Abstract
Endoscopic retrograde cholangiopancreatography (ERCP)‐related perforation leads to high morbidity and mortality. The Stapfer classification divides patients with different perforation locations and suggests management accordingly. The classification may be unknown if perforation is not detected during endoscopy. We classified patients with ERCP‐related perforation (ERP) through computed tomography (CT) and observed the clinical outcomes with varyingly invasive management. Fifty‐two cases of ERP between July 2009 and December 2017 were retrospectively reviewed. Of them, 41 who underwent CT for ERCP were included. According to their CT findings, we divided patients into air‐alone (n = 16), air‐fluid (n = 18), and fluid‐alone (n = 7) groups. Perforation severity was graded using the Clavien‐Dindo classification for surgical complications. Demographic data and clinical outcomes among different groups were analyzed. Fifteen patients (37%) had an unknown Stapfer classification. More than half of the patients in the air‐fluid group had a Clavien‐Dindo complication grade of >3. Four patients underwent surgical repair; all of them were from the air‐fluid group. All patients in the air‐ and fluid‐alone groups underwent medical treatment without need for subsequent salvage surgery. The air‐fluid group had the longest mean hospital stay (25.1 ± 21.9 days) and the exclusive two mortality cases in this study. Patients with ERCP can be divided into groups with different outcomes according to the presence of air or fluid on CT images. Because patients with both air and fluid have the worst clinical outcome, they may require more aggressive treatment than patients with either air or fluid alone.
Keywords: air, computed tomography, endoscopic retrograde cholangiopancreatography, fluid, perforation
1. INTRODUCTION
Complications from endoscopic retrograde cholangiopancreatography (ERCP) include cholangitis, pancreatitis, bleeding, and perforation. The likelihood of these complications is 5%‐12% in all cases, and they can increase morbidity and mortality.1 Among these complications, endoscopic retrograde cholangiopancreatography‐related perforation (ERP) remains a clinical challenge to endoscopists because it may be overlooked during these procedures and there is no consensus on its management. Overall, the incidence of ERP is approximately 0.1%‐1.6%2, 3 with a mortality rate of 3%‐35.7%.4, 5
Surgical debridement and repair is the mainstay of treatment options for ERP. It is, however, associated with high morbidity and mortality; recent studies have suggested more conservative treatment for certain ERP types.6, 7 According to the Stapfer classification, ERPs are grouped into four subtypes depending on the mechanism of injury and the anatomic location.8 Type 1 ERPs occur at the duodenal wall, away from the ampulla of Vater; they are usually caused by the shaft of the endoscope and are typically large with extensive leakage into the intraperitoneal or retroperitoneal space. Type 1 ERPs are usually diagnosed during endoscopy, which requires immediate surgery if endoscopic closure is impossible. Type 2 ERPs are perforation around the periampullary area caused by sphincterotomy and/or balloon dilation with wall disruption extending beyond the intramural portion of the bile duct. Although some clinicians consider it an absolute indication for surgery, others suggest that a nonsurgical approach may be effective in these patients.5 Type 3 ERPs are caused by guidewire‐related bile duct perforation and type 4 ERPs are probably caused by excessive air inflation. In general, both type 3 and 4 ERPs are diminutive and can be treated conservatively.
Although different treatments have been suggested according to the types of Stapfer classification, diagnosing the ERP type is relatively difficult if it is not found during the ERCP procedure. Computed tomography (CT) can be used to diagnose ERP based on the extraintestinal air and local fluid accumulation within the retroperitoneal or intraperitoneal space. Using CT findings to guide ERP treatment can be clinically practical. Thus, in this retrospective study, we grouped patients with ERP according to their CT findings and compared their clinical outcomes.
