Spontaneous bacterial peritonitis (SBP) is an acute bacterial infection of ascitic fluid. Generally, no source of the infecting agent is easily identifiable, but contamination of dialysate can cause the condition among those receiving peritoneal dialysis (PD).
Spontaneous bacterial peritonitis occurs in both children and adults and is a well-known and ominous complication in patients with cirrhosis.[1] Of patients with cirrhosis who have spontaneous bacterial peritonitis, 70% are Child-Pugh class C. In these patients, the development of spontaneous bacterial peritonitis is associated with a poor long-term prognosis. Patients with cirrhosis who develop spontaneous bacterial peritonitis in the face of septic shock are at high risk of mortality.
Once thought to occur only in those individuals with alcoholic cirrhosis, spontaneous bacterial peritonitis is now known to affect patients with cirrhosis from any cause. In addition, spontaneous bacterial peritonitis can occur as a complication of any disease state that produces the clinical syndrome of ascites, such as heart failure and Budd-Chiari syndrome. Children with nephrosis or systemic lupus erythematosus who have ascites have a high risk of developing spontaneous bacterial peritonitis.
The mechanism for bacterial inoculation of ascites has been the subject of much debate since Harold Conn first recognized the disorder in the 1960s. Enteric organisms have traditionally been isolated from more than 90% of infected ascites fluid in spontaneous bacterial peritonitis, suggesting that the GI tract is the source of bacterial contamination.
The preponderance of enteric organisms, in combination with the presence of endotoxin in ascitic fluid and blood, once favored the argument that spontaneous bacterial peritonitis was due to direct transmural migration of bacteria from an intestinal or hollow organ lumen, a phenomenon called bacterial translocation. However, experimental evidence suggests that direct transmural migration of microorganisms might not be the cause.
An alternative proposed mechanism for bacterial inoculation of ascites is hematogenous transmission in combination with an impaired immune system. Nonetheless, the exact mechanism of bacterial displacement from the GI tract into ascites fluid remains controversial.
A variety of factors contributes to peritoneal inflammation and bacterial growth in ascitic fluid. A key predisposing factor may be the intestinal bacterial overgrowth found in people with cirrhosis, mainly attributed to delayed intestinal transit time. Intestinal bacterial overgrowth, along with impaired phagocytic function, low serum and ascites complement levels, and decreased activity of the reticuloendothelial system, contributes to an increased number of microorganisms and decreased capacity to clear them from the bloodstream, resulting in their migration into and eventual proliferation within ascites fluid.
Interestingly, adults with spontaneous bacterial peritonitis typically have ascites, but most children with spontaneous bacterial peritonitis do not have ascites. The reason for and mechanism behind this is the source of ongoing investigation.
Traditionally, three fourths of spontaneous bacterial peritonitis infections have been caused by aerobic gram-negative organisms (50% of these being Escherichia coli). The remainder has been due to aerobic gram-positive organisms (19% streptococcal species). E coli is displayed in the image below.
View Image | Gram-negative Escherichia coli. |
However, some data suggest that the percentage of gram-positive infections may be increasing.[2, 3] One study cites a 34.2% incidence of streptococci, ranking in second position after Enterobacteriaceae.[3] Viridans group streptococci (VBS) accounted for 73.8% of these streptococcal isolates.
Anaerobic organisms are rare because of the high oxygen tension of ascitic fluid.
A single organism is noted in 92% of cases, and 8% of cases are polymicrobial.
Patients with cirrhosis who are in a decompensated state are at the highest risk of developing spontaneous bacterial peritonitis.[4] Bacterial translocation (viable microorganism passage from the intestinal lumen to mesenteric lymph nodes) is a key factor in the development of spontaneous bacterial peritonitis.[5] Low complement levels are associated with the development of spontaneous bacterial peritonitis. Patients at greatest risk for spontaneous bacterial peritonitis have decreased hepatic synthetic function with associated low total protein level or prolonged prothrombin time (PT).
Patients with low protein levels in ascitic fluid (< 1 g/dL) have a 10-fold higher risk of developing spontaneous bacterial peritonitis than those with a protein level greater than 1 g/dL.
