Keywords
ERC topics - Strictures - Pancreatobiliary (ERCP/PTCD) - Stones
Introduction
Antibiotic resistance is a growing problem worldwide, with antibiotic-resistant infections
associated with increased morbidity, mortality, and healthcare costs [1]
[2]. In the United States, antibiotic-resistant infections lead to approximately 35,000
deaths annually [2]. Mitigating antimicrobial drug resistance encompasses strategies such as drug discovery
and the control of antimicrobial drug-resistant pathogen development. Given the protracted
research timelines and substantial expenses involved in drug discovery, controlling
the development of antimicrobial drug-resistant pathogens is imperative [1]
[3].
The risk factors for developing infections caused by antimicrobial drug-resistant
pathogens include a history of antimicrobial exposure [4]
[5]. Other risk factors include exposure to medical devices and healthcare workers [1]
[5]
[6]
[7]
[8]
[9]
[10], regular use of proton pump inhibitors (PPIs) [11], implantation of biliary stents [12]
[13], and prior endoscopic sphincterotomy (EST) [14]
[15]. Moreover, the long-term administration of antimicrobial agents, especially in
patients with pneumonia, poses a risk for the emergence of multidrug-resistant bacteria
[16]
[17]
[18]
[19]. Therefore, global recommendations advocate for the short-term administration of
antimicrobial agents [20]
[21]. Acute cholangitis (AC), the second or fourth most prevalent cause of
community-acquired bacteremia [22]
[23], stands as one of the most common diseases necessitating antimicrobial therapy.
However, no study has explored the potential association between antimicrobial resistance
and
the duration of antimicrobial therapy recommended by the current AC guidelines.
The Tokyo Guidelines 2018 (TG18), the most prominent AC guideline, recommend a 4-
to 7-day
course of antibiotic therapy following biliary drainage; however, robust evidence
substantiating the requisite duration for curing AC remains elusive [24]. Recent studies suggested that a 1- to 3-day course or less of antibiotic therapy
following biliary drainage is a reasonable duration for antibiotic administration
to cure AC
[25]
[26]
[27]
[28]; thus, Dutch guidelines recommended a 3-day regimen after biliary drainage [29]. Moreover, the French Infectious Disease Society (SPILF) proposed the reduction
of
antimicrobial therapy duration to 3 days if successful drainage was achieved, including
those
with bacteremia [20]. Once the source of infection is controlled by biliary drainage, bacteremia is likely
to resolve, potentially obviating the need for further antibiotic therapy [30].
Shortening the administration of such antibiotics may effectively mitigate the development
of resistant bacteria in patients with AC. Nevertheless, whether a shortened course
of antimicrobial therapy, as opposed to the TG18 recommendation, indeed diminishes
the emergence of resistant organisms remains enigmatic. Therefore, we hypothesized
that administering antimicrobial therapy for up to 3 days following endoscopic retrograde
cholangiopancreatography (ERCP) could mitigate the emergence of resistant organisms
in AC patients. To explore this hypothesis, we analyzed the duration of previous antimicrobial
therapy in two distinct groups: patients with recurrent AC in whom resistant bacteria
emerged and those in whom susceptible bacteria emerged. By investigating this relationship,
we aimed to ascertain whether a treatment duration shorter than the standard duration
recommended by TG18 may be necessary to suppress the emergence of resistant bacteria
in patients with AC.
Patients and methods
Study population
This retrospective cohort study was conducted at the Shonan Kamakura General Hospital
in Japan. The medical records of patients treated at the hospital between January
2018 and June 2020 were searched to identify patients with AC who obtained positive
results on blood or bile culture tests; the patients successfully underwent ERCP and
were infected with organisms that were susceptible to the administered antimicrobials.
Patients who experienced AC recurrence by January 2022 and obtained positive blood
or bile culture test results at that time were included in the final analysis. The
patients were divided into two groups at the time of recurrence: those in whom the
causative organisms had become resistant to the antimicrobials administered at the
time of initial admission (resistant group) and those in whom the organisms remained
susceptible (susceptible group). This dichotomy enabled the researchers to investigate
whether the duration of antimicrobial treatment during the initial AC was correlated
with the acquisition of resistance by the causative organisms during the relapse AC
([Fig. 1]).
Fig. 1 Flow diagram. ERCP, endoscopic retrograde cholangiopancreatography; AC, acute cholangitis
Blood cultures were collected prior to the antibiotic administration, while bile cultures
were collected immediately after ERCP. We employed interpretive standards from the
Clinical and Laboratory Standards Institute (CLSI) for minimal inhibitory concentration
(MIC) or zone diameter testing to identify susceptible or resistant organisms [31].
