagonist, with greater activity at α2 than at α1 receptors (6–8). Fenoldopam exerts renal vasodilatory effects at doses of ≥0.01 μg · kg−1 · min−1 (9,10), plateauing at 0.5 μg · kg−1 · min−1 (10,11). In contrast, systemic vasodilatory effects occur at doses of ≥0.1 μg · kg−1 · min−1 (11). Glomerular filtration increases or is maintained during fenoldopam infusion (10).
The primary objective of this study was to evaluate the incidence of intraoperative and postoperative renal failure and/or dysfunction in patients undergoing LTx comparing fenoldopam to small-dose dopamine. The second objective was to evaluate the urine output and the number of “interventions” of furosemide and/or mannitol needed to maintain urine output >1 mL · kg−1 · h−1.
This was a randomized, active-controlled nonblinded trial in which patients were randomized into two treatment groups: group fenoldopam (n = 22) patients received fenoldopam 0.1 μg · kg−1 · min−1 and group dopamine (n = 21) received dopamine 2 μg · kg−1 · min−1. Both drugs were given as a continuous IV infusion starting after the induction of anesthesia and were discontinued 48 h after the end of surgery. Patients with CRE >1.4 mg/dL were excluded because the study was designed to evaluate patients with normal renal function.
After approval of the study protocol by the Institutional Ethics Committee and obtaining patients’ signed written informed consent, 43 patients (35 men and 8 women) scheduled for LTx were consecutively enrolled into the study. Patients with preexisting renal, (CRE >1.4 mg/dL) pulmonary (hypoxia and/or hypercarbia), and/or cardiac (coronary artery disease and/or previous myocardial infarction) diseases were excluded.
The same anesthetic management was applied to all individuals to mitigate the influence of anesthetic drugs.
All patients were monitored with a lead II/V5 electrocardiograph, pulse oximetry (Spo2), and radial artery catheterization for invasive mean arterial blood pressure (MAP) (PCM SpaceLabs, Inc., Redmond, WA), pulmonary artery catheter (Intellicath®; Edwards Laboratories, Irvine, CA) for intermittent conventional pulmonary artery thermodilution cardiac output, continuous cardiac output, and mixed venous oxygen saturation measurement and to monitor central venous pressure (CVP), mean pulmonary artery pressure (mPAP), pulmonary artery occlusion pressure (PAOP), and body temperature.
In all patients undergoing LTx, a 4F thermistor-tipped catheter (Pulsiocath PV2014L®; Pulsion Medical Systems, Munich, Germany) was placed through the right femoral artery, and connected to the PiCCO® System (Pulsion Medical Systems) to monitor transpulmonary cardiac index, intrathoracic blood volume index (ITBVI normal values 800–1000 mL/m2), and extravascular lung water index (EVLWI n.v. 4–7 mL/kg).
For guidance of volume loading, ITBVI <800 mL/m2 was corrected with volume replacement whereas ITBVI >1000 mL/m2 required the use of diuretic therapy; EVLWI was maintained <7 mL/kg. In addition, standard hemodynamic monitoring was used to optimize cardiovascular function (cardiac index >3 L · min−1 · m−2, CVP approximately 8 cm H2O, PAOP ≥12 mm Hg, mean arterial pressure >70 mm Hg, and mPAP >18 mm Hg). The intraoperative goal to maintain urine output at ≥1 mL · kg−1 · h−1 was: if urine output decreased to <1 mL · kg−1 · h−1, a mannitol 18% IV bolus (0.3–0.5 g/kg) was administered and, if the goal was still not reached, furosemide 0.3 mg/kg was given.
All volumetric and pressure-derived variables were indexed to the body surface area (BSA) to improve the interindividual comparison.
After the end of surgery, all patients were transferred to the intensive care unit (ICU).
Postoperative analgesia was provided with sufentanil continuous infusion 4 μg/h for 48 h. IV cyclosporine immunosuppression was begun on postoperative day (POD) 1 at a dose of 2 mg/kg per day. Goal cyclosporine trough levels were 250–350 ng/mL during the first several PODs.