A 2014 prospective study by Terg et al[6] demonstrated that proton pump inhibitor therapy (specifically evaluated at admission) was not associated with a higher risk of spontaneous bacterial peritonitis. This finding was contrary to findings by previously published studies (2012 review by Siple et al[7] and a 2013 study by Deshpande et al[8] , which had reported several case studies and cohorts of patients with cirrhosis and chronic liver disease who were on proton pump inhibitors (PPIs) for a prolonged duration who were at significantly increased risk for the development of spontaneous bacterial peritonitis. .
In patients with ascites, the frequency may be as high as 18%. This number has grown from 8% over the past 2 decades, most likely secondary to an increased awareness of spontaneous bacterial peritonitis and a lowered threshold to perform diagnostic paracentesis.
No race predilection is known for spontaneous bacterial peritonitis. In patients with ascites, both sexes are affected equally.
Although the etiology and incidence of hepatic failure differ between children and adults, in those individuals with ascites, the incidence of spontaneous bacterial peritonitis is roughly equal. Two peak ages for spontaneous bacterial peritonitis are characteristic in children: the first in the neonatal period and the second at age 5 years.
According to a 2015 study by Ge and Runyon, when the initial infection was spontaneous bacterial peritonitis, subsequent infections were more likely to be caused by drug-resistant organisms. The risk of subsequent infections was increased in older patients and in patients taking proton-pump-inhibitors (PPIs) or spontaneous bacterial peritonitis prophylaxis (ie, selective intestinal decontamination).[9]
The mortality rate in patients with spontaneous bacterial peritonitis ranges from 40-70% in adult patients with cirrhosis. Rates are lower in children with nephrosis. Patients with concurrent renal insufficiency have been shown to be at a higher risk of mortality from spontaneous bacterial peritonitis than those without concurrent renal insufficiency. Mortality from spontaneous bacterial peritonitis may be decreasing among all subgroups of patients because of advances in its diagnosis and treatment. In addition, nonselective beta-blockers increase the risk for hepatorenal syndrome and death in patients with cirrhosis and spontaneous bacterial peritonitis.[10]
Growing evidence supports early diagnostic paracentesis (defined as occurring within the first 11 hours of presentation). This, combined with early antibiotic treatment, leads to decreased ICU and hospital length of stay, in-hospital mortality, and 3-month mortality.[11, 12]
The mortality rate exceeds 80% in patients with cirrhosis who develop septic shock secondary to spontaneous bacterial peritonitis. In addition, each hour of delay in appropriate antimicrobial therapy increases the in-hospital mortality rate by 1.86 times.[13]
A broad range of signs and symptoms are seen in spontaneous bacterial peritonitis (SBP). A high index of suspicion must be maintained when caring for patients with ascites, particularly those with acute clinical deterioration. Completely asymptomatic cases have been reported in as many as 30% of patients.
Fever and chills occur in as many as 80% of patients. Abdominal pain or discomfort is found in as many as 70% of patients.
Other signs and symptoms may include the following:
Abdominal tenderness is found in more than 50% of patients with spontaneous bacterial peritonitis. Findings can range from mild tenderness to overt rebound and guarding. In some cases, the abdominal examination findings mimic an acute intra-abdominal catastrophe requiring emergency surgical evaluation. Physical examination may also disclose hypotension (5-14% of patients) or signs of hepatic failure such as jaundice and angiomata.
All patients suspected of having spontaneous bacterial peritonitis (SBP) must undergo peritoneal fluid analysis while in the emergency department. Diagnostic paracentesis should be performed in all patients who do not have an indwelling peritoneal catheter and are suspected of having spontaneous bacterial peritonitis. In peritoneal dialysis patients with a peritoneal catheter, fluid should be withdrawn with sterile technique. Ultrasonography may aid paracentesis if ascites is minimally detectable or questionable. Growing evidence supports early diagnostic paracentesis (defined as occurring within the first 11 hours of presentation). This, combined with early antibiotic treatment, leads to decreased ICU and hospital length of stay, in-hospital mortality, and 3-month mortality.[11, 12]
Blood and urine cultures should be obtained in all patients suspected of having spontaneous bacterial peritonitis. Blood culture results are positive for the offending agent in as many as 33% of patients with spontaneous bacterial peritonitis and may help guide antibiotic therapy. Urine culture may also prove useful, since asymptomatic bacteruria has been suggested to predispose to the development of spontaneous bacterial peritonitis.
If there is clinical suspicion of a perforated viscus, imaging should be strongly considered. Although plain radiographs (including abdominal flat plate, abdominal upright, and chest) may be obtained, CT scanning of the abdomen should be considered, as it is much more sensitive for a small perforation.
