ORIGINAL ARTICLE

Endoscopic Retrograde Cholangiopancreatography Is an Effective Treatment for Bile Leak After Severe Liver Trauma
Rahul J. Anand, MD, Paula A. Ferrada, MD, Peter E. Darwin, MD, Grant V. Bochicchio, MD, and Thomas M. Scalea, MD

Background: Biliary leak after severe hepatic trauma is a complex problem requiring multidisciplinary care. We report on our experience with endoscopic management of posttraumatic bile leaks and clarify the role of endoscopic retrograde cholangiopancreatography (ERCP). Methods: A retrospective analysis was performed on all patients who sustained liver injury and underwent ERCP from September 2003 to September 2009. Patients who had associated biliary leak were identied. Patient demographics, injury characteristics, liver operations, endoscopic treatment, and success of endoscopic intervention were reviewed. Liver injury was managed in an interdisciplinary fashion, including immediate or delayed operation or angiography or both for primary or adjunctive hemostasis. ERCP with stenting and sphincterotomy was used to treat biliary stulae. Sequelae of liver injury including biloma or other perihepatic uid collection were also managed by computed tomography scan-guided or ultrasound-guided drainage. Results: A total of 26 patients underwent ERCP for the management of biliary stula as a result of severe hepatic trauma. There were 14 (54%) blunt injuries. In every patient (100%), ERCP with stenting and sphincterotomy was successful in controlling bile leak. All patients eventually had removal of stents and drains, with resolution of leak. Two patients had concomitant treatment of associated pancreatic ductal injury. Conclusion: ERCP is useful as both a diagnostic and therapeutic tool for the safe treatment of biliary ductal injuries after severe liver trauma and should be part of a multidisciplinary treatment algorithm. Key Words: Endoscopic retrograde cholangiopancreatography (ERCP), Liver trauma, Bile leak. (J Trauma. 2011;71: 480 485)

phy (ERCP) with ampullary sphincterotomy and biliary stent insertion is useful in the management of biliary injury resulting from laparoscopic cholecystectomy, after elective hepatic surgery, and even after pancreatic trauma.4 6 Small series report the use of therapeutic ERCP in the management of traumatic bile leaks.7 Bridges et al.8 support the use of endobiliary stenting, whereas others support endoscopic placement of nasobiliary tubes.9 In the past 6 years, we have used ERCP to treat traumatic biliary stula after both blunt and penetrating hepatic injury. We retrospectively reviewed our experience with endoscopic management of traumatic bile leaks and injuries to analyze the role of ERCP in the management of this complex problem.

PATIENTS AND METHODS

The R Adams Cowley Shock Trauma database from September 2003 to September 2009 was reviewed to identify all patients during a 6-year interval who underwent ERCP. All charts were retrospectively reviewed after Institutional Review Board approval was obtained. Demographic data collected on the patients included age, gender, length of stay, and mechanism of injury. Liver injuries were graded using the American Association for the Surgery of Trauma Organ Injury Scale for hepatic injuries.10 Liver injury was managed in an interdisciplinary fashion, including immediate or delayed operation or angiography or both for primary or adjunctive hemostasis. Bile leak was suspected in patients if they had persistent or high-volume perihepatic drain output, along with the data from adjunctive imaging such as computed tomography (CT) scan. Hepatobiliary iminodiacetic acid (HIDA) scan was also used to diagnose and/or conrm the resolution of the bile leak. During the study period, ERCP with stenting and sphincterotomy was used as our preferred method to treat biliary stulae. Sequelae of liver injury including biloma or other perihepatic uid collection were also managed by CT scan-guided or ultrasound-guided drainage. ERCP reports were reviewed to identify location of leak and for interventions performed. Clinic records were also reviewed for patient follow-up. Success of ERCP for the treatment of biliary leak after liver injury was dened as resolution of bile leak, with removal of stents or any perihepatic drains. Leaks were dened as postoperative if on ERCP report location of the leak was consistent with a resection margin or with a hepatorrhaphy site. Leaks were dened as posttraumatic if the bile leak occurred in the absence of operative management or if biliary injury on ERCP was obviously caused by the index trauma. Resolution of bile leak

