8 Antibiotic Prophylaxis in Endoscopy
Mouen A. Khashab and Brooks D. Cash
8.1 Introduction
Antibiotic prophylaxis in endoscopy plays a critical role in minimization of infectious complications associated with endoscopic procedures. Although bacteremia is relatively common after both diagnostic and therapeutic procedures, the incidence of infectious endocarditis is very low. Therefore, administration of prophylactic antibiotics solely for the prevention of infectious endocarditis is not recommended for patients undergoing endoscopic procedures. Antibiotic prophylaxis may have an important role in the prevention of infectious complications resulting from certain endoscopic procedures and in specific clinical settings. These include endoscopic retrograde cholangiopancreatography (ERCP) in patients with biliary obstruction and incomplete ductal drainage, ERCP in liver transplant patients, endoscopic ultrasound (EUS)-guided fine needle aspiration (FNA) of cystic lesions, percutaneous endoscopic gastrostomy (PEG), and endoscopy in cirrhotic patients presenting with gastrointestinal (GI) bleeding, among others.
Bacterial translocation of GI microbial flora into the bloodstream may occur during endoscopy due to procedure-related trauma. Resulting bacteremia carries a minor risk of localization of infection in distant tissues. In addition, endoscopy may also result in local infections where a typically sterile space or tissue is breached and contaminated by an endoscopic accessory or by contrast injection. In this chapter, infectious complications related to endoscopy and the role of periprocedural antibiotic prophylaxis for the prevention of these complications are presented.
8.2 Bacteremia Related to Endoscopic Procedures
Bacteremia can arise following endoscopic procedures and is considered a surrogate marker for infective endocarditis (IE) risk. However, clinically important infections arising from GI endoscopy are uncommon, with only 25 cases of IE reported with temporal association to an endoscopic procedure. 1 , 2 , 3 Additionally, no data exist that validate a causal association between endoscopic procedures and IE. Likewise, there are no data indicating that antibiotic prophylaxis preceding endoscopic procedures prevents the occurrence of IE.
8.2.1 Procedures Associated with Low Risk of Bacteremia
Bacteremia after gastroscopy and colonoscopy occurs in approximately 4% of patients and is not associated with infectious complications. 4 , 5 , 6 Bacteremia is even uncommon (6.3%) with therapeutic colonic procedures such as colonic stent insertion for colonic obstruction. 7
The frequency of bacteremia after EUS, with or without FNA, is similar to that of upper endoscopy. Prospective studies in patients undergoing EUS-FNA of cystic or solid lesions along the upper GI indicate a bacteremia rate of 4.0 to 5.8%. 8 , 9 , 10 , 11 Likewise, EUS-FNA of solid rectal and perirectal lesions is associated with a low risk of bacteremia, with one study reporting a risk of 2%. 12
8.2.2 Procedures Associated with High Risk of Bacteremia
The highest rates of bacteremia have been reported with esophageal dilation, sclerotherapy of esophageal varices, and ERCP in patients with biliary obstruction. The rate of bacteremia following esophageal dilation ranged between 12 and 22% in three prospective trials, 13 , 14 , 15 and may be higher with dilation of malignant strictures 14 and with passage of multiple dilators. 14 The cultured organisms are typically commensal to the mouth. In one study, viridans streptococci was the organism isolated in 79% of cases. 13
The rate of bacteremia associated with variceal sclerotherapy is approximately 15%, 16 , 17 , 18 , 19 while that associated with endoscopic variceal ligation is about 9%. 20 , 21 , 22
ERCP in patients with a nonobstructed biliary tree is associated with a relatively low rate of bacteremia of 6%, rising to 18% in the setting of biliary obstruction due to stones or tumors. 23
8.3 Antibiotic Prophylaxis for the Prevention of Infective Endocarditis
The American Heart Association (AHA) recommends that administration of prophylactic antibiotics merely for the prevention of IE is not recommended for patients undergoing GI endoscopic procedures. 24 This recommendation is due to the absence of data demonstrating a conclusive link between endoscopic procedures and the development of IE, in addition to a lack of evidence that antibiotic prophylaxis prevents IE following endoscopy. 24 Similarly, the American Society for Gastrointestinal Endoscopy (ASGE) recommends against the routine administration of antibiotic prophylaxis solely for prevention of IE. 25
The AHA specifically recognized cardiac conditions associated with the highest risk of poor clinical outcome from IE, including: (1) prosthetic (mechanical or bioprosthetic) cardiac valve; (2) history of previous IE; (3) cardiac transplant recipients who develop cardiac valvulopathy; and (4) patients with congenital heart disease (CHD), including (a) those with unrepaired cyanotic CHD including palliative shunts and conduits, (b) those with completely repaired CHD with prosthetic material or devices, placed surgically or by catheter, for the first 6 months after the procedure, and (c) those with repaired CHD with residual defects at the site or adjacent to the site of a prosthetic patch or device. 24 The AHA suggests administration of antibiotics with coverage of enterococci in patients with these specific cardiac conditions who also have established infections of the GI tract where enterococci may be part of the infecting bacterial flora (such as cholangitis). 24 Although resulting infections are likely to be polymicrobial, coverage for enterococci is recommended because only enterococci are likely to result in IE.