In our hospital, ampicillin/sulbactam, cefmetazole, ceftriaxone, piperacillin/tazobactam,
meropenem, and ciprofloxacin are typically used as initial treatments. Mild cases
were primarily treated with cefmetazole, with other antibiotics administered based
on severity and results of previous cultures.
If plastic stent implantation was required during ERCP, one or two 7F stents were
generally implanted. Self-expandable metallic stents measuring 10 and 8 mm in diameter
were implanted in the common bile duct and hilar regions, respectively.
Research items
Data on the patient background, ERCP findings, antimicrobials used, duration of administration,
and blood and bile culture results at initial admission were collected. We also investigated
the antimicrobial agents used, blood and bile culture results, and clinical characteristics
at the second admission to assess for recurrent cholangitis.
According to the results of previous studies exploring the risks of developing resistant
bacteria, we also investigated the incidence of PPI use [11], implantation of biliary stents [12]
[13], history of EST [14]
[15], and exposure to medical devices and healthcare workers [1]
[5]
[6]
[7]
[8]
[9]
[10].
We also investigated the history of antimicrobial administration for other infections
not related to this study from the time of initial disease onset to the time of cholangitis
recurrence. If any pertinent information was not available from the electronic medical
record, we contacted the patient, his or her family, the nursing home in which he
or she resided, or the attending physician by telephone to inquire about the antimicrobial
administration history from the initial onset to the recurrence of cholangitis.
To the best of our knowledge, no existing study has reported long-term antimicrobial
therapy as a risk factor for the development of resistant organisms in patients with
AC. However, a relationship between long-term antimicrobial therapy and the development
of resistant organisms was reported in patients with other infectious diseases [16]
[17]
[32]. Therefore, we investigated the factors involved in the development of resistant
organisms in patients with recurrent cholangitis, including long-term antimicrobial
therapy.
Definitions
The diagnosis and assessment of cholangitis severity were based on the TG18 guidelines
[33]. The duration of hospitalization was defined as the number of days from admission
to discharge or death. The duration of antimicrobial therapy was investigated, with
4 to 7 days after ERCP as the standard treatment period according to the TG18 guidelines
and within 3 days after ERCP as the short-term treatment period.
Cholangitis is frequently attributed to polymicrobial infection. Blood cultures exhibit
low sensitivity and may fail to detect the causative organism, whereas bile cultures
possess low specificity and may detect enteric bacteria that are unrelated to the
causative organisms. This discrepancy complicates the accurate identification of the
causative organism of cholangitis. Therefore, in this study, all identified bacteria
were regarded as causative organisms.
The development of resistant organisms in recurrent cholangitis was defined as follows.
Organisms identified from the blood and bile cultures were susceptible to antimicrobials
administered during the initial hospitalization. On the second admission, owing to
the recurrence of AC, the bacteria identified from the blood and bile cultures had
developed resistance to the antimicrobial agent administered during the initial admission.
Resistance to the initial antibiotics was defined as in vitro resistance to these
antibiotics, with antibiotic efficacy determined in accordance with the CLSI interpretive
standards for MIC or zone diameter testing [31].
Statistical analyses
The Mann-Whitney U test was used to compare non-normally distributed continuous variables,
while the χ2 test or Fisher’s exact test was used to compare categorical variables.
Multivariate analysis was performed using logistic regression and included antibiotics
used within 3 days after ERCP in addition to the reported risk factors for the development
of bacterial resistance. “Regular use of PPI,” “first ERCP in their life,” “EST performed
during the initial hospitalization,” “antibiotics within 3 days after ERCP,” “piperacillin/tazobactam,”
and “history of antimicrobial administration from the initial onset to the recurrence
of cholangitis” were included in the multivariate analysis as independent variables.
Notably, the “history of antimicrobial administration from the initial onset to the
recurrence of cholangitis,” a facet seldom explored in previous studies to our knowledge,
seems to be an important independent variable signifying the interaction between patients
and medical interventions or antimicrobial agents. Piperacillin/tazobactam is the
most frequently used broad-spectrum antibacterial agent in our hospital. Previous
studies considered “bile duct stent placement from initial admission to second admission”
as a risk factor for the development of resistant pathogens [12]
[13]. However, these studies were single-center, retrospective studies; did not include
the period of antimicrobial administration, which is an important risk factor for
the development of bacterial resistance, in the multivariate analysis; or did not
sufficiently investigate the antimicrobial exposure history. Therefore, “bile duct
stent placement from initial admission to second admission” was excluded from the
multivariate analysis in favor of a more pivotal factor. A sensitivity analysis was
performed using more independent variables, including “bile duct stent placement from
initial admission to second admission.”
