Microbiology and Antibiotic Susceptibility of Organisms in Bile Cultures from Patients with and without Cholangitis at an Asian Academic Medical Center

Patients and Methods

Results

One hundred twenty-five patients with microbiologic and ISIS data were included. A detailed breakdown of the characteristics of the study population is given in Table 1. The distribution of patients in terms of gender and age was not statistically different between the groups. Choledocholithiasis was the most common underlying biliary disease in patients with and without cholangitis but was significantly more likely to be associated with cholangitis (p < 0.05).

Cultures were positive in 98 of 125 bile specimens (78%). Patients with a clinical diagnosis of cholangitis were significantly more likely to have positive bile cultures (p = 0.012). However, the two groups did not differ significantly in terms of bacterial isolates. The most frequent isolates (>5%) for both groups were Enterobacteriaceae (Table 2).

Table 3 shows the antimicrobial susceptibility of the organisms isolated most commonly from this study population. For both groups, the most commonly isolated gram-negative bacilli were highly sensitive to the empiric antimicrobial therapy recommended in the 2010 Surgical Infection Society/Infectious Diseases Society of America (SIS/IDSA) Guidelines [13]. Escherichia coli remained highly susceptible to amikacin, aztreonam, cefepime, ceftriaxone, gentamicin, imipenem-cilastatin meropenem, and piperacillin-tazobactam. Klebsiella pneumoniae showed high susceptibility to amikacin, aztreonam, cefepime, ceftriaxone, gentamicin, imipinem, meropenem, and piperacillin-tazobactam. Pseudomonas aeruginosa maintained high susceptibility to aztreonam, cefepime, gentamicin, imipenem-cilastatin meropenem, and piperacillin-tazobactam. Enterobacter cloacae exhibited high susceptibility to imipenem-cilastatin and meropenem only.

Discussion

In the present study, gram-negative bacteria (Enterobacteriaceae) were the most common bile culture isolates for patients both with and without cholangitis. Among them, E. coli was the predominant bacterium isolated from both groups (35% vs. 38%). Susceptibility rates to extended-spectrum antibiotics, especially the anti-pseudomonal carbapenems (100% for all Enterobacteriaceae, even Enterobacter cloacae), remained high for both groups of patients. There are indications that extended-spectrum beta-lactamase (ESBL)-producing gram-negative bacteria were represented among the pathogens in this sample of Filipino patients with cholangitic and non-cholangitic biliary infections.

Our main findings are consistent with the results of the Study for Monitoring Antimicrobial Resistance Trends (SMART), which reported culture and susceptibility data from intra-abdominal infections in six Asian countries, including the Philippines. Those investigators reported that Enterobacteriaceae constituted the large majority (82%) of aerobic and facultative gram-negative bacteria isolated. As in our study, SMART cited Escherichia coli (43%) as the species most commonly isolated. Likewise, the collective susceptibility rates of E. coli reported in SMART showed similarities to the results of this study, where the most active antibiotics were meropenem (99.6% and 100%, respectively), imipenem-cilastatin (99.2% and 100%), and amikacin (92.8% and 90%) [14,15].

The present study's findings also correspond to the bile culture isolates in a study involving cholangitis patients done by Boey and Way, where E. coli (34%), Klebsiella spp. (21%), and Pseudomonas spp. (13%) were the most common isolates [16]. Saik et al. reported similar findings, where E. coli (30%), Klebsiella spp. (23%), Streptococcus spp. (12%), Aerobacter spp. (9%), and Pseudomonas spp. (7%) were isolated from cholangitis patients [17].

The recommended empiric antimicrobial therapy for cholangitis has changed over the years as a result of the development of drug resistance and the increased incidence of anaerobic infection in a population with more co-morbidities and risk factors for multidrug-resistant infection. A study was done in Southeast Asia by Sheen-Chen et al. [8] documenting the antibiotic susceptibility of isolates from patients with confirmed hepatolithiasis in cholangitis. The gram-negative bacteria commonly isolated often were not sensitive to ampicillin (6–58%), ticarcillin (14–81%), or chloramphenicol (21–90%). For gram-positive bacteria, low sensitivity rates were documented for cephalothin (38–44%), tetracycline (44–55%), cefamandole (37–90%), and erythromycin (21–48%) [8]. In the regional data analysis from SMART, the lowest susceptibility rates of Enterobacteriaceae in Asia were seen for tobramycin (69.3%), levofloxacin (69.1%), ceftriaxone (68.5%), ceftazidime (67.7%), and ciprofloxacin (64.3%) [14]. These studies have influenced the empiric antimicrobial therapy recommendations in the Tokyo Guidelines.

