Changes in Microbial Profiles and Antibiotic Resistance Patterns in Patients with Biliary Tract Infection over a Six-Year Period

Abstract

Background:

Biliary tract infection (BTI) is a common complication in patients with biliary obstruction caused by various biliary tract disorders. To improve the management of patients with BTI, microbial profiles from bile cultures and antibiotic resistance patterns were evaluated in this six-year retrospective study.

Methods:

A total of 709 patients with various biliary tract disorders from January 2012 to December 2017 were enrolled in this study. Bile specimens were taken from intra-operative puncture or post-operative drainage for microbial culture under sterile conditions. Microbial culture, identification, and antibiotic resistance test were performed according to a standard routine.

Results:

Cultures were positive in 574 of 789 bile specimens (73%). Of all isolates, 386 were gram-negative bacilli (67%), 170 were gram-positive cocci (30%), and 18 were fungi (3%). The two most common micro-organisms were Escherichia coli (29%) and Klebsiella pneumoniae (16%), the extended-spectrum β-lactamases (ESBL) positivity rates of which were 52% and 21%, respectively. Both bacteria were highly resistant to commonly used antibiotic agents (penicillins, cephalosporins, and quinolones), and highly susceptible to carbapenems, amikacin, and piperacillin-tazobactam. Enterococcus (19%) was the next most common bacteria isolated from bile samples, mainly including Enterococcus faecium (8%) and Enterococcus faecalis (6%). These two bacteria were resistant to cefazolin and clindamycin but sensitive to teicoplanin, tigecycline, and vancomycin. Enterococcus faecium was more resistant than Enterococcus faecalis to most of the tested antibiotic agents. Annual statistical analysis showed that the frequency of gram-negative enteric bacteria was decreasing slowly, but that of gram-positive enterococci was increasing slowly. Moreover, the overall antibiotic resistance rates of the most common strains have been decreasing slowly over the past six years.

Conclusions:

Enterobacteriaceae (Enterococcus coli and Klebsiella pneumoniae) and Enterococcus (Enterococcus faecium and Enterococcus faecalis) were the common pathogenic bacteria causing BTI, which exhibited high resistance to routinely used antibiotic agents and were highly sensitive to piperacillin-tazobactam, carbapenem, amikacin, and vancomycin. The microbial profiles from bile and its antibiotic resistance patterns have changed, which will help in the empirical treatment of BTI.

Biliary tract infection (BTI) is often a bacterial infection in the biliary tract system, including acute or chronic cholecystitis and cholangitis [1,2]. Normally, bile is sterile in healthy individuals, however, in various predisposing conditions, bacteria can colonize and proliferate within the stagnant bile while biliary pressure increases [3]. Therefore, BTI is a common complication in patients with biliary obstruction caused by cholelithiasis, neoplasms, and benign strictures [4]. Without effective treatment, BTI can progress into severe abdominal infection, bacteremia, sepsis, and multiple organ dysfunction, which are often life-threatening especially in the elderly [5].

The investigation of biliary microflora and resistance to antibiotic agents will be helpful in the diagnosis of BTI, nosocomial infection, and choosing an empiric antibiotic regimen. Most studies over the past decade have shown that BTI was caused mainly by enteric organisms, such as Escherichia coli, Klebsiella pneumoniae, and Enterococcus faecalis [6 –11]. However, the distribution of biliary microflora may vary in different periods and regions. Moreover, because of the rapid development of multi-drug–resistant organisms, the choice of appropriate empiric antimicrobial therapy has become more complicated [12]. Therefore, monitoring dynamic changes of biliary microflora and resistance to antibiotic agents is important.

This study focused on patients undergoing biliary operations from January 2012 to December 2017. A total of 574 bacterial strains were isolated from 789 bile specimens prepared from intra-operative extraction or post-operative drainage, and the changes of its frequency and antibiotic resistance during the six-year period were analyzed.

