Trimethoprim-Sulfamethoxazole for the Prevention of Spontaneous Bacterial Peritonitis in Cirrhosis
Nina Singh,T. Gayowski,V. Yu,M. Wagener
DOI: https://doi.org/10.7326/0003-4819-122-8-199504150-00007
IF: 39.2
1995-04-15
Annals of Internal Medicine
Abstract:Spontaneous bacterial peritonitis is a serious complication that occurs in 8% to 25% of patients with cirrhosis and ascites [1]. The overall mortality rate is 30% to 50%. Antibiotic prophylaxis for the prevention of spontaneous bacterial peritonitis has been attempted using selective intestinal decontamination with nonabsorbable antibiotics [2] and partially absorbable antibiotics (for example, norfloxacin) [3-6]. Use of these antibiotics was not only ineffective against gram-positive (streptococcal) infections but was associated with the selection of highly resistant gram-negative isolates and with the emergence of streptococci as pathogens, not only in patients with cirrhosis but also in other clinical situations (for example, myelosuppressed patients, patients receiving mechanical ventilation, and transplant recipients) [3, 7-10]. Trimethoprim-sulfamethoxazole is a systemically absorbed antibiotic with in vitro activity against enteric gram-negative bacteria and streptococci, including pneumococci. Trimethoprim-sulfamethoxazole has been extensively used as a prophylactic agent in diverse patient populations (for example, patients with granulocytopenia [7], transplant recipients [11], and patients with human immunodeficiency virus infection [12]). We did a randomized trial to assess the efficacy and safety of trimethoprim-sulfamethoxazole for the prevention of spontaneous bacterial peritonitis in patients with cirrhosis and ascites, using the Bactec culture method to detect spontaneous bacterial peritonitis. Methods All patients with documented cirrhosis and ascites were considered eligible. Patients were excluded if they 1) were allergic to sulfonamides; 2) had renal failure with a creatinine clearance of less than 15 mL/min [use of trimethoprim-sulfamethoxazole is not recommended for patients with a creatinine clearance of <15 mL/min unless the patient is receiving dialysis]; or 3) had active spontaneous bacterial peritonitis or extraperitoneal infection at the time of enrollment. The patients were randomly assigned to receive either trimethoprim-sulfamethoxazole (one double-strength tablet five times a week, Monday through Friday) or no prophylaxis. Entry was stratified by serum bilirubin level, renal function, and ascitic fluid protein level. Patients at risk for spontaneous bacterial peritonitis or mortality have ascitic fluid protein levels less than 1 g/dL [13], serum bilirubin levels greater than 51 mol/L (>3 mg/dL) [14], and renal dysfunction (serum creatinine levels >177 mol/L [72 mg/dL]) [3, 15]. These criteria were selected so that high-risk patients would not be disproportionately allocated to one group. Paracentesis was done in both groups when ascitic fluid infection was suspected (that is, for the evaluation of fever, abdominal pain or tenderness, leukocytosis, or worsening encephalopathy). The primary end points were development of spontaneous bacteremia or spontaneous bacterial peritonitis, which was defined as either an ascitic fluid polymorphonuclear cell count of 250/mm3 or more and a positive ascitic fluid culture (culture-positive neutrocytic ascites) or as an ascitic fluid polymorphonuclear cell count of 250/mm3 or more and a negative ascitic fluid culture (culture-negative neutrocytic ascites). Ascitic fluid was cultured using the Bactec culture system (Becton Dickinson Diagnostic Instrument System, Sparks, Maryland) as previously described [16]. Extraperitoneal infections (for example, pneumonia and mortality) were also assessed. Clinical and laboratory data were entered into a database (Prophet, BBN Systems and Technologies, Cambridge, Massachusetts). Continuous variables (age and bilirubin, albumin, and cholesterol levels) were compared using the Student t-test or the Mann-Whitney test. Categorical variables (such as underlying liver disease, comorbid illnesses, and infections) were compared using the chi-square or the Fisher exact tests. Results Patient Characteristics Sixty-seven consecutive patients with cirrhosis and ascites were evaluated, but 7 were excluded on the basis of preset criteria; thus, 60 patients were ultimately enrolled. Of the 60 patients, 30 were randomly assigned to the group receiving trimethoprim-sulfamethoxazole and 30 were assigned to the group not receiving prophylaxis. The underlying causes of liver disease in the 30 patients in the trimethoprim-sulfamethoxazole group were hepatitis C virus (HCV) (n = 11), alcohol (n = 8), alcohol and HCV (n = 4), and hepatitis B virus (n = 5); cause was indeterminate in 2 patients. Liver diseases in the 30 patients not receiving prophylaxis were caused by HCV (n = 12), alcohol (n = 11), alcohol and HCV (n = 1), hepatitis B virus (n = 1), -1-antitrypsin deficiency (n = 2), hepatocellular carcinoma (n = 1), and autoimmune hepatitis (n = 1); cause was indeterminate in 1 patient. The two study