• Users Online: 200
  • Print this page
  • Email this page

 Table of Contents  
Year : 2022  |  Volume : 12  |  Issue : 4  |  Page : 145-151

Surgical outcomes after endoscopic retrograde cholangiopancreatography and sphincterotomy associated duodenal perforations: Experience from a tertiary care centre in India

1 Department of Surgical Gastroenterology and Liver Transplantation, Sir Ganga Ram Hospital, New Delhi; Department of Surgical Gastroenterology and HPB Surgery, Dhaara Speciality Hospital, Bengaluru, Karnataka, India
2 Department of Surgical Gastroenterology and Liver Transplantation, Sir Ganga Ram Hospital, New Delhi, India

Date of Submission05-Jun-2022
Date of Decision15-Jul-2022
Date of Acceptance02-Aug-2022
Date of Web Publication30-Aug-2022

Correspondence Address:
Dr. Srinivas Bojanapu
Department of Surgical Gastroenterology and HPB Surgery, Dhaara Speciality Hospital, #61, Jakkur Road, Yelahanka, Bengaluru, Karnataka - 560 064
Login to access the Email id

Source of Support: None, Conflict of Interest: None

DOI: 10.4103/cmrp.cmrp_51_22

Rights and Permissions

Background: The advent and progress of endoscopic retrograde cholangiopancreatography (ERCP) for hepato-pancreaticobiliary diseases have introduced a paradigm shift in their treatment but have also been associated with complications that have been difficult to manage. We examined the outcome of patients who had duodenal perforations from periampullary endoscopic procedures who needed surgical intervention.
Aim: To study the outcomes in patients who underwent surgical intervention for ERCP associated duodenal perforation.
Materials and Methods: Between January 2001 and November 2021, we retrieved the details of 20 patients who had been operated on for duodenal perforations following ERCP from a prospectively maintained database.
Results: There were nine males and 11 females whose mean age was 51.3 (range 29–81) years. The most common indication for the endoscopic procedure was choledocholithiasis in 16, followed by biliary stricture in 3 and malignancy in 1. The median duration between the diagnosis of perforation and surgery was 8.3 days (range 0–48). The mean hospital stay was 21.5 (range 5–60) days. Four (20%) patients underwent re-exploration. Seven (35%) patients died, in all of whom the diagnosis had been missed at the time of the procedure (P = 0.015). Eighty per cent of mortality occurred in patients with a Boey score of three.
Conclusion: ERCP-related duodenal perforations though uncommon, are dreaded complications and do not have universally accepted standard management. Patients with delayed diagnosis of duodenal perforation and higher Boey scores have higher mortality rates.

Keywords: Duodenal, endoscopic retrograde cholangiopancreatography, perforation, retroperitoneal air, sphincterotomy, Stapfer's classification

How to cite this article:
Bojanapu S, Vajpeyajula U, Das AP, Mehta N, Nundy S. Surgical outcomes after endoscopic retrograde cholangiopancreatography and sphincterotomy associated duodenal perforations: Experience from a tertiary care centre in India. Curr Med Res Pract 2022;12:145-51

How to cite this URL:
Bojanapu S, Vajpeyajula U, Das AP, Mehta N, Nundy S. Surgical outcomes after endoscopic retrograde cholangiopancreatography and sphincterotomy associated duodenal perforations: Experience from a tertiary care centre in India. Curr Med Res Pract [serial online] 2022 [cited 2023 Feb 5];12:145-51. Available from: http://www.cmrpjournal.org/text.asp?2022/12/4/145/355205

  Introduction Top

The advent of endoscopic retrograde cholangio pancreatography (ERCP), its initial use in the diagnosis, and its evolution into a predominantly therapeutic procedure has been a major advance in the management of patients with hepato-pancreatobiliary (HPB) diseases. ERCP has complemented and replaced surgical treatment for conditions such as choledocholithiasis and biliary stricture, among others. Nevertheless, ERCP is technically demanding and is associated with a higher risk of adverse events when compared with other endoscopic procedures.[1] The widespread availability of ERCP and its use has been accompanied by some morbidity and occasional mortality, which are now being more often recognised and reported.[2],[3] Its success rates vary markedly, as do its associated complications. A systematic survey in 2007[4] reported an overall complication rate of 6.85% with a 0.33% mortality.

The incidence of duodenal perforation in the published literature following ERCP is 0.3%–1.3%, with a mortality rate of 18%. Whether primary surgical therapy[5] is better than conservative management is still controversial. Recently, the trend has been towards conservative management of ERCP-related perforations, with close monitoring and surgery indicated only in certain types of perforations and clinical situations.[6] In this study, we present our experience with patients who underwent surgical management for duodenal perforation following ERCP and their outcomes.

