A randomized unblinded pilot study comparing albumin versus hydroxyethyl starch in spontaneous bacterial peritonitis


  • See Editorial on Page 533

  • Potential conflict of interest: Nothing to report.


The administration of albumin improves circulatory function, prevents hepatorenal syndrome, and reduces hospital mortality in patients with cirrhosis and spontaneous bacterial peritonitis. This randomized unblinded pilot study compared the effect of albumin (10 patients) and the synthetic plasma expander hydroxyethyl starch 200/0.5 (10 patients) on the systemic hemodynamics of patients with spontaneous bacterial peritonitis. Baseline measurements were performed within 12 hours after diagnosis of infection. Patients then received 2 doses of the volume expander (1.5 g/kg body weight after baseline measurements and 1 g/kg body weight on day 3). Measurements were repeated after infection resolution. Treatment with albumin was associated with a significant increase in arterial pressure and a suppression of plasma renin activity, indicating an improvement in circulatory function. This occurred in the setting of a significant expansion of central blood volume (increase in cardiopulmonary pressures and atrial natriuretic factor) and an increase in systolic volume and systemic vascular resistance. In contrast, no significant changes were observed in these parameters in patients treated with hydroxyethyl starch. Von Willebrand–related antigen plasma levels significantly decreased in patients treated with albumin but not in those treated with hydroxyethyl starch. Serum nitrates and nitrites increased in patients treated with hydroxyethyl starch but not in those treated with albumin. These data suggest an effect of albumin on endothelial function. In conclusion, albumin but not hydroxyethyl starch improves systemic hemodynamics in patients with spontaneous bacterial peritonitis. This effect is due not only to volume expansion but also to an action on the peripheral arterial circulation. (HEPATOLOGY 2005.)

Patients with cirrhosis and spontaneous bacterial peritonitis (SBP) frequently develop a severe impairment in circulatory function, the pathogenesis of which is not well understood. Initial studies suggested that it was due to an accentuation of the arterial vasodilation, already present in patients with decompensated cirrhosis, related to a cytokine-mediated release of endothelial vasodilators, mainly nitric oxide (NO). However, recent investigations suggest that a decrease in cardiac output could also be involved.1–3 In approximately one third of the patients, this abnormality leads to type 1 hepatorenal syndrome, which is associated with high hospital mortality rate.1 A recent randomized controlled trial has shown that plasma expansion with albumin at infection diagnosis improves circulatory function, markedly reduces the incidence of hepatorenal syndrome and hospital mortality, and increases the 3-month probability of survival in patients with SBP.4 Albumin administration is, therefore, currently considered standard of care in SBP in our center.5 However, the use of albumin presents several problems. First, its supply is limited because it is obtained from human plasma. Second, it involves a potential risk of infectious disease transmission. Finally, it is very expensive.

This article reports the results of a randomized unblinded pilot investigation comparing the hemodynamic effects of albumin and of a synthetic plasma expander, hydroxyethyl starch 200/0.5 (HES 200/0.5), in patients with SBP. The doses of albumin and HES 200/0.5 and the schedule of treatment were similar to those used previously for the prevention of hepatorenal syndrome in SBP.4 The aim of the study was to determine whether albumin could be substituted by a synthetic plasma expander in patients with SBP. A previous investigation from our group showed that albumin administration in SBP is associated with an improvement in effective arterial blood volume, indicated by a significant increase in arterial pressure and a suppression of the renin-angiotensin system, related to an increase in central blood volume, cardiac work, and systemic vascular resistance.6 Accordingly, the end point of the current study was to investigate whether HES 200/0.5 induces an improvement in systemic hemodynamics comparable to that produced by albumin administration.