A two-tailed P <0.05 was considered significant. All statistical analyses were performed using EZR
(Saitama Medical Center, Jichi Medical University, Saitama, Japan), a graphical user
interface for R version 4.2.3 (R Foundation for Statistical Computing, Vienna, Austria).
It is a modified version of R commander, designed to allow additional biostatistical
functions [34].
Results
Patient characteristics
[Table 1] summarizes the patient characteristics. The data from 89 patients with recurrent
AC and positive blood or bile culture results were analyzed. Of these, 43 patients
(48.3%) had causative organisms that developed resistance to the administered antimicrobials
at the time of initial admission (resistant group), while 46 (51.7%) had organisms
that remained susceptible (susceptible group). No significant differences were observed
between the two groups in terms of age, sex, cause of cholangitis, severity of cholangitis,
underlying medical conditions, nursing home residence, or PPI use.
Table 1 Patient characteristics.
|
Resistant group n=43
|
Susceptible group n=46
|
P value
|
|
Some cases overlapped.
CKD, chronic kidney disease; CHF, chronic heart failure; DM, diabetes mellitus; IQR,
interquartile range; LC, liver cirrhosis; PPI, proton pump inhibitor.
|
|
Age, median (IQR)
|
83.00 (74.00, 86.00)
|
77.50 (70.00, 85.00)
|
0.17
|
|
Age > 80 years, n (%)
|
26 (60.5)
|
21 (45.7)
|
0.2
|
|
Sex ratio
|
Male 19: Female 24
|
Male 27: Female 19
|
0.21
|
|
Cause of cholangitis, n (%)
|
|
Malignant stricture
|
18 (41.9)
|
18 (39.1)
|
0.94
|
|
Bile duct stone
|
18 (41.9)
|
21 (45.7)
|
|
|
Benign stricture
|
3 (7.0)
|
4 (8.7)
|
|
|
Other
|
4 (9.3)
|
3 (6.5)
|
|
|
Stent obstruction, including malignant and benign, n (%)
|
12 (27.9)
|
18 (39.1)
|
0.37
|
|
Severity, n (%)
|
|
Mild
|
19 (44.2)
|
18 (39.1)
|
0.67
|
|
Moderate
|
19 (44.2)
|
22 (47.8)
|
0.83
|
|
Severe
|
5 (11.6)
|
6 (13.0)
|
>0.99
|
|
Underlying medical conditions, n (%)
|
|
CKD
|
3 (7.0)
|
1 (2.2)
|
0.35
|
|
CHF
|
3 (7.0)
|
2 (4.3)
|
0.67
|
|
LC
|
2 (4.7)
|
1 (2.2)
|
0.61
|
|
DM
|
6 (14.0)
|
8 (17.4)
|
0.77
|
|
Malignant tumor
|
18 (41.9)
|
18 (39.1)
|
0.83
|
|
Hemodialysis
|
1 (2.3)
|
0 (0.0)
|
0.48
|
|
Gastrostomy
|
0 (0.0)
|
0 (0.0)
|
1
|
|
Constant placement of urinary catheter
|
0 (0.0)
|
0 (0.0)
|
1
|
|
Aspiration pneumonia
|
0 (0.0)
|
0 (0.0)
|
1
|
|
Residence nursing home
|
9 (20.9)
|
7 (15.2)
|
0.58
|
|
Immunosuppressant user
|
2 (4.7)
|
3 (6.5)
|
>0.99
|
|
Regular use of PPI
|
20 (46.5)
|
21 (45.7)
|
>0.99
|
ERCP findings
[Table 2] outlines the ERCP findings during the initial admission. No significant differences
were found between the groups in terms of ERCP treatment history, technical success
rates of biliary drainage, ERCP drainage procedures, or complications.
