| 2008 JASN IMPACT FACTOR 7.505 | HOME AUTHOR INFO EDITORIAL BOARD SUBSCRIBE FEEDBACK ALERTS HELP | |||
| CURRENT ISSUE | ARCHIVES | JASN Express | ONLINE SUBMISSION | |





*
Division of Gastroenterology, Department of Medicine, Indiana University
School of Medicine, Indianapolis, Indiana.
Division of Clinical Pharmacology, Department of Medicine, Indiana
University School of Medicine, Indianapolis, Indiana.
Division of Biostatistics, Department of Medicine, Indiana University
School of Medicine, Indianapolis, Indiana.
Correspondence to Dr. D. Craig Brater, Indiana University School of Medicine, 1120 South Drive, Fesler Hall 302, Indianapolis, IN 46202-5114. Phone: 317-274-8157; Fax: 317-274-8439; E-mail: dbrater{at}iupui.edu
| Abstract |
|---|
|
|
|---|
| Introduction |
|---|
|
|
|---|
Clinical trials assessing this strategy are conflicting, and most have not been rigorously performed (13, 17,18,19,20). Despite the uncertainty of efficacy, many physicians administer furosemide/albumin mixtures to enhance diuresis in hypoalbuminemic patients, particularly those with nephrotic syndrome or cirrhosis (13). Examination of this issue among patients with nephrotic syndrome is confounded by the proteinuria of such patients, which can lead to urinary binding of diuretic agents (21). Hypoalbuminemic patients with cirrhosis thus represent a better model to study this issue. We therefore conducted a randomized crossover study to determine the effects of albumin/furosemide mixtures on the response to furosemide in cirrhotic subjects with ascites.
| Materials and Methods |
|---|
|
|
|---|
|
Protocol
Two weeks before admission, patients who were not receiving spironolactone
began to be treated with a dose of 50 mg twice daily. This strategy was
selected in preference to discontinuing spironolactone treatment for all
patients because of uncertainty regarding the length of spironolactone
treatment cessation needed to allow all effects to dissipate. Participants
were admitted to the General Clinical Research Center at Indiana University
Medical Center, where they remained until completion of the study. At the time
of admission, they were placed on a metabolic diet containing 30 mEq of sodium
and 60 to 80 mEq of potassium, with 3 L/d of dietary fluids. Administration of
all other diuretic agents was discontinued at the time of admission. This
sodium restriction allowed safe discontinuation of these diuretic agents
without weight gain throughout the study. Full chemistry panel and complete
blood count analyses, urinalysis, and baseline weight assessments were
performed for each subject at the time of admission. Thereafter, individuals
were weighed and serum electrolyte and creatinine concentrations were measured
each morning. In addition, 24-h urine samples were collected each day for
measurement of electrolytes and creatinine. Patients were equilibrated on the
metabolic diet until they attained sodium balance, as defined by two
consecutive 24-h urinary sodium excretion values that varied
20% and no
change in two consecutive daily weights of >0.5 kg. After sodium balance
was attained, participants underwent one of the four phases of the study, in
random order, as follows: (1) 25 g of albumin alone administered
intravenously in 30 min, (2) 40 mg of furosemide alone administered
intravenously in 30 min, (3) albumin (25 g) and furosemide (40 mg)
premixed ex vivo for 10 min and infused intravenously in 30 min,
which duplicated the method used by Inoue et al.
(18), or (4) albumin
(25 g) and furosemide (40 mg) infused intravenously, into opposite forearms
simultaneously, in 30 min.
Patients fasted, except for distilled water, from midnight until 4 h after administration of the study medication. A 10 ml/kg distilled water load was administered orally before the start of the 30-min infusion of furosemide and/or albumin, to ensure the ability to produce frequent urine samples. Urine and serum samples were collected before the dose and 0.5, 1, 1.5, 2, 2.5, 3, 4, 6, 8, and 24 h after the dose. Urine losses were replaced for 8 h with equal volumes of one-half normal saline solution administered intravenously, to prevent volume depletion and the development of acute diuretic tolerance. The concentration of urinary sodium elicited by loop diuretics is reasonably well approximated by 0.45% normal saline solution; therefore, this method maintained volume status in each subject.
After finishing the first phase of the protocol, subjects continued to receive the metabolic diet until they reached sodium balance, as previously defined. The second phase of the protocol was then performed in a manner identical to the first. Similarly, the third and fourth phases of the protocol were conducted after each volunteer attained sodium balance. At least 48 h separated each phase. This design allowed individual subjects to achieve comparable states of sodium balance before each phase of the study, so that they could serve as their own control subjects. Table 2 demonstrates that the participants were in comparable clinical conditions before each phase of the study.
|
Analyses
Serum and urine samples were assayed for sodium, furosemide, and creatinine
using techniques described in detail elsewhere
(22,23,24,25).