Hemodynamic-volumetric data (heart rate, MAP, CVP, PAOP, mPAP, cardiac index, ITBVI, and EVLWI), fluid-balance, and urine output (mL · kg−1 · h−1) were collected at baseline after the induction of anesthesia, during the anhepatic phase, and at the end of surgery. Use of mannitol and furosemide was recorded during all intraoperative and postoperative phases. Blood urea nitrogen (BUN) and CRE were collected preoperatively (Preop), at the end of surgery (Final), and on the third POD (3POD), and creatinine clearance (CreCl) was recorded Preop and on the 3POD.
CreCl was calculated using the following formula:
where UC is CRE concentration in the urine (mg/dL), UV is urine volume (mL/min), and CRE is the concentration in serum (mg/dL).
Based on the work of Bellomo et al. (3), we defined acute renal injury as a CRE >1.4 mg/dL and BUN >22 mg/dL or urine output <800 mL/24 h or <200 mL/6 h within 72 h of surgery. ARF syndrome (ARFS) was defined as a CRE >2.9 mg/dL and BUN >45 mg/dL or urine output <400 mL/24 h or <100 mL/6 h and severe ARFS was defined by the need for renal replacement therapy in the presence of acute renal injury or ARFS criteria within 72 h of surgery (3).
For this pilot study, a formal sample size was 20 patients per group. The power analysis revealed that the actual sample size had a 79% power to detect the observed difference in CRE change from baseline at 3POD with a 5% α error.
Descriptive statistics were obtained for all variables. Mean and standard deviation were used for continuous normally distributed variables, median and interquartile range (IQR) for continuous non-normally distributed variables, count and percentage for categorical variables.
The differences between Preop and 3POD (as well as differences between Preop and Final) in CRE, CreCl, and BUN were calculated.
The total number of mannitol and furosemide interventions needed to maintain adequate urine output for each patient was determined by adding all doses administered in each phase.
Because the measurements we made were objective, the study was not blinded throughout all the study.
All tests were two-tailed. Because of the small sample size, all between-groups comparisons were made by means of the Mann-Whitney test. Exact P values are presented.
To take into account repeated measures over time, generalized least squares models with autocorrelation of order 1 were used to assess differences between groups in terms of intraoperatory diuresis.
Analyses were performed with STATA (release 7.0, 2001; StataCorp, College Station, TX). The randomization list was generated using STATA, and the treatment allocation was concealed to investigators.
The two groups were similar in terms of age, BSA, sex, and Child-Pugh class (Table 1). Anesthesia time (Table 1) was similar in both groups. No differences were noted between groups for any of these variables considered at the Preop phase: CreCl, CRE, and BUN (Table 2). Hemodynamic and volumetric data (MAP, CVP, PAOP, mPAP, cardiac index, ITBVI, and EVLWI) were similar in both groups. Input and urine output were similar (Tables 2 and 3).
Mean cyclosporine trough levels were 310 (20) and 325 (25) mL/kg, respectively, in the fenoldopam and dopamine groups on 3POD. CRE, BUN, input, urine output, and CreCl are reported as mean and standard deviation, and the number of interventions of furosemide and mannitol are reported in Table 2.
Differences between Preop and 3POD in CRE, CreCl, and BUN are presented in Figure 1. Fenoldopam and dopamine patients were different only in terms of median CRE change from Preop at 3POD 0.2 (IQR −0.2–0.5) versus 0.5 (IQR 0.3–0.9) mg/dL, respectively (P = 0.004) and median BUN change from Preop at 3POD 2 (IQR −2–8) versus 8.5 (IQR 5–12) mg/dL, respectively (P = 0.01) (Fig. 1). CreCl was not different between the two groups (Table 2 and Fig. 1). Urine output in both groups was maintained to similar values (Table 2), however, to do so, significantly fewer fenoldopam patients required furosemide compared with dopamine patients; the median total number of furosemide interventions needed was 1 (IQR 0–3) in the fenoldopam versus 3 (IQR 2–4) in the dopamine group (P = 0.003), whereas the total mannitol doses needed were not different at 1 (IQR 0–1) versus 1 (IQR 1–1), respectively (P = 0.38) (Fig. 2). ICU and hospital length of stay were similar in the two groups (Table 1). All patients but one were tracheally extubated in the operating room at the end of surgery. Only the 5 patients who died stayed in ICU >48 h. Five patients, three in the dopamine group and two in the fenoldopam group, died because of primary nonfunction/multiple organ failure and/or sepsis. The data from the five patients who died were excluded from the results.
