Albumin for SBP

Published

Sort P, et al. "Effect of intravenous albumin on renal impairment and mortality in patients with cirrhosis and spontaneous bacterial peritonitis". The New England Journal of Medicine. 1999. 341(6):403-409.
PubMed • Full text • PDF

Clinical Question

Among patients with cirrhosis and spontaneous bacterial peritonitis treated with antibiotics, does intravenous albumin reduce incident renal impairment and mortality when compared to antibiotics alone?

Bottom Line

Among patients with cirrhosis and spontaneous bacterial peritonitis on IV antibiotics, IV albumin reduces the incidence of renal impairment and mortality when compared to antibiotics alone.

Major Points

Spontaneous bacterial peritonitis (SBP) is a common infection associated with cirrhosis that carries significant morbidity and mortality. While antibiotics mitigate some harm, SBP survivors experience a high incidence of renal dysfunction. The underlying mechanism of SBP's harm are thought to be related to a decrease in effective arterial blood volume from the infection. As individuals with cirrhosis have a low baseline oncotic pressure, renal perfusion is reduced, leading to renal insufficiency. Albumin increases oncotic pressure over the short term and its administration may improve renal perfusion, renal outcomes, and mortality among those with SBP.

Published in 1999 by Sort and colleagues, this study randomized 126 patients with cirrhosis and SBP to cefotaxime alone or cefotaxime and intravenous albumin. Resolution of infection was similar in both treatment groups (P=0.36). However, the cefotaxime+albumin group had a lower incidence of renal impairment (10% vs. 33%; P=0.002), in-hospital mortality (10% vs. 29%; P=0.01), and all-cause mortality at 3 months (22% vs. 41%; P=0.03).

A 2013 meta-analysis of 4 studies with 288 patients was performed to determine the effects of albumin infusion on renal impairment and mortality rates among patients with SBP. Albumin administration was associated with a pooled odds ratio of 0.21 (95% CI 0.11-0.42) for renal impairment and 0.34 (95% CI 0.19-0.60) for mortality, supporting the practice of albumin infusion among adults with cirrhosis and SBP. Included in this meta-analysis was a 2007 retrospective study by Sigal and colleagues published as a letter in in GUT. This letter reported that none of the 15 low-risk patients (bilirubin <4 mg d/L [<68.4 μmol/L] and creatinine <1 mg/dL [<88.4 μmol/L]) developed renal impairment or died. In contrast, the 1999 manuscript by Sort and colleagues, report a 7% (cefotaxime) and 0% (cefotaxime+albumin) incidence of renal impairment in this low-risk group. No participants in either arm with a BUN <30 mg/dL and bilirubin <4 mg/dL died in Sort's 1999 RCT. As such, low-risk individuals with SBP may not require albumin, though these conclusions are based on limited data.

Guidelines

European Association for the Study of the Liver (EASL) clinical practice guidelines (2010, adapted)

• Administration of albumin (1.5 g/kg at diagnosis and 1 g/kg on day 3) is associated with a reduction in hepatorenal syndrome and improvement in survival (Level A1)
  • It is unclear whether albumin is useful in the subgroup of patients with baselines serum bilirubin <68μmol/L (4 mg/dL) and serum creatinine <88μmol/L (1 mg/dL) (Level B2)
• Recommendation to treat all patients with SBP with broad spectrum antibiotics and intravenous albumin (Level A2)

American Association for the Study of Liver Diseases (AASLD) practice guidelines (2012, adapted)

• Patients with ascitic fluid PMN counts >250 cells/mm³ (0.25 x 10⁹/L) and clinical suspicion of SBP and either a serum creatinine >1 mg/dL, blood urea nitrogen >30 mg/dL, or total bilirubin >4 mg/dL should receive 1.5 g albumin per kg body weight within 6 hours of presentation and 1 g/kg on day 3. (Class IIa, Level B)
• If there is a suspicion of SBP and PMNs are >250 cells/mm³, check ascitic proteini, LDH, glucose, Gram's stain, CEA, and alk phos to help differentiate it from secondary peritonitis. Also perform a CT scan. (Class IIa, level B)

Design

• Multi-center, open-label (analyst-blinded), randomised controlled trial
• N=126
  • Cefotaxime (n=63)
  • Cefotaxime+albumin (n=63)
• Settling: 7 university hospitals in Spain
• Enrollment: 1995-1997
• Follow-up: 90 days
• Analysis: Intention-to-treat
• Primary outcome: Not clearly defined. The WJC editors are considering renal impairment, hospital mortality, and 90 day all-cause mortality as the primary outcomes.