8.3.1 Antibiotic Prophylaxis for the Prevention of Procedural-Related Infections (Other Than IE)
Antibiotic prophylaxis may have an important role in the prevention of infectious complications resulting from certain endoscopic procedures and in specific clinical settings.
ERCP
Patients with acute cholangitis are typically treated with antibiotics and biliary drainage via ERCP. 26 Additional antibiotic administration for procedural prophylaxis is not recommended. Cholangitis and sepsis are known complications of ERCP, occurring in up to 3% of patients. 27 , 28 , 29 , 30 , 31 , 32 , 33 , 34 Antibiotic prophylaxis has been shown to reduce the incidence of bacteremia associated with ERCP, 35 , 36 but this does not clearly translate into a reduction in the incidence of cholangitis. A recent Cochrane systematic review that comprised 9 randomized clinical trials and 1,573 patients established that prophylactic antibiotics reduced bacteremia and seemed to prevent cholangitis and septicemia in patients undergoing elective ERCP. 37 However, in the subgroup of patients with uncomplicated ERCP, the effect of antibiotics was less prominent. 37 In one study, incomplete biliary drainage was predictive of 91% of all cases of sepsis. 38 Antibiotic therapy may therefore have particular value where drainage achieved at ERCP is incomplete or achieved with difficulty, such as with hilar cholangiocarcinoma and primary sclerosing cholangitis. 38 , 39 , 40 In one of the few trials that indicated a benefit of antibiotics in patients undergoing ERCP for biliary obstruction, prophylactic antibiotics were continued postprocedurally for several days. 41 The ASGE recommends against antibiotic prophylaxis prior to ERCP where obstructive biliary tract disease is not suspected or when complete biliary drainage is expected. 25 On the other hand, the ASGE recommends that antibiotic prophylaxis be administered prior to ERCP in patients following liver transplantation and/or known or suspected biliary obstruction, where there is a possibility that complete biliary drainage may not be achieved. 25 Antibiotics should also be continued postprocedure when biliary drainage is incomplete.
There are no studies that have assessed the value of antibiotic prophylaxis in patients undergoing ERCP who have pancreatic cystic lesions that communicate with the main pancreatic duct. However, the incidence of infectious complications in this setting seems to be uncommon given that ERCP is commonly performed in patients with such cystic lesions (e.g., intraductal papillary mucinous neoplasms, pseudocysts) without reports of cyst infections.
Acute cholecystitis may result from placement of biliary self-expandable metallic stents (SEMSs) and is believed to be due to cystic duct obstruction. This occurs in 2 to 12% of cases. 42 In two meta-analyses, the incidence of cholecystitis was similar between covered and uncovered SEMSs. 43 , 44 The majority of reported cases of cholecystitis after biliary SMES placement occurred in patients with malignant biliary obstruction, 42 and tumor involvement of the cystic duct orifice is an independent risk factor for acute cholecystitis after SEMS insertion. 45 The role of prophylactic antibiotics has not been studied, but may help prevent this complication, especially since surgery is frequently needed to manage this untoward event.