The Tokyo Guidelines, which formulate clinical guidance on diagnosis, severity assessment, and treatment of biliary infection, were developed by an expert panel composed of gastroenterologists, surgeons with expertise in biliary tract surgery, critical care medicine specialists, epidemiologists, and experts in laboratory medicine. The draft was finalized through discussions among the Working Group and the members of the scientific societies relevant to clinical practice in acute biliary infections. Global consensus on the management of acute biliary infection was established during the 2006 Tokyo meeting, which was attended by an international consensus panel. The recommendations were based on the best available evidence and expert consensus. The guidelines provide uniform diagnostic criteria and a flowchart for the appropriate treatment of acute biliary infection in accordance with its severity [1,2].

Comparing the 2010 SIS/IDSA guidelines with the Tokyo Guidelines, both emphasize that the selection of antimicrobial drugs should be in accordance with disease severity. In the 2010 SIS/IDSA guidelines for intra-abdominal infections, the choice of empiric antimicrobial agent is based on the severity of the infection, which was defined as a composite of patient age, physiologic derangements, and co-morbidities; whereas in the Tokyo Guidelines, the selection of empiric antimicrobial agent is dependent on the severity of cholangitis [9,13]. Although not reflected in the 2010 SIS/IDSA Guidelines because of lack of supporting evidence, the consensus panel of Tokyo Guidelines cited the importance of biliary penetration in selecting empiric antimicrobial agents in acute cholangitis [9].

With regard to empiric antimicrobial treatment for biliary infections, the guidelines differ in their recommended severity-based antibiotic regimens. In the 2010 SIS/IDSA Guidelines, the recommended treatment for community-acquired acute cholecystitis of mild-to-moderate severity is cephalosporin-based monotherapy such as cefazolin, cefuroxime, or ceftriaxone. For both community-acquired acute severe cholecystitis and acute cholangitis of any severity following bilio-enteric anastomosis, the recommended single-agent regimens are imipenem-cilastatin, meropenem, doripenem, piperacillin-tazobactam, whereas the recommended combination regimens are ciprofloxacin, levofloxacin, or cefepime, each in combination with metronidazole. The same guidelines also distinguish cases of health care-associated biliary infection, in which vancomycin is recommended to be added to the same regimen prescribed for severe conditions until definitive culture results are available [13].

On the other hand, the Tokyo Guidelines suggest, for mild (grade I) cases of acute cholangitis, the use of a beta-lactam/beta-lactamase inhibitor such as piperacillin-tazobactam or ampicillin-sulbactam as monotherapy aside from the usual selections of cephalosporin-based monotherapy such as cefazolin, cefmetazole, or cefotiam. With moderate (grade II) and severe (grade III) disease, the recommended drugs of first choice are broad-spectrum beta-lactam/beta-lactamase inhibitors such as piperacillin-tazobactam and ampicillin-sulbactam or third- or fourth-generation cephalosporins with wide antimicrobial spectra such as cefoperazone-sulbactam, ceftriaxone, ceftazidime, and cefepime. To cover any anaerobes, the addition of metronidazole is recommended. Depending on the susceptibility patterns, the second-line drugs of choice are fluoroquinolones (ciprofloxacin, levofloxacin, and pazufloxacin) and carbapenems (meropenem, imipenem-cilastatin, and doripenem) [9].

In contrast to the Tokyo Guidelines, the 2010 SIS/IDSA Guidelines do not recommend the use of ampicillin-sulbactam because of the documented high rates of resistance to this agent among community-acquired E. coli in the same way that quinolones are not recommended unless local antibiograms indicate > 90% susceptibility to these antibiotics [13]. With regard to monobactam-based therapy, the 2010 SIS/IDSA Guidelines recommended aztreonam plus metronidazole as an alternative regimen in cases of high-risk community-acquired infection in adults, whereas the Tokyo Guidelines consider it one of the first options for moderate-to-severe acute cholangitis [9,13].