2. PATIENTS AND METHODS
2.1. Patients
We retrospectively analyzed patients who underwent ERCP between July 2009 and December 2017 at a referral center in Southern Taiwan. In 4909 sessions of ERCP, 52 ERPs (1.1%) were noted. Of them, 26 patients were diagnosed as having ERP based on a duodenum wall disruption on endoscopy imaging (n = 5) or free air/contrast extravasation on fluoroscopy imaging (n = 21), whereas the remaining 26 were diagnosed because they had abdominal pain or fever after ERCP, with subsequent confirmation through noncontrast CT. We excluded patients who did not have CT examination because they underwent surgical repair directly after ERCP diagnosis (n = 3) and those who were considered to have guidewire‐related ductal perforation and received medical treatment directly (n = 8). In total, 41 patients were examined for imaging and data analysis (Figure 1).
Figure 1.
Study algorithm of patients with ERCP perforation. Patients were divided according to their treatment methods. The distribution of CT imaging patterns in patients who underwent different treatments are depicted at the bottom of each group
2.2. Computed tomography
CT images were retrospectively reviewed by a radiologist (Tsai), who has more than 20 years of experience of imaging diagnosis; the reviewer was blinded to the other clinical information of the patients. The time elapsed from the ERCP to CT examination was recorded and patients were divided into those who had a CT examination done within 12 hours, between 12 and 48 hours, and after 48 hours of ERCP. The patients were diagnosed as having ERP if their CT imaging showed extraintestinal air and/or local fluid accumulation without the coexistence of pancreatic swelling or peripancreatic fat stranding (ie, pancreatitis). According to the CT findings (Figure 2), the patients were divided into three groups: those with (1) free air (n = 16; air alone), (2) both free air and fluid accumulation (n = 18; air‐fluid), and (3) fluid accumulation (n = 7; fluid alone).
Figure 2.
(A) Air‐alone group. Asterix = extraluminal air that has accumulated in the retroperitoneal space. (B) Air‐fluid group. White arrow = fluid, white arrow head = air. (C) Fluid‐alone group. White arrow = fluid accumulation
2.3. Perforation management
Treatment decisions were made according to patient condition (eg, the probable Stapfer type of ERP, pain severity, and sepsis) and the consensus between the supervising clinicians and patients. Of the 41 patients we examined, four (9.7%) underwent surgical repair. The remaining 37 (90.3%) were treated conservatively through bowel rest, parenteral nutrition and broad‐spectrum antibiotic treatment, or nasogastric tube decompression. Moreover, 10 patients received CT‐guided percutaneous drainage due to their progressive symptoms and signs of ERP (abdominal pain and/or fever) after initial medical treatment.
2.4. Clinical data and course
The patients’ demographic data, including age, sex, body mass index, ERCP indication, and diagnosis of associated comorbidities (presented as Charlson comorbidity index) were collected.9 Intervention procedures performed during ERCP, ERP manifestations, the probable Stapfer type, and time lag between ERCP and CT imaging were also recorded. Signs of systemic inflammatory response syndrome, including temperature, heart and respiratory rate, and leukocyte count, were recorded, and hospitalization treatment course indicators—including duration of fever, that of abdominal pain, time to first bowel movement, length of hospitalization, and status upon discharge—were noted to discern treatment outcomes. Finally, ERP severities were stratified according to the Dindo‐Clavien classification, comprising seven grades, namely I, II, IIIa, IIIb, IVa, IVb, and V, ranging from complications resulting in any deviation from the normal postoperative course without the need for pharmacological treatment or surgical, endoscopic, or radiological intervention (I) to death (V).
2.5. Statistical analysis
Variables of the air‐alone, air‐fluid, and fluid‐alone groups were compared using the chi‐square or Fisher's exact test, and Kruskal‐Wallis tests for categorical and numerical ones, respectively. Data were analyzed on IBM SPSS Statistics (version 22.0; IBM, Armonk, New York); two‐tailed analytical results with P < .05 were considered significant.