Our findings suggest that fenoldopam may have a role as a renal vasodilator in patients undergoing LTx. Patients receiving dopamine had significant increases in CRE and BUN after LTx but this was not observed in fenoldopam-treated patients. In our study, fenoldopam 0.1 μg · kg−1 · min−1 did not change systemic and pulmonary hemodynamics; therefore, the dose selected was renal-specific. Our findings are consistent with those of Halpenny et al. (12,13) who found that in patients undergoing coronary artery bypass grafting and in patients undergoing elective aortic surgery, CreCl decreased significantly (P < 0.01) in the groups randomized to placebo but not in those randomized to fenoldopam 0.1 μg · kg−1 · min−1 in the perioperative period (12,13). Furthermore, neither study by Halpenny et al. found significant changes in systemic hemodynamics with fenoldopam at a dose of 0.1 μg · kg−1 · min−1. In another randomized, controlled trial in patients undergoing LTx, Ramsay et al. (14) reported significant increases in CRE with placebo versus fenoldopam 0.03–0.1 μg · kg−1 · min−1 on the first POD. Furthermore, at this same time point, glomerular filtration rate was significantly increased with fenoldopam versus placebo. In this trial, the study drug was started at anesthesia induction and continued for 24 hours after LTx.
Urine output was maintained in both groups to similar values; however, to do so, significantly fewer fenoldopam patients required furosemide compared with dopamine patients; median 1 (IQR 0–3) versus 3 (IQR 2–4) (P = 0.003). Maintenance of urine output is protectant because a nonoliguric renal failure has been shown to have a survival benefit (15). The present prospective study suggests that the selective DA1 agonist, fenoldopam, results in significantly reduced need for loop diuretics compared with low-dose dopamine. Dopamine 2 μg · kg−1 · min−1 and not 3 μg · kg−1 · min−1 was chosen in this study because, at this dosage, the dopaminergic effects tend to predominate, although even at this small dose, wide variability can exist among patients and clinical conditions (16–18). Despite the difference in CRE and in BUN, there were no differences between ICU lengths of stay between the two groups.
To avoid confounding in this renal functional study by volume and systemic hemodynamics, care was taken to equalize hemodynamic-volumetric assessments based on PiCCO® monitoring, such that the effective intravascular blood volume was optimized in both groups. The essential elements of perioperative renal preservation are perioperative optimization of fluid status and cardiovascular performance, maintenance of renal perfusion, and avoidance of nephrotoxins.
Cyclosporine causes renal vasoconstriction and reduces renal blood flow that may contribute to chronic and acute nephrotoxicity. In our study, preservation of CRE was seen in patients even though they received cyclosporine. Our data are consistent with those of Jorkasky et al. (19) who demonstrated that fenoldopam reverses the renal vasoconstriction caused by cyclosporine in renal transplant patients, with a significant increase of renal plasma flow.
