Sodium bicarbonate to prevent increases in serum creatinine after ...

Sodium bicarbonate to prevent increases in serum creatinine after cardiac surgery: A pilot double-blind, randomized controlled trial* Michael Haase, MD; Anja Haase-Fielitz, BPharm; Rinaldo Bellomo, MD, FRACP; Prasad Devarajan, MD; David Story, MD, FANZCA; George Matalanis, FRACS; Michael C. Reade, MD, PhD, FANZCA; Sean M. Bagshaw, MD, MSc, FRCCP; Narelle Seevanayagam, RN; Siven Seevanayagam, FRACS; Laurie Doolan, MD, FANZCA; Brian Buxton, FRACS; Duska Dragun, MD

LEARNING OBJECTIVES On completion of this article, the reader should be able to: 1. Explain potential renal effects of cardiac surgery. 2. Describe the effect of sodium bicarbonate on patient outcomes after cardiopulmonary bypass. 3. Use this information in a clinical setting. As a consequence of this study, Dr. Haase, Dr. Bellomo, and the Austin Hospital have sought patent rights for the use of bicarbonate as renal protective treatment in cardiac surgery, which have been successfully granted in Australia and are under examination internationally. Dr. Devarajan has disclosed that he was/is the recipient of grants/research funds from the National Institutes of Health, was/is a consultant/advisor for Abbott and Biosite/Inverness, and was/is on the speaker’s bureau for Abbott and Biosite/Inverness. The remaining authors have disclosed that they have no financial relationships with or interests in any commercial companies pertaining to this educational activity. The authors have disclosed that the U.S. Food and Drug Administration has not approved sodium bicarbonate for the treatment of acute kidney injury discussed in this article. Please consult the products labeling information for approved indications and usage. All faculty and staff in a position to control the content of this CME activity have disclosed that they have no financial relationship with, or financial interests in, any commercial companies pertaining to this educational activity. Lippincott CME Institute, Inc., has identified and resolved all faculty conflicts of interest regarding this educational activity. Visit the Critical Care Medicine Web site (www.ccmjournal.org) for information on obtaining continuing medical education credit.

Objective: To test whether perioperative sodium bicarbonate infusion can attenuate postoperative increases in serum creatinine in cardiac surgical patients. Design: Double-blind, randomized controlled trial. Setting: Operating rooms and intensive care unit at a tertiary hospital. Patients: Cohort of 100 cardiac surgical patients at increased risk of postoperative acute renal dysfunction. Intervention: Patients were randomized to either 24 hrs of intravenous infusion of sodium bicarbonate (4 mmol/kg) or sodium chloride (4 mmol/kg). Measurements and Main Results: The primary outcome measure was the proportion of patients developing acute renal dysfunction defined as a postoperative increase in plasma creatinine concentration >25% of baseline within the first five postoperative days. Secondary outcomes included changes in plasma creatinine, plasma urea, urinary neutrophil gelatinase-associated lipocalin, and urinary neutrophil gelatinase-associated lipocalin/urinary creatinine ratio. Patients were well balanced for baseline characteristics. Sodium bicarbonate infu-

W

ith over one million operations a year, cardiac surgery with cardiopulmonary bypass is one of the most common major surgical procedures worldwide (1). Acute renal dysfunction is a common and serious postoperative complication of cardiopulmonary bypass and may affect 25% to 50% of patients (2– 4). Acute renal dysfunction carries significant costs (4) and is independently Crit Care Med 2009 Vol. 37, No. 1

sion increased plasma bicarbonate concentration (p < 0.001), base excess (p < 0.001), plasma pH (p < 0.001), and urine pH (p < 0.001). Fewer patients in the sodium bicarbonate group (16 of 50) developed a postoperative increase in serum creatinine compared with control (26 of 50) (odds ratio 0.43 [95% confidence interval 0.19 – 0.98]), (p ⴝ 0.043). The increase in plasma creatinine, plasma urea, urinary neutrophil gelatinase-associated lipocalin, and urinary neutrophil gelatinase-associated lipocalin/urinary creatinine ratio was less in patients receiving sodium bicarbonate, (p ⴝ 0.014; p ⴝ 0.047; p ⴝ 0.009; p ⴝ 0.004). There were no significant side effects. Conclusions: Sodium bicarbonate loading and continuous infusion was associated with a lower incidence of acute renal dysfunction in cardiac surgical patients undergoing cardiopulmonary bypass. The findings of this pilot study justify further investigation. (ClinicalTrials.gov, NCT00334191). (Crit Care Med 2009; 37:39 – 47) KEY WORDS: cardiopulmonary bypass; prevention; cardiac surgery; acute renal dysfunction; creatinine; neutrophil gelatinase associated lipocalin

associated with increased morbidity and mortality (2, 3). Even minimal increments in plasma creatinine are associated with an increase in mortality (5, 6). Multiple causes of cardiopulmonary bypass-associated acute renal dysfunction have been proposed, including ischemiareperfusion, generation of reactive oxygen species, hemolysis, and activation of inflammatory pathways (7–10). To date, no simple, safe, and effective intervention

to prevent cardiopulmonary bypassassociated acute renal dysfunction in a broad patient population has been found (11–14). Urinary acidity may enhance the generation and toxicity of reactive oxygen species induced by cardiopulmonary bypass (10, 15). Activation of complement during cardiac surgery (16) may also participate in renal injury. Urinary alkalinization may protect from renal injury 39

induced by oxidant substances, ironmediated free radical pathways, complement activation, and tubular hemoglobin cast formation (9, 17, 18). Of note, increasing urinary pH—in combination with N-acetylcysteine (19, 20) or without (21)— has recently been reported to attenuate acute renal dysfunction in patients undergoing contrast-media infusion. Accordingly, we hypothesized that urinary alkalinization might protect kidney function in patients at increased risk of acute renal dysfunction undergoing cardiopulmonary bypass and conducted a pilot double-blind, randomized controlled trial of intravenous sodium bicarbonate in this setting.