onoperative management of liver trauma has become the standard of care for hemodynamically stable patients, and the use of damage control has become standard for those who are hemodynamically unstable.1 Improvements in transport to trauma centers, and ICU care have increased survival in patients suffering from severe liver trauma, making secondary bile leak from the intrahepatic biliary tree a more common complication.2,3 Biliary leaks complicate liver trauma with a frequency of 0.5% to 21%.2,3 Endoscopic retrograde cholangiopancreatograSubmitted for publication April 7, 2010. Accepted for publication June 30, 2010. Copyright 2011 by Lippincott Williams & Wilkins From the R Adams Cowley Shock Trauma Center, University of Maryland School of Medicine, Baltimore, Maryland. Address for reprints: Thomas M. Scalea, MD, R Adams Cowley Shock Trauma Center, University of Maryland Medical System, 22 South Greene Street, Baltimore, MD 21201-1595; email: tscalea@umm.edu. DOI: 10.1097/TA.0b013e3181efc270

480

The Journal of TRAUMA Injury, Infection, and Critical Care Volume 71, Number 2, August 2011

The Journal of TRAUMA Injury, Infection, and Critical Care Volume 71, Number 2, August 2011

ERCP After Liver Trauma

was documented on follow-up ERCP or HIDA scan. Data are presented as mean standard deviation.

RESULTS
During the study period, 36,076 trauma patients were admitted, in which 1,548 patients had liver injury. Thirty-one patients (2%) with severe liver injuries had an ERCP for treatment of suspected biliary stula. Of these 31 patients, 5 patients had a normal biliary anatomy or no bile leak on ERCP. These patients were excluded from the study, leaving a study population of 26 patients. The study population included 21 men (70%), with an average age of 29 years. Fourteen patients (54%) sustained blunt trauma, whereas the remainder sustained penetrating trauma. Mean hospital length of stay was 33 days 21 days. ERCP was successful in all patients, and in all patients, bile leaks resolved after ERCP and associated treatment, eventually allowing for drains and stents to be removed (Fig. 1). Of the 26 patients, 24 (92%) were primarily managed operatively with bile leak suspected based on high output from perihepatic drains placed either at operation or percutaneously. These patients went on to get ERCP for therapeutic intervention. One patient (patient 18, Table 1) was managed nonoperatively after assault, and the bile leak was suspected based on high output from a percutaneously placed drain. Another patient (patient 5) was managed nonoperatively after motor vehicle crash, and the bile leak was suspected based on CT imaging after a positive nding of the HIDA scan. In all patients except one (25 of 26), ERCP occurred within 3 weeks of the operative exploration or injury. In two patients (patients 14 and 25), ERCP occurred 24 hours after the index trauma. In these cases, there was high-volume bile leak either in the operating room or in the immediate postoperative period. In the last patient (patient 4), ERCP occurred 8 months after the index trauma. It was difcult to ascertain whether the bile leaks were postoperative or posttraumatic. In the penetrating group, 58% (7 of 12) of patients had bile leaks related directly to the index trauma as opposed to postoperative complications after surgery. Twelve of the 14 patients with blunt trauma had an exploration before ERCP. Of these, 10 had postoperative leaks. ERCP therapy resulted in decreased perihepatic drain output. On average, ERCP resulted in a 72% decrease in perihepatic drain output within 2 days of stenting and sphincterotomy. In 19 of 26 patients, follow-up ERCP was conducted within 2 months of the rst ERCP to show closure of the biliary stula. In 23 of 26 patients, follow-up ERCP within 3 months of the rst ERCP showed closure of the stula with an average of 47 days. In all patients, biliary stula was closed within 7 months of ERCP. No patient suffered complication as a result of ERCP, including bleeding or perforation. Amylase and lipase were not routinely checked after the ERCP to monitor for periprocedural pancreatitis. In the blunt group, 50% (7 of 14) of patients underwent a pancreatogram in addition to cholangiogram. The same was true in the penetrating group (6 of 12). The decision to perform pancreatogram was largely operator dependentalthough in a few cases, the pancreatic duct was
2011 Lippincott Williams & Wilkins

Figure 1. Patient 17A 51-year-old man status after motor vehicle crash and hepatorrhaphy. Bile leak was suspected based on high postoperative bilious Jackson-Pratt drain output and HIDA scan. Initial ERCP (A) confirmed bile leak from right hepatic ductal system (black oval). Patient underwent sphincterotomy and placement of a 10-Fr biliary stent (B, white oval). Follow-up ERCP 1 month later demonstrated resolution of leak.

cannulated and injected to aid in successful cannulation of the biliary system.