Table 2 ERCP findings.
|
Resistant group n=43
|
Susceptible group n=46
|
P value
|
|
EST, endoscopic sphincterotomy; ERCP, endoscopic retrograde cholangiopancreatography;
ENBD, endoscopic nasobiliary drainage
*One case overlapped.
|
|
First ERCP in their life, n (%)
|
21 (48.8)
|
13 (28.3)
|
0.053
|
|
EST, n (%)
|
0.17
|
|
EST was performed in the past
|
19 (44.2)
|
29 (63.0)
|
|
|
EST performed during initial hospitalization
|
16 (37.2)
|
10 (21.7)
|
|
|
No EST was performed
|
5 (11.6)
|
2 (4.3)
|
|
|
Choledochojejunostomy
|
3 (7.0)
|
5 (10.9)
|
|
|
ERCP drainage procedure, n (%)
|
0.79
|
|
Stent placement
|
27 (62.8)
|
30 (65.2)
|
|
|
Self-expandable metallic stent
|
10 (37.0)
|
5 (15.7)
|
|
|
Plastic stent
|
17 (63.0)
|
25 (83.3)
|
|
|
ENBD
|
1 (2.3)
|
0 (0.0)
|
|
|
Stone extraction
|
13 (30.2)
|
15 (32.6)
|
|
|
Other
|
2 (4.7)
|
1 (2.2)
|
|
|
Technical success of ERCP
|
42 (97.7)
|
45 (97.8)
|
>0.99
|
|
Technical success of biliary drainage procedure
|
43 (100)
|
46 (100)
|
1
|
|
Complications*, n (%)
|
|
Pancreatitis
|
2 (4.7)
|
0 (0.0)
|
0.23
|
|
Bleeding
|
0 (0.0)
|
1 (2.2)
|
>0.99
|
|
Perforation
|
0 (0.0)
|
0 (0.0)
|
1
|
|
Cholecystitis
|
4 (9.3)
|
0 (0.0)
|
>0.99
|
|
No complications
|
38 (88.4)
|
45 (97.8)
|
0.1
|
Antimicrobials used on initial and second admission
[Table 3] summarizes the antibiotics administered during the initial admission. Cefmetazole
(CMZ) and ampicillin/sulbactam and piperacillin/tazobactam (PIPC/TAZ) were the most
commonly used antibiotics. CMZ was predominantly used in the resistant group (51.2%)
and PIPC/TAZ in the susceptible group (52.2%).
Table 3 Antimicrobials used on initial admission.
|
Resistant group n=43
|
Susceptible group n=46
|
P value
|
|
One patient in the resistant group and one in the susceptible group were administered
two antimicrobial agents.
|
|
Cefmetazole, n (%)
|
22 (51.2)
|
11 (23.9)
|
0.01
|
|
Piperacillin/tazobactam, n (%)
|
5 (11.6)
|
24 (52.2)
|
<0.01
|
|
Ampicillin/sulbactam, n (%)
|
11 (25.6)
|
6 (13.0)
|
0.18
|
|
Ceftriaxone, n (%)
|
4 (9.3)
|
1 (2.2)
|
0.19
|
|
Meropenem, n (%)
|
1 (2.3)
|
3 (6.5)
|
0.62
|
|
Ciprofloxacin, n (%)
|
0 (0.0)
|
1 (2.2)
|
>0.99
|
|
Vancomycin, n (%)
|
0 (0.0)
|
1 (2.2)
|
>0.99
|
|
Others
|
1 (2.3)
|
0 (0.0)
|
0.48
|
[Table 4] summarizes the antibiotics administered during the second admission. CMZ, ampicillin/sulbactam,
and PIPC/TAZ were the most commonly used antibiotics. CMZ was predominantly used in
the resistant group (51.2%) and PIPC/TAZ in the susceptible group (47.8%). Approximately
half of patients in both groups received the same antimicrobials during their initial
and second hospitalizations.
Table 4 Antimicrobials used on second admission.
|
Resistant group n=43
|
Susceptible group n=46
|
P value
|
|
Cefmetazole, n (%)
|
22 (51.2)
|
12 (26.1)
|
0.02
|
|
Piperacillin/tazobactam, n (%)
|
7 (16.3)
|
22 (47.8)
|
< 0.01
|
|
Ampicillin/sulbactam, n (%)
|
4 (9.3)
|
8 (17.4)
|
0.36
|
|
Ceftriaxone, n (%)
|
3 (7.0)
|
2 (4.3)
|
0.67
|
|
Meropenem, n (%)
|
3 (7.0)
|
1 (2.2)
|
0.35
|
|
Ciprofloxacin, n (%)
|
4 (9.3)
|
1 (2.2)
|
0.19
|
|
Vancomycin, n (%)
|
0 (0.0)
|
0 (0.0)
|
N/A
|
|
Other, n (%)
|
0 (0.0)
|
0 (0.0)
|
N/A
|
|
Use of same antimicrobials on initial and second hospitalizations, n (%)
|
21 (48.8)
|
26 (56.5)
|
0.53
|
Characteristics of antimicrobial use on initial admission
[Table 5] summarizes the characteristics of antimicrobial use. No significant differences
were observed between the two groups according to the time from first physician contact
to antibiotic administration. However, the resistant group had a significantly longer
duration of antimicrobial therapy compared with the susceptible group (median, 6 vs.