3. RESULTS
Gallstone disease was the main ERCP indication in our study. Most of our patients were old and had several age‐associated comorbidities, such as diabetes, hypertension, and cardiovascular disease (Table 1). The demographic data demonstrated no significant difference among the three groups, whether due to sex, age, body mass index, or Charlson comorbidity index. Juxtapapillary diverticulum was more frequent in patients with air alone ERP (62.5%) than in those with fluid alone ERP (38.9%) or air‐fluid ERP (42.9%); however, these results were nonsignificant. ERP can occur to any patient who underwent ERCP with or without further intervention procedure. Half of the patients in the air alone group had CT examination within 12 hours of ERCP and most of the patients (71.4%) in the fluid alone group had CT examination after 48 hours of ERCP. Although there was a trend that more air‐alone CT finding found in the immediately diagnosed patients while more air‐fluid or fluid‐alone CT findings, in the late‐diagnosed patients, the difference was not statistically significant (P = .05). Any of CT findings including air, air‐fluid, or fluid can be found through patients who had ERP diagnosed immediately after the ERCP to those who diagnosed later. Notably, 15 of the 41 patients had an unknown Stapfer classification. All patients with Stapfer type 1 ERPs demonstrated both air and fluid on their CT images, whereas those with Stapfer type 4 perforations showed air alone. In addition, the ERPs in all patients in the fluid‐alone group were of Stapfer type 3, whereas those in the air‐fluid group were of any Stapfer type, except type 4.
Table 1.
Baseline characteristics of the study population
| Group | Air (N = 16) | Air‐fluid (N = 18) | Fluid (N = 7) | P valuea | |
|---|---|---|---|---|---|
| Male, n (%) | 11 (68.8) | 10 (55.6) | 1 (14.3) | 0.054 | |
| Age, years | 70.69 ± 14.97 | 66.33 ± 12.13 | 70.57 ± 19.17 | 0.645 | |
| Body mass index | 25.00 ± 5.11 | 24.51 ± 3.55 | 24.71 ± 3.06 | 0.943 | |
| Charlson comorbidity index | 4.06 ± 2.35 | 3.44 ± 2.30 | 4.14 ± 3.23 | 0.715 | |
| Juxtapapillary diverticulum, n (%) | 10 (62.5) | 7 (38.9) | 3 (42.9) | 0.366 | |
| Billroth‐II gastrojejunostomy, n (%) | 0 | 1 | 0 | 0.520 | |
| Indication for ERCP: Gallstone, n (%) | 15(93.8) | 18 (100) | 6 (85.7) | 0.313 | |
| Timing of CT imaging study, n (%) | <12 h | 8 (50) | 3(16.7) | 0 | 0.05 |
| 12–48 h | 4 (25) | 9 (50) | 2 (28.6) | ||
| >48 h | 4 (25) | 6 (33.3) | 5 (71.4) | ||
| Intervention during ERCP | |||||
| Scope intubation | 0 | 1 | 0 | 0.520 | |
| Cannulation with/without drainage | 5 | 5 | 2 | 0.975 | |
| Sphincterotomy or sphincter dilationb | 11 | 12 | 5 | 0.973 | |
| Stapfer classification | |||||
| Type 1 | 0 | 3 | 0 | 0.126 | |
| Type 2 | 4 | 3 | 0 | 0.341 | |
| Type 3 | 0 | 4 | 6 | 0.014 | |
| Type 4 | 6 | 0 | 0 | 0.371 | |
| Unknown | 6 | 8 | 1 | 0.371 | |
Note: All data are presented with mean ± SD.
χ 2 test, Fisher's exact test, or Kruskal‐Wallis test.
It includes endoscopic sphincterotomy or/and endoscopic papillary balloon dilation.
As shown in Table 2, 63.4% (26/41) of the ERP patients were not diagnosed during ERCP examination. Since fluid cannot be detected via fluoroscopy, none of the patients in the fluid‐alone group was diagnosed during ERCP examination. Abdominal pain was noted in all of the ERP patients. Most of the ERP patients had fever and leukocytosis on presentation. Among the presentations of ERP, tachycardia and systemic inflammatory response syndrome were more frequently found in patients in the air‐fluid and fluid groups than patients in the air alone group.
