Spontaneous bacterial peritonitis
Spontaneous bacterial peritonitis is an infection of abdominal fluid, called ascites, that does not come from an obvious place within the abdomen, such as a hole in the intestines or a collection of pus. It is seen with few exceptions in patients with ascites caused by chronic liver disease. Translocation of enteric bacteria across the gut wall or mesenteric lymphatics leads to seeding of the ascitic fluid, as may bacteremia from other sites. Approximately 20–30% of cirrhotic patients with ascites develop spontaneous peritonitis; however, the incidence is greater than 40% in patients with ascitic fluid total protein less than 1 g/dL, probably due to decreased ascitic fluid opsonic activity.
Virtually all cases of spontaneous bacterial peritonitis are caused by a monomicrobial infection. The most common pathogens are enteric gram-negative bacteria (E coli, Klebsiella pneumoniae) or gram-positive bacteria (Streptococcus pneumoniae, viridans streptococci, Enterococcus species). Anaerobic bacteria are not associated with spontaneous bacterial peritonitis.
The most important diagnostic test is abdominal paracentesis. Ascitic fluid should be sent for cell count with differential, and blood culture bottles should be inoculated at the bedside; Gram stain and reagent strips are insensitive.
In the proper clinical setting, an ascitic fluid PMN count of greater than 250 cells/mcL (neutrocytic ascites) is presumptive evidence of bacterial peritonitis. The percentage of PMNs is greater than 50–70% of the ascitic fluid white blood cells and commonly approximates 100%. Patients with neutrocytic ascites are presumed to be infected and should be started—regardless of symptoms— on antibiotics. Although 10–30% of patients with neutrocytic ascites have negative ascitic bacterial cultures (“culture-negative neutrocytic ascites”), it is presumed that these patients have bacterial peritonitis and should be treated empirically. Occasionally, a positive blood culture identifies the organism when ascitic fluid is sterile.
Spontaneous bacterial peritonitis must be distinguished from secondary bacterial peritonitis, in which ascitic fluid has become secondarily infected by an intra-abdominal infection. Even in the presence of perforation, clinical symptoms and signs of peritonitis may be lacking owing to the separation of the visceral and parietal peritoneum by the ascitic fluid. Causes of secondary bacterial peritonitis include appendicitis, diverticulitis, perforated peptic ulcer, and perforated gallbladder. Secondary bacterial infection accounts for 3% of cases of infected ascitic fluid.
Ascitic fluid total protein, LD, and glucose are useful in distinguishing spontaneous bacterial peritonitis from secondary infection. Up to two-thirds of patients with secondary bacterial peritonitis have at least two of the following: decreased glucose level (less than 50 mg/dL), an elevated LD level (greater than serum), and total protein greater than 1 g/dL. Ascitic neutrophil counts greater than 10,000/mcL also are suspicious; however, most patients with secondary peritonitis have neutrophil counts within the range of spontaneous peritonitis. The presence of multiple organisms on ascitic fluid Gram stain or culture is diagnostic of secondary peritonitis.
If secondary bacterial peritonitis is suspected, abdominal CT imaging of the upper and lower gastrointestinal tracts should be obtained to look for evidence of an intra-abdominal source of infection. If these studies are negative and secondary peritonitis still is suspected, repeat paracentesis should be performed after 48 hours of antibiotic therapy to confirm that the PMN count is decreasing. Secondary bacterial peritonitis should be suspected in patients in whom the PMN count is not below the pretreatment value at 48 hours.
Neutrocytic ascites may also be seen in some patients with peritoneal carcinomatosis, pancreatic ascites, or tuberculous ascites. In these circumstances, however, PMNs account for less than 50% of the ascitic white blood cells.
Up to 70% of patients who survive an episode of spontaneous bacterial peritonitis will have another episode within 1 year. Oral once-daily prophylactic therapy—with norfloxacin, 400 mg, ciprofloxacin, 250–500 mg, or trimethoprim-sulfamethoxazole, one double-strength tablet—has been shown to reduce the rate of recurrent infections to less than 20% and is recommended.
Prophylaxis should be considered also in patients who have not had prior bacterial peritonitis but are at increased risk of infection due to low-protein ascites (total ascitic protein less than 1 g/dL). Although improvement in survival in cirrhotic patients with ascites treated with prophylactic antibiotics has not been shown, decision analytic modeling suggests that in patients with prior bacterial peritonitis or low ascitic fluid protein, the use of prophylactic antibiotics is a cost-effective strategy.
Empiric therapy for spontaneous bacterial peritonitis should be initiated with a third-generation cephalosporin (such as cefotaxime, 2 g intravenously every 8–12 hours, or ceftriaxone, 1–2 g intravenously every 24 hours) or a combination beta-lactam/beta-lactamase agent (such as ampicillin/sulbactam, 2 g/1 g intravenously every 6 hours). Because of a high risk of nephrotoxicity in patients with chronic liver disease, aminoglycosides should not be used.
A repeat paracentesis is recommended after 48 hours of treatment in patients without clinical improvement. If the ascitic neutrophil count has not decreased by 25%, antibiotic coverage should be adjusted (guided by culture and sensitivity results, if available) and secondary causes of peritonitis excluded. Although the optimal duration of therapy is unknown, a course of 5–10 days is sufficient in most patients, or until the ascites fluid PMN count decreases to less than 250 cells/mcL.
Kidney injury develops in up to 40% of patients and is a major cause of death. Intravenous albumin increases effective arterial circulating volume and renal perfusion, decreasing the incidence of kidney injury and mortality. Intravenous albumin, 1.5 g/kg on day 1 and 1 g/kg on day 3, should be administered to patients at high risk for hepatorenal failure (ie, patients with baseline creatinine greater than 1 mg/dL, blood urea nitrogen [BUN] more than 30 mg/dL, or bilirubin more than 4 mg/dL).
Nonselective beta-blockers increase the risk of hepatorenal syndrome in patients with bacterial peritonitis and should be discontinued permanently due to their adverse impact on cardiac output and renal perfusion in advanced cirrhosis which is associated with decreased long-term survival.
Patients with suspected secondary bacterial peritonitis should be given broad-spectrum coverage for enteric aerobic and anaerobic flora with a third-generation cephalosporin and metronidazole pending identification and definitive (usually surgical) treatment of the cause.