Prophylaxis of post-ERC infectious complications in patients with biliary obstruction by adding antimicrobial agents into ERC contrast media- a single center retrospective study

The presented study demonstrates several important findings that may give cause to modify the current practice of antibiotic prophylaxis to prevent ERC-related infectious complications. These include: (1) addition of antimicrobial agents into the ERC contrast media significantly reduces the incidence of post-ERC infection in patients with biliary obstruction; (2) combination of different antibiotics and antifungal regiments might be even more effective; (3) the benefit of local application of antimicrobials into obstructed bile ducts is most obvious if cholangitis is already present before ERC; (4) secondary sclerosing cholangitis represents the most eligible biliary disorder which takes particular profit from locally administered antimicrobials during ERC.

The routine administration of systemic antibiotic prophylaxis to all patients undergoing ERC has been left in favor of a selective use only in those patients with suspected or known biliary obstruction for whom complete endoscopic drainage will presumably not be achieved. This concerns particularly patients with hilar strictures and PSC [19]. Patients with post-transplant biliary strictures undergoing ERC represent other feasible candidates for systemic antibiotic prophylaxis [21]. Since systemically administered antibiotics poorly penetrate into the bile in case of biliary obstruction [20, 25, 26], a theoretical benefit of injecting antimicrobial agents directly into the bile tracts during ERC is assumable. Several studies have investigated the effect of antibiotics applied in contrast media on preventing post-ERC cholangitis with conflicting results. In vitro studies have proven that aminoglycosides retain their antibacterial properties when mixed to ERC contrast media. Thus, the aminoglycosides tobramycin and gentamicin efficiently eliminated common biliary bacteria such as E. coli, Klebsiella pneumonia, Proteus vulgaris and Pseudomonas aeruginosa [21, 27]. In line with these findings, we observed a significantly reduced post-ERC infection rate in patients with biliary obstruction when antimicrobial agents were added into the ERC media. Patients not receiving intraductal antibiotics into ERC contrast media exhibited a 2.33-fold increased risk to develop post-ERC cholangitis.

In contrast to our results, 3 prior prospective randomized-controlled studies could not demonstrate a beneficial effect on the rate of post-ERC infectious complications by adding antibiotics into ERC contrast media [23, 28, 29]. To explain these discrepancies, one has to take into account that only an aminoglycoside was used in the three studies, and that the analyzed study population strongly differed in matters of endoscopic procedures and subtype of biliary disorders. In the two randomized controlled studies published in 1980 and 1986 [28, 29], 51% of the study population underwent solely diagnostic ERC and did not exhibit any biliary disorder, whereas in our study all patients suffered from mainly severe obstructive biliary disease. In the most recent study [23] 114 patients with non-calculous biliary obstruction were enrolled, 57 of them receiving gentamicin 10 mg into ERC contrast media. In addition, all of them received a peri-interventional systemic antibiosis. Biliary obstruction was mainly caused by malignant strictures (79% of cases vs. 9.5% in our study), whereas in our study sclerosing cholangitis (75%) was the most prevalent cause. In contrast to our study, all patients underwent endoscopic biliary stenting (vs. 9.5% in our study). Biliary obstruction was relieved resulting in an adequate drainage in all patients, whereas in our study only in 49.3% of therapeutic ERC adequate drainage was achieved. In the mentioned study, no significant difference in the incidence of post-ERC cholangitis in each group with and without gentamicin added to contrast media (8.8% each) was detected. In contrast, in our study the incidence of post-ERC infection was significantly lower when adding antimicrobial agents into the ERC contrast media (14.3% vs. 33% in the control group; p = 0.045). The absolute risk reduction was 19% when adding antimicrobial agents into the ERC contrast media. We suggest that patients with secondary sclerosing cholangitis, who presented 52.4% of our study population, are particularly prone to post-ERC infectious complications. Presumably, the ERC-related infectious risk in these patients is even more pronounced than in patients with malignant strictures. Thus, 65% (13/20) of the patients with post-ERC infection suffered from SC in our study. SC is a chronic cholestatic biliary disease characterized by PSC-like biliary lesions apparent on ERC, namely multifocal biliary strictures with interposed normal or dilated bile ducts [24]. The most frequent cause (72.7%) of SC in our study population was SC in critically ill patients (SC-CIP). SC-CIP is an emerging disease entity with unfavorable outcome, mostly observed in patients who have survived life-threatening illnesses and who received aggressive treatment on intensive-care units. The median survival of patients with SC-CIP who are not liver-transplanted was reported to be only 13 months [30]. ERC reveals severe bile-duct damage with extensive biliary casts and multiple irregular strictures. Recurrent episodes of bacterial cholangitis are typically observed in patients with SC [31, 32]. In 68.2% of our patients with SC, bile fluid was tested positive for bacteria before ERC with no statistical difference between the two study groups. However, infectious complication rate after ERC was significantly higher in patients with SC not given antibiotics into the contrast media (50% vs. 9%; p = 0.03). In these SC-patients, ERC-related infectious complications were observed even despite a systemic antibiotic treatment. Regarding the other subgroups of the study population, addition of antibiotics to the contrast media seemed to have no effect on the post-ERC infectious rate, although patient numbers are too small for valid statistical analysis.