Box 1. Inclusion and exclusion criteria Inclusion criteria. Cardiac surgical patients in whom the use of cardiopulmonary bypass was planned and having one ore more of the following risk factors for postoperative acute kidney injury (12): ● Age above 70 years ● Preexisting renal impairment (preoperative plasma creatinine concentration ⬎120 ␮mol/L ● New York Heart Association class III/IV or impaired left ventricular function (left ventricular ejection fraction ⬍35%) ● Valvular surgery or concomitant valvular and coronary artery bypass graft surgery ● Redo cardiac surgery ● Insulin-dependent Type 2 diabetes mellitus Exclusion criteria. ● End stage renal disease (plasma creatinine concentration ⬎300 ␮mol/L) ● Emergency cardiac surgery ● Planned off-pump cardiac surgery ● Known blood-bourne infectious disease ● Chronic inflammatory disease on immunosuppression ● Chronic moderate to high dose corticosteroid therapy (⬎10 mg/day prednisone or equivalent) ● Enrolled in conflicting research study ● Age ⬍18 years

MATERIALS AND METHODS Design Overview. This study was a pilot double-blind, single center, randomized controlled trial. The Human Research Ethics Committee of the Austin Hospital approved this study and the Therapeutic Goods Administration of the Australian Government, Department of Health and Aging approved the use of sodium bicarbonate for this indication. Written informed consent was obtained from each patient. The study was reported using the recommendations of the CONSORT statement (22). Setting and Participants. Patients were identified in out patient departments and on hospital wards. We enrolled patients deemed at increased risk of postoperative acute renal dysfunction (12) who were scheduled for elective or urgent cardiac surgery necessitating the use of cardiopulmonary bypass at a university tertiary referral hospital as long as defined exclusion criteria were not present (inclusion and exclusion criteria listed in Box 1). Clinical practice was not changed or modified for the purpose of the study and followed the same standard protocols as previously described (13). All preoperative medications were routinely omitted on the day of surgery. Aspirin was stopped 1 week before. Angiotensin-converting enzyme inhibitors were withdrawn on hospital admission (generally 1 day

before surgery). All surgical approach was by median sternotomy. The cardiopulmonary bypass circuit consisted of a roller pump (Stoeckert S3, Munich, Germany), tubing (Lovell Uncoated Tubing, Lovell, Melbourne, Australia; or Cobe Smart Tubing, Cellplex, Melbourne, Australia), a membrane oxygenator (Affinity Trillium Coated Oxygenator, Medtronic, Minneapolis, MN; or MonolythPro, Sorin, Saluggia, Italy), and an arterial filter (Pall, Melbourne, Australia). The pump prime solution was made up with 2 L of crystalloid solution (Plasmalyte, Baxter, Sydney, Australia). Perioperative hemodynamic management targeted a cardiac output of ⬎2.5 L/min/m2 and a mean arterial blood pressure of ⬎70 mm Hg. The use of intraoperative fluid management was left to the discretion of the anesthesiologist and was documented. For fluid resuscitation, crystalloid solution (Compound Sodium Lactate, Baxter, Old Toongabbie, Australia) or gelatin-based colloid solutions (Gelofusine, Braun, Bella Vista, Australia) were used. Blood glucose levels between 5 and 8 mmol/L (90 –144 mg/dL) were achieved using insulin bolus or infusions. Analgesia was achieved with acetaminophen and morphine or tramadol with avoidance of nonsteroidal anti-inflammatory drugs. Urinary output was

*See also p. 333. Research Fellow (MH, AH-F), Department of Nephrology and Intensive Care, Charité University Medicine, Potsdam, Germany; Director of Research (RB), Austin Health, Melbourne, Australia; Louise M. Williams Endowed Chair (PD), Professor of Pediatrics and Developmental Biology, Director of Nephrology and Hypertension, Cincinnati Children’s Hospital Medical Center, Cincinnati, OH; Head of Research (DS), Department of Anesthesiology, Austin Health, Melbourne, Australia; Director (GM), Cardiac Surgery, Austin Hospital, Melbourne, Australia; Associate Professor (MCR), Austin Hospital and University of Melbourne, Melbourne, Australia; Assistant Professor (SMB), Department of Nephrology, University of Alberta, Edmon-

ton, Alberta, Canada; Critical Care Nurse (NS), Nurse Unit Manager, Cardiac Surgery Department, Austin Health, Heidelberg, Victoria, Australia; Professor (SS), Austin Hospital, Department of Cardiac Surgery, Heidelberg, Melbourne, Australia; Staff Specialist (LD), Department of Anesthesiology and Intensive Care, Austin Hospital, Melbourne, Australia; Professor (BB), Cardiac Surgery, University of Melbourne, Melbourne, Australia; Professor (DD), Department of Nephrology and Intensive Care, Charité University Medicine, Berlin, Germany. M. Haase holds a postdoctoral Feodor-Lynen Research Fellowship from the Alexander von HumboldtFoundation. The Alexander von Humboldt Foundation is a nonprofit foundation established by the Federal Re-

40

maintained at ⬎0.5 mL/kg/hr using furosemide if necessary. Clinical data were prospectively collected, coded, and entered into a computerized database by a research nurse.