Patients and Methods

Study population

This retrospective study was conducted in the Department of General Surgery, Zhuji People's Hospital (a tertiary care center), the affiliated hospital of Wenzhou Medical University, eastern China, from January 2012 to December 2017. During this period, a total of 709 patients were included in this study, whose pre-operative diagnosis included cholelithiasis with cholecystitis or cholangitis, bile duct carcinoma, pancreatic cancer, obstructive jaundice, etc. The diagnosis of BTI was based on the presence of clinical evidence, including fever, abdominal pain, leukocytosis, elevated alkaline phosphatase, and ultrasound images. Written informed consent was obtained from all patients and the trial was approved by the local ethics committee.

Bacteriology

Specimen preparation, bacterial culture, isolation, identification, and antimicrobial susceptibility test were conducted in accordance with the National Guide to Clinical Laboratory Procedures of China. Briefly, bile specimens were prepared by intra-operative extraction or post-operative drainage under aseptic conditions and transported to the clinical microbiology laboratory within two hours. These samples were then immediately inoculated onto the blood agar plates and cultured at 37°C for 24–48 hours. Bacterial isolates were identified at the species level and its antibiotic susceptibility was tested using a VITEK 2 Compact instrument (bioMérieux, Lyon, France). The minimal inhibitory concentrations (MICs) were interpreted in accordance with the Clinical Laboratory Standards Institute (CLSI) guidelines. Escherichia coli ATCC 25922, Enterobacter cloacae ATCC 700323, Staphylococcus aureus ATCC 25923, and Streptococcus pneumoniae ATCC 49619 were used as standard strains for quality control.

Statistics

Categorical variables are presented as proportions or rates. The descriptive statistical analysis was performed using GraphPad Prism 7.0 (GraphPad Software, San Diego, CA).

Results

Patients characteristics and microbiologic profile of bile

Of the 709 patients, 602 were diagnosed with cholelithiasis (85%), including choledocholithiasis (n = 440; 62%), intra-hepatic stones (n = 128; 18%) and cholecystolithiasis (n = 34; 5%). The majority of patients with cholelithiasis had cholecystitis (52%) or cholangitis (44%). The remaining patients were diagnosed as bile duct carcinoma, pancreatic cancer, obstructive jaundice, etc.

Cultures were positive in 574 of 789 bile specimens (73%). The frequency of different micro-organisms in positive bile cultures was shown in Table 1. Of all isolates, 386 were gram-negative bacilli (67%), 170 were gram-positive cocci (30%), and 18 were fungi (3%). The most common micro-organisms were Escherichia coli (29%), Klebsiella pneumoniae (16%), Enterococcus faecium (8%), and Enterococcus faecalis (6%). Moreover, 81 of 156 (52%) of all Escherichia coli strains and 18 of 85 (21%) of all Klebsiella pneumoniae had shown extended-spectrum β-lactamases (ESBL)-positive phenotype. Annual statistical analysis showed that the frequency of Escherichia coli and Klebsiella pneumoniae had a slow downward trend, but that of Enterococcus faecium and Enterococcus faecalis had gradually increased in recent years (Fig. 1A). In addition, ESBL-producing rates of Escherichia coli and Klebsiella pneumoniae decreased during the study period (Fig. 1C).

FIG. 1.

Annual change in frequency and antibiotic resistance of the most common bacterial isolates from bile culture. (A) Proportions, (B) total resistance rates to a list of antibiotic agents, (C) ESBL-producing rates, and (D) specific resistance rates to cephalosporins of main bile isolates are summarized on an annual basis.

Table 1.

Distribution of Micro-Organisms Isolated from Bile Specimens

	Microbial isolates	Total number	%
Gram-negative bacilli	Escherichia coli	165	28.75
(n = 386; 67%)	Klebsiella spp.	96	16.72
	Klebsiella pneumoniae	90	15.68
	Klebsiella oxytoca	6	1.05
	Enterobacter spp.	35	6.10
	Enterobacter cloacae	27	4.70
	Enterobacter aerogenes	7	1.22
	Enterobacter asburiae	1	0.17
	Pseudomonas spp.	28	4.88
	Pseudomonas aeruginosa	27	4.70
	Pseudomonas alcaligenes	1	0.17
	Aeromonas spp.	14	2.44
	Proteus spp.	9	1.57
	Acinetobacter baumannii	8	1.39
	Citrobacter spp.	6	1.05
	Other	25	4.36
Gram-positive cocci	Enterococcus spp.	110	19.16
(n = 170; 30%)	Enterococcus faecium	44	7.67
	Enterococcus faecalis	35	6.10
	Enterococcus casseliflavus	12	2.09
	Enterococcus gallinarum	10	1.74
	Enterococcus raffinosus	4	0.70
	Enterococcus avium	3	0.52
	Enterococcus durans	2	0.35
	Staphylococcus spp.	19	3.31
	Streptococcus spp.	30	5.23
	Other	11	1.92
Fungi (n = 18; 3%)	Candida spp.	18	3.14
	Candida albicans	10	1.74
	Candida tropicalis	4	0.70
	Candida parapsilosis	3	0.52
	Candida glabrata	1	0.17