Blunt Trauma
Of the blunt trauma patients (Table 1), 10 were involved in motor vehicle crashes, 2 sustained crush injuries, 1 was a pedestrian struck by a car, and 1 was an assault. Most
481

482
Hepatic Interventions After ERCP Stents Placed Biliary 7 Fr, pancreatic 7 Fr 1 122 Yes ERCP Liver resection bed Biliary Leak Location Follow-Up (d) Bile Leak Resolved? IR drainage of liver abscess Biliary 10 Fr, pancreatic 5 Fr, sphincterotomy; biliary 10 Fr Biliary 10 Fr 2 sphincterotomy Biliary 10 Fr, sphincterotomy Biliary 10 Fr, sphincterotomy 2 2 2 451 302 800 4 143 Yes Pancreatic duct; right intrahepatic BD, right main hepatic duct Right lobe Left hepatic, cystic ducts Right intrahepatic BD Yes Yes Yes Right hepatic lobectomy Cholecystectomy Right hepatic lobectomy Drainage of subhepatic abscess Right hepatic lobectomy Right intrahepatic BD Right intrahepatic BD Right intrahepatic BD Biliary 10 Fr, sphincterotomy Biliary 10 Fr, sphincterotomy Biliary 10 Fr, sphincterotomy 2 776 Yes 2 408 Yes 1 715 Yes Right intrahepatic BD Right intrahepatic BD Biliary 10Fr, sphincterotomy Biliary 10Fr, sphincterotomy 3 88 Yes Ex laparotomy, washout Ultrasound drain of hepatic abscess IR drainage of perihepatic liver abscess/stula to small bowel 2 114 Yes Right intrahepatic BD Cystic duct ? Intrahepatic Right intrahepatic BD Biliary 10 Fr, sphincterotomy Biliary 10 Fr, sphincterotomy Biliary 10 Fr, sphincterotomy Biliary 10 Fr, sphincterotomy 1 151 Yes 2 2 2 1096 97 35 Yes Yes Yes

TABLE 1.

Characteristics of Blunt Group

Anand et al.

Patient/Age (yr)/Sex

Mechanism

Hepatic Interventions Before ERCP

1/19/M

MVC vs. pedestrian

2/17/F

MVC

IRcoil right hepatic artery Right hepatic lobectomy IR drainage of perihepatic uid HIDAleak Ex laparotomy hepatorrhaphy, cholecystectomy

5/31/F

MVC

6/19/M

MVC

8/23/M

Crush

9/18/M

MVC

IR liverno bleed HIDAbiliary leak Ex laparotomy hepatorrhaphy IR embolization right hepatic artery Ex laparotomy hepatorrhaphy Gelfoam embolization of right and accessory right hepatic artery, embolization of pseudoaneurysm IR embolization right hepatic Gelfoam and coil

10/24/M

MVC

11/25/M

MVC

14/39/F

MVC

Ex laparotomy for compartment syndrome Ex laparotomy hepatorrhaphy Gelfoam embolization of right hepatic artery Right hepatic partial lobectomy Ex laparotomy hepatorrhaphy Inadvertent right hepatic a coil embolization Cholecystectomy Right hepatic lobectomy HIDAleak OSH Ex laparotomy hepatorrhaphy Ex laparotomy hepatorrhaphy

17/51/M

MVC

OSH Ex laparotomy hepatorrhaphy HIDAbiliary leak

18/35/M

Assault

21/53/M

Crush

24/55/F

MVC

25/18/F

MVC

IR Gelfoam embolization of branch right hepatic artery Drainage of biloma Ex laparotomy, damage control Cholecystectomy Ex laparotomy, partial hepatic lobectomy HIDAbiliary leak Ex laparotomy, hepatorrhaphy

The Journal of TRAUMA Injury, Infection, and Critical Care Volume 71, Number 2, August 2011

2011 Lippincott Williams & Wilkins

MVC, motor vehicle crash; IR, interventional radiology; OSH, outside hospital; BD, bile duct.