Furosemide, mannitol, and dopamine have all been used for renal protection with controversial results (20,21). Dopamine proved to be renoprotective in animal models of renal failure, but it failed in the clinical applications. In a retrospective study on patients undergoing LTx, prophylactic small-dose dopamine decreased the incidence of dialysis-dependent ARF from 27% to 9.5% (22) of patients, but later, Swygert et al. (23) failed to demonstrate that it had any beneficial effect on intraoperative urine flow, postoperative renal clearance, ARF, or mortality. Lassnigg et al. (21) demonstrated that continuous infusion of renal-dose dopamine failed to exert any advantage over placebo for renal vasodilatation in well hydrated patients after cardiac surgery; continuous infusion of furosemide was not only ineffective, but was even detrimental and induced renal dysfunction. A possible explanation for these controversial results is that in all animal models of renal failure there was a very high vasoconstrictive tone, because these models used either large norepinephrine infusion, or total occlusion of the renal artery to produce renal failure (24). However, in most clinical situations, kidneys never experience such an extreme vasoconstriction and, under these conditions, dopamine exerts its effects by acting on both DA1 and DA2 receptors, resulting in decreasing renal blood flow, glomerular filtration rate, and sodium excretion (25). These two classes of receptors have opposing effects on renal vasculature and this may explain the conflicting results in clinical trials. In fact, DA1 receptors are postsynaptic and, when activated, elicit vasodilation and inhibition of sodium-potassium adenosine triphosphatase consequently promoting diuresis and natriuresis, whereas DA2 receptors are less well understood. They are located presynaptically and, when activated, inhibit adenylate cyclase activity (in contrast to DA1 receptors) and norepinephrine release (18,26). Blockade of DA2 receptors increases renal blood flow and glomerular filtration rate, indicating that DA2 receptors activation decreases renal blood flow (26). In a meta-analysis of 15 studies containing 970 subjects, Marik (27) found that the incidence of renal dysfunction was 31% in the small-dose dopamine group compared with 33% in the control group. The study concluded that dopamine has no renal vasodilator effect; moreover, considering the potential side effects of the drug, there is little justification for the continued use of small-dose dopamine as a renal vasodilator drug. Holmes and Walley (28) found that, in addition to the lack of renal efficacy, small-dose dopamine worsens splanchnic oxygenation, impairs gastroenteric function, impairs the endocrine and immunologic system, and blunts ventilatory drives.
In this study, we compared fenoldopam to dopamine without a placebo group because during the last decades small-dose dopamine has been used as the standard clinical approach for major surgery. An investigation with a larger number of patients, however, should include a control (placebo) group.
Another limitation may be the lack of consistent investigator blinding to the study group assignment. However, when the measurements made are objective, as in this case, to blind the study is less problematic.
An additional limitation could be the normal preoperative renal function of the patients we studied; it could be useful to study patients with abnormal preoperative renal function.
We did not address the question of the titration of fenoldopam and dopamine administration, which may have led to different results.
Finally, a simple size with >79% power in larger trials might lead to more consistent results.
In conclusion, fenoldopam, at a dose that does not interfere with hemodynamics, seems to preserve CRE and BUN values in patients undergoing LTx. The fenoldopam dose of 0.1 μg · kg−1 · min−1 should be the best compromise of tubular and renal vascular effects. We conclude that fenoldopam may have a role as a new pharmacological option for perioperative renal vasodilation in high-risk surgical patients. A larger study population is necessary to validate the preliminary hypothesis of the renal vasodilatory effect of fenoldopam in patients undergoing LTx.
1. Fraley DS, Burr R, Bernardini J, et al. Impact of acute renal failure on mortality in end-stage liver disease with or without transplantation. Kidney Int 1998;54:518–24.
2. Bilbao I, Charco L, Balsells J, et al. Risk factors for acute renal failure requiring dialysis after liver transplantation. Clin Transplant 1998;12:123–9.
3. Bellomo R, Kellum J, Ronco C. Acute renal failure: time for consensus. Intensive Care Med 2001;27:1685–8.
4. Mathur V, Swan SK, Lambrecht LJ, et al. The effects of fenoldopam, a selective dopamine receptor agonist, on systemic and renal hemodynamics in normotensive subjects. Crit Care Med 1999;27:1832–7.
5. Mathur VS, O’Connell DO, Carey RG. Renal hemodynamic effects of very low dose fenoldopam. Anesth Analg 1999;88:SCA 85.
6. Kohli JD, Glock D, Goldberg LI. Relative DA1-dopamine-receptor agonist and α-adrenoreceptor antagonist activity of fenoldopam in the anesthetized dog. J Cardiovasc Pharmacol 1988;11:123–6.