Population

Inclusion Criteria

• Age 18-80 years
• PMN cell count in the ascitic fluid > 250 mm³ in the absence of findings suggestive of secondary peritonitis
• Serum creatinine no more than 3 mg/dL
• No antibiotic treatment in the week prior to SBP diagnosis, except norfloxacin prophylaxis

Exclusion Criteria

• Age >80 years
• Treatment with antibiotics at time of diagnosis
• Any medical of the following:
  • Other infections
  • Shock
  • Gastrointestinal bleeding
  • Ileus
  • Grade 3 or 4 hepatic encephalopathy (Conn and Lieberthal scale)
  • Cardiac failure
  • Evidence of nephropathy: Proteinuria, hematuria, or abnormal renal ultrasound
  • HIV infection
  • Any disease affecting short-term prognosis (e.g., advanced neoplasia)
• Potential causes of dehydration within 1 week before the diagnosis of peritonitis (diarrhea or an intense response to diuretics) or known CVP <4 mm Hg

Baseline Characteristics

Cefotaxime+albumin group

• Demographics: age 60 years, male sex 68%
• Cirrhosis details: Alcoholic 29%, with HCC 16%, with HE 21%
• Labs:
  • Serum: WBC 7.8k/mm³, bilirubin 4 mg/dL, albumin 2.7 g/dL, PT 55% of control
  • Ascitic: PMNs 5.2k/mm³
  • Isolated organisms: Any 51%, E. coli 31%, other GNR 11%, other bacteria 8%
• Child-Pugh score: 10
• Renal failure: 40%
• Diuretic treatment: 71%
• Prior norfloxacin prophylaxis: 10%

Interventions

• Enrollees were randomized to a group:
  • Cefotaxime - no albumin.
  • Cefotaxime+albumin - IV albumin at 1.5 g/kg in the first 6 hours then 1 g/kg on day 3.
• Intravenous cefotaxime was renally dosed (see details on page 404). If there was no response to cefotaxime, alternative antibiotics could be modified based upon results of sensitivities or local expertise.
• Patients with tense ascites were initially treated with total paracentesis and administration of albumin.
• Diuretic treatment or therapeutic paracentesis were not allowed until there was a resolution of the infection.
• All patients with ascites were treated with sodium restriction and diuretics and were treated with prophylactic norfloxacin after resolution of SBP.

Outcomes

Presented as cefotaxime vs. cefotaxime+albumin. Note: Primary outcomes were not explicitly defined by the investigators. Below are the primary outcomes as assumed by the WJC editors.

Primary Outcomes

Renal impairment

33% vs. 10% (P=0.002; NNT=4)

Renal failure at baseline was defined as BUN >30 mg/dL (11 mmol/L) or creatinine >1.5 mg/dL. Renal impairment was a non-reversible deterioration of renal function during hospitalization. For those without renal failure at baseline, this was diagnosed with an increase in BUN or creatinine >50% of the pretreatment level and with resultant levels of BUN >30 mg/dL or creatinine 1.5 mg/dL. For those with renal failure at baseline, this was diagnosed with an increase in BUN or creatinine >50% of the pretreatment level.

Hospital mortality

29% vs. 10% (P=0.01; NNT=5)

All-cause mortality at 3 months

41% vs. 22% (P=0.03; NNT=5)

Secondary Outcomes

Resolution of infection

94% vs. 98% (P=0.36)

Duration of antibiotic

6 vs. 5 days (P=0.48)

Paracentesis for ascites after resolution of infection

25% vs. 22% (P=0.83)

Hospital stay

13 vs. 14 days (P=0.48)

Additional Analyses

Changes in laboratory results

Plasma renin: Lower in the cefotaxime+albumin group at days 3, 6, and 9 (fig 1, pg 407)

BUN: Lower in the cefotaxime+albumin group at days 3, 6, and 9 (table 3, pg 406)

Creatinine: Lower in the cefotaxime+albumin group at days 6 and 9 (table 3, pg 406)

Sodium: Higher (closer to normal) in the cefotaxime+albumin group at days 3, 6, and 9 (table 3, pg 406)

Changes in MAP

No difference during follow-up between the groups (table 3, pg 406)

Subgroup Analysis

Incidence of renal impairment according to independent predictive value

Baseline serum bilirubin level ≥4 mg/dL: 48% vs. 12%

Baseline serum bilirubin level <4 mg/dL and serum creatinine level ≥1 mg/dL: 32% vs. 14%

Baseline serum bilirubin level <4 mg/dL and serum creatinine level <1 mg/dL: 7% vs. 0%.

Mortality rate related to renal function (renal impairment vs. no renal impairment)

In-hospital mortality rate: 78% vs. 3% (P<0.001)

All-cause mortality at 3 months: 89% vs. 16% (P<0.001)

Predictors of Outcome

Independent predictors of renal impairment

Baseline serum bilirubin level (P<0.001)

Baseline serum creatinine level (P=0.01)

Treatment with cefotaxime alone (OR 4.6; 95% CI 1.3-16.1; P=0.02)

Independent predictors of mortality

Baseline serum bilirubin level (P=0.01)

Blood urea nitrogen level (P=0.001)

Baseline prothrombin level (P=0.01)

Treatment with cefotaxime alone (OR 4.5; 95% CI 1.0-20.9; P=0.05)

Adverse Events

• No adverse effects of the albumin infusion.

Criticisms

• This study was not blinded, which may have introduced bias in treatment and ascertainment of outcomes.
• Renal impairment is related to the compound effects of sepsis and cirrhosis on renal function. Patients with cirrhosis are likely to develop renal failure if they have sepsis. Albumin infusion therefore may only serves the role of fluid resuscitation instead of providing a benefit of preventing any further deterioration in renal function.
• No details of fluid management and standard medical treatment for control group (cefotaxime alone) are provided. The details of vasoconstrictor therapy also not provided.

Funding

Fondo de Investigación Sanitaria (FIS 94/0956 and FIS 96/1723) and the Hospital Clínic.