Table 2.
Features of ERCP‐related perforation
| Group | Air (n = 16) | Air‐fluid (n = 18) | Fluid (n = 7) | P valuea |
|---|---|---|---|---|
| Delay diagnosis, n (%) | 5 (31.2) | 14 (77.8) | 7 (100) | 0.002 |
| Delayed time, hour (range) | 43.4 ± 23.9 (22.5‐70.8) | 96.5 ± 195.0 (2.6‐762.8) | 115.7 ± 111.5 (41.9‐361.1) | 0.083 |
| Presentation on delay diagnosis | ||||
| Abdominal pain, n (%) | 5 (100) | 14 (100) | 7 (100) | 1.00 |
| Temperature >38, n (%) | 4 (80) | 13 (92.9) | 5 (71.4) | 0.417 |
| Respiratory rate >22/min, n (%) | 0 (0) | 4 (28.6) | 1 (14.3) | 0.352 |
| Heart rate >90/min, n (%) | 0 (0) | 11 (78.6) | 6 (85.7) | 0.003 |
| Leukocytosis >12 000 or <4000, n (%) | 3 (60) | 7 (78.6) | 5 (71.4) | 0.640 |
| SIRS, n (%)b | 3 (60) | 14 (100) | 7 (100) | 0.011 |
Note: Data are presented with mean ± SD.
χ 2 test, or Kruskal‐Wallis test.
SIRS, systemic inflammatory reaction syndrome.
The patients in the air‐fluid group had the poorest clinical outcomes: they had longer hospital stays and prolonged perforation symptoms compared with those in the air‐ or fluid‐alone groups (Table 3). All patients except one in the air‐alone and fluid‐alone groups were classified as Dindo‐Clavien grade 2 complications, which implies most of them were successfully managed with medical treatment (Figure 1). By contrast, more than half of the patients (13/18) in the air‐fluid group required CT‐guided drainage or surgical repair (Figure 1). Two patients with Stapfer type 1 ERP and two with Stapfer type 2 ERP underwent surgery immediately on ERP diagnosis; all four were in the air‐fluid group. No patient who received medical treatment with or without CT‐guided drainage as the initial treatment subsequently required salvage surgery. In total, two patients died because of their ERPs: one with Stapfer type 1 ERP, who had already been admitted into the intensive care unit because of sepsis‐related respiratory failure, and one with Stapfer type 2 ERP who had underlying advanced liver cirrhosis.
Table 3.
Clinical course of patients with ERCP‐related perforation
| Group | Air (N = 16) | Air‐fluid (N = 18) | Fluid(N = 7) | P valuea |
|---|---|---|---|---|
| Hospital stay (day)b | 10. 6 ± 4.1 | 27.5 ± 21.8 | 15.6 ± 9.5 | 0.002 |
| Duration of fever (day)b | 2.6 ± 2.8 | 9.8 ± 10.0 | 4.3 ± 6.7 | 0.005 |
| Abdominal pain (day)b | 2.4 ± 2.5 | 6.6 ± 4.3 | 4.9 ± 1.9 | 0.004 |
| Bowel movement (day)b | 2.4 ± 1.6 | 5.9 ± 7.1 | 2.0 ± 1.4 | 0.088 |
| Grading of complicationc | ||||
| Grade 2 | 16 | 5 | 6 | <0.001 |
| Grade 3 | 0 | 10 | 1 | 0.001 |
| Grade 4/5 | 0 | 3 | 0 | 0.126 |
| Treatment options | ||||
| Conservative | 16 | 5 | 6 | <0.001 |
| CT‐guide drainage | 0 | 9 | 1 | <0.001 |
| Surgery | 0 | 4 | 0 | 0.059 |
| Outcome | ||||
| Mortality | 0 | 2 | 0 | 0.261 |
Kruskal‐Wallis test or Chi‐square test.
