Rosario Forlano

Incidence rates of post-ERCP complications: A systematic survey of prospective studies

OBJECTIVES:To provide health-care providers, patients, and physicians with an exhaustive assessment of prospective studies on rates of complications and fatalities associated with endoscopic retrograde cholangiopancreatography (ERCP).METHODS:We searched MEDLINE (1977–2006) for prospective surveys on adult patients undergoing ERCP. “Grey literature” was sought by looking at cited references to identify further relevant studies. Data on postprocedural pancreatitis, bleeding, infections, perforations, and miscellaneous events as well as their associated fatalities were extracted independently by two reviewers. Sensitivity analysis was performed to test for data consistency between multicenter versus single center studies, and old (1977–1996) versus recent (1997–2005) reports.RESULTS:In 21 selected surveys, involving 16,855 patients, ERCP-attributable complications totaled 1,154 (6.85%, CI 6.46–7.24%), with 55 fatalities (0.33%, CI 0.24–0.42%). Mild-to-moderate events occurred in 872 patients (5.17%, CI 4.83–5.51%), and severe events in 282 (1.67%, CI 1.47–1.87%). Pancreatitis occurred in 585 subjects (3.47%, CI 3.19–3.75%), infections in 242 (1.44%, CI 1.26–1.62%), bleeding in 226 (1.34%, CI 1.16–1.52%), and perforations in 101 (0.60%, CI 0.48–0.72%). Cardiovascular and/or analgesia-related complications amounted to 173 (1.33%, CI 1.13–1.53%), with 9 fatalities (0.07%, CI 0.02–0.12%). As compared with old reports, morbidity rates increased significantly in most recent studies: 6.27% versus 7.51% (Pc = 0.029).CONCLUSIONS:ERCP remains the endoscopic procedure that carries a high risk for morbidity and mortality. Complications continue to occur at a relatively consistent rate. The majority of events are of mild-to-moderate severity.

Pancreatic Duct Stents in the Prophylaxis of Pancreatic Damage after Endoscopic Retrograde Cholangiopancreatography: A Systematic Analysis of Benefits and Associated Risks

Methods: The efficacy of pancreatic stenting in the prevention of pancreatitis following endoscopic retrograde cholangiopancreatography (ERCP) was evaluated by a meta-analysis of 6 controlled studies; 12 additional uncontrolled studies were analyzed for rates of associated risk. Results: Post-ERCP pancreatitis (PEP) developed in 16.5% of controls, and in 5.1 or 9.6% of the stent group at the per-protocol (PP) or intention-to-treat (ITT) analyses. By analyzing only the 4 randomized trials, PEP developed in 24.1% of controls, and in 6.1 or 12.0% of the stented patients at the PP or ITT analyses. Risk was significantly lower in the stent group when compared with controls: OR 0.44 (95% CI 0.24–0.81). The absolute risk reduction is 12.0 (95% CI 3.0–21.0), the number needed to treat 8 (95% CI 5–34), and the publication bias 2. ORs for mild to moderate PEP were reduced in the stent group (OR = 0.537, 95% CI 0.283–1.021), as were those for severe PEP (OR = 0.123, 95% CI 0.021–0.726). Non-pancreatic complications were 4.2%, and included early stent migration (1.4%), perforations (0.4%), bleeding (1.4%), and infections (1.0%). Conclusion: Available trials show benefit for pancreatic stenting in the prophylaxis of PEP, but more randomized studies are needed before endorsing a routine use of this endoscopic procedure.

Fatal systemic venous air embolism during endoscopic retrograde cholangiopancreatography.

Hepatic portal venous air embolism is the rarest complication of gastrointestinal endoscopy, resulting from penetration of gas into the portal veins, and may occur during endoscopic retrograde cholangiopancreatography and endoscopic biliary sphincterotomy. The likely mechanism is intramural dissection of insufflated air into the portal venous system through duodenal vein radicles transected during the procedure. Hepatic portal air embolism may be fatal. Cerebral air embolism may also occur. So far 13 cases of air embolism after endoscopic retrograde cholangiopancreatography have been reported, with 4 cases of systemic spread that proved fatal. Death was due to pulmonary air embolism in 2 cases, and cerebral air embolism in another 2. We report on an additional such fatal case, concerning a 78-year-old male patient, who several years previously had undergone surgical gastroduodenal resection with cholecystectomy and papillotomy, and was admitted for recurrent ascending cholangitis secondary to bile duct stones. During the third endoscopic cholangioscopic procedure for removal of bile duct stones, sudden cardiopulmonary arrest occurred. Death was due to massive pulmonary air embolism. Cerebral air embolism was also found. Autopsy was performed. A spontaneous duodenobiliary fistula was found. On the basis of bench radiologic investigation (retrograde suprahepatic venography and anterograde portography), it was demonstrated that the air insufflated during duodenal endoscopy, which entered through the spontaneous duodeno-biliary fistula, penetrated into intrahepatic vein radicles injured secondarily to prolonged impaction of biliary sand and stones and infection, resulting in portal and hepatic venous gas and systemic air embolism.