Randomization and Intervention The hospital pharmacy clinical trials coordinator used a Microsoft Excel-based (Microsoft, Redmond, WA) random number generator to create the randomization list using a permuted block strategy with blocks of ten. Infusion bags were each delivered in separate shrink-wrapped black plastic bags that were identical in appearance. Allocation concealment to patients, anesthesiologists, cardiac surgeons, intensive care specialists, bedside nurses, and investigators was ensured by central randomization through the Department of Pharmacy at the Austin Hospital. Treatment allocation was only revealed after data analysis had been performed. We used the same dosing of bolus and maintenance infusion of sodium bicarbonate and sodium chloride recently published in the prevention of contrast-induced nephropathy (21). However, in our study, maintenance infusion was extended from 6 to 23 hrs. Specifically, in all patients, body weight adjusted

public of Germany for the promotion of international research cooperation. This study was supported, in part, a grants from the Australian and New Zealand College of Anaesthetists and by the Austin Hospital Anesthesia and Intensive Care Trust Fund, both federal non-profit foundations. For information regarding this article, E-mail: [email protected] Copyright © 2008 by the Society of Critical Care Medicine and Lippincott Williams & Wilkins DOI: 10.1097/CCM.0b013e318193216f

Crit Care Med 2009 Vol. 37, No. 1

dose of study medication was achieved by infusion of sodium bicarbonate (Pfizer, Bentley, WA, Australia) or sodium chloride (Astra Zeneca, North Ryde, New South Wales, Australia) at a dose of 0.5 mmol/kg body weight (⫽bolus) diluted in 250 mL of 5% dextrose (Baxter, Sydney, New South Wales, Australia) and H2O over 1 hr immediately after the induction of anesthesia, before the first surgical incision followed by continuous intravenous infusion of 0.15 mmol/kg/hr (⫽maintenance) diluted in 1000 mL of 5% dextrose and H2O over 23 hrs (total dose of 4 mmol/kg over 24 hrs). Because of this approach the concentration of bicarbonate or chloride varied according to patient weight however each patient received the same amount of fluid volume. For comparison, the use of sodium bicarbonate in Intensive Care is usually in bottles of 100 mL containing 100 mmol. In our study solution, sodium bicarbonate was diluted to approximately a quarter of the usual concentration. Outcome Measures and Follow-Up. The primary study outcome measure was the number of patients who developed acute renal dysfunction after cardiopulmonary bypass. This was defined prospectively as an increase in plasma creatinine concentration greater than 25% from baseline to peak value at any time within the first 5 days after cardiopulmonary bypass as previously published in this setting (12). The renal secondary outcomes were an increase in plasma creatinine concentration greater than 50%, greater than 100%, changes in plasma creatinine and plasma

urea concentrations within the first 5 days after cardiopulmonary bypass and in urinary neutrophil gelatinase-associated lipocalin (NGAL) concentration and urinary NGAL/ urinary creatinine ratio within the first 24 hrs after commencement of cardiopulmonary bypass. Also, we compared the proportion of patients between both groups developing acute renal dysfunction according to the network classification (Acute Kidney Injury Network [AKIN] total, AKIN stage 1, AKIN stage 2 and AKIN stage 3) (23). Plasma creatinine and plasma urea were sampled daily. Urinary NGAL and urinary creatinine were sampled before induction of anesthesia, at 6 and at 24 hrs after commencement of cardiopulmonary bypass. To adjust urinary NGAL concentration to urinary creatinine concentration we calculated urinary NGAL/ urinary creatinine ratio (ng/mg) dividing urinary NGAL concentration (ng/mL) by the urinary creatinine concentration (mg/mL). Plasma and urinary creatinine was measured using the modified Jaffe´ method, plasma urea using the urease reaction and urinary NGAL by enzyme-linked immunosorbent assay (24). The inter- and intra-assay coefficient of variation were ⬍5%. Urinary NGAL measurements by enzyme-linked immunosorbent assay are stable over a urinary pH range of 4 – 8. The laboratory investigators were blinded to the sample sources and clinical outcomes until the end of the study. Additional secondary outcomes included changes in acid– base status, clinical outcomes such as need for renal replacement therapy,

duration of ventilation, length of intensive care unit and hospital stay, and adverse events. Specific adverse events targeted for detection included the incidence of hypernatremia ([Na⫹] ⬎150 mmol/L), hypokalemia ([K⫹] ⬍3.5 mmol/L), alkalemia (pH ⬎7.50), postoperative atrial fibrillation, and other postoperative arrhythmias (supraventricular arrhythmias, ventricular tachycardia, and ventricular fibrillation) during study treatment. Power of the Study and Statistical Analysis. Using data available from the institutional cardiac surgery database and from the literature (3), we estimated that in the control group, 50% of our patients at increased risk would develop an increase in plasma creatinine greater than 25% from baseline. Given a reduction in the incidence of acute renal dysfunction by 30% we calculated that 100 patients would be needed to have a 90% power to detect a difference between the control and the intervention group at an alpha of 0.05 (MedCalc for Windows Version 9.3.9.0, Mariakerke, Belgium). All data were analyzed according to the intention-to-treat principle. Continuous data were tested for normal distribution using histograms. Between-group comparisons for continuous data were performed with the use of the Student’s t test or the Mann-Whitney U test and for categorical data with the use of Fisher’s exact test or chi-square test where appropriate. All tests were two-tailed and we considered a p value ⬍0.05 to indicate statistical significance. We report values as means with SD and mean difference, or medians with interquartile ranges or odds ratio with 95% confidence interval estimate as

Figure 1. Patient enrollment into the study (using CONSORT recommendations).


41

appropriate. We performed analyses by using SPSS, version 15.0 (SPSS Inc., Chicago, IL).

Table 1. Preoperative characteristics in patients receiving either sodium bicarbonate or sodium chloride infusion

RESULTS Between June 2006 and December 2006, we randomized 100 patients to receive intravenous sodium bicarbonate (n ⫽ 50) or sodium chloride (n ⫽ 50) (Fig. 1). All patients were Caucasians. One patient in the sodium bicarbonate group received 9 hrs of study treatment until the infusion bag was damaged by accident. One patient in the control group received 7 hrs of study treatment until the treating intensive care specialist stopped the infusion due to a cardiac arrest of the patient. There were no significant differences between the groups in baseline characteristics including plasma creatinine, plasma urea, and urinary NGAL (Table 1). Most patients had elective cardiac surgery (Table 2). Treatment groups did not differ significantly in duration of cardiopulmonary bypass and in hemodynamic and fluid management during and after cardiac surgery (Table 2). Mean dose of sodium bicarbonate was 307 ⫾ 57 mmol and mean dose of sodium chloride was 309 ⫾ 68 mmol (p ⫽ 0.89).