Exclude 4 patients that underwent surgery.
Clavien‐Dindo classification of surgical complication.
4. DISCUSSION
ERP remains an inevitable complication of ERCP, even for experienced endoscopists. In this study, the perforation rate was 1.1%, comparable to that reported previously (0.1%‐1.6%).2, 3 The Stapfer classification stratifies patients into different perforation types and suggests further management accordingly. However, intraprocedural detection of perforation for further determining the Stapfer classification is not always possible. In total, 15/41 (36.6%) of our patients had unknown Stapfer classification; moreover, our literature review indicated that the diagnosis rate of Stapfer type 2 ERP was 26%‐100%.10, 11 Therefore, we noted that using the presence of air or fluid on CT images can readily divide patients with ERPs into air‐alone, air‐fluid, and fluid‐alone groups of different severity and may be helpful in guiding further management.
CT scans detect perforations by showing extraintestinal air and/or retroperitoneal fluid collection.12 Our study showed that the presence of one or both of these imaging characters may be related to the location and size of the perforation. As biliary and pancreatic ducts only have fluid in the lumen, diminutive ductal perforation caused by the guidewire usually results in retroperitoneal fluid collection. As demonstrated in our research, all perforations in the patients of the fluid‐alone group were Stapfer type 3 perforations. Meanwhile, the presence of both air and fluid probably indicates a significant perforation over the bowel or ductal wall that allows intestinal air and fluid to pass into the retroperitoneal space, and which may be classified as a Stapfer type 1, 2, or 3 perforation. Intestinal air passes through small perforations more easily than does fluid. Therefore, a CT image showing extraintestinal air alone may indicate a diminutive perforation in the intestinal wall. As noted in our study, the air‐alone group included patients with Stapfer type 2 and type 4 perforations.
The intraprocedural detection rate of perforation in our study was 36.6% (15/41), which is comparable to the rate found in other studies, which has ranged from 10% to 79%.5, 13, 14, 15 Notably, perforation without retroperitoneal air shown on fluoroscopy is difficult to diagnose during the procedure. It is therefore not surprising to see that all of our patients in the fluid‐alone group were diagnosed after the procedure and has more time elapsed before the diagnosis. While it is well documented that early perforation diagnosis is associated with better prognosis,11, 16, 17 in our study, we found that although all the patients in the fluid‐alone group were diagnosed late, most of them only required medical treatment and had a shorter mean hospital stay than those in the air‐fluid group. We thus suggest that, in addition to early diagnosis, the prognosis of an ERP depends on the location and the size of the perforation.
Medical treatment is usually sufficient to address Stapfer types 3 and 4 perforations (as they usually close spontaneously), whereas surgical or endoscopic repair is required for type 1 perforations. Although several studies have suggested that patients with retroperitoneal fluid collection have poorer prognosis and require surgical intervention.8, 18, 19 In our study, all except one patient in the fluid‐alone group were successfully treated with medical treatment, and we therefore suggest that fluid collection alone should not be an indication for surgery. Similarly, all our patients of air alone group were also successfully treated with medical treatment. Increasing evidence including the ESGE guideline has pointed out that retroperitoneal fluid collections or peri‐ampullary perforations can be managed with medical treatment with or without CT‐guided percutaneous drainage.20, 21 In our study, 10 patients (one in the fluid‐alone group and nine in the air‐fluid group) required CT‐guided drainage in addition to medical treatment. While the process of treating our patients was not randomized, the fact that most of our patients (37/41) could be treated nonsurgically supports the recent trend of using medical treatment and percutaneous drainage for ERPs.21, 22 However, because the findings of CT imaging may change from “air alone” which mostly requires conservative treatment to “air‐fluid” which may require percutaneous drainage along with the time passed after ERCP. We suggested repeating CT scan examination if the clinical course of patients progresses after the initial management.