Peritoneal fluid analysis must be performed in any patient in whom spontaneous bacterial peritonitis (SBP) is considered. In patients undergoing peritoneal dialysis (PD), this can be accomplished by obtaining a sample of the dialysate. In patients without a peritoneal catheter, diagnostic paracentesis must be performed.
The examination of ascitic fluid for SBP has routinely involved sending the fluid for cell count, differential, and culture. It has been accepted that the results of aerobic and anaerobic bacterial cultures, used in conjunction with the cell count, are beneficial in guiding therapy for those with SBP.[14]
Recent data, though, suggest that ascitic fluid cultures have generally been shown to be of low yield with respect to altering management of patients with ascites. In addition, positive culture and sensitivity results obtained from emergency department testing have not been shown to result in appropriate adjustment of antibiotic therapy by inpatient physicians. The reasons for this may include inpatient physicians' distrust of the culture results and the difficulty in determining what constitutes a true pathogen in ascitic fluid cultures.[15]
The sensitivity of microbiologic studies has been reported to increase significantly with the direct inoculation of routine blood culture bottles at the bedside with 10 mL of ascitic fluid.
An ascitic fluid neutrophil count of more than 500 cells/µL is the single best predictor of spontaneous bacterial peritonitis, with a sensitivity of 86% and specificity of 98%. Lowering the ascitic fluid neutrophil count to more than 250 cells/µL results in an increased sensitivity of 93% but a lower specificity of 94%. (For simplicity, a threshold of 250 cells/µL is used for the remainder of this discussion.)
An exciting new development in the rapid diagnosis of spontaneous bacterial peritonitis is the proposed use of reagent strips that detect leukocyte esterase, which can be read at the bedside using a portable spectrophotometric device. In a pilot study that compared the reagent strips with the manual laboratory polymorphonuclear leukocyte count, the strips achieved a 100% sensitivity in diagnosis of spontaneous bacterial peritonitis.[16]
This diagnostic method holds promise in replacing manual cell counting, which is time-consuming and is often unavailable in many laboratories "after hours". Use of these reagent strips may result in a significant reduction of the time from paracentesis to presumptive diagnosis and antibiotic treatment of spontaneous bacterial peritonitis.
In a small cohort, the average time saved from dipstick to laboratory results ranged from 2.73 hours (dipstick to validated result from automated counter) to 3 hours (dipstick to validated manual cell count of ascitic fluid). Although promising, this diagnostic method has not been investigated in a large-scale study.
Other studies of ascitic fluid to be considered include the following:
An ascites lactate level of more than 25 mg/dL was found to be 100% sensitive and specific in predicting active spontaneous bacterial peritonitis in a retrospective analysis. In the same study, the combination of an ascites fluid pH below 7.35 and polymorphonuclear neutrophil count above 500 cells/µL was 100% sensitive and 96% specific for spontaneous bacterial peritonitis.
A 2012 study investigated using leukocyte reagent strips in the emergency department as a means of expediting the diagnosis of spontaneous bacterial peritonitis.[17] In this prospective study, 223 patients presenting with ascites and who had paracentesis performed in the emergency department had their peritoneal fluid sent for the usual diagnostic tests, but they also had the fluid dipped with both a Uri-Quick Clini 10 strip and Multistix 10SGA. Both had at least 90% positive predictive value and 94% negative predictive value for spontaneous bacterial peritonitis when compared with the criterion standard of peritoneal fluid Gram stain and culture—thus allowing a shorter interval between diagnosis and initiation of treatment.
Combining the results of the ascitic fluid polymorphonuclear neutrophil (PMN) count and the ascitic fluid culture yields the following subgroups:
Spontaneous bacterial peritonitis is noted when the PMN count is 250 cells/µL or higher, in conjunction with a positive bacterial culture result. As mentioned previously, one organism is usually identified on the culture in most cases. Obviously, these patients should receive antibiotic therapy.
Culture-negative neutrocytic ascites (probable spontaneous bacterial peritonitis) is noted when the ascitic fluid culture results are negative, but the PMN count is 250 cells/µL or higher. This may happen in as many as 50% of patients with SBP and may not actually represent a distinctly different disease entity. It may be the result of poor culturing techniques or late-stage resolving infection. Nonetheless, these patients should be treated just as aggressively as those with positive culture results.
