Endoscopic Management of Pancreaticobiliary Injuries: A Level 1 US Trauma Center Experience

Background: Traumatic pancreaticobiliary injuries are challenging to diagnose and manage. Endoscopic retrograde cholangiopancreatography (ERCP) has potential diagnostic and therapeutic utility in cases of traumatic pancreaticobiliary injuries. Methods: In this single-center retrospective study, we assessed 25 cases of abdominal trauma in which the patients underwent ERCP for management of suspected pancreaticobiliary injuries. We analyzed basic patient demographics, mechanism of trauma, method of diagnosis, ERCP results, surgical treatments, and outcomes. Results: Of the 25 assessed patients, 12 (48%) had pancreatic injuries, 12 (48%) had biliary injuries, and 1 (4%) patient had both. The median age was 28 years [IQR 25-35], and 84% of patients were males. Fifty-six percent of injuries were from blunt trauma, while 44% were from penetrating trauma. In cases of ERCP-confirmed biliary leaks (n=11), 100% of leaks were resolved in the 8 patients who underwent repeat ERCP after initial ERCP with stenting. In cases of ERCP-confirmed pancreatic duct leaks (n=10), 57% of duct leaks were resolved in the 7 patients who underwent repeat ERCP after initial ERCP with stenting. One patient in the biliary trauma cohort developed post-ERCP pancreatitis and sepsis. Conclusion: ERCP was a useful diagnostic and therapeutic intervention in this population of patients with pancreaticobiliary trauma.


INTRODUCTION
Traumatic pancreatic and biliary injuries present diagnostic and therapeutic dilemmas.2][3][4][5] Additionally, computed tomography (CT) has a relatively low sensitivity for identifying pancreatic duct trauma and biliary leaks, further increasing the risk for delayed diagnosis. 1,2,6,7Magnetic resonance imaging may not be feasible in the trauma setting, when multiple injuries are present, and when metallic implants are present.][10][11] In cases of pancreatic trauma, the rate of significant morbidity doubles when the diagnosis is delayed. 43][14][15][16] While assessment of the utility of ERCP for pancreaticobiliary injuries specifically in the trauma setting is limited, available studies suggest that ERCP is safe and is an effective diagnostic and treatment option for this population. 1,4,8,12In this study, we examined the role of ERCP in diagnosis and management of pancreatic and biliary injuries in patients with blunt and penetrating trauma.

METHODS
This retrospective analysis included all patients who underwent ERCP for suspected biliary or pancreatic injury after abdominal trauma from January 1, 2010, through September 30, 2020, at a Level 1 trauma center in a tertiary care institution.Cases of both blunt and penetrating abdominal trauma were included.Individuals with traumatic Goble, SR pancreaticobiliary injuries who did not undergo ERCP were excluded.We reviewed patient notes, laboratory results, imaging reports, operative reports, and ERCP reports and recorded basic demographics for each patient.We recorded the circumstances and initial presentation for each injury, with the severity of injury characterized by the Injury Severity Score (ISS).The ISS is calculated by assigning a numeric score that correlates with injury severity (0 to 5, ranging from no injury to unsurvivable injury) to 6 different body systems, squaring the 3 highest scores, and adding those scores for a final score of 0 to 75.We also reviewed and summarized the diagnostic workup for each patient, including imaging and laboratory results.
Patients were categorized by the management of their pancreaticobiliary trauma: surgical, endoscopic, percutaneous, and combinations of the different modalities.Procedural adverse events were assessed from procedural reports and daily progress notes.For assessments of management and outcomes, patients were categorized as having a pancreatic injury, biliary injury, or both.Patient demographics, diagnostics, management, and outcomes are all summarized using standard descriptive statistics.STATA version 17.0 (StataCorp LLC) was used for statistical computations.
The study was approved by the Hennepin Healthcare Institutional Review Board, and all study procedures were in accordance with the 1964 Declaration of Helsinki and its subsequent amendments.