This being a retrospective and descriptive study, doubts can be raised about the extent of the result's applicability to other settings. This study was not intended to be the final but rather the initial work that will document the antimicrobial susceptibilities of biliary pathogens in the Philippines. At this time, it is the best evidence available to inform and guide empiric antimicrobial choices for biliary infections. It is hoped that subsequent work will be pursued for this and other surgical infections.

Because the study relied on historical records, data collection was limited by the availability of medical charts. In the case of patients' medical records, 20% of patients with bile cultures did not have an ISIS record, thus prompting their exclusion from the study. The investigation of the microbiology of bile cultures is constrained by institutional technical limitations: They are not routinely performed except in clinical trials because of the limited availability of culture media. However, anaerobes are usually covered by the anti-anaerobic coverage utilized empirically in our institution. Second, antimicrobial susceptibility of organisms, being a dynamic parameter, is influenced by the limited number of isolates tested for inhibition by specific antibiotics. This is secondary to the inconsistent composition of the tested antibiotic panel used for all the isolates, as the antibiotic discs are subject to availability. Third, the researchers relied only on the antimicrobial sensitivity patterns in identifying the possible presence of ESBL strains. A more definitive phenotypic identification of ESBL strains could have been included. Finally, the small study population seen over the three-year period may not be truly representative of the general population with biliary infections/cholangitis. Only surgical patients were included in this study; patients undergoing medical management (by a non-surgical team) of their biliary infection were not included. The relatively small number of patients with both microbiologic and clinical data limits the capacity to extrapolate the findings to the general population of patients. In this study, 62% of patients were found to have cholangitis, and this may skew the results to be more representative of severe biliary infections. Further studies designed prospectively should be done to collect more consistent and thorough data.

The higher rates of susceptibility to ceftriaxone and ceftazidime observed in the present study compared with the SMART data may be accounted for by the larger number of ESBL-producing isolates found in the Asian region as a whole; Philippine data do not contribute a large proportion of the aggregate data reported in the SMART study (13% of the total isolates came from the Philippines). It also was reported that the ESBL isolates from the Philippines accounted for 1% of the total ESBL isolates detected in the Asian countries included in the surveillance study [14].

In a number of previous studies, aminoglycosides, alone or as part of a combination regimen, were deemed effective for biliary infection. One study showed that amikacin, gentamicin, cefotaxime, ceftazidime, and a cefoperazone-sulbactam combination had good activity against E. coli and K. pneumoniae [18]. In another study among patients with gallstone diseases, the suggested empirical therapy was ampicilllin in combination with sulbactam and an aminoglycoside [19]. Another study of patients with cholangitis and choledocholithiasis showed that the bacterial isolates were sensitive to a combination of ampicillin and gentamicin [20]. However, the 2010 SIS/IDSA Guidelines do not recommend the routine use of an aminoglycoside or another second agent against gram-negative facultative and aerobic bacilli unless culture-proved resistant organisms necessitate such therapy. As shown in the susceptibility profile of the present study, aminoglycosides remain highly effective against the target pathogens. However, their narrow therapeutic range, need for monitoring of serum concentrations, and associated side effects of ototoxicity and nephrotoxicity and the availability of safer alternatives make aminoglycosides less than ideal as recommended empiric antimicrobials [13]. The monitoring of serum aminoglycoside concentrations is not done routinely in the Philippines, so aminoglycoside use should be limited to the treatment of culture-proved aminoglycoside-sensitive infections when no safer alternatives exist.

Although the present study did not document anaerobes because of technical limitations, it was reported in a previous study [21] that anaerobic bacteria were more likely to be present in the bile of elderly patients and those with complex biliary problems. It is recommended that anaerobic therapy be given in the presence of a biliary–enteric anastomosis [13].