Hydroxyethyl starches (HESs) are modified natural polymers of amylopectin. The physical and chemical characteristics of these starches are defined by the degree of hydroxyethylation, which is the major determinant of circulating half-life, as well as by their molecular weight, which determines colloidal activity.7 As the molecular weight and number of substitutions of HESs increase, the side effects, mainly coagulation disorders and renal impairment, also increase.8 Several reasons justify the selection of HES 200/0.5 for this study. First, it is a compound with a medium molecular weight widely used in intensive care units because of its effectiveness, safety, and low cost.8, 9 It has been studied in depth in perioperative volume replacement, cardiac surgery, trauma, and sepsis.7–10 HES 200/0.5 is metabolized by alpha amylase, has an intravascular half-life of 4 to 6 hours, and is completely eliminated by the kidneys within 3 days.7, 8 Second, dextrans and polygeline solutions are less effective than albumin in the prevention of circulatory dysfunction after paracentesis in tense ascites.11 Third, a randomized trial has shown that HES 200/0.5 is as effective as albumin in the prevention of paracentesis-induced circulatory dysfunction.12 Fourth, HES 200/0.5 in combination with terlipressin has been found to be effective in the treatment of hepatorenal syndrome.13 Finally, HES 200/0.5 was as effective as albumin in the prevention of ovarian hyperstimulation syndrome, a condition also characterized by arterial vasodilation, marked decrease in arterial blood volume, and ascites.14–16

Von Willebrand factor is synthesized by endothelial cells and released in response to mechanisms that also stimulate NO release, such as cytokines and shear stress. It is considered to be a marker of endothelial activation.17, 18 In the current study, the effect of albumin and HES 200/0.5 on the plasma levels of this endothelial factor and on the serum levels of the NO metabolites, nitrates, and nitrites was also studied to assess possible differences in the effect of these solutions on endothelial function.


SBP, spontaneous bacterial peritonitis; NO, nitric oxide; HES 200/0.5, hydroxyethyl starch 200/0.5; RI, renal resistive index; HESs: Hydroxyethyl starches; TNF-α, tumor necrosis factor alpha; IL-6, interleukin-6; vWF:Ag, von Willebrand-related antigen.

Patients and Methods


Twenty consecutive patients with cirrhosis and SBP (polymorphonuclear cell count in ascitic fluid > 250/mm3 in the absence of findings suggestive of secondary peritonitis)19 treated with ceftriaxone (2 g intravenously immediately after diagnosis followed by 1 g intravenously every 24 hours) were included in the study (January 2002-February 2003). Exclusion criteria were: age below 18 or over 70 years; antibiotic administration (except for prophylactic norfloxacin), β-blocker or paracentesis treatment within 1 week before inclusion; cardiac, pulmonary, or renal disease; advanced hepatocellular carcinoma or human immunodeficiency virus infection; gastrointestinal hemorrhage within 1 month before the study; grade 3 to 4 hepatic encephalopathy20; other types of infection or septic shock. Renal disease was excluded by standard urine analysis (fresh urine sediment and 24-hour protein excretion) and abdominal ultrasonography.21 Cardiac disease was excluded by physical examination, chest radiograph, and electrocardiography. Patients or their relatives, in cases of hepatic encephalopathy, gave written informed consent to participate in the study, which was approved by the Investigation Committee of the hospital.


Physical examination, laboratory tests, ascitic fluid cultures, and chest radiograph were performed in all patients at SBP diagnosis. Patients were then treated with ceftriaxone. Baseline hemodynamic, ultrasonographic, and laboratory measurements were performed within 12 hours after diagnosis. Subsequently, patients were randomized to receive intravenous albumin (Albúmina 20%, Instituto Grífols, Barcelona, Spain; Group 1) or hydroxyethyl starch (Hesteril 6% 200/0.5, Fresenius Kabi, Barcelona, Spain; Group 2). Randomization was performed using sealed envelopes containing the treatment assignments, which were based on random numbers generated by the SAS statistical package. Ten patients were included in each group. Both plasma expanders were given at the same dose (1.5 g/kg body weight after baseline measurements and 1 g/kg body weight on day 3). The duration of infusion was 6 hours for albumin and 18 hours for HES 200/0.5 owing to the higher volume of fluid required for the administration of this latter compound. Diagnostic paracentesis was repeated every 2 days. SBP was considered resolved when clinical signs of infection had disappeared, the polymorphonuclear cell count in ascitic fluid decreased to less than 250 cells/mm3, total and differential white blood cell count normalized, and blood and ascitic fluid cultures were negative.19 Antibiotic treatment was maintained for 48 hours after infection resolution. Laboratory, ultrasonographic, and hemodynamic measurements were then repeated. Diuretics or therapeutic paracentesis were only allowed after the end of the study.