The Journal of TRAUMA Injury, Infection, and Critical Care Volume 71, Number 2, August 2011

ERCP After Liver Trauma

(13 of 14) of these liver injuries were high grade on CT, ranging from grades 3 to 5. All of these patients underwent some intervention to treat the liver injury before ERCP was used to control biliary stula. Eight of these patients went to interventional radiology either in addition to or instead of surgery, with seven undergoing placement of either Gelfoam (Pzer, New York, NY) or coil embolization devices to the right hepatic artery. Interestingly, six of these patients (86%) ultimately went on to have formal hepatic lobectomy or partial lobectomy secondary to liver necrosis, suggesting considerable morbidity associated with selective arterial embolization in the treatment of blunt hepatic injury. One of the patients (patient 11) underwent right hepatic lobectomy after inadvertent placement of a coil into the common hepatic artery contributing to the need for right hepatic lobectomy, whereas another (patient 9) developed abdominal compartment syndrome after nonoperative management of his grade 4 liver laceration. One patient (patient 8) underwent concomitant treatment of a hepatic vein pseudoaneurysm. One patient (patient 21) developed bile leak after cholecystectomy. This was done as part of a damage control strategy for severe abdominal crush injury. Twelve patients, two at an outside hospital, also underwent abdominal exploration with hepatorrhaphy before ERCP. HIDA scan was used in ve patients before ERCP as an aid to the diagnosis of biliary stula. Six patients also had either inerventional radiology- or ultrasound-assisted drainage of perihepatic uid collections secondary to blunt liver injury, including one patient (patient 17) with a stula to the small bowel. ERCP successfully dened the location of leak in all but two (85%) of the patients and every patient who had a biliary leak shown on ERCP underwent sphincterotomy and placement of (10 Fr.) biliary stents. Patient 1 did not have a leak anatomically dened on ERCP but was shown to have a leak on preprocedure HIDA scan. Leak from liver resection bed was suspected at ERCP, due to bilious drainage from an operative drain. This biliary leak resolved after placement of a 7-Fr biliary stent. Similarly, patient 24 also clinically had a leak, although on ERCP, the precise anatomic distribution could not be dened. This patient also underwent placement of a 10-Fr biliary stent. All patients in the blunt trauma group had resolution of stula demonstrated on follow-up ERCP or HIDA. All patients had successful removal of indwelling biliary stents and also removal of perihepatic drains as outpatients. Two patients (patients 1 and 2) had simultaneous treatment of pancreatic stula at the same time as biliary stenting. Patients in the blunt trauma group required on average two ERCPs for treatment of biliary stula, with all eventually having their stent removed or passed. One patient (patient 2) underwent four ERCPs with placement of one pancreatic duct stent and two biliary stents in addition to sphincterotomy.

tient 4). Four patients underwent subsequent embolization of the right hepatic artery, with one (patient 7) undergoing the procedure to treat a hepatic artery to portal vein stula. Similar to the blunt trauma group, two patients required subsequent lobectomy or partial lobectomy after arterial embolization secondary to hepatic necrosis. Interventional radiology or ultrasound-guided techniques were used to drain perihepatic uid collections in six patients. In seven patients, HIDA was used to aid in diagnosis of biliary leak. In all the patients, ERCP detected bile leak although in one (patient 4)the location was not clearly dened. All patients were treated with sphincterotomy, and in all the patients where a specic ductal injury was dened (92%), a biliary stent was placed. One patient (patient 22) had successful treatment of biliary leak through a combination of ERCP and percutaneous transhepatic biliary stenting but ultimately required biliary reconstruction secondary to stricture. Patients in the penetrating trauma group required on average 2.2 ERCPs for treatment of biliary stula. The leak resolved in all patients. All patients had stents and perihepatic drains removed. One patient (patient 15) had a total of four ERCPswith demonstration of a biliary vascular stula at the second procedure, which was successfully treated with placement of biliary stents. Interestingly, HIDA scan disclosed nothing abnormal before ERCP in two patients (patients 23 and 26). ERCP was obtained in these patients because of high suspicion for biliary leak based on high output from perihepatic drains.