Renal Outcomes Fewer patients in the sodium bicarbonate group (16 of 50) developed acute renal dysfunction as defined compared to the control group (26 of 50) (odds ratio 0.43; 95% confidence interval 0.19 – 0.98), (p ⫽ 0.043); Figure 2. Other renal outcomes are also presented in Figure 2 including the number of patients developing acute kidney injury according to the network classification (23). Patients receiving sodium bicarbonate experienced a significant attenuation in their absolute and relative increase in plasma creatinine compared to patients receiving sodium chloride (Fig. 3A and B). Plasma creatinine concentration reached its peak value at postoperative day 3 after commencement of cardiopulmonary bypass in both groups. The absolute and relative increase in plasma urea concentration was less in patients receiving sodium bicarbonate (1.9 ⫾ 3.6 mmol/L vs. 4.1 ⫾ 6.3 mmol/L, p ⫽ 0.047 and 36.4 ⫾ 71.4% vs. 75.1 ⫾ 109.1%, p ⫽ 0.041, respectively). The increase in urinary NGAL concentration and in urinary NGAL/urinary creatinine ratio was attenuated in patients receiving sodium bicarbonate (p ⫽ 0.009 and p ⫽ 0.004; Fig. 4A and B). 42

Characteristics Demographic data Age, yr, mean (SD) Male, n (%) Weight, kg, mean (SD) Inclusion criteria Age ⬎70 yr, n (%) Undergoing valve surgery, n (%)a Left ventricular dysfunction, n (%)b Plasma creatinine ⬎120 ␮mol/L (⬎1.4 mg/dL), n (%) Insulin dependent diabetes mellitus, n (%) Previous cardiac surgery, n (%) Comorbidities Arterial hypertension, n (%) Hypercholesterolemia, n (%) Atrial fibrillation, n (%) Recent myocardial infarction, n (%) Chronic obstructive pulmonary disease, n (%) Vascular disease, n (%)c Euro Score (24), points, mean (SD) Medication Beta-receptor blockers, n (%) Calcium channel blockers, n (%) ACE inhibitors or angiotensin blockers, n (%) Platelet inhibitors, n (%) HMG-CoA reductase inhibitor, n (%) Measures of renal function Plasma creatinine, ␮mol/L, mean (SD) Plasma urea, mmol/L, mean (SD) Urinary NGAL, ng/mL, mean (SD) Urinary NGAL/urinary creatinine ratio, ng/mg, mean (SD)

Sodium Bicarbonate (n ⴝ 50)

Sodium Chloride (n ⫽ 50)

71.5 (9.2) 30 (60) 77.7 (14.4)

70.6 (9.5) 33 (66) 78.2 (17.1)

32 (64) 35 (70) 13 (26) 3 (6) 5 (10) 5 (10)

31 (62) 38 (76) 10 (20) 5 (10) 2 (4) 4 (8)

0.84 0.50 0.48 0.72 0.44 ⬎0.99

46 (92) 31 (62) 13 (26) 8 (16) 6 (12) 10 (20) 6.7 (2.7)

42 (84) 29 (58) 13 (26) 9 (18) 8 (16) 6 (12) 6.0 (2.1)

0.22 0.68 ⬎0.99 0.79 0.56 0.28 0.15

27 (54) 14 (28) 40 (80) 34 (68) 28 (56)

19 (38) 9 (18) 32 (64) 28 (56) 21 (42)

0.11 0.24 0.08 0.22 0.16

91.9 (27.3) 7.1 (2.6) 17.8 (22.4) 21.8 (24.3)

89.5 (24.2) 6.8 (2.5) 13.0 (22.1) 12.1 (14.8)

0.77 0.39 0.21 0.01

p

0.58 0.53 0.54

ACE, angiotensin converting enzyme; HMG-CoA reductase inhibitor, 3-hydroxy-3-methyl-glutarylCoA reductase inhibitor; NGAL, neutrophil gelatinase-associated lipocalin. a Valve surgery with or without additional coronary revascularisation; bNew York Heart Association class III/IV or impaired left ventricular function (defined as left ventricular ejection fraction; cvascular disease summarizes peripheral vascular disease and carotid artery disease.

There was no difference in requirement for renal replacement therapy, duration of mechanical ventilation and hospital mortality (Table 3).

Other Outcomes We found marked group differences in plasma bicarbonate concentration, base excess, and pH from baseline to 24 hrs (all p ⬍ 0.001) (Table 3). Sodium bicarbonate infusion induced urinary alkalinization at 6 and 24 hrs after commencement of study drug infusion, whereas urine pH level decreased in control patients (p ⬍ 0.001; Fig. 5). At baseline, plasma sodium concentration was 138.5 ⫾ 2.7 mmol/L in patients randomized to sodium bicarbonate treatment and 138.5 ⫾ 2.5 mmol/L in patients randomized to sodium chloride treatment (p ⫽ 0.97). At 24 hrs after commencement of cardiopulmonary bypass, plasma sodium concentration was 141.5 ⫾ 3.5

mmol/L in sodium bicarbonate treated patients and 142.3 ⫾ 4.2 mmol/L in patients randomized to sodium chloride treatment (p ⫽ 0.33). We found no differences in other hospital outcomes (Table 3). One patient in each study group died from treatmentresistant cardiogenic shock within the same hospital admission, both outside the study treatment period.