Hemodynamic Studies.

A venous catheter introducer was placed into the right jugular vein by the Seldinger technique. Blood samples for laboratory determinations were obtained with the patient being supine for at least 1 hour. A Swan-Ganz catheter (Edwards Lifesciences, Irvine, CA) was then advanced into the pulmonary artery for measurement of cardiopulmonary pressures and cardiac output (thermal dilution). Afterwards, a 7 F balloon catheter (MediTech, Boston Scientific Corp., Natick, MA) was advanced into the main right hepatic vein for measurement of wedged and free hepatic venous pressures. A solution of indocyanine green (Pulsion Medical Systems, Munich, Germany) containing 2% serum albumin was then infused intravenously at a rate of 0.1 or 0.2 mg/min (Child C and Child B patients, respectively). After an equilibration period of 40 minutes, 4 separate sets of samples of peripheral and hepatic venous blood were obtained for measurement of hepatic blood flow.22 A hepatic extraction greater than 10% was required for the calculation of the hepatic blood flow. All hemodynamic measurements were performed in triplicate. Mean arterial pressure was measured with an automatic sphygmomanometer (Mac-Lab, K74A0841J; Marquette, Milwaukee, WI). Heart rate was derived from continuous electrocardiogram monitoring. Systemic vascular resistance (dyn·sec/cm5) was calculated as (mean arterial pressure − right atrial pressure)/cardiac output × 80. Pressures are reported in mmHg and cardiac output as cardiac index (L/min/m2). The left ventricular stroke work index was calculated as (systolic volume/body surface area) × (mean arterial pressure − pulmonary capillary pressure) × 0.0136) (g·m/m2).23

Normal values in our laboratory are: cardiac index, 2.5 to 4.0 L/min/m2; systolic volume, 60 to 100 mL/beat; left ventricular stroke work index, 45 to 75 g·m/m2/beat; systemic vascular resistance, 900 to 1,600 dyn·sec/cm5; mean arterial pressure, 80 to 95 mmHg, right atrial pressure, 2 to 10 mmHg, mean pulmonary artery pressure, 10 to 20 mmHg; pulmonary capillary pressure, 6 to 14 mmHg; hepatic venous pressure gradient, 1 to 5 mmHg.

Ultrasonographic Studies.

The renal resistive index (RI) was measured by color-Doppler ultrasonography using a multifrequency convex transducer (3-6 MHz Toshiba Powervision 6000, Tokyo, Japan). It was defined as peak systolic frequency shift − minimum diastolic frequency shift/peak systolic frequency shift and determined from the Doppler waveform by hand measurement with a caliper obtained in the arcuate arteries of the right and left kidney.24, 25 The RI for each patient was calculated as an average value obtained from one measurement recorded in each kidney. Normal values are below 0.7.

Laboratory Measurements.

Plasma renin activity and plasma concentration of atrial natriuretic factor, tumor necrosis factor alpha (TNF-α), and interleukin-6 (IL-6) were measured in blood samples taken during the hemodynamic studies in baseline conditions and after infection resolution. Samples were placed on ice, centrifuged at 4°C, and stored at −30°C until assayed. Plasma renin activity and atrial atriuretic factor were measured by RIA (Clinical Assay, Baxter, Cambridge, MA, and Euro-Diagnostica, Malmö, Sweden)26, 27 and IL-6 and TNF-α by ELISA (Medgenix Diagnostics SA, Fleurus, Belgium).2 Normal values in our laboratory are: plasma renin activity, 1.4 ± 0.9 ng·mL/h; atrial natriuretic factor, 19 ± 5 fmol/mL; IL-6, < 5 pg/mL, TNF-α, < 20 pg/mL.