The findings of our study on the relatively decreased susceptibility to ceftazidime of P. aeruginosa and to ceftazidime, aztreonam, and ceftriaxone of E. cloacae may be suggestive of the emergence of bacterial resistance to broad-spectrum cephalosporins, possibly secondary to the development of ESBL strains. Aztreonam is not readily available in the Philippines and therefore should be excluded from any recommended empiric formulary for biliary infections in this country.

According to a review by Paterson and Bonomo, national surveys reported the presence of ESBL in 5–8% of E. coli isolates from Korea, Japan, Malaysia, and Singapore and 12–24% in Thailand, Taiwan, the Philippines, and Indonesia [22]. This finding is confirmed by a subsequent surveillance study done in the Asia-Pacific region showing high rates of ESBL-producing isolates of E. coli (19%). Furthermore, carbapenems were the most active antimicrobial agent in vitro regardless of whether the pathogen produced an ESBL [14].

With findings from this study suggesting the presence of ESBL-producing isolates, the authors recommend routine conduct of antimicrobial surveillance and implementation of standards in detecting and reporting ESBL resistance patterns. Timely detection and reporting of ESBL-producing strains and their susceptibilities is indicated. In the absence of such a system of timely feedback, recommendations for empiric therapy must be based on the most recent local susceptibility data.

The recommendations for antibiotic treatment of both cholangitis and non-cholangitis biliary conditions are similar. Although cholangitis patients have significantly more positive bile cultures than do patients without cholangitis, the frequencies of gram-negative bacilli and gram-positive cocci are not statistically different. A common empiric antibiotic recommendation can be based on the susceptibility profile of the bacterial isolates in this Philippine institution. It must be noted, however, that this recommendation is based on limited microbiological resources. Likewise, our recommendations for a broad-spectrum empiric antimicrobial drug with documented effectiveness against the pathogens most common in cholangitis and non-cholangitic biliary disease are underpinned by the principles that providing late and inappropriate antimicrobials significantly increases mortality rates and that the step-down of the intensity of antimicrobial therapy, guided by culture results, as soon as possible reduces the development of multidrug-resistant bacteria while retaining a survival advantage [13,23].

In summary, in the present study, gram-negative bacteria (Enterobacteriacae) were the most common bile culture isolates from patients with and without cholangitis. Among them, E. coli was the predominant bacterium isolated from both groups (35% vs. 38%). Antimicrobial susceptibility to extended-spectrum antibiotics, especially the anti-pseudomonal carbapenems (100% for all Enterobacteriaceae, even E. cloacae), remains high for both groups of patients.

Conclusions

This study reviewed the microbiology of bile cultures and antimicrobial sensitivity of organisms isolated from bile samples from patients with and without cholangitis seen at an Asian university medical center during a three-year period. Cholangitis patients are significantly more likely to have positive bile cultures than are patients without cholangitis but with biliary disease (p = 0.012). The frequency of gram-negative bacilli and gram-positive cocci were not statistically different for cholangitis and no cholangitis. Gram-negative bacilli were the most common microbial isolates.

Antimicrobial susceptibility data on P. aeruginosa and E. cloacae demonstrated the possible development of ESBL infections among patients with biliary infections. This seminal study has provided evidence for formulating recommendations for the choice of empiric antimicrobial therapy for biliary infections in the Philippines. The resistance rate (≥20%) of the most commonly grown organisms to the lower-generation cephalosporins disqualifies them as first-line empiric agents for biliary infections [13].

On the basis of the evidence from this study, our recommendation for presumptive or empiric antimicrobial therapy for biliary infections in the Philippines can be any of the following: Cefepime or ceftriaxone in combination with metronidazole or imipenem-cilastatin meropenem, or piperacillin-tazobactam, given as monotherapy, the drugs being administered intravenously as soon as possible after a clinical diagnosis is made. Shifting to a narrower-spectrum antibacterial agent should be initiated as soon as culture data or the clinical situation allows. It must be noted, however, that the study's preliminary findings suggest that meropenem and imipenem-cilastatin remain the most active agents against E. cloacae. Obtaining biliary cultures routinely will fulfill two vital needs:

It will supply the practitioner with the necessary data to inform a choice of step-down therapy if/when it is appropriate; and