As indicated above, two patients (4.9%) died during the course of our study, which is comparable to the mortality rate reported elsewhere (3%‐35.7%).4, 5 Both these patients had poor body reserves, which precluded them from surgical repair or made postsurgery survival impossible. Our limited data indicate the importance of endoscopic closure for Stapfer type 1 ERPs and of percutaneous drainage for patients with impaired body reserves.
In summary, our retrospective study showed that the severity of ERP can be triaged on the basis of the presence of air or fluid in abdominal CT images. The presence of both fluid and air in CT images is associated with clinical significant bowel wall perforation, and therefore, is associated with a more serious clinical outcome. Although more data are required, our findings support the current trend of nonsurgical treatment of patients. Except for patients with free duodenal wall rupture, patients who demonstrated air or fluid‐alone on their CT images could initially be treated conservatively, whereas patients presenting with both air and fluid on their CT images but failing to respond to the initial treatment should receive CT‐guided percutaneous drainage.
CONFLICT OF INTEREST
The authors declare no potential conflict of interest.
ACKNOWLEDGEMENTS
The study was approved by Institutional Review Board of National Cheng Kung University Hospital, a tertiary referral medical center in southern Taiwan.
Wu J‐H, Tsai H‐M, Chen C‐Y, Wang Y‐S. Computed tomography classification of endoscopic retrograde cholangiopancreatography‐related perforation. Kaohsiung J Med Sci. 2020;36:129–134. 10.1002/kjm2.12138
Contributor Information
Chiung‐Yu Chen, Email: chiungyu@mail.ncku.edu.tw.
Yao‐Sheng Wang, Email: yaoko110@gmail.com.
REFERENCES
- 1. Committee ASoP , Chandrasekhara V, Khashab MA, Muthusamy VR, Acosta RD, Agrawal D, et al. Adverse events associated with ERCP. Gastrointest Endosc. 2017;85(1):32–47. [DOI] [PubMed] [Google Scholar]
- 2. Cotton PB, Garrow DA, Gallagher J, Romagnuolo J. Risk factors for complications after ERCP: A multivariate analysis of 11,497 procedures over 12 years. Gastrointest Endosc. 2009;70(1):80–88. [DOI] [PubMed] [Google Scholar]
- 3. Miller R, Zbar A, Klein Y, Buyeviz V, Melzer E, Mosenkis BN, et al. Perforations following endoscopic retrograde cholangiopancreatography: A single institution experience and surgical recommendations. Am J Surg. 2013;206(2):180–186. [DOI] [PubMed] [Google Scholar]
- 4. Koc B, Bircan HY, Adas G, Kemik O, Akcakaya A, Yavuz A, et al. Complications following endoscopic retrograde cholangiopancreatography: Minimal invasive surgical recommendations. PLoS One. 2014;9(11):e113073. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Kumbhari V, Sinha A, Reddy A, Afghani E, Cotsalas D, Patel YA, et al. Algorithm for the management of ERCP‐related perforations. Gastrointest Endosc. 2016;83(5):934–943. [DOI] [PubMed] [Google Scholar]
- 6. Avgerinos DV, Llaguna OH, Lo AY, Voli J, Leitman IM. Management of endoscopic retrograde cholangiopancreatography: Related duodenal perforations. Surg Endosc. 2009;23(4):833–838. [DOI] [PubMed] [Google Scholar]
- 7. Lee TH, Han JH, Park SH. Endoscopic treatments of endoscopic retrograde cholangiopancreatography‐related duodenal perforations. Clin Endosc. 2013;46(5):522–528. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Stapfer M, Selby RR, Stain SC, Katkhouda N, Parekh D, Jabbour N, et al. Management of duodenal perforation after endoscopic retrograde cholangiopancreatography and sphincterotomy. Ann Surg. 2000;232(2):191–198. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: Development and validation. J Chronic Dis. 1987;40(5):373–383. [DOI] [PubMed] [Google Scholar]