Because creatinine clearance values did not change during the study, those
data are not reported. Responses were analyzed in several ways. First, total
sodium excretion was compared for different treatments. Second, the
sensitivity of the nephron to furosemide was determined, as described
previously, by relating the urinary furosemide excretion rate to the sodium
excretion rate
(22,23,24,25).
Third, the pharmacokinetics of furosemide were examined, because of the
potential of albumin to alter these parameters. Standard model-independent
methods were used to determine the pharmacokinetic parameters of interest
(Win-Nonlin version 1.1; Scientific Consulting, Apex, NC). The terminal
elimination rate constant (ß) was determined by linear regression. The
elimination half-life was determined as t1/2 = 0.693/ß. The area
under the serum concentration versus time curve
(AUC0
) was determined by a combination of linear
and logarithmic trapezoidal methods, with extrapolation to infinity from the
last measured serum concentration, using the terminal elimination rate
constant. The clearance of furosemide was calculated as
dose/AUC0
.
Statistical Analyses
Statistical analyses were performed using SAS software (SAS, Inc. Cary,
NC). Univariate repeated-measures ANOVA models for a crossover design were
used to analyze the urine excretion measures. Treatment effects were tested in
the primary analysis model with control for the period of study
(i.e., the time the subject had been in the study). The inclusion of
the period main effect accounted for the effects of metabolic diet duration on
the treatment response. Person was treated in the model as a random effect, to
account for the four repeated measurements obtained for each person (one with
each treatment). Separate models were used for 6- and 24-h urine measurements.
Because diuretic responses returned to baseline levels within 6 h (see below)
and because 24-h results did not differ from 6-h findings, the 6-h data are
reported. When an overall treatment effect was significant, pairwise
comparisons between the treatments were tested with P adjustment
using Sidak's multiple-comparison procedure. The effects of the baseline serum
albumin concentration on the response to different treatments were tested by
adding serum albumin level-treatment interaction to the primary analysis
model. Albumin was tested as a continuous variable as well as a categorical
variable, i.e., <3 g/dl (n = 7) versus >3
g/dl (n = 6). Carryover effect was tested by adding the treatment
from the previous period to the primary analysis model. P values of
0.05 were considered statistically significant.
| Results |
|---|
|
|
|---|
|
|
|
|
Despite a wide range of serum albumin concentrations (2.1 to 4.3 g/dl), we also failed to detect any interaction between serum albumin concentrations and the response to furosemide, whether the albumin concentration was analyzed as a continuous variable or as a categorical variable (P > 0.05).
Furosemide Pharmacokinetics
Figure 4 demonstrates serum
furosemide concentrations versus time, and
Table 4 lists the estimated
pharmacokinetic parameters. Albumin infusion would be predicted to potentially
have two effects on the pharmacokinetics of furosemide. First, its binding of
furosemide might result in increased AUC and decreased clearance and volume of
distribution values. This did not occur in our study. Second, albumin might
enhance urinary furosemide excretion, as occurred in the animal study of Inoue
et al. (18).
Table 3 demonstrates a lack of
effect of albumin on the total amount of furosemide excreted into urine;
Figure 2 demonstrates that
albumin infusion had no effect on the time course of furosemide excretion.
|
|
| Discussion |
|---|
|
|
|---|
There has long been interest in the use of intravenously administered albumin to enhance diuresis in hypoalbuminemic patients. After salt-poor human albumin became available in 1944, several anecdotal reports suggested that albumin infusions could enhance diuresis in cirrhotic patients (4,5,6,7,8). In contrast, a randomized study published in 1962 demonstrated that repeated albumin infusions failed to decrease the diuretic needs of cirrhotic subjects with refractory ascites (9). Similar studies of patients with nephrotic syndrome have demonstrated no utility of albumin alone in the treatment of this disorder (13,14,15,16). Moreover, data from this study indicate no diuretic effect of albumin alone (Table 3 and Figure 1). More recently, however, Gentilini et al. (17) demonstrated that albumin infusion produced a 26% increase in sodium excretion caused by furosemide in hospitalized cirrhotic patients with ascites. That study did not characterize the pharmacokinetics or pharmacodynamics of furosemide and therefore did not offer any mechanistic clues regarding why albumin would enhance the efficacy of furosemide. It did reinforce the uncertainty regarding the utility of albumin in enhancing diuretic responses.