Demographic and Clinical Characteristics
Twenty-five patients were evaluated: 12 patients (48%) had a traumatic pancreatic injury, 12 (48%) had a traumatic biliary injury, and 1 (4%) had both (Table 1).Age at presentation ranged from 14 to 60 years, with a median age of 28 years [IQR 25-35]; most of the cohort was male (84%) and White (64%).Blunt force injury was common in the pancreatic trauma cohort (8/12, 67%), while penetrating injury was common in the biliary trauma cohort (7/12, 58%).Gunshot was the most common mechanism of injury, accounting for 40% of presentations.Associated injuries in addition to pancreatic and biliary trauma were common (24/25, 96%).Intraabdominal hematoma or active vessel bleeding was noted in 7 patients (28%), including 1 patient with an injury to the inferior vena cava and another with an injury to the superior mesenteric artery; both patients were managed surgically.Two patients (8%) required partial colectomy, 5 patients (20%) required splenectomy, and 3 patients (12%) required cholecystectomy.Ten patients (40%) had serious intrathoracic injuries, including pneumothorax and pulmonary contusion.

Biliary Trauma Cohort
The median age of the 13 patients who had documented concern for biliary injury was 28 years [IQR 24-43], and 85% (11/13) were male.Of the 7 cases of penetrating abdominal injury, 6 (86%) were due to gunshot wounds, with the remaining case due to a stab wound.Of the 6 cases of blunt force injuries, 2 (33%) were secondary to being struck with a blunt object, 2 (1 was the patient with both biliary and pancreatic trauma) resulted from motor vehicle accidents, while a snowmobile accident and being struck by a train accounted for the remaining 2 cases.a The patient who had both biliary and pancreatic injuries sustained those injuries in a motor vehicle accident.The patient is included in both the biliary trauma cohort and the pancreatic trauma cohort in the Results section but is counted once here.b One patient received both packed red blood cells and fresh frozen plasma.
Note: Data are presented as n (%) unless otherwise indicated.ICU, intensive care unit; IQR, interquartile range.a The number of bile leaks in the biliary trauma cohort = 10, while the total number of bile leaks (including the patient with biliary and pancreatic trauma) = 11.
b The number of pancreatic duct leaks in the pancreatic trauma cohort = 9, while the total number of pancreatic duct leaks (including the patient with biliary and pancreatic trauma) = 10.Note: Data are presented as n (%) unless otherwise indicated.ERCP, endoscopic retrograde cholangiopancreatography; IQR, interquartile range; N/A, not applicable.
Concern for biliary injury was first established by CT as part of the initial trauma workup in 77% of patients (10/13), and the remaining 23% of patients (3/13) had concern first established during exploratory laparotomy done on initial presentation.
CT was performed prior to ERCP in 77% of patients (10/13), and hepatic lacerations were noted on all 10 CTs.In 5 of the 10 cases (50%), perihepatic fluid collections were seen on the CTs, and in all 5 of those cases (100%), biliary leak was later confirmed on ERCP.A bile leak was later confirmed in 3 of the 5 cases (60%) where initial CT showed no perihepatic fluid collection.One patient underwent magnetic resonance cholangiopancreatography (MRCP) prior to ERCP.The MRCP did not demonstrate any biliary leak, and none was noted on the ERCP.
Six of the 13 patients (46%) with documented concern for biliary injury had hepatobiliary iminodiacetic acid (HIDA) scans, and 5 of the HIDA scans (83%) documented bile leaks, all of which were later confirmed on ERCP.One HIDA scan was negative for any biliary leak; however, ERCP later demonstrated a bile leak in that patient.
Eight of the 11 patients with confirmed biliary leak underwent repeat ERCP, and resolution of the biliary leak was demonstrated in all 8 patients (100%).Repeat ERCP was recommended for the other 3 patients with a confirmed biliary leak, but they were lost to follow-up.Consequently, 73% of patients (8/11) with confirmed biliary leak had resolution confirmed by ERCP.
Eight of the 13 patients (62%) with biliary trauma underwent exploratory laparotomy prior to any endoscopic intervention and had ERCP performed after surgery (Figure 1).All 8 surgeries were performed to manage associated intraabdominal injuries; biliary leaks were not repaired during these surgeries.Of these 8 patients, none required further surgeries after the ERCP.The other 5 patients (38%) had an ERCP without prior surgery, and none required subsequent surgery.
Of the 13 patients with a biliary injury, 9 (69%) required 2 or fewer ERCPs during the study period, and the median number of ERCPs in patients with only a biliary injury was 2 (Table 2).One patient developed sepsis and pancreatitis shortly after ERCP, but no other procedural adverse events were documented.Seven of the 13 patients (54%) eventually developed hepatic abscesses, 6 of whom (86%) required percutaneous drainage.Two patients developed infected bilomas.No deaths occurred in the biliary trauma cohort.