Patients presenting with fever or elevated leucocytes prior to ERC were excluded from all previous studies that evaluated the benefit of intraductal antibiotics on post-ERC complications [23, 28, 29]. In contrast, 59.5% of our study population suffered already at the time of ERC from acute cholangitis (defined as bacterial colonization of the biliary system and elevated leucocytes > 12 000/μl/CRP > 5 mg/l). Acute cholangitis at the time of ERC was present in both, the case- and control group without statistical difference. Injection of ERC contrast media into obstructed and infected bile tracts will most likely result in bacteremia [8]. This will particularly be the case when complete biliary drainage is not achieved by ERC. On the other hand, addition of antimicrobial agents to the ERC contrast media should reduce biliary bacterial growth and decrease the risk of bacteremia. Indeed, acute cholangitis, present at the time of ERC, was calculated as a risk factor for developing post-ERC infectious complications in our study. Hence, the risk of infectious complications after ERC was 2.72-fold increased when acute cholangitis was present compared to patients without cholangitis. The absolute risk reduction was 29,3% in patients with cholangitis when adding antibiotics to contrast media. In line with our data, Motte et al. identified leukocytosis and prior cholangitis as significant risk factors for septicemia following endoscopic biliary stenting of biliary obstruction [6].

Most patients in our study received a combination of antimicrobial agents into the ERC contrast media. Only 16.6% received solely gentamicin, as used in the previous studies [23, 29]. The most frequently isolated organism in bile samples taken from patients given intraductal antibiosis were gram-positive with Enterococcus spp. found in 71.4%. Gram-negative organisms found in the collected bile samples were E. coli in 25.7%, Klebsiella spp. in 11.4% and Pseudomonas spp. in 11.4%. Of note, only in 10.7% of positive bile cultures, all detected bacterial strains were sensitive to gentamicin. Combination of gentamicin with vancomycin increased the response rate to 89.3%. These data question the effectiveness of adding solely gentamicin into ERC contrast media for prevention of post-ERC infectious complications. Instead, the choice of the administered antibiotic regiments should be based on the sensitivity of the isolated bacteria and the local pattern of antibiotic resistance. Noteworthy, we found Candida species in 25.7% of the fungal cultures of taken bile samples. All Candida species were sensitive to fluconazole. Candida spp. were shown to be predominantly detected in bile fluids of patients with primary and secondary sclerosing cholangitis, immunosuppressive therapy, after placement of plastic biliary stents, and after liver transplantation [33‐36]. Our data on fungal bile cultures are in line with these findings, as our study population comprises all the mentioned entities above. In conclusion, collection of bile fluid during ERC for microbiological analysis should be considered in all patients with a high risk for post-ERC infectious complications. When adding antimicrobial agents into ERC contrast media, we recommend a combination of antibiotic and antimycotic agents instead of mono-therapy suggesting a more potent effect on preventing post-ERC infectious complications.

The main limitations of our study are the retrospective study design and the rather small number of patients. Moreover, the combination of antimicrobial agents added to the contrast media was not standardized in a uniform protocol, but was recommended to the respective investigator. This explains the number of patients receiving solely gentamicin, or an antibiotic regiment without antimycotic agents. Despite these limitations, our data are of particular interest for the clinical practice of antibiotic prophylaxis in ERC. This is the first study that evaluates ERC-related infectious complications in patients with SC. Pre-procedural cholangitis and incomplete endoscopic drainage due to multifocal biliary strictures are common findings in patients with SC, defining them as a high risk-population for post-ERC infectious complications. Injection of ERC contrast media might increase the intraductal pressure and incomplete drainage of already infected bile might then facilitate bacteremia in SC. A benefit of locally applied antibiotic agents is therefore highly assumable. Our preliminary data should now be prospectively evaluated in a larger patient cohort to improve the safety of ERC, especially in patients with SC.