To assess the effects of albumin and HES 200/0.5 on the production of von Willebrand–related antigen factor (vWF:Ag) and NO, the coagulant activity of factor VIII, the plasma levels of vWF:Ag, and the serum concentration of the NO metabolites, nitrates, and nitrites, were measured at baseline and within 8 hours after the second dose of each plasma volume expander. vWF:Ag was measured by ELISA (DG-EIA-vWF, Corgenix Inc, Westminster, CO; limit of detection 1.1 U/dL, intra- and interassay variation coefficients <5.0%) and factor VIII by a chromogenic substrate technique (Coamatic Factor VIII, Chromogenix Instrumentation Lab SpA, Milano, Italy). To measure serum levels of nitrates and nitrites (NO2− and NO3−), samples were ultrafiltered (PL-10 Ultrafree-MC centrifugal filter units; Millipore Corp., Bedford, MA) at 1,200g for 1 hour to remove proteins. Filtered serum was refluxed in glacial acetic acid containing sodium iodide. Under these conditions, NO2− and NO3− are reduced to NO, which, after reacting with ozone, can be quantified by a chemoluminescence detector (Nitric Oxide Analyzer, NOA 280, Sievers Instruments, Boulder, CO).28 Normal values in our laboratory are: vWF:Ag, 60 to 160 U/dL; factor VIII coagulant activity, 60 to 160 U/dL; and NO2− and NO3−, 37 ± 14 nmol/mL.


Renal failure was defined as blood urea nitrogen or serum creatinine greater than 25 mg/dL or 1.5 mg/dL, respectively. Circulatory dysfunction was defined as a 100% increase in plasma renin activity over baseline level up to a value greater than 4 ng·mL/h (upper value in control subjects on a 50-mEq sodium diet). Hepatorenal syndrome was defined according to the International Ascites Club.21 The systemic inflammatory response syndrome was defined according to the Consensus Conference on Sepsis and Multiple Organ Failure29 and diagnosed when 2 or more of the following features were present: temperature greater than 38°C or less than 36°C, heart rate greater than 90 beats/min, tachypnea greater than 20 breaths/min, white cell count greater than 12 × 109/L or less than 4 × 109/L or more than 10% of immature neutrophils.

Statistical Analysis.

The Mann-Whitney U test and the Wilcoxon test for quantitative variables, and the chi square test for qualitative variables, applying Yates's correction when required, were used. Results are expressed as mean ± SD. Differences were considered significant at the level of P < .05. Analysis was performed with the SPSS Statistical package (SPSS Inc. version. 10.0, 2000, Chicago, IL).


Clinical, Hemodynamic, and Laboratory Data at Infection Diagnosis.

No significant differences were found between patients treated with albumin (group 1) and those treated with HES 200/0.5 (group 2) in baseline characteristics except for a lower serum albumin in group 2 (Tables 1–3). SBP was associated with an intense inflammatory response, as indicated by the large number of cases with systemic inflammatory response syndrome and the high concentration of polymorphonuclear leucocytes in ascitic fluid (Table 1) and of IL-6 and TNF-α in plasma in both groups (Table 2).

Table 1. Clinical Data of Patients From Groups 1 and 2 at Diagnosis of Infection
 Group 1 (Albumin)Group 2 (Hydroxyethyl Starch)
  1. NOTE. No significant differences were observed between groups.

Age (yr)62 ± 960 ± 10
Sex (male)65
Etiology (HCV/non HCV)6/46/4
Previous variceal bleeding (n)23
Hepatic encephalopathy (n)33
Renal impairment (n)53
Systemic inflammatory response syndrome (n)65
Peripheral leukocyte count (/109)6,967 ± 3,5837,379 ± 2,819
Ascitic fluid polymorphonuclear count (/mm3)3,719 ± 2,8725,407 ± 6,342
Total serum protein (g/L)63 ± 556 ± 11
Ascitic fluid protein (g/L)10 ± 69 ± 6
Community-acquired/nosocomial SBP (n)7/34/6
Culture positive SBP (n)74
Table 2. Changes in Hepatic and Renal Function and Cytokines After Resolution of SBP in Groups 1 and 2
 Group 1 (Albumin)Group 2 (Hydroxyethyl Starch)
At DiagnosisAt ResolutionPAt DiagnosisAt ResolutionP
  1. NOTE. P < .05 group 2 vs. group 1.