Pancreatic Trauma Cohort
The median age of the 13 patients who had documented concern for pancreatic injury was 29 years [IQR 26-35], and 85% (11/13) were male.All 4 cases of penetrating abdominal injury were due to a gunshot wound.Of the 9 cases of blunt force injuries, 3 (33%) were secondary to being struck with a blunt object; 3 (1 was the patient with both biliary and pancreatic trauma) were due to motor vehicle accidents; and a fall from a building, a go-cart accident, and a machinery accident (crush injury) at work accounted for the remaining 3 cases.
Concern for pancreatic injury was first established by CT as part of the initial trauma workup in 92% of cases (12/13), and the remaining patient had concern first established during exploratory laparotomy.
Four of the 13 patients (31%) had peripancreatic fluid collections found on initial CT, and each patient had a pancreatic duct leak later confirmed on ERCP.Three of the 13 patients (23%) had pancreatic lacerations noted on CT, and each patient was found to have a pancreatic duct leak on ERCP.Three of the 13 patients (23%) had pancreatic transections on CT; however, only 2 of the 3 (67%) were found to have a duct leak on ERCP.Two of the 13 patients (15%) had peripancreatic hematomas on initial CT, but neither had a pancreatic duct leak on subsequent ERCP.The remaining patient had no signs of pancreatic trauma on CT but was later found to have a pancreatic duct leak on ERCP.
Four of the 13 patients (31%) had MRCPs prior to ERCP.One MRCP correctly identified a pancreatic duct disruption that was later confirmed on ERCP.Another MRCP visualized an intact duct that ERCP confirmed did not have a leak.The duct was not well visualized on the MRCP for the third patient, but ERCP later identified a leak in this patient.The MRCP for the fourth patient showed pancreatic transection, but no leak was found on subsequent ERCP.
In the 10 cases of ERCP-confirmed pancreatic duct leak, 7 patients (70%) underwent repeat ERCP, and resolution of the duct leak was demonstrated in 4 of the 7 patients (57%).Among the remaining 3 patients who underwent repeat ERCP, 2 patients (67%) were lost to follow-up after the second ERCP.The third patient underwent a third ERCP, the duct leak was still present, and the patient was then lost to follow-up.Repeat ERCP was recommended for the other 3 of 10 patients with a confirmed pancreatic duct leak, but all 3 patients were lost to follow-up.In total, 40% of patients (4/10) with a confirmed pancreatic duct leak had resolution confirmed by ERCP.
Ten of the 13 patients (77%) with pancreatic trauma underwent surgery prior to any endoscopic intervention and had ERCP performed after surgery (Figure 2).Three patients had distal pancreatectomies during the initial exploratory laparotomy.All 3 patients were subsequently found to have distal pancreatic duct leaks on ERCP, but the patient notes did not specify if the leaks were secondary to the initial trauma or to surgical adverse events.Of the 10 patients with pancreatic injuries who underwent surgery prior to ERCP, 3 (30%) required repeat surgery after ERCP to manage intra-abdominal injuries, including 1 patient who required further pancreatic resection after previous pancreatic resection.
Three of the 13 patients (23%) with pancreatic trauma had an ERCP without prior surgery.One patient required surgical intervention after ERCP to repair a diaphragm injury, but no specific surgical procedure for the pancreatic trauma was needed.The other 2 patients were successfully managed with endoscopy and percutaneous drainage alone.
No ERCP-related adverse events were noted in the pancreatic cohort; however, 4 patients (31%) developed pancreatic pseudocysts, and 1 patient (8%) developed a peripancreatic abscess.Two of the patients with pancreatic pseudocysts required percutaneous drainage, while the other 2 patients were managed solely with endoscopic stenting of the pancreatic duct.The peripancreatic abscess was successfully managed with percutaneous drainage.No deaths occurred in the pancreatic trauma cohort.