Safety No patient developed a plasma sodium concentration ⬎150 mmol/L. One patient in the control group had a pH level below 7.30 (p ⬎ 0.99) and three patients receiving sodium bicarbonate had a pH level greater than 7.50 (p ⫽ 0.24). More patients receiving sodium bicarbonate had a plasma potassium ⬍3.5 mmol/L compared to control (5 of 50 vs. 1 of 50, p ⫽ 0.20). During treatment, there was no difference in the incidence of newCrit Care Med 2009 Vol. 37, No. 1

onset atrial fibrillation (5 of 50 vs. 4 of 50, p ⬎ 0.99). One sodium bicarbonate and one patient in the control group had an episode of supraventricular

tachycardia. Another patient in the control group had ventricular fibrillation during study drug infusion, which was initially successfully treated. The latter

Table 2. Intraoperative characteristics and intraoperative and postoperative interventions in patients receiving either sodium bicarbonate or sodium chloride infusion

Characteristics Cardiopulmonary bypass time, mins, mean (SD) Aortic cross-clamp time, mins, mean (SD) Valve surgery Valve surgery only, n (%) Aortic valve replacement, n (%) Mitral valve replacement, n (%) Mitral valve repair, n (%) Double valve surgery, n (%) Coronary artery bypass grafting surgery Coronary artery bypass grafting surgery only, no. (%) Average of vessels bypassed, mean (SD) Other cardiac surgery Valvular and coronary artery bypass graft surgery, n (%) Aortic valvular and arch surgery, n (%) Hemodynamic and fluid managementa Lowest cardiac index, L/m2, mean (SD) Lowest mean arterial pressure, mm Hg, mean (SD) Intravenous fluids, mL, mean (SD) Urine output, mL, mean (SD) Drain output, mL, mean (SD) Furosemide, mg, mean (SD) a


Sodium Chloride (n ⴝ 50)

132.1 (61.1) 105.3 (47.8)

142.6 (69.4) 112.6 (54.7)

25 (50) 14 (28) 3 (6) 4 (8) 4 (8)

25 (50) 12 (24) 2 (4) 6 (12) 5 (10)

15 (30.0) 2.6 (1.2)

12 (24.0) 3.0 (0.8)

0.50 0.22

7 (14.0) 3 (6.0)

8 (16.0) 5 (10.0)

0.78 0.72

2.1 (0.4) 63.4 (6.4) 6350 (2750) 3850 (1500) 750 (570) 37.1 (52.7)

2.2 (0.5) 65.4 (6.7) 6150 (2700) 3900 (1400) 600 (350) 55.8 (85.4)

0.15 0.19 0.71 0.95 0.50 0.19

Within the first 24 hours after commencement of cardiopulmonary bypass.

p 0.19 0.17 ⬎0.99 0.65 ⬎0.99 0.74 ⬎0.99

patient, however, died as described in the earlier section.

DISCUSSION We conducted a pilot double-blind, randomized controlled clinical trial to investigate whether sodium bicarbonate infusion with preoperative intravenous loading to achieve urinary alkalinization could attenuate the creatinine rise associated with cardiopulmonary bypass in cardiac surgical patients at increased risk. We found that sodium bicarbonate treatment successfully alkalinized both blood and urine. Sodium bicarbonate infusion was associated with an absolute risk reduction for acute renal dysfunction of 20% and with a significant attenuation in the postoperative increase of plasma urea, urinary NGAL (a marker of oxidative stress) and urinary NGAL/urinary creatinine ratio. Previous double-blind randomized controlled trials attempting to prevent or attenuate acute renal dysfunction after cardiopulmonary bypass targeting oxidative stress, renal adenosine triphosphate consumption, and improvement of perioperative hemodynamic stability have

Figure 2. Number of patients receiving sodium bicarbonate (white bars) developing acute kidney injury after cardiac surgery compared to patients receiving sodium chloride (black bars). From left to right: increase in plasma creatinine ⬎25%, increase in plasma creatinine ⬎50%, increase in plasma creatinine ⬎100% within the first five postoperative days. Acute Kidney Injury Network (AKIN) stage 1, 2, 3 is based on the definition of acute kidney injury by the AKIN (24). OR, odds ratio.


43

Figure 3. Absolute and relative changes in plasma creatinine concentration after cardiac surgery. A, Absolute changes in plasma creatinine concentration (␮mol/L) after cardiac surgery from baseline to peak value at any time within the first five postoperative days displayed as mean (SEM). Patients receiving sodium bicarbonate experienced a significant attenuation in their absolute increase in plasma creatinine (median 11.5, interquartile range [IQR] 26.0 ␮mol/L) compared with patients receiving sodium chloride (median 27.5 [IQR 45.0] ␮mol/L); p ⫽ 0.014 (Mann-Whitney U test). To convert plasma creatinine in ␮mol/L to mg/dL, divide by 88.4. B, Relative changes in plasma creatinine concentration (%) after cardiac surgery from baseline to peak value at any time within the first five postoperative days displayed as mean (SEM). Patients receiving sodium bicarbonate had a significant attenuation in their relative increase in plasma creatinine (median 15.5 [IQR 29.5] %) compared with patients receiving sodium chloride (median 30.5 [IQR 55.0] %); p ⫽ 0.017 (Mann-Whitney U test).