  2. Abbreviation: BUN, blood urea nitrogen.

Hepatic function      
 Serum bilirubin (mg/dL)6.6 ± 8.66.2 ± 9.0.656.5 ± 7.86.1 ± 7.7.45
 Serum albumin (g/L)28 ± 531 ± 4.0322 ± 4#22 ± 3#.51
 Prothrombin index (%)45 ± 1644 ± 22.3243 ± 1646 ± 10.31
 Child-Turcotte-Pugh (score)10.0 ± 2.49.3 ± 2.7.1710.9 ± 2.810.4 ± 2.0.19
Renal function      
 Serum creatinine (mg/dL)1.6 ± 0.81.0 ± ± 0.51.0 ± 0.2.02
 BUN (mg/dL)34 ± 2326 ± 17.0526 ± 1820 ± 11.26
 Serum sodium (mEq/L)131 ± 6134 ± 7.17128 ± 6130 ± 8.10
 Plasma IL-6 (pg/mL)719 ± 101249 ± 48.02784 ± 119778 ± 83.02
 Plasma TNF-α (pg/mL)133 ± 12699 ± 28.3795 ± 6970 ± 16#.88
Table 3. Changes in Systemic, Splanchnic and Renal Hemodynamics and Hormonal Systems After Resolution of SBP in Groups 1 and 2
 Group 1 (Albumin)Group 2 (Hydroxyethyl Starch)
At DiagnosisAt ResolutionPAt DiagnosisAt ResolutionP
  • *

    NOTE. Hepatic blood flow was measured in 5 patients from group 1 and in 4 from group 2 (ICG extraction <10% in 11 patients).

  • P < .05 group 2 vs. group 1.

Systemic hemodynamics      
 Mean arterial pressure (mmHg)76 ± 985 ± 13.0180 ± 1581 ± 8.36
 Cardiac index (L/min/m2)4.9 ± 0.74.6 ± ± 1.04.9 ± 1.4.44
 Systemic vascular resistance (dyn·sec/cm5)668 ± 134803 ± 197.03777 ± 239778 ± 290.96
 Systolic volume (mL)97 ± 20104 ± 17.0583 ± 1592 ± 23.12
 Heart rate (beats/min)91 ± 1480 ± 13.0190 ± 1084 ± 9.01
 Stroke work index (g·m/m2)51 ± 1759 ± 17.0148 ± 1353 ± 13.24
 Right atrial pressure (mmHg)8 ± 29 ± 2.037 ± 37 ± 31.0
 Pulmonary artery pressure (mmHg)17 ± 421 ± 5.0118 ± 618 ± 7.88
 Pulmonary capillary pressure (mmHg)11 ± 314 ± 4.0310 ± 411 ± 5.51
Plasma hormonal systems      
 Plasma renin activity (ng·mL/h)5.7 ± 4.73.1 ± ± 7.316.8 ± 24.6.65
 Atrial natriuretic factor (fmol/mL)55 ± 2185 ± 37.0545 ± 3228 ± 18.51
Splanchnic and renal hemodynamics      
 Wedged hepatic venous pressure (mmHg)30.6 ± 2.931.9 ± 4.0.3130.6 ± 5.631.6 ± 5.2.29
 Free hepatic venous pressure (mmHg)12.8 ± 3.013.4 ± 2.1.4710.5 ± 3.511.0 ± 2.7.57
 HVPG (mmHg)17.8 ± 2.518.5 ± 3.1.3120.1 ± 4.320.6 ± 4.6.59
 Hepatic blood flow (L.min−1)*0.8 ± 0.21.0 ± ± 0.41.2 ± 0.5.71
 Renal resistive index0.80 ± 0.040.78 ± ± 0.020.81 ± 0.04.44