groups were similar at baseline for all clinical and laboratory data outlined in Table 1, except that more patients in the trimethoprim-sulfamethoxazole group had anasarca (P = 0.09) and more had a history of encephalopathy before enrollment (P = 0.04) (Table 1). The median duration of follow-up was 90 days (range, 7 to 682 days). Table 1. Clinical and Laboratory Characteristics of Patients at Baseline* Incidence and Type of Infections Overall, infectious complications developed in 9 patients (30%) not receiving prophylaxis and in 1 patient (3%) receiving trimethoprim-sulfamethoxazole (P = 0.012) (Table 2). Spontaneous bacterial peritonitis or spontaneous bacteremia developed in 8 patients (27%) receiving no prophylaxis and in 1 patient (3%) receiving trimethoprim-sulfamethoxazole (P = 0.025). Table 2. Incidence and Types of Infections The eight episodes of spontaneous bacterial peritonitis or spontaneous bacteremia in the group not receiving prophylaxis comprised Escherichia coli peritonitis with bacteremia (two episodes), spontaneous Klebsiella pneumoniae bacteremia (one episode), enterococcal peritonitis (two episodes), and culture-negative neutrocytic ascites (three episodes). These infections developed a mean of 37 days (range, 3 to 114 days) after enrollment. Only one patient in the trimethoprim-sulfamethoxazole group developed an episode of culture-negative neutrocytic ascites 111 days after enrollment. Only one patient in the group not receiving prophylaxis developed an extraperitoneal infection (Haemophilus influenzae pneumonia). Adverse Effects None of the patients developed adverse effects; hematologic toxicity caused by trimethoprim-sulfamethoxazole was notably absent. Diarrhea possibly related to trimethoprim-sulfamethoxazole was seen in one patient; Clostridium difficile colitis was not evident, and the diarrhea resolved despite continuing therapy with trimethoprim-sulfamethoxazole. Mortality Two patients (7%) receiving trimethoprim-sulfamethoxazole and 6 patients (20%) not receiving prophylaxis died (P = 0.15). The median time to death after enrollment was 45 days in the group not receiving prophylaxis (range, 7 to 509 days). The cause of death in these patients was gastrointestinal bleeding (two patients), hepatic insufficiency (three patients), and hepatic insufficiency with respiratory failure (one patient). Although none of the deaths could be directly attributed to spontaneous bacterial peritonitis, two of six patients died during the hospitalization in which spontaneous bacterial peritonitis was diagnosed. Two patients in the trimethoprim-sulfamethoxazole group died 85 and 167 days after enrollment because of gastrointestinal bleeding and hepatic insufficiency, respectively. Discussion Several studies [3-5] have shown that norfloxacin reduces the incidence of spontaneous bacterial peritonitis in patients with cirrhosis, and this antibiotic is now commonly used as prophylaxis in these patients [3]. However, prolonged use of norfloxacin and quinolones has been associated with the selection of gram-positive isolates and with the eventual emergence of methicillin-resistant Staphylococcus aureus and enterococci as pathogens. In fact, it has been proposed that long-term prophylaxis with norfloxacin would probably shift the spectrum of microbial agents causing peritonitis from gram-negative aerobic bacilli to gram-positive organisms [17]. Dupeyron and colleagues [6] showed that the use of norfloxacin to prevent spontaneous bacterial peritonitis in cirrhosis led to colonization with gram-negative bacilli that were resistant to fluoroquinolones and with staphylococci that were resistant to methicillin in 52% of patients a median of 25 days after therapy. All isolates of Enterobacteriaceae after therapy were also resistant to aminoglycosides and -lactams [6]. Another study [18] also reported that quinolones can promote cross-resistance to unrelated classes of antibiotics. Although acquisition of resistance to trimethoprim-sulfamethoxazole has also been reported, the selection of streptococci appears to be uniquely more prevalent with quinolones. In a study [7] of patients with granulocytopenia, bacteremia caused by gram-positive isolates occurred significantly more often in patients receiving norfloxacin (29%) than in those receiving trimethoprim-sulfamethoxazole (6%). Therefore, we selected trimethoprim-sulfamethoxazole as a potentially useful agent for the prophylaxis of spontaneous bacterial peritonitis. In this randomized trial, trimethoprim-sulfamethoxazole was compared with no prophylaxis for the prevention of spontaneous bacterial peritonitis in 60 consecutive patients with documented cirrhosis and ascites. The incidence of spontaneous bacterial peritonitis or spontaneous bacteremia (3% compared with 27%) and overall infection (3% compared with 30%) was lower in patients receiving trimethoprim-sulfamethoxazole compared with patients not receiving prophylaxis (P < 0.05 for both values). Emergence of enterococci as pathogens did not occur in any patient receiving trimethoprim-sulfamethoxazole, whereas enterococci cau