Antibiotic resistance patterns of bacterial isolates

Antibiotic susceptibility testing was performed for all bacterial isolates. The results of the antibiotic resistance test for most common isolated gram-negative and gram-positive bacteria were shown in Table 2 and Table 3, respectively. In general, Escherichia coli and Klebsiella pneumoniae were highly resistant to penicillins, cephalosporins, and quinolones. All isolates of Klebsiella pneumoniae and 75% of Escherichia coli were resistant to ampicillin. With the addition of sulbactam, the rates of antibiotic resistance decreased to 60% and 48%, respectively. Similarly, both bacteria are susceptible to amoxicillin, piperacillin, and cefoperazone by the addition of a β-lactamase inhibitor. Carbapenems and amikacin showed high activity against gram-negative isolates. All of the isolates of Enterococcus faecium and Enterococcus faecalis were resistant to cefazolin and clindamycin but were sensitive to teicoplanin, tigecycline, and vancomycin. Moreover, compared with Enterococcus faecalis, Enterococcus faecium was more resistant to most of the tested antibiotics. In addition, annual statistical analysis showed that total antibiotic resistance rates of all four strains decreased slowly year by year (Fig. 1B). The specific resistance rate of Escherichia coli and Klebsiella pneumoniae to cephalosporins decreased more in the last few years (Fig. 1D).

Table 2.

Antibiotic Resistance of Most Frequently Isolated Gram-Negative Bacteria (%)

Antibiotic agents	Escherichia coli		Klebsiella pneumoniae
Antibiotic agents	n	%	n	%
Ampicillin	125/165	75.76	89/89	100.00
Ampicillin/sulbactam	32/53	60.38	12/25	48.00
Amoxicillin/clavulanic acid	26/111	23.42	10/64	15.63
Piperacillin/tazobactam	6/160	3.75	8/89	8.99
Cefazolin	74/101	73.27	23/36	63.89
Ceftazidime	26/52	50.00	9/25	36.00
Ceftriaxone	87/164	53.05	26/89	29.21
Cefotetan	4/52	7.69	3/25	12.00
Cefepime	43/163	26.38	19/88	21.59
Cefoperazone/sulbactam	9/151	5.96	9/86	10.47
Imipenem	1/164	0.61	6/89	6.74
Ertapenem	7/161	4.35	4/86	4.65
Meropenem	1/40	2.50	0/20	0.00
Aztreonam	69/159	43.40	23/89	25.84
Gentamicin	48/163	29.45	10/87	11.49
Tobramycin	17/164	10.37	6/88	6.82
Amikacin	7/164	4.27	3/90	3.33
Ciprofloxacin	87/163	53.37	26/89	29.21
Levofloxacin	81/163	49.69	15/89	16.85
Nitrofurantoin	15/165	9.09	54/90	60.00
Trimethoprim-sulfamethoxazole	87/164	53.05	26/88	29.55
Tigecycline	1/111	0.90	8/57	14.04

Table 3.