Most patients were severely ill, with poor hepatic and renal function (Tables 1 and 2). Eight patients had renal failure (5 from group 1 and 3 from group 2); and six, hepatic encephalopathy (3 from each group) (Table 1). Systemic hemodynamics showed the characteristic hyperdynamic circulation (arterial hypotension, low peripheral vascular resistance, and high cardiac index) associated with increased levels of plasma renin activity in the 2 groups (Table 3). Cardiopulmonary pressures were within normal limits. All patients had severe portal hypertension and increased RI.

Changes in Systemic, Splanchnic, and Renal Hemodynamics and in Hepatic and Renal Function (Tables 2 and 3).

SBP resolved in all patients. The mean time to infection resolution was 4 ± 1 days and 5 ± 1 days in groups 1 and 2, respectively (range, 2-6 days). SBP resolution was associated with a significant decrease in the plasma concentration of IL-6 in both groups (Table 2).

SBP resolution in group 1 was associated with a significant increase in mean arterial pressure (9.05 ± 8.01 mmHg, 12%; P = .01), right atrial pressure (1.7 ± 1.8 mmHg, 21%; P = .03), pulmonary capillary pressure (2.7 ± 3.2 mmHg, 25%; P = .03), pulmonary artery pressure (3.6 ± 4.6 mmHg, 21%; P = .01), systolic volume (7 ± 9 mL/beat, 7%; P = .05), left ventricular stroke work index (11 ± 9 g·m/m2, 19%; P < .01), and systemic vascular resistance (135 ± 135 dyn·sec/cm5, 20%; P = .03) and a reduction in heart rate (11 ± 9 bpm, 12%; P = .01). The cardiac index remained unmodified. These changes were associated with a suppression of plasma renin activity (2.6 ± 3.8 ng·mL/h, 46%; P = .04) and increase in plasma concentration of atrial natriuretic factor (30 ± 38 fmol/mL, 55%, P = .05). In contrast, SBP resolution in group 2 was not associated with significant changes in any of these parameters except for a decrease in heart rate (5 ± 6 beats.min, 6%; P = .01). No changes were observed in wedged and free hepatic venous pressures, hepatic venous pressure gradient, hepatic blood flow, and RI in the 2 groups. Plasma albumin increased significantly in group 1 (Table 2).

No patient from group 1 developed SBP-induced circulatory dysfunction or renal failure. In contrast, 3 patients from group 2 developed SBP-induced circulatory dysfunction and 1 an acute renal failure characterized by a decrease in urine volume (from 525 mL/24 hours to 310 mL/24 hours), sodium excretion (from 5 to 3 mEq/24 hours) and mean arterial pressure (from 76 to 72 mmHg), increase in plasma renin activity (from 18.5 to 74 ng·mL/h), reduction in cardiac index (from 3.2 to 2.5 L/min/m2) and increase in systemic vascular resistance (from 1,263 to 1,488 dyn·sec/cm5).

Effects on Factor VIII Coagulant Activity, vWF:Ag, and NO Metabolites.

Patients from both groups presented high baseline plasma levels of factor VIII and vWF:Ag (Fig. 1). Albumin administration was associated with a significant decrease in the plasma levels of vWF:Ag (from 331 ± 35 to 257 ± 65 U/dL; −22%, P = .01) and factor VIII (from 214 ± 33 to 155 ± 36 U/dL; −28%, P = .02). The serum levels of nitrates and nitrites remained unchanged (from 61 ± 30 to 78 ± 55 nmol/mL; P = .17). In contrast, HES 200/0.5 administration was not associated with significant changes in vWF:Ag (from 297 ± 44 to 278 ± 47 U/dL; −6%, P = .12) or factor VIII (from 220 ± 53 to 212 ± 52 U/dL; −8%, P = .21), but with a significant increase in nitrates and nitrites (from 39 ± 13 to 63 ± 32 nmol/mL; +64%, P = .03).