  Materials and Methods Top

We performed a retrospective review from our prospectively maintained electronic database of all patients who underwent surgery for ERCP-related duodenal perforation between January 2001 and November 2021 at Sir Ganga Ram Hospital, New Delhi. Their demographic details, indication for ERCP, papillotomy, time of identification, conservative management, time to surgery, type of surgery, intraoperative findings, length of hospital stay, clinical course and outcomes were recorded. We classified duodenal perforations according to the Stapfer system; Type I – Lateral or medial wall perforation, Type II – PeriVaterian injuries, Type III – Distal bile duct injuries and Type IV – Retroperitoneal air alone.[7] We used the Boey score for perforated duodenal ulcer, i.e., accompanying major medical illness, pre-operative shock and longstanding perforation of more than 24 h, to predict morbidity and mortality outcomes. The Clavien–Dindo classification was used to grade post-operative complications (mild – Grade I and II, severe – Grade III and IV). The study was cleared by the Institutional Ethical Committee of Sir Ganga Ram Hospital, New Delhi. EC No.: EC/04/21/1873. Date of Approval: 04th June 2021.

Statistical analysis

Continuous variables are presented as mean ± standard deviation or median interquartile range for non-normally distributed data. Categorical variables expressed as frequencies and percentages were compared using Fisher's exact test. For continuous variables, the Mann–Whitney U-test was used. Receiver operating characteristic (ROC) curve analysis was used to estimate the predictive ability of the Boey score. The area under the curve (AUC) falling under ROC area indicates the probability of post-operative morbidity or mortality and actual post-operative condition and is considered perfect (1), good (>0.8), moderate (0.6-0.8) and poor (<0.6) for AUC. Statistical testing was conducted with the statistical package for the social science system version (SPSS) 25.0 for Windows (IBM Corp. Released 2017. IBM SPSS Statistics for Windows, Version 25.0. Armonk, NY: IBM Corp).

The details of the patients and operations performed are listed in [Table 1] and [Table 2].
Table 1: Cohort Details- Demographics, ERCP, Pre-operative and intraoperative variables

Click here to view
Table 2: Cohort Details- Intraoperative and postoperative variables

Click here to view

  Results Top

Pre-operative variables

The most common indication for ERCP was choledocholithiasis (n = 16, 80%), followed by biliary stricture (n = 3, 15%) and pancreatic head cancer (n = 1, 5%). Papillotomy was performed in 18 (90%) patients. Duodenal perforation was identified during the procedure in 8 (40%) patients either through a guidewire or a leak of dye and missed in 12 (60%) patients.

The mean duration between the identification of perforation and surgical intervention was 8.3 days (range 0–48). In 9 (45%) patients, initial conservative management was by stopping oral intake, inserting a nasogastric or nasojejunal tube insertion and standard supportive measures. Surgery was performed on patients who did not improve or showed clinical deterioration. However, 11 (55%) patients in whom a significant leak was detected underwent surgery without initial conservative treatment. Five patients had acute kidney injury, but none of them needed dialysis. [Figure 1] shows computed tomography of abdomen in a patient with ERCP associated duodenal perforation.
Figure 1: Computed tomography of abdomen in a patient with ERCP associated duodenal perforation.

Click here to view

Intraoperative variables and surgery

Of the 20 patients, 15 (75%) patients had a perforation in the second portion of the duodenum (D2), and one each in the third part of the duodenum (D3) and afferent limb of a gastrojejunostomy (patient with a previous Billroth II procedure). In 3 (15%) patients, no perforation was found at laparotomy. Perforations identified intraoperatively were designated according to Stapfer's classification.[7] Nine (45%) patients had Type I, followed by Type II in 8 (40%), and in three patients, no perforation was detected. The median size of bowel perforation was 10 mm (range 5–25 mm). Three patients had evidence of pancreatitis at operation.

Sixteen (80%) patients underwent primary closure of the perforation with an omental patch. Three (15%) patients with no perforation had paraduodenal collection, which was drained. Distal gastrectomy was performed in a patient with a history of gastrojejunostomy; in the cohort, 18 (90%) patients had a diversion such as a gastrojejunostomy and a pyloric exclusion in 4 (20%) patients. In addition, a feeding jejunostomy was done in 18 (90%) patients. Of the remaining two patients, one patient had a previous gastrojejunostomy and nasojejunal tube, and the other had drainage alone.

Out of 20 patients, 3 (15%) patients had had their gallbladders removed before the ERCP. Fifteen (75%) patients had cholecystectomy at the time of surgery. However, in two patients, cholecystectomy was not done as one had TB lymphadenitis in the porta with a stent in situ, and the other was on a high inotropic support.


Out of 20 patients, 4 (20%) underwent re-exploration. The mean duration between the primary and re-operation was 16.5 days (range 10–24 days). In addition, three patients were reexplored for an enteric content leak and one patient for recurrent collection with sepsis. Unfortunately, all four patients succumbed, with the mean time to death after ERCP being 53.2 days (range 36–60 days).