been found to be ineffective (11–13), inconclusive or studied in specific cardiac surgical subpopulations (14). The simultaneous alkalinizing effect and apparent renal protection achieved with intravenous sodium bicarbonate is consistent with the biological rationale for our trial. According to this rationale, a creatinine rise after cardiopulmonary bypass might represent a combination of tubular injury induced by reactive oxygen species (7, 8), complement activation (16) and free hemoglobin release (9, 10). For example, experimental studies show that sodium bicarbonate protects from oxidant injury by slowing pHdependent Haber-Weiss free radical production (15). It also directly scavenges peroxynitrite and other reactive species generated from nitric oxide (25). Also, in all previous randomized controlled trials, urinary alkalinization with intravenous sodium bicarbonate in patients undergoing infusion of contrast media has been found to attenuate acute kidney injury (19 –21, 26, 27)—another condition where free oxygen radical generation may be involved. Sodium bicar44

bonate administration decreases the urinary excretion rate of complement activation products (17). In addition, urinary alkalinization with sodium bicarbonate might have protected patients from free hemoglobinmediated kidney injury. Hemoglobin infusion causes acute renal failure (9) and urinary alkalinization or hemoglobin blockade with haptoglobin attenuates free hemoglobin-induced kidney injury (9, 18, 28). Furthermore, aciduria converts hemoglobin to methemoglobin, which precipitates, forms distal casts, and induces acute kidney injury (9). Animal experiments show that red blood cell hemolysate is a potent mitogen for renal tubular epithelial cells (29), that free ferrous ions causes hydroxyl radical formation and lipid peroxidation during reperfusion of ischemic kidneys (18, 25), and that free-radical production catalyzed by free ferrous ions is most active at acid pH (15). In contrast, at neutral or alkaline pH induced by sodium bicarbonate, more free ferric ions precipitate as insoluble ferric hydroxides, reducing the production of injurious hydroxyl radicals (15).

We are the first to find that intravenous sodium bicarbonate might attenuate the rise in serum creatinine seen in cardiac surgical patients and that this treatment appears safe. In patients receiving sodium bicarbonate, three independent surrogate parameters of kidney function or injury suggest the possibility of an attenuation of renal dysfunction (plasma creatinine, plasma urea, and urinary NGAL). Emerging evidence from experimental and human studies indicate that urinary NGAL is derived from massive synthesis and release from the distal nephron in the setting of a rapid response to acute kidney injury (30). Urinary NGAL has been reported to be a sensitive, specific, and highly predictive early biomarker for acute kidney injury after cardiac surgery (24). The appearance of NGAL in the urine is related to the dose and duration of renal injury and precedes the appearance of other urinary markers (23). Its expression is induced by hydrogen peroxide making it a useful biomarker of oxidative stress (31). Our results suggest a novel utility of urinary NGAL measureCrit Care Med 2009 Vol. 37, No. 1

Figure 4. Changes in urinary neutrophil gelatinase-associated lipocalin (NGAL) concentration (A) and in urinary NGAL/urinary creatinine ratio (B). A, Absolute changes in urinary NGAL concentration (ng/mL) after cardiac surgery from baseline to peak value at any time within the first 24 hrs after the commencement of cardiopulmonary bypass displayed as mean (SD). Patients receiving sodium bicarbonate experienced a significant attenuation in their absolute increase in urinary NGAL (median 56.5 interquartile range [IQR] 198.5 ng/mL) compared with patients receiving sodium chloride (median 246.5 [IQR 429.0] ng/mL); p ⫽ 0.009 (Mann-Whitney U test). B, Absolute changes in urinary neutrophil gelatinase-associated lipocalin (NGAL)/urinary creatinine ratio (ng/mg) after cardiac surgery from baseline to peak value at any time within the first 24 hrs after the commencement of cardiopulmonary bypass displayed as mean (SD). Patients receiving sodium bicarbonate experienced a significant attenuation in their absolute increase in urinary NGAL/urinary creatinine ratio (median 111.5 [IQR 455.0] ng/mg) compared with patients receiving sodium chloride (median 527.0 [IQR 1130.5] ng/mg); p ⫽ 0.004 (Mann-Whitney U test).

ments as potential monitor of the efficacy of therapy for acute kidney injury. We enrolled patients with a mix of cardiac surgery procedures. Thus, our results may have implications for a variety of cardiac surgery patients. Sodium bicarbonate is inexpensive (cost: US$ 25–30 per patient) and simple to administer. This is the first randomized controlled trial directed at the prevention of an acute increase in serum creatinine after cardiac surgery reporting the incidence of this complication according to the network classification (23). There were nonsignificant between-group differences in age (patients receiving sodium bicarbonate were 1 year older) and Euro Score (32) (patients receiving sodium bicarbonate had a higher score by one point predicting a higher risk of death). On the other hand, we found nonsignificant between-group differences in the duration of cardiopulmonary bypass (shorter by 10 min in patients receiving sodium bicarbonate). Patients in the sodium bicarbonCrit Care Med 2009 Vol. 37, No. 1

ate group appeared to receive slightly more preoperative cardiovascular medication. However, when comparing patients who developed a creatinine rise ⬎25% vs. patients who never did, preoperative medication was not found to differ (all p ⬎ 0.5). There was a group difference in urinary NGAL/urinary creatinine ratio at baseline which appeared to be associated with decreased preoperative urinary creatinine concentration in patients receiving sodium bicarbonate infusion compared to those receiving sodium chloride. Given similar age, sex distribution, and preoperative serum creatinine concentration, this small imbalance at baseline might have occurred by chance. We carefully suggest that, within the limitations of a 100 patients study, both groups were reasonably well balanced with respect to pre- and intraoperative characteristics and risk factors. We found a similar pattern of risk factors for postoperative acute renal dysfunction as in the largest randomized

controlled trial in the prevention of this postoperative complication (12). This observation confirms that our patient population was at increased risk of acute renal dysfunction. Possibly because of differences in surgical technique (coronary revascularization using mammary or radial arteries likely contributing to prolonged duration of cardiopulmonary bypass) the proportion of patients developing acute kidney injury was slightly higher compared to the cohort of Burns et al (12). However, the mortality rates in both patient cohorts were similar and reflect values frequently found in nonemergency cardiac surgery. The internal validity of our results was strengthened by double-blinding and central randomization. Sodium bicarbonate was infused as a loading dose followed by a maintenance infusion immediately before, during and for 24 hrs after the commencement of cardiopulmonary bypass in order to deliver sufficient concentrations through the period of maximal 45

Table 3. Acid base status and hospital outcomes in patients receiving either sodium bicarbonate or sodium chloride infusion