Studies of patients with nephrotic syndrome have been similarly conflicting. The seminal study by Inoue et al. (18) in analbuminemic rats also reported the effects of an ex vivo mixture of furosemide and albumin in four patients with nephrotic syndrome. All patients exhibited an increase in urine volume, compared with furosemide alone. No information was provided with respect to the design of this clinical component of their study or the results in terms of sodium excretion. Akicek et al. (19) studied the effects of albumin alone, furosemide alone, and the combination in eight hypoalbuminemic patients with nephrotic syndrome. Each patient received each treatment, in random order, but there was no re-equilibration between phases of the study. Albumin alone had negligible natriuretic effects (13 ± 8 mEq/4 h) and had no effect on the response to furosemide. Fliser et al. (20) also studied nine patients with nephrotic syndrome. Their study included dietary equilibration, and they measured the amount of furosemide that reached the urinary site of action. There was no effect on urinary furosemide levels. Albumin increased the response to furosemide by 20%. The mechanism seemed to involve an increase in renal blood flow. Those authors concluded that the effects were statistically significant but likely not clinically relevant.
On the basis of the aforementioned data for analbuminemic rats and the data presented above for patients with hypoalbuminemia, it was unclear whether the hypothesis constructed from the animal data could be extrapolated to hypoalbuminemic patients, leading to the motivation for our study. We think that patients with cirrhosis represent the best clinical model for examination of the principles underlying the potential utility of albumin/furosemide mixtures. In nephrotic syndrome, results are likely confounded by the rapid excretion of administered albumin into the urine and the ability of albumin to bind loop diuretics in the urine (21). Our data convincingly demonstrated that coadministration of albumin did not enhance the diuretic response to furosemide. Correspondingly, the pharmacokinetics and pharmacodynamics of furosemide were not altered by concomitant albumin administration (Tables 3 and 4 and Figures 2 and 4). This lack of effect has several possible explanations. First, the dose of albumin infused may not have been sufficient. We doubt that this is a reasonable explanation, because the dose of albumin used in our study was twice the amount used by Gentilini et al. (17) and was the same as that used by Inoue et al. (18). Moreover, the use of larger doses of albumin would not be practical and would be expensive. Second, it may be necessary to administer repeated doses of albumin to produce benefits. This issue was not addressed by our study, but previous studies with repeated doses of albumin yielded mixed results (4,5,6,7,8,9). Third, it is possible that the baseline serum albumin concentration in our study population was not low enough to yield a benefit from albumin infusion. However, we consider this possibility to be unlikely, because similar albumin concentrations were observed in the patients described by Gentilini et al. (17) and in our patients (3.0 ± 0.7 and 3.0 ± 0.6 g/dl, respectively). The range of albumin concentrations for our patients was 2.1 to 4.3 g/dl. Therefore, we included patients with substantial hypoalbuminemia. In addition, we tested for a relationship between diuretic response and serum albumin concentrations and found none. Fourth, the sample size may not have been sufficient for detection of a significant effect. We observed a mean difference in the 6-h urinary sodium excretion produced by furosemide with and without albumin of 9.7 mEq, with a SD of 81.7 mEq. On the basis of these data, we would need to study 563 patients to demonstrate a difference with 80% power at the 5% significance level. Even if we proved a significant effect with such a sample size, it is apparent that the magnitude of that effect would not be clinically relevant.
Although our crossover design decreased interindividual variability, the design presents a potential risk of carryover effects. We minimized such effects by including a washout period and by attaining sodium balance before each phase of the study. Moreover, the statistical absence of any effect of previous treatment argues against carryover effects.
In conclusion, albumin administered in an ex vivo mixture with furosemide or administered simultaneously with furosemide did not enhance diuretic effects in patients with cirrhosis and ascites. In addition, the administration of albumin did not alter the pharmacokinetics or pharmacodynamics of furosemide. These data argue against the clinical use of this therapeutic strategy. It is likely that these results can be extrapolated to other hypoalbuminemic disorders, such as nephrotic syndrome.
| Acknowledgments |
|---|
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
G. Kapur, R. P. Valentini, A. A. Imam, and T. K. Mattoo Treatment of Severe Edema in Children with Nephrotic Syndrome with Diuretics Alone -- A Prospective Study Clin. J. Am. Soc. Nephrol., May 1, 2009; 4(5): 907 - 913. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. J Elwell, A. P Spencer, and G. Eisele Combined Furosemide and Human Albumin Treatment for Diuretic-Resistant Edema Ann. Pharmacother., May 1, 2003; 37(5): 695 - 700. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. S. Shankar and D. C. Brater Loop diuretics: from the Na-K-2Cl transporter to clinical use Am J Physiol Renal Physiol, January 1, 2003; 284(1): F11 - F21. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
HOME
CURRENT ISSUE
ARCHIVES
JASN Express
ONLINE SUBMISSION
AUTHOR INFO
EDITORIAL BOARD SUBSCRIBE FEEDBACK ALERTS HELP |