Figure 1.

Mean values of von Willebrand–related antigen (vWF:Ag), coagulant activity of factor VIII, and nitric oxide (NO) at diagnosis (black columns) and on day 4 (gray columns) in patients receiving albumin (A) and in those who received hydroxyethyl starch (B).


Investigations performed during the last decade have led to a better understanding of the pathogenic mechanisms associated with SBP.1–3 The current concept is that SBP in patients with cirrhosis, particularly in those with severe liver failure, renal insufficiency, and intense inflammatory response, produces an accentuation of the arterial vasodilation already present in these patients and a marked decrease in cardiac output.2, 3 Arterial vasodilation is thought to be caused by the release of vasodilator molecules, such as carbon monoxide and NO.30, 31 The decrease in cardiac output is related to a reduction in cardiac preload, an impairment in the chronotropic function of the heart, and perhaps to the cardiomyopathy of cirrhosis aggravated by the deleterious effect of sepsis on myocardial function.3 As a result, a severe reduction in effective arterial blood volume develops, leading to marked homeostatic activation of the renin-angiotensin and sympathetic nervous systems. These systems maintain arterial pressure within limits compatible with life. However, they produce vasoconstriction in essential organs such as the kidneys, brain, and liver, leading to renal hypoperfusion, reduction in hepatic and cerebral blood flow, and increase in intrahepatic vascular resistance. The clinical consequences of these changes are hepatorenal syndrome (generally type 1), rapid progression of liver failure, hepatic encephalopathy, aggravation of portal hypertension, and a high risk of variceal bleeding.3 Interestingly, once these changes are initiated, they progress in most cases despite a rapid resolution of the infection. A recent study in patients with cirrhosis has shown that bacterial infections other than SBP are also associated with circulatory dysfunction, hepatorenal syndrome, and a poor short-term survival.32

The concept that the severity of the clinical course of patients with cirrhosis with serious bacterial infection is related to an impairment of circulatory function has led to new and effective approaches in the prevention and treatment of these complications. Plasma volume expansion with albumin at infection diagnosis in patients with cirrhosis with SBP, bilirubin over 4 mg/dL, or serum creatinine over 1 mg/dL prevents circulatory dysfunction and is associated with a drastic reduction (60%) in the incidence of type 1 hepatorenal syndrome and hospital mortality.4 Moreover, simultaneous treatment with vasoconstrictors (terlipressin, noradrenaline, midodrine) and intravenous albumin has been shown to reverse type 1 hepatorenal syndrome and to improve the survival of these patients.33–35

The current study represents the first investigation aimed at exploring whether albumin can be substituted by synthetic plasma expanders in patients with SBP. The oncotic capacity of 1 g albumin is identical to that of 1 g HES 200/0.5. However, the pharmacokinetics and pharmacodynamics of both substances as well as the characteristics of the solutions are markedly different.7, 8, 36 In healthy subjects, the half-life of albumin is of 19 days, although it decreases to 9 days in patients with sepsis and perhaps even more in patients with SBP accumulating ascites.37 In contrast, in healthy subjects, the half-life of HESs ranges from 6 hours to 3 days.7, 8 Albumin presents numerous biological effects in addition to its oncotic action, and this is probably not the case with HESs.37 Finally, albumin solutions are prepared at high concentrations (20%) in a saline solution containing 130 160 mEq/L sodium in the form of caprylate and acetyltryptophanate salts. Therefore, high doses of albumin can be given within a short period without inducing a significant water and sodium overload. In contrast, HES 200/0.5 is prepared at a 6% concentration in a sodium chloride solution. This limits the administration of high doses of this compound within a short period. For an identical amount of the active compound, in the current study the amount of water and sodium infused was 3.3 times higher and the amount of oncotic equivalents per unit of time 3 times lower in patients treated with HES 200/0.5. The role played by these differences in our findings is difficult to ascertain.