5 days, respectively; P=0.04). Additionally, the resistant group had a longer antimicrobial therapy duration
following ERCP (median, 6 vs. 5 days, respectively; P=0.02), along with a reduced proportion of patients receiving a shorter dosing regimen
(within 3 days after ERCP) (9.3% vs. 30.4%; P=0.02).
Table 5 Characteristics of antimicrobial use.
|
Resistant group n=43
|
Susceptible group n=46
|
P value
|
|
IQR, interquartile range; ERCP, endoscopic retrograde cholangiopancreatography.
|
|
Time from first physician contact to antibiotic administration, median hours (IQR)
|
3.0 (2.0–5.0)
|
4.0 (2.0–7.0)
|
0.61
|
|
Total duration of antimicrobial therapy, median days (IQR)
|
6.0 (5.0–10.5)
|
5.0 (4.0–8.0)
|
0.04
|
|
Antimicrobial duration after ERCP, median days (IQR)
|
6.0 (5.0–10.5)
|
5.0 (3.0–7.0)
|
0.02
|
|
Antimicrobial agents used within 3 days after ERCP, n(%)
|
4 (9.3)
|
14 (30.4)
|
0.02
|
|
Total duration of antimicrobial therapy in patients with positive results on blood
culture tests, median days (IQR)
|
8.00 [5.75–11.00]
|
6.50 [4.25–10.00]
|
0.31
|
|
Antimicrobial duration after ERCP in patients with positive results on blood culture
tests, median days (IQR)
|
8.00 [5.00–11.00]
|
6.00 [4.00–10.00]
|
0.3
|
|
Antimicrobial agents used within 3 days after ERCP in patients with positive results
on blood culture tests, n(%)
|
0 (0.0)
|
3 (16.7)
|
0.23
|
Laboratory findings regarding microbial cultures on initial and second admission
The laboratory findings regarding the microbial cultures from patients during the
initial admission are summarized in [Table 6]. The proportions of patients with sampling blood and bile cultures were 74.4% and
93.0% in the resistant group and 71.7% and 97.8% in the susceptible group, respectively.
The proportions of patients with positive results on blood and bile culture tests
were 37.2% and 90.7% in the resistant group and 39.1% and 97.8% in the susceptible
group, respectively. No significant differences were observed in the proportion of
patients with positive results on blood or bile culture tests. Escherichia coli, Klebsiella spp., and Enterococcus spp. were the most commonly detected pathogens. Enterococcus spp. in bile cultures was detected more frequently in the susceptible group than
in the resistant group.
Table 6 Laboratory findings from microbial cultures on initial admission.
|
Resistant group n=43
|
Susceptible group n=46
|
P value
|
|
Blood culture, n (%)
|
|
Sampling a blood culture
|
32 (74.4)
|
33 (71.7)
|
0.82
|
|
Positive rate
|
16 (37.2)
|
18 (39.1)
|
0.9
|
|
Pathogens
|
|
Escherichia coli
|
8 (18.6)
|
10 (21.7)
|
0.88
|
|
Klebsiella sp.
|
6 (14.0)
|
4 (8.7)
|
0.74
|
|
Enterococcus sp.
|
0 (0.0)
|
2 (4.3)
|
0.57
|
|
Enterobacter sp.
|
1 (2.3)
|
1 (2.2)
|
0.91
|
|
Citrobacter sp.
|
0 (0.0)
|
0 (0.0)
|
0.82
|
|
Streptococcus sp.
|
1 (2.3)
|
1 (2.2)
|
0.91
|
|
Pseudomonas sp.
|
0 (0.0)
|
0 (0.0)
|
0.82
|
|
Anaerobes
|
0 (0.0)
|
2 (4.3)
|
0.57
|
|
Others
|
1 (2.3)
|
1 (2.2)
|
0.91
|
|
Bile culture, n (%)
|
|
Sampling a bile culture
|
40 (93.0)
|
45 (97.8)
|
0.35
|
|
Positive rate
|
39 (90.7)
|
45 (97.8)
|
0.19
|
|
Pathogens
|
|
Escherichia coli
|
18 (41.9)
|
21 (45.7)
|
0.83
|
|
Klebsiella sp.
|
17 (39.5)
|
16 (34.8)
|
0.67
|
|
Enterococcus sp.