DISCUSSION
ERCP was well tolerated in this population of 25 patients, with adverse events (pancreatitis and sepsis) noted in only 1 patient (4%).ERCP also appeared to be effective, particularly for patients with biliary trauma, as 100% of biliary leaks were resolved on repeat ERCP after stenting (8 of 8 patients), and 57% of pancreatic duct leaks were resolved on repeat ERCP (4 of 7 patients).
Diagnosis of biliary injuries requires a high index of suspicion as routine trauma imaging with abdominal CT may not demonstrate the injury. 17In our study, CT was performed in most patients with biliary trauma (77%), but each case required further diagnostic testing for confirmation of the injury.MRCP is a useful diagnostic tool in cases of biliary leaks. 18,19Aduna et al found that MRCP was 95% sensitive and 100% specific for biliary leaks in patients after cholecystectomy. 19However, MRCP may not always be feasible in the trauma setting given the high acuity of the patient.Additionally, associated injuries may complicate the clinical picture, and suspicion for biliary injury may not be high enough to pursue dedicated imaging. 3Despite the effectiveness of MRCP, the diagnosis of traumatic biliary leaks is usually made during exploratory laparotomy or after the development of late complications. 20,21Our study demonstrates this diagnostic dilemma as diagnosis was often delayed (median 7 days from injury), and MRCP was only performed prior to ERCP in 1 patient with biliary trauma.Of note, MRCP did not accurately diagnose the biliary leak in that case.
While small intrahepatic bile leaks may respond well to conservative treatment, larger bile leaks are prone to complications. 22][25][26][27][28][29] Infected bilomas occurred twice in our population: in a patient who underwent ERCP 4 days after trauma and in a patient who underwent ERCP 18 days after trauma.4][25][26][27][28][29] Our study supports this assertion, as ERCP was effective in our biliary trauma cohort with 100% of bile leaks resolved in the patients who underwent repeat ERCP and no significant morbidity or mortality secondary to the bile leak.
Traumatic injury to the pancreas is less common than biliary tract injury and may be more difficult to diagnose.Main pancreatic duct disruptions portend a worse prognosis than other pancreatic trauma complications, and key to the diagnostic evaluation of pancreatic trauma is the ability to evaluate the integrity of the main pancreatic duct. 9,30T has demonstrated poor sensitivity for the evaluation of pancreatic duct injuries. 1,2,7Of the 12 patients with pancreatic trauma without biliary trauma in this cohort, 9 (75%) had pancreatic duct leaks found on ERCP.Each of those 9 was evaluated with CT prior to ERCP, and radiology interpreted the initial CT as concerning for pancreatic injury in 8 cases (89%), with concern for pancreatic duct injury specifically documented in 6 cases (67%).
MRCP has become the noninvasive imaging method of choice when evaluating for pancreatic duct injury; however, MRCP use in our cohort was fairly limited with only 2 of the 9 patients with ERCP-confirmed pancreatic duct leaks undergoing MRCP prior to ERCP.While prior studies have suggested MRCP to be effective at diagnosing pancreatic duct injuries, 1,9 we cannot confirm this finding given the low sample size of our study and that MRCP was diagnostic in just 1 of the 2 patients with later confirmed duct leaks.
ERCP is the most accurate method for assessing the location and extent of pancreatic duct injury and can be performed prior to, during, or after surgical management. 31RCP in stable patients who do not require emergent surgery allows for nonoperative treatment in the absence of ductal injury and earlier operative treatment or primary therapy such as stent placement in the presence of ductal injury. 7,32,33RCP can also help to manage local complications of pancreatic trauma as both endoscopic transpapillary and transmural drainage can effectively manage pseudocysts and fistulas. 2,4,9,34In our cohort, repeat ERCP with pancreatic duct stenting led to resolution of the pancreatic duct leak in 57% of patients (4/7).ERCP was also generally well tolerated with no noted procedural adverse events in the pancreatic trauma cohort.Given the potential for morbidity related to surgical management, ERCP appears to be a reasonable management strategy for stable trauma patients with suspected pancreatic duct leaks.We agree with the necessity of surgery in unstable patients which was the case for numerous patients in this study.
Our study has some noteworthy limitations.The small sample size limits generalizability as does the retrospective nature of the study.Although we were assessing the efficacy of ERCP, numerous interventions, including percutaneous and surgical interventions, were used to manage the traumatic injuries in these patients, so determining the efficacy of one intervention by itself is difficult.

CONCLUSION
ERCP was an effective diagnostic and therapeutic tool in this population of patients who sustained pancreaticobiliary injuries after abdominal trauma.Performing early ERCP in stable trauma patients with suspected pancreaticobiliary injuries may be warranted.

Figure 1 .
Figure 1.Initial management of patients with traumatic biliary injury.One patient had both biliary and pancreatic injuries.ERCP, endoscopic retrograde cholangiopancreatography.

Figure 2 .
Figure 2. Initial management of patients with traumatic pancreatic injury.One patient had both biliary and pancreatic injuries.ERCP, endoscopic retrograde cholangiopancreatography.