Our results show clear differences in the effect of the 2 plasma expanders on central blood volume. The right atrial pressure and the pulmonary artery and capillary pressures increased in patients treated with albumin but not in those receiving HES 200/0.5. This feature explains why the plasma levels of atrial natriuretic peptide, a hormone that is released in response to central blood volume expansion, only increased in patients treated with albumin. Albumin administration was also associated with an increase in systolic volume, probably as a consequence of an increase in left ventricular filling. This was not seen in patients treated with HES 200/0.5. Secondarily, cardiac work increased in patients treated with albumin but not in those receiving HES 200/0.5. These data suggest that albumin induces a more effective expansion of the central blood volume, probably because of the longer half-life of albumin as compared with HES 200/0.5.

An interesting observation, which confirms a previous investigation by our group,6 is that the administration of albumin in patients with SBP increased not only cardiopulmonary pressures and cardiac work, but also mean arterial pressure. This effect was attributable to an increase in peripheral vascular resistance, suggesting that albumin administration has an effect on the peripheral arterial circulation. This was not seen in patients treated with HES 200/0.5.

The current study was aimed at comparing changes in cardiovascular hemodynamics induced by the administration of albumin and HES 200/0.5 in SBP. The design of the study and the number of cases included, therefore, preclude the drawing of conclusions on the clinical effects of the 2 plasma expanders in patients with SBP. Nevertheless, there were clear differences in the incidence of circulatory dysfunction between the 2 groups. Albumin infusion was associated with a significant reduction in plasma renin activity, indicating an improvement in circulatory function, and no patient in this group developed SBP-induced circulatory dysfunction or renal failure. In contrast, plasma renin activity increased albeit not significantly; in patients treated with HES 200/0.5, 3 patients developed SBP-induced circulatory dysfunction, and 1 developed acute renal failure with the characteristics of type 1 hepatorenal syndrome. Serum creatinine concentration decreased significantly in both groups (from 1.6 ± 0.8 to 1.0 ± 0.3 mg/dL in patients treated with albumin and from 1.2 ± 0.5 to 1.0 ± 0.2 mg/dL in patients receiving HES 200/0.5). The significance of these small changes in the setting of a severe bacterial infection is, however, difficult to ascertain.

The production of NO, as estimated by the serum levels of nitrates and nitrites, is higher in patients with SBP than in those with uncomplicated ascites and increases during several days following the resolution of the infection. This feature is also observed when the concentrations of nitrates and nitrites are measured in ascitic fluid.31, 38 There is, therefore, a clear dissociation between the time course of cytokines in plasma and ascitic fluid, the concentration of which decreases markedly after the resolution of the infection, and the plasma and ascitic fluid levels of nitrates and nitrites, which reach the highest levels several days after SBP resolution.2, 31, 38 This dissociation is not surprising considering that the inducible form of NO synthase is expressed in many cell types and that the production of NO is maintained over relatively long periods after challenging the cells with immunological or inflammatory stimuli.38 In our patients treated with HES 200/0.5, we observed a significant increase in the serum levels of nitrates and nitrites despite the resolution of the infection. In contrast, in patients receiving albumin, no significant increase in the serum levels of nitrates and nitrites was observed. Another interesting observation was that albumin but not HES 200/0.5 administration was associated with a significant decrease in the plasma levels of factor VIII and vWF:Ag. These data suggest that albumin, but not HES 200/0.5 administration, decreases endothelial activation in patients with SBP. Further studies in larger series of patients are needed to confirm these observations.

In summary, the administration of albumin improves circulatory function in patients with SBP. The mechanism of this effect includes a sustained expansion of the circulating blood volume, an increase in cardiac work, and an increase in peripheral vascular resistance. An inhibitory effect of albumin on endothelial function is also possible. These effects were not observed after the administration of the synthetic plasma expander HES 200/0.5. This pilot study of small sample size, therefore, suggests that HES 200/0.5 is not a suitable substitute for albumin in the management of patients with SBP.