- 10. Vezakis A, Fragulidis G, Polydorou A. Endoscopic retrograde cholangiopancreatography‐related perforations: Diagnosis and management. World J Gastrointest Endosc. 2015;7(14):1135–1141. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Jin YJ, Jeong S, Kim JH, Hwang JC, Yoo BM, Moon JH, et al. Clinical course and proposed treatment strategy for ERCP‐related duodenal perforation: A multicenter analysis. Endoscopy. 2013;45(10):806–812. [DOI] [PubMed] [Google Scholar]
- 12. Kuhlman JE, Fishman EK, Milligan FD, Siegelman SS. Complications of endoscopic retrograde sphincterotomy: Computed tomographic evaluation. Gastrointest Radiol. 1989;14(2):127–132. [DOI] [PubMed] [Google Scholar]
- 13. Alfieri S, Rosa F, Cina C, Tortorelli AP, Tringali A, Perri V, et al. Management of duodeno‐pancreato‐biliary perforations after ERCP: Outcomes from an Italian tertiary referral center. Surg Endosc. 2013;27(6):2005–2012. [DOI] [PubMed] [Google Scholar]
- 14. Srivastava S, Sharma BC, Puri AS, Sachdeva S, Jain L, Jindal A. Impact of completion of primary biliary procedure on outcome of endoscopic retrograde cholangiopancreatographic related perforation. Endosc Int Open. 2017;5(8):E706–E709. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Weiser R, Pencovich N, Mlynarsky L, Berliner‐Senderey A, Lahat G, Santo E, et al. Management of endoscopic retrograde cholangiopancreatography‐related perforations: Experience of a tertiary center. Surgery. 2017;161(4):920–929. [DOI] [PubMed] [Google Scholar]
- 16. Motomura Y, Akahoshi K, Gibo J, Kanayama K, Fukuda S, Hamada S, et al. Immediate detection of endoscopic retrograde cholangiopancreatography‐related periampullary perforation: Fluoroscopy or endoscopy? World J Gastroenterol. 2014;20(42):15797–15804. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Li G, Chen Y, Zhou X, Lv N. Early management experience of perforation after ERCP. Gastroenterol Res Pract. 2012;2012:657418. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Sarr MG, Fishman EK, Milligan FD, Siegelman SS, Cameron JL. Pancreatitis or duodenal perforation after peri‐Vaterian therapeutic endoscopic procedures: Diagnosis, differentiation, and management. Surgery. 1986;100(3):461–466. [PubMed] [Google Scholar]
- 19. Zissin R, Shapiro‐Feinberg M, Oscadchy A, Pomeranz I, Leichtmann G, Novis B. Retroperitoneal perforation during endoscopic sphincterotomy: Imaging findings. Abdom Imaging. 2000;25(3):279–282. [DOI] [PubMed] [Google Scholar]
- 20. Paspatis GA, Dumonceau JM, Barthet M, Meisner S, Repici A, Saunders BP, et al. Diagnosis and management of iatrogenic endoscopic perforations: European Society of Gastrointestinal Endoscopy (ESGE) Position Statement. Endoscopy. 2014;46(8):693–711. [DOI] [PubMed] [Google Scholar]
- 21. Theopistos V, Theocharis G, Konstantakis C, Kitrou P, Kehagias I, Triantos C, et al. Non‐operative Management of Type 2 ERCP‐related retroperitoneal duodenal perforations: A 9‐year experience from a single center. Gastroenterology Res. 2018;11(3):207–212. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. McCarthy CJ, Butros SR, Dawson SL, Arellano RS. Image‐guided percutaneous management of duodenal perforation following endoscopic retrograde cholangiopancreatography (ERCP): Assessment of efficacy and safety. Clin Radiol. 2018;73(3):319 e9–319 e15. [DOI] [PubMed] [Google Scholar]