End-Stage Liver Disease
Elijah J. Mun, MD
Tyler J. Albert, MD
You are called by the ED to admit a 55-year-old man with history of cirrhosis secondary to chronic alcoholism complicated by hepatic encephalopathy (HE) and ascites who is brought in by his caregiver with increasing confusion, lethargy, and abdominal pain over 5 days. His caregiver reports the patient has completely abstained from alcohol for the past 4 months and is adherent to lactulose with daily bowel movements.
You perform a bedside ultrasound exam in the ED, which demonstrates the following (Figure 21.1).
You perform abdominal paracentesis and diagnose the patient with spontaneous bacterial peritonitis (SBP). Other laboratory and radiographic evaluations do not reveal any other potential causes of his symptoms. You initiate empiric IV antibiotics while awaiting ascitic fluid culture results and consider the role of albumin infusion.
What is the role of albumin infusion in the treatment of SBP?
In cirrhotic patients with SBP and Cr > 1 mg/dL, BUN > 30 mg/dL, or total bilirubin > 4 mg/dL, albumin infusion is associated with lower risk of renal failure and death.
This question was studied in a multicenter randomized controlled trial of 126 adult patients1 diagnosed with SBP (based on ascitic fluid Polymorphonuclear cell count >250 cells/mm3 in the absence of findings concerning for secondary bacterial peritonitis). Patients were randomized to receive either antibiotics alone (cefotaxime) or antibiotics (cefotaxime) plus IV albumin. Antibiotics were dosed based on patients’ serum Cr and albumin was given twice—at the time of diagnosis (1.5 g/kg) and again 48 hours later on day 3 (1 g/kg). Exclusion criteria included antibiotic exposure within 1 week of SBP diagnosis, other infection, shock, gastrointestinal bleeding, ileus, grade 3 to 4 HE, or cardiac failure. Primary outcomes were renal impairment (defined as nonreversible deterioration of renal function during hospitalization) and mortality (in-hospital and 3-month).
Renal impairment was lower among patients in the antibiotics plus albumin group (10% vs. 33%; P = .002). Patients receiving antibiotics and albumin also had lower in-hospital (10% vs. 29%; P = .01) and 3-month (22% vs. 41%; P = .03) mortality. Baseline total bilirubin and Cr levels independently predicted renal impairment and in-hospital mortality, and among patients with total bilirubin <4 g/dL and Cr < 1 g/dL, renal impairment was less frequent in the antibiotics plus albumin group (0% vs. 7%; P-value not reported). Caveats include absent data about adequateness of fluid resuscitation or vasoconstrictor therapy among control group patients.
Other work has built upon these findings by demonstrating the acceptability of therapeutic protocols that utilize a restrictive strategy of administering albumin only to patients with high risk of renal failure (Cr>1, BUN >30, or total bilirubin >4)2—the strategy recommended in the 2014 AASLD practice guidelines (Class IIa, Level B recommendation).3
Based on a serum Cr of 2.0 mg/dL, you treat the patient with IV albumin in addition to antibiotics. The patient gradually improves over the next few days, and you are asked by his caregiver if anything can be done to prevent such infections in the future.
In which situations should prophylactic antibiotics be considered for the prevention of SBP?
Prophylactic antibiotics should be initiated in cirrhotic patients with either gastrointestinal bleeding, previous SBP, or ascites if ascitic fluid protein <1 g/dL or ascitic fluid protein <1.5 g/dL and Child class C disease.
A meta-analysis evaluated five randomized controlled trials investigating the role of prophylactic antibiotics in cirrhotic patients with gastrointestinal bleeding.4 Trials without control groups or those comparing two or more antibiotics were excluded. Four main efficacy outcomes were identified: proportion of patients free of infection, bacteremia and/or SBP, SBP, and death. The five trials accounted for a total of 534 patients, 264 of whom were treated with antibiotics for 4 to 10 days and 270 who did not receive antibiotics. Antibiotic prophylaxis was associated with a greater proportion of patients free of infection (32% mean rate difference, 95% CI 22%-42%; P < .001), bacteremia and/or SBP (19% mean rate difference, 95% CI 11%-26%; P < .001), SBP (7% mean rate difference, 95% CI 2.1%-12.6%; P = .006), and death (9.1% mean rate difference, 95% CI 2.9%-15.3%; P = .004).
Subsequent trials have also shown prophylactic antibiotics to be effective in the prevention of SBP in cirrhotic patients with previous SBP (SBP recurrence rate 35% among controls vs. 12% in those treated with norfloxacin; P = .014)5 and ascites with ascitic fluid protein <1 g/dL or ascitic fluid protein <1.5 g/dL and Child-Pugh score >9, i.e., Child class C disease (comparing those treated with norfloxacin vs. controls, 1-year incidence of SBP 7% vs. 61%, P < .001; hepatorenal syndrome [HRS] 28% vs. 41%, P = .02; 3-month survival 94% vs. 62%, P = .003).6
Consequently, AASLD guidelines recommend the administration of prophylactic antibiotics for any cirrhotic patients with gastrointestinal bleeding, previous SBP, and ascites if ascitic fluid protein <1 g/dL, or ascitic fluid protein <1.5 g/dL and Child class C disease (Class I, Level A).3