Monomicrobial nonneutrocytic bacterascites exists when a positive culture result coexists with a PMN count of 250 cells/µL or fewer. Although this may often be the result of contamination of bacterial cultures, one study found that 38% of these patients subsequently develop spontaneous bacterial peritonitis.[18] Therefore, monomicrobial nonneutrocytic bacterascites may represent an early form of spontaneous bacterial peritonitis.
All study patients described that eventually developed spontaneous bacterial peritonitis were symptomatic.[18] For this reason, any patient suspected clinically of having spontaneous bacterial peritonitis in this setting must be treated.
The American Association for the Study of Liver Diseases (AASLD) has issued updated guidelines for adult patients with ascites due to cirrhosis.[19]
A 2009 guideline from the American Association for the Study of Liver Diseases recommends that adult cirrhotic patients with ascitic fluid polymorphonuclear neutrophil (PMN) counts of 250 cells/µL or greater in a community-acquired setting (in the absence of recent beta-lactam antibiotic exposure) should receive empiric antibiotic therapy (eg, an intravenous third-generation cephalosporin, preferably cefotaxime 2 g every 8 hours). Patients with cirrhosis who have PMN counts of 250 cells/µL or more in a nosocomial setting or patients who have recently received beta-lactam antibiotics should receive empiric antibiotic therapy based on local susceptibility testing of bacteria.[4, 19] As an alternative to intravenous cefotaxime, inpatients with cirrhosis can be treated with oral ofloxacin (400 mg twice per day), if none of the following contraindications are present[4] :
Patients with a peritoneal fluid PMN count greater than 500 cells/µL should universally be admitted and treated for spontaneous bacterial peritonitis, regardless of peritoneal fluid Gram stain result. Antibiotics should be initiated as soon as possible. The regimen can be chosen empirically, unless microbiologic studies further guide treatment. Evidence supports early antibiotic administration (combined with early diagnostic paracentesis) to help reduce ICU and hospital length of stay, in-hospital mortality, and 3-month mortality.[11, 12]
A retrospective study by Mazer et al. showed that patients who received ceftriaxone 2 grams daily (as opposed to 1 gram daily) had fewer ICU days and may have improved survival.[20]
For patients with a peritoneal fluid PMN count below 250 cells/µL, management depends upon the results of ascitic fluid cultures. All symptomatic patients should be admitted. Patients whose culture results are positive should be treated for spontaneous bacterial peritonitis. A select subset of patients who are completely asymptomatic yet have positive culture results may be managed without treatment but must undergo a follow-up paracentesis within 24-48 hours.
All symptomatic patients with a peritoneal fluid PMN count of 250-500 cells/µL should be admitted and treated for spontaneous bacterial peritonitis.
Probiotic therapy in conjunction with antimicrobial treatment does not improve efficacy in the treatment of spontaneous bacterial peritonitis, as was found in a double-blind, placebo-controlled, randomized-controlled trial.[21] In this study, Pande et al found over a 28-month period that 110 patients who were randomized to either norfloxacin 400 mg with probiotics or placebo did not have improved efficacy in primary or secondary prophylaxis or in reducing mortality in cirrhotic patients with ascites.
A 2013 meta-analysis by Salerno et al showed that albumin infusion in patients with spontaneous bacterial peritonitis reduced the incidence of renal impairment and mortality.[22]
For spontaneous bacterial peritonitis (SBP), a 10- to 14-day course of antibiotics is recommended. Although not required, a repeat peritoneal fluid analysis is recommended to verify declining PMN counts and sterilization of ascitic fluid.
If improvement in ascitic fluid or clinical condition does not occur within 48 hours, further evaluation is required to rule out bowel perforation or intra-abdominal abscess. Evaluation may include a combination of radiography, CT scanning, intraluminal contrast studies, or surgical exploration.