Antibiotic Resistance of Most Frequently Isolated Gram-Positive Bacteria (%)

Antibiotic agents	Enterococcus faecium		Enterococcus faecalis
Antibiotic agents	n	%	n	%
Penicillin G	26/43	60.47	5/35	14.29
Ampicillin	22/43	51.16	4/35	11.43
Cefazolin	13/13	100.00	6/6	100.00
Erythromycin	30/43	69.77	8/35	22.86
Clindamycin	41/41	100.00	34/34	100.00
Quinupristin-dalfopristin	0/41	0.00	34/34	100.00
Streptomycin, high	12/43	27.91	6/35	17.14
Gentamicin, high	12/43	27.91	5/35	14.29
Teicoplanin	0/15	0.00	0/6	0.00
Vancomycin	0/43	0.00	1/35	2.86
Tigecycline	0/40	0.00	0/34	0.00
Linezolid	1/40	2.50	1/32	3.13
Nitrofurantoin	11/43	25.58	1/35	2.86
Tetracycline	15/43	34.88	16/35	45.71
Ciprofloxacin	24/43	55.81	2/35	5.71
Levofloxacin	19/41	46.34	2/34	5.88
Moxifloxacin	23/41	56.10	2/34	5.88

Discussion

The main cause of BTI is bile duct obstruction caused by a gallstone, especially with choledocholithiasis, which will increase bile tract pressure and leads to microbial colonization and proliferation. In the present study, most cases of BTI were older (median age, 67) and had gallstones in the common bile duct (62%). Only a small number of patients with BTI had malignancies, including bile duct cancer, pancreatic cancer, and liver cancer. A survey from Taiwan reported that of 112 patients with cholangitis, 55% had choledocholithiasis and 25% had malignancy [13]. Another survey from India reported that 63% of 185 patients with acute cholangitis had choledocholithiasis and 28% had malignancy [14]. In this study, of the 397 patients with cholangitis, choledocholithiasis was responsible for 48%, intra-hepatic stones in 25%, and malignancy in 17%. Therefore, choledocholithiasis is more likely to be associated with cholangitis (80%).

Evaluation of microbial profiles and antibiotic sensitivity patterns in bile cultures can help guide effective empirical antibiotic therapy. Most studies reported that the positive rate of bile culture was approximately 70%. Similar to these surveys, we found 73% of the bile specimens were positive for bacterial or fungal culture. In all microbial isolates, gram-negative bacilli (67%) were predominant, followed by gram-positive cocci (30%), and a small number of fungi (3%). The top five most frequently isolated bacteria were Escherichia coli, Klebsiella pneumoniae, Enterococcus, Enterobacter, and Pseudomonas aeruginosa, which were consistent with most previous observations [15]. However, a study from Germany showed that among all the bacterial isolates from bile culture, more were gram-positive (57%) than gram-negative (43%), and Enterococcus species were predominant (494/1150 samples; 33%) [16]. In the past six years, the microbial profiles seem to have changed, the frequency of gram-negative enteric bacteria was declining slowly, but that of gram-positive enterococci was increasing slowly, which is a noteworthy change.

The number of drug-resistant strains has been increasing dramatically as a result of the widespread and inappropriate use of antibiotic agents, which poses a challenge to the choice of optimal empirical antibiotic therapy. Local antibiotic resistance patterns can be used as supporting data to optimize the selection of empiric antibiotic therapy and increase the appropriate use of antibiotic agents, thereby reducing mortality and healthcare costs. Similar to most previous studies, Escherichia coli and Klebsiella pneumoniae were the most common micro-organisms isolated from patients with BTI, and their frequency remained stable during the study period. In 2011, a study from the Philippines showed that Escherichia coli and Klebsiella pneumoniae isolates obtained from bile culture remained highly susceptible to cefepime, ceftriaxone, aztreonam, and gentamicin [9]. However, a recent study from South Korea showed that the drug-resistance rate was 14% to the fourth-generation cephalosporins and over 20% to the first- through third-generation cephalosporins. The fluoroquinolones and penicillin resistance were also observed in 20% or more of cases [11]. In the present study, Escherichia coli and Klebsiella pneumoniae isolates had resistance to cefepime (26% and 22%), aztreonam (43% and 26%), fluoroquinolones (>50% and >17%), and higher resistance to the first-/third-generation cephalosporins (>50% and >29%). However, these two bacteria remained highly susceptible to piperacillin with tazobactam, cefoperazone with sulbactam, cephamycins, amikacin, carbapenem, and tigecycline.