DISCUSSION
This report demonstrates that in patients with traumatic biliary injury, ERCP is a useful tool. All patients treated with ERCP (100%) in the study had eventual resolution of their biliary leak. Previous signicantly smaller retrospective reviews have documented the use of ERCP in the treatment of traumatic biliary injury.7,11,12 There are also scattered case reports documenting the feasibility of this technique.13,14 The current series is the largest series documenting the use of this modality in the treatment of traumatic biliary stula in both blunt and penetrating hepatic injuries. ERCP is well documented as a treatment for pancreatic stula after trauma.6 Indeed, in the current series, two patients underwent simultaneous treatment of pancreatic stula at time of biliary intervention, highlighting the versatility of this technique. Importantly, no patient suffered untoward complication as a result of ERCP including bleeding or perforation. We believe this is a safe technique. All patients with an anatomically dened injury underwent a combination of sphincterotomy and biliary stenting because this is the technique of choice for treatment at our institution. This combination allows the reduction of the pressure gradient between the bile duct and the duodenum by bypassing the sphincter of Oddi. Therefore, bile ows preferentially down the path of least resistance and allows for spontaneous healing of the bile leak. This may be accomplished by stenting of the bile duct, sphincterotomy, or by a combination of the two.11,15 Controversy exists over which
483

Penetrating Trauma
Of the patients with penetrating mechanism (Table 2), 10 were secondary to gunshot wounds and 2 secondary to stab wounds. All patients had laparotomy and hepatorrhaphy at admission, with one undergoing hepatic tractotomy (pa 2011 Lippincott Williams & Wilkins

484
Hepatic Interventions After ERCP Biliary Leak Location Left intrahepatic BD ? Intrahepatic Sphincterotomy 1 944 Yes Biliary 10 Fr, sphincterotomy 2 128 Yes Stents Placed ERCP Follow-Up (d) Bile Leak Resolved? Liver abscess drainage Right intrahepatic BD Biliary 10 Fr, sphincterotomy 2 104 Yes Aberrant left duct leak; aberrant right hepatic duct lea Biliary 10 Fr, sphincterotomy 3 232 Yes Right and Left intrahepatic branches Left hepatic ductat second ERCPbiliary vascular stula seen Biliary 10 Fr, sphincterotomy Biliary 10 Fr sphincterotomy; stent placement, biliary 10 Fr 2 2 112 Yes 4 278 Yes Ex laparotomy washout, JP drain HIDAongoing biliary Abdominal wall dehiscence Abdominal wall abscess drainage Perihepatic uid collection Evacuation of perihepatic hematoma Perihepatic biloma drainage Left hepatic duct; right and left intrahepatic branches Left intrahepatic BD Drainage of intrahepatic uid collection Right intrahepatic BD Common hepatic duct Biliary 10Fr, Sphincterotomy, then Biliary 10Fr Biliary 10Fr, Sphincterotomy Biliary 7 Fr, sphincterotomy Biliary 10 Fr, sphincterotomy 3 170 Yes 2 2 2 110 1145 1803 Yes Yes Yeseventual operation for stricture Percutaneous transhepatic internal drainage catheter Operative placement of a biliary catheter Right intrahepatic BD Cystic duct Biliary 10 Fr, sphincterotomy Biliary 10 Fr, sphincterotomy 2 2 55 30 Yes Yes

TABLE 2.

Characteristics of Penetrating Group

Anand et al.