|
11 (25.6)
|
22 (47.8)
|
0.05
|
|
Enterobacter sp.
|
5 (11.6)
|
6 (13.0)
|
>0.99
|
|
Citrobacter sp.
|
2 (4.7)
|
6 (13.0)
|
0.27
|
|
Streptococcus sp.
|
10 (23.3)
|
5 (10.9)
|
0.16
|
|
Pseudomonas sp.
|
1 (2.3)
|
1 (2.2)
|
>0.99
|
|
Anaerobes
|
2 (4.7)
|
6 (13.0)
|
0.27
|
|
Other
|
5 (11.6)
|
2 (4.3)
|
0.26
|
The laboratory findings of the microbial cultures from the patients during the second
admission are summarized in [Table 7]. The proportions of patients with sampling blood and bile cultures were 44.2% and
97.7% in the resistant group and 47.8% and 100% in the susceptible group, respectively.
The proportions of patients with positive results on blood and bile culture tests
were 20.9% and 97.0% in the resistant group and 19.6% and 100% in the susceptible
group, respectively. No significant differences were observed in the proportion of
patients with positive results on positive blood or bile culture tests. Escherichia coli, Klebsiella spp., and Enterococcus spp. were the most common pathogens. Enterococcus sp. and Enterobacter sp. in bile cultures were more frequently detected in the resistant group than in
the susceptible group. Klebsiella spp. and anaerobes in bile culture were detected more frequently in the susceptible
group than in the resistant group.
Table 7 Laboratory findings of microbial cultures on second admission.
|
Resistant group n=43
|
Susceptible group n=46
|
P value
|
|
Blood culture, n (%)
|
|
Sampling a blood culture
|
19 (44.2)
|
22 (47.8)
|
0.83
|
|
Positive rate
|
9 (20.9)
|
9 (19.6)
|
0.85
|
|
Pathogens
|
|
Escherichia coli
|
4 (9.3)
|
4 (8.7)
|
0.95
|
|
Klebsiella sp.
|
1 (2.3)
|
4 (8.7)
|
0.51
|
|
Enterococcus sp.
|
3 (7.0)
|
2 (4.3)
|
0.78
|
|
Enterobacter sp.
|
0 (0.0)
|
1 (2.2)
|
0.91
|
|
Citrobacter sp.
|
0 (0.0)
|
0 (0.0)
|
0.83
|
|
Streptococcus sp.
|
0 (0.0)
|
0 (0.0)
|
0.83
|
|
Pseudomonas sp.
|
0 (0.0)
|
0 (0.0)
|
0.83
|
|
Anaerobes
|
1 (2.3)
|
1 (2.2)
|
0.91
|
|
Other
|
0 (0.0)
|
0 (0.0)
|
0.83
|
|
Bile culture, n (%)
|
|
Sampling a bile culture
|
42 (97.7)
|
46 (100.0)
|
0.48
|
|
Positive rate
|
42 (97.7)
|
46 (100.0)
|
0.48
|
|
Pathogens
|
|
Escherichia coli
|
16 (37.2)
|
22 (47.8)
|
0.39
|
|
Klebsiella sp.
|
12 (27.9)
|
24 (52.2)
|
0.03
|
|
Enterococcus sp.
|
32 (74.4)
|
20 (43.5)
|
<0.01
|
|
Enterobacter sp.
|
9 (20.9)
|
3 (6.5)
|
0.04
|
|
Citrobacter sp.
|
8 (18.6)
|
8 (17.4)
|
0.89
|
|
Streptococcus sp.
|
3 (7.0)
|
7 (15.2)
|
0.32
|
|
Pseudomonas sp.
|
5 (11.6)
|
3 (6.5)
|
0.37
|
|
Anaerobes
|
1 (2.3)
|
8 (17.4)
|
0.03
|
|
Other
|
2 (4.7)
|
8 (17.4)
|
0.09
|
Clinical features associated with the second hospitalization
[Table 8] summarizes the clinical features during the second admission. There was no significant
difference in the duration until recurrence of AC between the two groups. The administered
antimicrobials were less effective in the resistant group than in the susceptible
group (11.6% vs. 76.1%, P <0.01). However, no significant differences were found between the two groups in
terms of in-hospital mortality due to cholangitis or duration of hospitalization.
In the process of bacteria developing resistance to antimicrobials, it was more common
to observe a phenomenon of microbial substitution than the original causative bacteria
acquiring resistance directly.
Table 8 Clinical features associated with second hospitalization.
|
Resistant group n=43
|
Susceptible group n=46
|
P value
|
|
*One case overlapped.