The most common ESBL-producing bacteria are Escherichia coli and Klebsiella pneumoniae. However, the proportion of these two ESBL-positive bacteria from bile culture varied in different regions: 30% and 6% in a Korean university medical center [17], 66% and 73% in an Indian medical college hospital [18], 44% and 56% in Iranian university hospitals [19], and 52% and 21% in our study. Extended-spectrum β-lactamases can break down penicillin derivatives and cephalosporins and render them ineffective for treatment. Thus, because of the high incidence of ESBL-producing Enterobacteriaceae, the antibiotic agents containing β-lactamase inhibitors are an appropriate choice of empirical antibiotic therapy, although carbapenems are more sensitive. Ng et al. [20] compared 30-day prognosis of patients treated with empiric piperacillin-tazobactam versus carbapenem for bacteremia caused by ESBL-producing Enterobacteriaceae (Escherichia coli and Klebsiella pneumoniae) in a multi-center retrospective cohort study in Singapore. They observed that the use of empiric piperacillin-tazobactam was not associated with increased 30-day mortality and may result in fewer multi-drug–resistant and fungal infections compared with carbapenem [20]. Darkahi et al. [6] reported that piperacillin-tazobactam was sensitive to all strains isolated from bile in a retrospective study in Sweden. In this study, the resistance rates of piperacillin-tazobactam to Escherichia coli and Klebsiella pneumoniae were 3.75% and 8.99%, respectively, which were lower than those of other penicillins and cephalosporins. Recently, piperacillin-tazobactam was also on the recommended list of antimicrobial agents for community-acquired and health-care–associated acute BTI in the Tokyo Guidelines 2018 [21].

Enterococcus spp. is another important pathogen causing biliary infection in the present study. In general, compared with Enterococcus faecalis, Enterococcus faecium had higher resistance to penicillins, cephalosporins, aminoglycosides, and fluoroquinolones. Almost all of Enterococcus faecium and Enterococcus faecalis isolates were resistant to cefazolin and clindamycin but were susceptible to vancomycin, teicoplanin, tigecycline, and linezolid. Thus, for BTI caused by Enterococcus, vancomycin is the drug of choice for empirical therapy and was also on the recommended list of Tokyo Guidelines 2018 [21].

Fortunately, the overall antibiotic resistance rate of both gram-negative bacilli and gram-positive enterococci has been declining slowly over the past six years. Especially for the two most important bacteria (Escherichia coli and Klebsiella pneumoniae), the ESBL-positive rate has shown a substantial downward trend, and correspondingly, its resistance to cephalosporins has also decreased. These observations may indicate that the risk of inappropriate use of antibiotic agents is decreasing as a result of a series of strict policies and regulations on the clinical use of antibiotic agents in China.

Finally, some limitations should be noted. First, current results come from a single-center retrospective study. Despite the inherent selectivity bias, there are enough number of patients, bile specimens, and microbial isolates to obtain these reliable results. Second, the study population did not include all patients with BTI but only those treated surgically. Therefore, the results may be more representative of relatively severe BTI. Third, incomplete patient information resulted in the study population not being classified according to the Tokyo Guidelines 2018 to draw more valuable conclusions.

In conclusion, the present study showed that almost three-quarters of bile specimens from patients with various biliary disorders during the past six-year period were positive for microbial culture. The most frequently isolated bacteria were Enterobacteriaceae including Escherichia coli and Klebsiella pneumoniae, followed by Enterococcus including Enterococcus faecium and Enterococcus faecalis. These major bacterial isolates exhibited high resistance to routinely used antibiotic agents and were highly sensitive to piperacillin-tazobactam, carbapenem, amikacin, vancomycin, etc. Although the microbial profiles and its drug-resistance patterns were basically stable in the past six years but substantially different from other countries and regions, which will assist clinicians in the empirical treatment of BTI.

Footnotes

Acknowledgments

The authors are grateful to Dr. Fangjun Luo and Ms. Yujia Fei (Department of Clinical Laboratory, Zhuji People's Hospital, China) for granting access to bile culture records, and to Dr. Nansong Yu (Department of Hepatobiliary Surgery, Shaoxing People's Hospital, China) for his valuable suggestions.

Author Disclosure Statement

No competing financial interests exist.