7. Oparil S, Aronson S, Deeb GM, et al. Fenoldopam: a new parenteral antihypertensive—consensus roundtable on the management of perioperative hypertension and hypertensive crises. Am J Hypertens 1999;12:653–64.
8. Garwood S, Hines R. Perioperative renal preservation: dopexamine and fenoldopam—new agents to augment renal performance. Semin Anesth Periop Med Pain 1998;17:308–18.
9. Allison NL, Dubb JW, Ziemniak JA, et al. The effect of fenoldopam, a dopaminergic agonist, on renal hemodynamics. Clin Pharmacol Ther 1987;41:282–8.
10. Tumlin JA, Dunbar LM, Oparil S. Fenoldopam, a dopamine agonist, for hypertensive emergency: a multicenter randomized trial. Acad Emerg Med 2000;7:653–62.
11. Mathur VS. The role of the DA1 receptor agonist fenoldopam in the management of critically ill, transplant and hypertensive patients. Rev Cardiovasc Med 2003;4(Suppl 1):S35–40.
12. Halpenny M, Lakshmi S, O’Donnell A, et al. Fenoldopam: renal and splanchnic effects in patients undergoing coronary artery bypass grafting. Anaesthesia 2001;56:953–60.
13. Halpenny M, Rushe C, Breen P, et al. The effects of fenoldopam on renal function in patients undergoing elective aortic surgery. Eur J Anaesthesiol 2002;19:32–9.
14. Ramsay MA, Jones CC, Emmett ME. The effect of a fenoldopam infusion on postoperative renal function in patients undergoing liver transplantation. Anesthesiology 2002;A242.
15. Star RA. Treatment of acute renal failure. Kidney Int 1998;54:1817–31.
16. Dishart MK, Kellum JA. An evaluation of pharmacological strategies for the prevention and treatment of acute renal failure. Drugs 2000;59:79–91.
17. MacGregor DA, Prielipp RC, Black CS, et al. Renal dose dopamine does not alter the response to β-adrenergic stimulation by isoproterenol in healthy human volunteers. Chest 1997;112:40–4.
18. MacGregor DA, Smith TE, Prielipp RC, et al. Pharmacokinetics of dopamine in healthy male subjects. Anesthesiology 2000;92:338–46.
19. Jorkasky DK, Audet P, Shusterman N, et al. Fenoldopam reverses cyclosporine-induced renal vasoconstriction in kidney transplant recipients. Am J Kidney Dis 1992;19:567–72.
20. Sirivella S, Gielchinsky I, Parsonnet V, et al. Mannitol, furosemide and dopamine infusion in postoperative renal failure complicating cardiac surgery. Ann Thorac Surg 2000;69:501–6.
21. Lassnigg A, Donner E, Grubhofer G, et al. Lack of renoprotective effects of dopamine and furosemide during cardiac surgery. J Am Soc Nephrol 2000;11:97–104.
22. Polson RJ, Park GR, Lindop MJ, et al. The prevention of renal impairment in patients undergoing orthotopic liver grafting by infusion of low dose dopamine. Anesthesia 1987;42:15–9.
23. Swygert TH, Roberts LC, Valek TL, et al. Effect of intraoperative low dose dopamine on renal function in liver transplant recipients. Anesthesiology 1991;75:571–6.
24. Myles PS, Buckland MR, Schenk NJ, et al. Effects of “renal-dose” dopamine on renal function following cardiac surgery. Anaesth Intensive Care 1993;21:56–61.
25. Carey RM, Siragy HM, Ragsdale NV, et al. Dopamine-1 and dopamine-2 mechanism in control of renal function. Am J Hypertens 1990;56:59s–63s.
26. Bailey JM. Dopamine: one size does not fit all. Anesthesiology 2000;92:303–5.
27. Marik PE. Low-dose dopamine: a systematic review. Intensive Care Med 2002;28:877–83.
© 2004 International Anesthesia Research Society
28. Holmes CL, Walley KR. Bad medicine: low-dose dopamine in the ICU. Chest 2003;123:1266–75.