†One patient did not undergo ERCP as he was in the terminal stage of cholangiocarcinoma.
AC, acute cholangitis; ERCP, endoscopic retrograde cholangiopancreatography; IQR,
interquartile range.
|
|
Duration to AC recurrence, median weeks (IQR)
|
16.0 (6.0, 40.5)
|
20.0 (8.0, 51.5)
|
0.63
|
|
Bile duct stent placement from initial admission to second admission
|
28 (65.1)
|
30 (65.2)
|
> 0.99
|
|
History of antimicrobial administration from the initial onset to the recurrence of
cholangitis, n (%)
|
14 (32.6)
|
9 (19.6)
|
0.23
|
|
Mechanisms of resistance to initial antimicrobials*, n (%)
|
|
Microbial substitution
|
39 (90.7)
|
|
|
|
The same bacteria as before resistance developed
|
6 (14.0)
|
|
|
|
Cases in which the antimicrobials used during the second admission were effective,
n (%)
|
5 (11.6)
|
35 (76.1)
|
< 0.01
|
|
Technical success of ERCP, n (%)
|
42/42† (100.0)
|
46/46 (100.0)
|
1
|
|
In-hospital mortality due to cholangitis, n (%)
|
1 (2.3)
|
0 (0.0)
|
0.48
|
|
Duration of hospitalization, median days (IQR)
|
8.0 (5.5, 13.5)
|
6.5 (5.0, 9.0)
|
0.21
|
Association between antimicrobial therapy within 3 days after ERCP and development
of bacterial resistance
Multivariate analysis showed that the use of antibiotics, completed within 3 days
after ERCP, was associated with a lower risk of developing bacterial resistance (odds
ratio [OR]: 0.17, 95% confidence interval [CI]: 0.04–0.65; P=0.01) ([Table 9]).
Table 9 Multivariate analysis of the risk factors for the development of bacterial resistance.
|
Resistant group n=43
|
Susceptible group n=46
|
Odds ratio
|
Confidence interval
|
P value
|
|
ERCP, endoscopic retrograde cholangiopancreatography; EST, endoscopic sphincterotomy;
PPI, proton pump inhibitor
|
|
Regular use of PPI, n (%)
|
20 (46.5)
|
21 (45.7)
|
1.64
|
0.56–4.77
|
0.36
|
|
First ERCP in their life, n (%)
|
21 (48.8)
|
13 (28.3)
|
1.12
|
0.18–6.92
|
0.91
|
|
EST performed during initial hospitalization, n (%)
|
16 (37.2)
|
10 (21.7)
|
1.04
|
0.15–7.02
|
0.97
|
|
History of antimicrobial administration from the initial onset to the recurrence of
cholangitis, n (%)
|
14 (32.6)
|
9 (19.6)
|
2.54
|
0.74–8.78
|
0.14
|
|
Antibiotics completed within 3 days after ERCP, n (%)
|
4 (9.3)
|
14 (30.4)
|
0.17
|
0.04–0.65
|
0.01
|
|
Piperacillin/tazobactam used on initial admission, n (%)
|
5 (11.6)
|
24 (52.2)
|
0.09
|
0.02–0.33
|
<0.01
|
PIPC/TAZ also emerged as a factor linked to a low incidence of resistant bacteria.
No significant differences were noted in other parameters such as “regular use of
PPI”, “first ERCP in their life”, “EST performed during the initial hospitalization”,
and “history of antimicrobial administration from the initial onset to the recurrence
of cholangitis.”
A sensitivity analysis was also performed using more independent variables, including
“age over 80,” “bile duct stent placement from initial admission to second admission”,
“residence nursing home”, and “cefmetazole used on initial admission.” We decided
to include cefmetazole, which was the most frequently used narrow-spectrum antimicrobial
agent in this study, in the sensitivity analysis. The results in this sensitivity
analysis mirrored those of the main analysis, with “antibiotics used within 3 days
after ERCP” and “PIPC/TAZ used on initial admission” extracted as independent inhibitors
(Supplementary file 1).
Discussion
This study primarily investigated whether a treatment duration shorter than the duration
recommended by the TG18 guidelines (4–7 days after ERCP) affected the development
of resistant pathogens that caused AC. The multivariate analysis included “antimicrobial
therapy within 3 days after ERCP” in addition to the established risk factors for
bacterial resistance development. The results showed a significantly low odds ratio
and suggested that a 3-day antimicrobial therapy or less during the initial AC may
have suppressed the development of resistant organisms when cholangitis recurred.