References

Jain

, Jain

. Acute bacterial cholangitis. Curr Treat Options Gastroenterol, 2006; 9:113–121.

Kujiraoka

, Kuroda

, Asai

, et al. Comprehensive diagnosis of bacterial infection associated with acute cholecystitis using metagenomic approach. Front Microbiol, 2017; 8:685.

Ahmed

. Acute cholangitis—An update. World J Gastrointest Pathophysiol, 2018; 9:1–7.

Ortega

, Marco

, Soriano

, et al. Epidemiology and prognostic determinants of bacteraemic biliary tract infection. J Antimicrob Chemother, 2012; 67:1508–1513.

Kinney

. Management of ascending cholangitis. Gastrointest Endosc Clin N Am, 2007; 17:289–306.

Darkahi

, Sandblom

, Liljeholm

, et al. Biliary microflora in patients undergoing cholecystectomy. Surg Infect, 2014; 15:262–265.

Tajeddin

, Sherafat

, Majidi

, et al. Association of diverse bacterial communities in human bile samples with biliary tract disorders: A survey using culture and polymerase chain reaction-denaturing gradient gel electrophoresis methods. Eur J Clin Microbiol Infect Dis, 2016; 35:1331–1339.

Kim

, Park

, et al. Clinical impact of extended-spectrum beta-lactamase-producing Enterobacteriaceae in patients with biliary tract infection. Dig Dis Sci, 2013; 58:841–849.

Salvador

, Lozada

, Consunji

. Microbiology and antibiotic susceptibility of organisms in bile cultures from patients with and without cholangitis at an Asian academic medical center. Surg Infect, 2011; 12:105–111.

10.

Kaya

, Bestas

, Bacalan

, et al. Microbial profile and antibiotic sensitivity pattern in bile cultures from endoscopic retrograde cholangiography patients. World J Gastroenterol, 2012; 18:3585–3589.

11.

Yun

, Seo

. Clinical aspects of bile culture in patients undergoing laparoscopic cholecystectomy. Medicine, 2018; 97:e11234.

12.

Baughman

. Considerations in the choice and administration of agents for empiric antimicrobial therapy. Surg Infect, 2005; 6 Suppl 2:S71–S82.

13.

Lee

, Chang

, Lai

, et al. Epidemiology and prognostic determinants of patients with bacteremic cholecystitis or cholangitis. Am J Gastroenterol, 2007; 102:563–569.

14.

Sahu

, Chacko

, Dutta

, Prakash

. Microbial profile and antibiotic sensitivity pattern in acute bacterial cholangitis. Indian J Gastroenterol, 2011; 30:204–208.

15.

Gomi

, Solomkin

, Takada

, et al. TG13 antimicrobial therapy for acute cholangitis and cholecystitis. J Hepatobiliary Pancreat Sci, 2013; 20:60–70.

16.

Rupp

, Bode

, Weiss

, et al. Microbiological assessment of bile and corresponding antibiotic treatment: A strobe-compliant observational study of 1401 endoscopic retrograde cholangiographies. Medicine, 2016; 95:e2390.

17.

Kwon

, Han

, Kim

, et al. Changes in causative pathogens of acute cholangitis and their antimicrobial susceptibility over a period of 6 years. Korean J Gastroenterol, 2014; 63:299–307.

18.

Shenoy

, Shenoy

, Gopal

, et al. Clinicomicrobiological analysis of patients with cholangitis. Indian J Med Microbiol, 2014; 32:157–160.

19.

Zaniani

, Meshkat

, Naderi Nasab

, et al. The prevalence of TEM and SHV genes among extended-spectrum beta-lactamases producing Escherichia coli and Klebsiella pneumoniae. Iran J Basic Med Sci, 2012; 15:654–660.

20.

, Khong

, Harris

, et al. Empiric piperacillin-tazobactam versus carbapenems in the treatment of bacteraemia due to extended-spectrum beta-lactamase-producing Enterobacteriaceae. PLoS One, 2016; 11:e0153696.

21.

Gomi

, Solomkin

, Schlossberg

, et al. Tokyo Guidelines 2018: Antimicrobial therapy for acute cholangitis and cholecystitis. J Hepatobiliary Pancreat Sci, 2018; 25:3–16.