Patient/ Age/Sex

Mechanism

Hepatic Interventions Before ERCP

3/29/M

GSW

4/28/M

GSW

7/39/M

GSW

12/22/M

GSW

13/19/M

GSW

15/29/M

GSW

Ex laparotomy, hepatorrhaphy Drainage of a biloma Ex laparotomy hepatorrhaphy Hepatic tractotomy HIDAleak Ex laparotomy, hepatorrhaphy Gelfoam embolization of hepatic artery to portal vein stula, right hepatic artery embolization HIDAbile leak Ex laparotomy hepatorrhaphy Coil embolization of branch of right hepatic artery Right hepatic lobectomy HIDAleak Ex laparotomy hepatorrhaphy Gelfoam embolization right hepatic Right hepatic lobectomy OSH Ex laparotomy, hepatorrhaphy

16/18/M

GSW

Ex laparotomy hepatorrhaphy

Ex laparotomy, hepatorrhaphy

19/47/M

Stab

20/19/M

GSW

22/17/M

GSW

Ex laparotomy hepatorrhaphy IR drainage of bile leak (HIDA) Ex laparotomy, hepatorrhaphy HIDAbile leak Ex laparotomy, hepatorrhaphy

Embolization of right hepatic artery

23/24/M

Stab

26/25/M

GSW

Partial right hepatic lobectomy Ex laparotomy, hepatorrhaphy HIDAnegative for leak Ex laparotomy, cholecystectomy HIDAnegative for leak IR drain of perihepatic uid

The Journal of TRAUMA Injury, Infection, and Critical Care Volume 71, Number 2, August 2011

2011 Lippincott Williams & Wilkins

GSW, gunshot wound.

The Journal of TRAUMA Injury, Infection, and Critical Care Volume 71, Number 2, August 2011

ERCP After Liver Trauma

specic ERCP maneuver should be used.16 Another approach to the problem, although using the same physiologic principle, is endoscopic placement of a nasobiliary catheter. A series from Poland reports on the treatment of three patients who had biliary leak after blunt hepatic injury detected on ERCP, all of who had subsequent placement of a nasobiliary catheter to allow for drainage of bile.9 We think that an indwelling stent results in greater patient comfort. Interestingly, in the current series, 86% of the patients in the blunt group and 50% of the patients in the penetrating group who had arterial embolization went on to require formal hepatic lobectomy or partial lobectomy because of hepatic necrosis. As we have previously reported, major hepatic necrosis is a common complication of angioembolization and is associated with high-grade liver injuries, and our results are in concert with those in the literature.17 The high rate in the blunt population suggests the need for super-selective coiling to minimize liver injury postprocedure. In one patient (patient 1), the bile leak was seen to be coming from the liver resection bed after right hepatic lobectomy. This puts forth the possibility that the bile leak in this case could have been postoperative rather than truly a posttraumatic injury. Nonetheless, ERCP in this situation also proves to be both a diagnostic and therapeutic tool. Treatment of liver injury has evolved signicantly with a shift from denitive repair and resection to damage control in unstable patients and nonoperative therapy in patients who are hemodynamically stable.18 The use of nonoperative management for liver injury and improvements in ICU care has made bile leak a more common complication.2,3 Previous authors have commented that in patients treated nonoperatively, adjunctive procedures are needed to be selectively used for successful management.19 A recent study from Kozar et al.12 comments that the nonoperative management of blunt high-grade liver injuries is associated with signicant morbidity that correlates with the grade of liver injury. In their series, 14% of patients developed hepatic complications of which 5% were with grade 3 injuries, 22% were with grade 4 injures, and 52% were inpatients with grade 5 injuries.12 24-hour blood transfusion requirement was also predictive of complications by multivariate analysis.12 Christmas et al.19 reported on their selective management of blunt hepatic trauma and noted that 11.1% of patients managed nonoperatively for blunt hepatic trauma required adjunctive operative treatment proceduresincluding arterial embolization, drainage of biloma, and ERCP. Other authors note that, in the case of penetrating liver trauma, the use of techniques such as damage control, packing, hemostatic agents, and angioembolization all lend themselves to the development of hepatic complications.20 Our ndings that a signicant proportion of bile leaks in our study may have been postoperative rather than posttraumatic in nature support these ndings. It is difcult to know for sure as some of these patients, particularly those with high-grade lesions, may have leaked without operation. However, these ndings suggest that in place of nonoperative therapy, angioembolization, and hepatorrhaphy,