Outcomes Acid base status Plasma bicarbonate, mmol/L, mean (SD) Baselinea 6 hoursb 24 hoursb Change from baseline to 24 hours postoperatively Plasma pH, mean (SD) Baselinea 6 hoursb 24 hoursb Change from baseline to 24 hours postoperatively Hospital outcomes Renal replacement therapy, n (%) Length of ventilation, min, mean (SD) Length of stay in intensive care unit, hrs, mean (SD) Length of stay in hospital, days, mean (SD) Hospital deaths, n (%)


Sodium Chloride (n ⴝ 50)

p

Mean Difference (95% CI)

25.8 (1.9) 27.0 (2.0) 29.4 (3.4) 3.7 (3.2)

25.2 (1.7) 24.4 (2.1) 23.7 (2.2) ⫺1.5 (2.3)

0.20 ⬍0.001 ⬍0.001 ⬍0.001

7.41 (0.05) 7.43 (0.05) 7.45 (0.06) 0.04 (0.08)

7.41 (0.05) 7.39 (0.05) 7.38 (0.04) ⫺0.03 (0.06)

0.68 ⬍0.001 ⬍0.001 ⬍0.001

0.002 (⫺0.018–0.022) 0.038 (0.016–0.059) 0.071 (0.050–0.092) 0.07 (0.04–0.10)

2 (4) 1657 (2115) 73.8 (98.7) 10.4 (7.2) 1 (2)

2 (4) 1756 (2600) 74.8 (92.9) 11.2 (8.5) 1 (2)

⬎0.99 0.83 0.96 0.60 ⬎0.99

— ⫺99 (⫺1040–841) ⫺1.0 (⫺39.0–37.0) ⫺0.8 (⫺3.9–2.3) —

0.5 (⫺0.2–1.3) 2.6 (1.7–3.4) 5.7 (4.6–6.8) 5.2 (4.1–6.4)

CI, confidence interval; pH, potentia hydrogenii. a First blood gas analysis prior to cardiopulmonary bypass; bafter commencement of study infusion.

Figure 5. Changes in urine pH. Urine pH before (0 hr) and at 6 and 24 hrs after commencement of sodium bicarbonate or sodium chloride infusion displayed as mean (SD). p values for group comparisons before and at 6 and 24 hrs: *p ⫽ 0.60 and †p ⬍ 0.001. (Student’s t test).

risk. We did not find any significant group differences in safety markers including potassium concentration or occurrence of arrhythmias. Because of our protocol the concentration of bicarbonate or chloride varied according to body weight, while the amount of fluid did not. 46

Whether this aspect of the protocol influenced outcome remains uncertain. The number of participants enrolled was not large. However, it was appropriate for a single center pilot phase II equivalent efficacy trial. We had calculated the sample size on the basis of in-

formation from our institutional Cardiac Surgery database and from the literature (3) and our findings were consistent with predictions. This study, like all pilot studies, entails some chance of a type I error and the presence of only a marginally significant result (p ⫽ 0.043) might also be consistent with this possibility. However, the cost of a large multicenter phase III study is only justified if findings of pilot investigations support it. Preexisting renal impairment was not present in all patients to identify a very high-risk cohort. On the other hand, we observed an overall high incidence of increases in serum creatinine within the first 5 days after cardiopulmonary bypass. Thus, we correctly identified a cohort of cardiac surgical patients at increased risk of acute renal dysfunction. The study was not powered to detect group differences in the need for renal replacement therapy and mortality. However, this study was a phase II equivalent and was conducted to test whether there is justification for larger (phase III) studies addressing those outcomes. Such studies are only justified if preliminary evidence such as ours suggests benefit. We used plasma creatinine as the primary outcome measure of our study. Plasma creatinine is the marker used in daily practice to guide clinical decisions worldwide and has served as surrogate parameter of renal function in previous large randomized controlled trials directed at preventing acute kidney injury (12, 21). One may speculate that sodium Crit Care Med 2009 Vol. 37, No. 1

bicarbonate might lower the plasma creatinine, plasma urea, and NGAL concentrations by simple biochemical interference with the measurement methods. However, this is not likely given that both concentrations reached their maximum values several days after completion of study infusion. Urinary NGAL measurements by enzyme-linked immunosorbent assay are stable over a urinary pH range of 4 – 8. The clinical meaning of our findings is uncertain. Decreasing the change in serum creatinine by an average of ⬃20 ␮mol/L may have no beneficial clinical consequences at all. Larger studies are necessary to both confirm our findings and allow a better appreciation of their clinical value.

CONCLUSION We report the findings of a pilot double-blind, randomized controlled trial of intravenous sodium bicarbonate to attenuate cardiopulmonary bypassassociated acute renal dysfunction. We found sodium bicarbonate to be efficacious, safe, inexpensive and easy to administer. These findings now need to be confirmed or refuted by further clinical investigations in other geographic and institutional settings.

ACKNOWLEDGMENTS We thank our nurse coordinators for their assistance, the cardiac surgeons, anesthesiologists and nursing staff for their cooperation, and the patients for their participation in this study.

6.

7.

8.

9.

10.

11.

12.

13.

14.

REFERENCES 1. Albert MA, Antman EM: Preoperative evaluation for cardiac surgery. In: Cardiac Surgery in the Adult. Cohn LH, Edmunds LH Jr (Eds). New York, McGraw-Hill, 2003, pp 235–248 2. Chertow GM: Independent associations between acute renal failure and mortality following cardiac surgery. Am J Med 1998; 104: 343–348 3. Stafford-Smith M, Podgoreanu M, Swaminathan M, et al: Association of genetic polymorphisms with risk of renal injury after coronary bypass graft surgery. Am J Kidney Dis 2005; 45:519 –530 4. Ghotkar SV, Grayson AD, Fabri BM, et al: Preoperative calculation of risk for prolonged intensive care unit stay following coronary artery bypass grafting. J Cardiothorac Surg 2006; 1:14 5. Lassnigg A, Schmidlin D, Mouhieddine M, et


15.