To the best of our knowledge, no study has examined the association between short-course
antimicrobial therapy during initial cholangitis and the development of resistant
organisms during recurrent cholangitis. Moreover, few previous studies reporting the
occurrence of resistant organisms in cholangitis have mentioned the duration of antimicrobial
therapy.
Reuken et al. investigated whether the total duration of antimicrobial therapy (14
days) for AC was associated with resistant organisms and concluded that it had no
discernible impact [12]. However, in other severe infections and ventilator-associated pneumonia, the incidence
of resistant organisms increased as the duration of antimicrobial therapy increased,
which was consistent with the findings of our study [16]
[17]
[18]
[19]. Another study reported that the hazard ratio for the emergence of resistant organisms
increased by 1.04 for every additional day of exposure to β-lactam antibiotics [19]. Therefore, the overutilization of antimicrobial agents contributes to the proliferation
of resistant organisms [35]. The absence of an association between the duration of antimicrobial administration
and the emergence of resistant organisms observed in Reuken et al.’s study may be
attributed to the extended predefined duration of “14 days.” Resistant organisms might
have emerged in groups with a treatment duration of less than 14 days, which was almost
as frequently as those with a treatment duration of 14 days or more.
Another novelty of this study is that the results of blood and bile cultures of patients
with recurrent AC were compared with the blood and bile cultures of those with initial
AC. Previous studies examining the relationship between AC and resistant organisms
only relied on the data from a single culture. To the best of our knowledge, no study
has compared the cultures obtained during the initial AC episode and those from recurrent
AC, as conducted in this study [12]
[13]
[14]
[15]. In bile cultures derived from the resistant group at the time of AC recurrence,
the rate of Enterococcus spp. detection was higher than that in bile cultures from the initial AC episode.
Consequently, if the duration of antibiotic administration following ERCP during the
initial episode of cholangitis exceeds 4 days, the likelihood of encountering Enterococcus spp. upon recurrence tends to be higher, compared with the cases where the administration
period is only ≤3 days. However, the number of patients in whom Enterococcus spp. was one of the treatment-resistant pathogens was insufficient for multivariate
analysis; therefore, a detailed study of Enterococcus spp. was not possible.
One major limitation of this study is its single-center, retrospective nature, which
led to insufficient adjustment for confounding factors. For example, the PIPC/TAZ
use rate was higher in the susceptible group. Based on the multivariate analysis,
PIPC/TAZ use was associated with a lower likelihood of resistant bacterial development.
However, the risk of bacterial resistance related to the use of other antimicrobials
has not been examined; to the best of our knowledge, no study has reported the difference
in the incidence of resistant bacteria associated with PIPC/TAZ use. A difference
in the incidence of resistant bacteria depending on the type of antimicrobial agent
used is possible [19]
[36]; however, several studies that compared antibiotics head to head found no evidence
that the type of antibiotic contributed to the development of bacterial resistance
[37]. Furthermore, although we confirmed that the patients were receiving regular PPI
prescriptions at the time of admission, we were unable to accurately assess the impact
of PPI due to the lack of data on the treatment duration or adherence prior to admission.
Therefore, a multicenter prospective study is needed to validate the results of this
study, a plan for which is in development.
Second, this study exceeded the commonly advised limit of one-tenth of the event count
for the number of independent variables included in the multivariate analysis, as
stipulated by the events per variable (EPV) rule. Specifically, our model incorporated
six independent variables against an event count of 43. This deviation from the EPV
rule potentially increases the risk of overfitting, which might lead to biased estimates
and may affect the generalizability and reliability of the findings. Despite this
limitation, we explored the complex relationships among the variables and to gain
preliminary insights. Future studies with a larger event count are warranted to validate
our findings.
Conclusions
In conclusion, the administration of antimicrobials within 3 days after ERCP may suppress
the development of resistant bacteria in patients with AC. Furthermore, in cases of
AC requiring more than 4 days of antimicrobial therapy, the possibility of Enterococcus spp. as the causative organism at the time of recurrence should be considered. In
addition, in alignment with the reports of previous studies on other infectious diseases,
clinicians treating cholangitis should avoid the unnecessary use of antimicrobial
agents.
Although the current international guideline, TG18, recommends a standard post-ERCP
antimicrobial therapy period of 4 to 7 days, several recent studies have shown that
clinical
outcomes are not inferior when the duration of therapy is 1 to 3 days. The present
study is
significant in that it presents the possibility that an antimicrobial treatment duration
of 3
days or less after ERCP may reduce the risk of the development of resistant pathogens.