there may be a certain subset of patients with liver injury who will benet from a denitive resectional therapy early on. In conclusion, we think that ERCP is a useful as both a diagnostic and therapeutic tool for the safer treatment of both biliary and pancreatic ductal injuries after severe abdominal trauma. REFERENCES
1. Croce MA, Fabian TC, Menke PG, et al. Nonoperative management of blunt hepatic trauma is the treatment of choice for hemodynamically stable patients: results of a prospective trial. Ann Surg. 1995;221:744 753. 2. Singh V, Narasimhan KL, Verma GR, Singh G. Endoscopic management of traumatic hepatobiliary injuries. J Gastroenterol Hepatol. 2007; 22:12051209. 3. Asensio JA, Demetriades D, Chahwan S, et al. Approach to the management of complex hepatic injuries. J Trauma. 2000;48:66 69. 4. ORahilly A, Duignan JP, Lennon JR, OMalley E. Successful treatment of a post-operative external biliary stula by endoscopic papillotomy. Endoscopy. 1983;15:68 69. 5. Smith AC, Schapiro RH, Kelsey PB, Warshaw AL. Successful treatment of nonhealing biliary-cutaneous stulas with biliary stents. Gastroenterology. 1986;90:764 769. 6. Thomas H, Madanur M, Bartlett A, Marangoni G, Heaton N, Rela M. Pancreatic trauma12 year experience from a tertiary center. Pancreas. 2009;38:113116. 7. Harrell DJ, Vitale GC, Larson GM. Selective role for endoscopic retrograde cholangiopancreatography in abdominal trauma. Surg Endosc. 1998;12:400 404. 8. Bridges A, Wilcox CM, Varadarajulu S. Endoscopic management of traumatic bile leaks. Gastrointest Endosc. 2007;65:10811085. 9. Graia P, Skrzywanek P, Sowier A. Biliary stulas resulting from blunt hepatic injury treated by endoscopic diversion of the bile ow. Acta Chir Belg. 2009;109:4751. 10. Moore EE, Cogbill TH, Jurkovich GJ, Shackford SR, Malangoni MA, Champion HR. Organ injury scaling: spleen, liver and kidney (1994 revision). J Trauma. 1995;38:323324. 11. Lubezky N, Konikoff FM, Rosin D, Carmon E, Kluger Y, Ben-Haim M. Endoscopic sphincterotomy and temporary internal stenting for bile leaks following complex hepatic trauma. Br J Surg. 2006;93:78 81. 12. Kozar RA, Moore FA, Cothren C, et al. Risk factors for hepatic morbidity following nonoperative management. Arch Surg. 2006;141: 451 458; discussion 458 459. 13. Jenkins MA, Ponskiy JL. Endoscopic retrograde cholangiopancreatography and endobiliary stenting in the treatment of biliary injury resulting from liver trauma. Surg Laparosc Endosc. 1995;5:118 120. 14. Scioscia PJ, Dillon PW, Cilley RE, Hoover WC, Krummel TM. Endoscopic sphincterotomy in the management of posttraumatic biliary stula. J Pediatr Surg. 1994;29:3 6. 15. Baillie J, ed. Endoscopic retrograde cholangiopancreatography problems and solutions. In: Gastrointestinal EndoscopyBeyond the Basics. Boston, MA: Butterworth-Heinemann; 1997:65134. 16. Bajaj JS, Dua KS. The role of endoscopy in noniatrogenic injuries of the liver. Curr Gastroenterol Rep. 2007;9:147150. 17. Dabbs DN, Stein DM, Scalea TM. Major hepatic necrosis: a common complication after angioembolization for treatment of high-grade liver injuries. J Trauma. 2009;66:621 627. 18. Stein DM, Scalea TM. Nonoperative management of spleen and liver injuries. J Intensive Care Med. 2006;21:296 304. 19. Christmas AB, Wilson AK, Manning B, et al. Selective management of blunt hepatic injuries including nonoperative management is a safe and effective strategy. Surgery. 2005;138:606 610; discussion 610 611. 20. Asensio JA, Roldan G, Petron P, Crisco JJ, Moore DC, Ehrlich MG. Operative management and outcomes in 103 AAST-OIS grades IV and V complex hepatic injuries: trauma surgeons still need to operate, but angioembolization helps. J Trauma. 2003;54:647 654.

2011 Lippincott Williams & Wilkins

485