16.

17.

18.

19.

al: Minimal changes of serum creatinine predict prognosis in patients after cardiothoracic surgery: A prospective cohort study. J Am Soc Nephrol 2004; 15:1597–1605 Zanardo G, Michielon P, Paccagnella A, et al: Acute renal failure in the patient undergoing cardiac operation. Prevalence, mortality rate, and main risk factors. J Thorac Cardiovasc Surg 1994; 107:1489 –1495 Doi K, Suzuki Y, Nakao A, et al: Radical scavenger edaravone developed for clinical use ameliorates ischemia/reperfusion injury in rat kidney. Kidney Int 2004; 65: 1714 –1723 McCord JM: Oxygen-derived free radicals in postischemic tissue injury. N Engl J Med 1985; 312:159 –163 Zager RA, Gamelin LM: Pathogenetic mechanisms in experimental hemoglobinuric acute renal failure. Am J Physiol 1989; 256: F446 –F455 Paller MS: Hemoglobin- and myoglobin induced acute renal failure in rats: Role of iron in nephrotoxicity. Am J Physiol 1988; 255: F539 –F544 Bove T, Landoni G, Calabro MG, et al: Renoprotective action of fenoldopam in high-risk patients undergoing cardiac surgery: A prospective, double-blind, randomized clinical trial. Circulation 2005; 111: 3230 –3235 Burns KE, Chu MW, Novick RJ, et al: Perioperative N-acetylcysteine to prevent renal injury in high-risk patients undergoing CABG surgery: A randomized controlled trial. JAMA 2005; 294:342–350 Haase M, Haase-Fielitz A, Bagshaw SM, et al: A phase II randomized controlled trial of high-dose N-acetylcysteine in high-risk cardiac surgery patients. Crit Care Med 2007; 35:1324 –1331 Mentzer RM Jr, Oz MC, Sladen RN, et al: Effects of perioperative nesiritide in patients with left ventricular injury undergoing cardiac surgery. J Am Coll Cardiol 2007; 49: 716 –726 Halliwell B, Gutteridge JM: Role of free radicals and catalytic metal ions in human disease: An overview. Methods Enzymol 1990; 186:1– 85 Chello M, Mastroroberto P, Romano R, et al: Complement and neutrophil activation during cardiopulmonary bypass: A randomized comparison of hypothermic and normothermic circulation. Eur J Cardiothorac Surg 1997; 11:162–168 Morita Y, Ikeguchi H, Nakamura J, et al: Complement activation products in the urine from proteinuric patients. J Am Soc Nephrol 2000; 11:700 –707 Atkins JL: Effect of sodium bicarbonate preloading on ischemic renal failure. Nephron 1986; 44:70 –74 Briguori C, Airoldi F, D’Andrea, et al: Renal Insufficiency Following Contrast Media Administration Trial (REMEDIAL): A random-

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

ized comparison of 3 preventive strategies. Circulation 2007; 115:1211–1217 Recio-Mayoral A, Chaparro M, Prado B, et al: The reno-protective effect of hydration with sodium bicarbonate plus N-acetylcysteine in patients undergoing emergency percutaneous coronary intervention: The RENO Study. J Am Coll Cardiol 2007; 49: 1283–1288 Merten GJ, Burgess WP, Gray LV, et al: Prevention of contrast-induced nephropathy with sodium bicarbonate: A randomized controlled trial. JAMA 2004; 291: 2328 –2334 Moher D, Schulz KF, Altman DG: The CONSORT statement: Revised recommendations for improving the quality of reports of parallel-group randomised trials. Lancet 2001; 357:1191–1194 Mehta RL, Kellum JA, Shah SV, et al: Acute kidney injury Network: Report of an initiative to improve outcomes in acute kidney injury. Crit Care 2007; 11:R31 Mishra J, Dent C, Tarabishi R, et al: Neutrophil gelatinase-associated lipocalin (NGAL) as a biomarker for acute renal injury after cardiac surgery. Lancet 2005; 365:1231–1238 Caulfield JL, Singh SP, Wishnok JS, et al: Sodium bicarbonate inhibits N-nitrosation in oxygenated nitric oxide solutions. J Biol Chem 1996; 271:25859 –25863 Ozcan EE, Guneri S, Akdeniz B, et al: Sodium bicarbonate, N-acetylcysteine, and saline for prevention of radiocontrast-induced nephropathy. A comparison of 3 regimens for protecting contrast-induced nephropathy in patients undergoing coronary procedures. A single-center prospective controlled trial. Am Heart J 2007; 154:539 –544 Masuda M, Yamada T, Mine T, et al: Comparison of usefulness of sodium bicarbonate versus sodium chloride to prevent contrast-induced nephropathy in patients undergoing an emergent coronary procedure. Am J Cardiol 2007; 100:781–786 Epub 2007 Jun 13 Tanaka K, Kanamori Y, Sato T, et al: Administration of haptoglobin during cardiopulmonary bypass surgery. ASAIO Trans 1991; 37: M482–M483 Anderson RJ, Ray CJ, Burke TJ: Human red blood cell hemolysate is a potent mitogen for renal tubular epithelial cells. Ren Fail 2000; 22:267–281 Schmidt-Ott KM, Mori K, Li JY, et al: Dual action of neutrophil gelatinase-associated lipocalin. J Am Soc Nephrol 2007; 18:407– 413 Roudkenar MH, Kuwahara Y, Baba T, et al: Oxidative stress induced lipocalin 2 gene expression: Addressing its expression under the harmful conditions. J Radiat Res (Tokyo) 2007; 48:39 – 44 Nashef SA, Roques F, Michel P, et al: European system for cardiac operative risk evaluation (EuroSCORE). Eur J Cardiothorac Surg 1999; 16:9 –13

47