Outpatient prophylaxis, although not recommended routinely, has been shown to prevent spontaneous bacterial peritonitis in the following high-risk groups:
Suggested outpatient prophylactic regimens include the following:
A 2012 guideline from the American Association for the Study of Liver Diseases recommends that adult cirrhotic patients who have survived an episode of SBP should receive long-term prophylaxis with norfloxacin or trimethoprim-sulfamethoxazole.[19] However, long-term prophylaxis with norfloxacin is a risk factor for infection with a multi-resistant organism.[23]
However, mounting evidence shows that use of PPIs and selective-intestinal decontamination leads to development of increased pathogenic and drug-resistant flora.[24] In addition, despite evidence suggesting that primary prophylaxis of spontaneous bacterial peritonitis delays the development of hepatorenal syndrome and improves survival, a prospective study showed that patients with spontaneous bacterial peritonitis on long-term norfloxacin subsequently developed quinolone-resistant spontaneous bacterial peritonitis.[25]
The goals of pharmacotherapy in patients with spontaneous bacterial peritonitis (SBP) are to reduce morbidity and prevent complications. Antibiotics are initially chosen empirically, as these patients may die from overwhelming infection if treatment is delayed until culture results become available.[4]
Probiotic therapy in conjunction with antimicrobial treatment does not improve efficacy in the treatment of spontaneous bacterial peritonitis, as was found in a double-blind, placebo-controlled, randomized-controlled trial.[21] In this study, Pande et al found over a 28-month period that 110 patients who were randomized to either norfloxacin 400 mg with probiotics or placebo did not have improved efficacy in primary or secondary prophylaxis or in reducing mortality in cirrhotic patients with ascites.
Clinical Context: A third-generation cephalosporin with broad gram-negative spectrum, cefotaxime has lower efficacy against gram-positive organisms and higher efficacy against resistant organisms. Thus, it provides excellent empiric coverage of SBP. By binding to 1 or more penicillin-binding proteins, cefotaxime arrests bacterial cell wall synthesis and inhibits bacterial growth.
Multiresistant bacteria (including extended-spectrum beta-lactamases (ESBL)–producing Enterobacteriaceae) are increasingly being isolated in nosocomial (and, to a lesser extent, healthcare-acquired) infections among patients with cirrhosis, which renders third-generation cephalosporins clinically ineffective. This calls for new antimicrobial strategies with a tailored approach according to local epidemiological patterns in order to best determine empirical treatment.[21]
Risk factors for multiresistant organisms include nosocomial origin of infection, long-term norfloxacin prophylaxis, recent infection with multiresistant bacteria, and recent use of beta-lactam antibiotics. Infections with resistant organisms are associated with higher mortality. It is also prudent to limit prophylactic antibiotics to patients with well-defined criteria for spontaneous bacterial peritonitis prophylaxis, to limit duration of antibiotic treatment for infections, and to narrow the spectrum of coverage once susceptibility testing results are available.[22]
Clinical Context: Gentamicin is an aminoglycoside antibiotic effective against Pseudomonas aeruginosa; E coli; and Proteus, Klebsiella, and Staphylococcus species. Dosing regimens are numerous; adjust dose based on creatinine clearance (CrCl) and changes in volume of distribution. Gentamicin may be given IV or IM.
Clinical Context: Ampicillin interferes with bacterial cell wall synthesis during active multiplication, causing bactericidal activity against susceptible organisms.
Clinical Context: Norfloxacin is used for prophylaxis in the outpatient setting (400 mg/d). It is a fluoroquinolone with activity against pseudomonads, streptococci, MRSA, S epidermidis, and most gram-negative organisms, but it has no activity against anaerobes. It inhibits bacterial DNA synthesis and, consequently, growth.
Clinical Context: Ciprofloxacin is used for prophylaxis in the outpatient setting (750 mg weekly). It is a fluoroquinolone that inhibits bacterial DNA synthesis and, consequently, growth, by inhibiting DNA gyrase and topoisomerases, which are required for replication, transcription, and translation of genetic material. Quinolones have broad activity against gram-positive and gram-negative aerobic organisms. This agent has no activity against anaerobes.
Clinical Context: This agent is used as prophylaxis in the outpatient setting (5 doses of double-strength trimethoprim-sulfamethoxazole per week (Monday through Friday). It inhibits bacterial growth by inhibiting the synthesis of dihydrofolic acid.
Traditionally, a combination of an aminoglycoside and ampicillin was used to treat spontaneous bacterial peritonitis (SBP). This regimen affords excellent empiric coverage of more than 90% of cases of spontaneous bacterial peritonitis caused by gram-negative aerobes or gram-positive cocci.
More recently, the third-generation cephalosporin cefotaxime has been demonstrated to be as efficacious as the ampicillin/aminoglycoside combination, and it does not carry the increased risk of nephrotoxicity in cirrhotic patients. Cefotaxime does not cover enterococci (up to 5% of cases).