Outcome variables–hospital stay, morbidity and mortality

The mean hospital stay for the whole cohort was 21.5 (5–60) days; for those who died, it was 36 (range 22–60) days. Three patients who were not re-explored had a mean hospital stay of 17.6 days (range 5–33 days) to death. According to the Clavien–Dindo classification, 9 (45%) patients had mild complications, and four had severe complications. A total of 7 (35%) patients died during the study, 5 (71%) were females, and 2 (29%) were males. Of these, 4 (57%) patients had been re-explored. Three were re-explored for enteric content leak and one patient for recurrent collection with persistent fever spikes. We found that patients with a higher Boey score had higher mortality, with death occurring in 4 (80%) patients having a score of 3 and 3 (50%) patients with a score of 2. ROC curve analysis demonstrated AUC of .863 (0.701–1.00) for the cohort with P = 0.041.

Association of mortality with pre-operative, intraoperative and post-operative variables

We also compared the association of mortality with various pre-operative, intraoperative and post-operative variables. Most importantly, the diagnosis of duodenal perforation at the time of ERCP was the only significant (P = 0.015) variable influencing outcomes in such patients. The type of perforation, surgical procedure, the time interval between diagnosis and surgery, primary surgery versus conservative management, followed by surgery, and the presence of pancreatitis did not significantly affect mortality. Although patients with organ failure had higher mortality, this did not reach statistical significance (P = 0.06).

A subgroup analysis comparing mortality in the primary surgery group with surgery following conservative management did not yield any statistically significant difference.

  Discussion Top

ERCP, as a diagnostic and therapeutic tool in HPB diseases, is now widely available, and there has been a major increase in the number of procedures performed. The associated complications are, expectedly, now more evident, and so are the management strategies for dealing with them. The incidence of ERCP-associated duodenal perforations ranges between 0.01% and 2.1%.[8],[9]

However, the treatment of this life-threatening condition lacks standardisation owing to its infrequent occurrence, different patient characteristics and divergent management practices. Duodenal perforation carries high mortality if not managed early and rigorously. Overall, ERCP-associated mortality is approximately 1.5%.[10] Studies quote up to 46% mortality in patients undergoing surgical repair for ERCP-associated duodenal perforation.[8],[11]

The classification of ERCP-associated perforations put forward by Stapfer et al.[7] is now widely accepted and guides clinicians in deciding on the preferred type of management. Though used preoperatively to stratify patients, we found only Type I and II intraoperatively, which may not represent the actual type of perforation in the presence of inflammation and local sepsis.

We found ERCP-related perforations more common in females than in other studies,[11],[12] probably due to their higher prevalence of gallstone disease. For example, Kim et al.[13] found that females comprised 53% of the study group, similarly, in the present study, females represented 55% (n = 11) of the cohort, but the gender difference was not statistically significant.

The mean age of the patients in the present series was 51.3 (range 29–81) years, similar to other studies[11] where the mean age was 49.7 and 44.6 years in the early and delayed surgery group and in a similar study by Chaudhary and Aranya,[5] where the age ranged between 36 and 72 years. Others, however, reported a higher median age of 67 and 72.5 years in their studies.[14],[15]

The indications for ERCP vary widely, with one study[11] reporting 68% (n = 34) for choledocholithiasis, whereas in another,[14] it was done for bile duct stones in only 19% (n = 71). In some reports, the indication for ERCP was malignancy in 33% (n = 123) patients.

Duodenal perforation is usually identified at the time of ERCP. However, studies report different time intervals from immediate identification to delayed diagnosis, varying from 46.2% to 100% in their studies.[2],[8],[9],[11],[13],[16] In the present study, only 40% (n = 8) of patients had their duodenal perforations identified at the time of ERCP. We found this to impact the outcomes of patients undergoing surgical intervention significantly (P = 0.015). Delayed diagnosis and surgery may cause significant morbidity and mortality due to ongoing sepsis and result in organ failure.[8]

The median time interval between ERCP perforation and surgery was 3 days (range 0–49) in a study,[14] and in another study,[11] the median time to surgery was 1 day[1],[2],[3],[4],[5],[6] for primary surgery group and 20.5 days (range 13–35) for the group needing surgery after failed conservative management. However, in our study, it was 8.3 days (range 0–48), and the cohort included both primary surgery and delayed surgery patients together and did not significantly influence mortality. Time to surgery differs based on the treatment strategy adopted; hence, a wide range exists in the literature and the present study.

Stapfer's perforation type and size did not influence the surgical outcomes in our patients, which was similar to a study by Patil et al.[11] though an accurate classification of a perforation in an inflamed field may not have been correct.

The type of surgical intervention in the present study was laparotomy, perforation closure, cholecystectomy, exploration of the common bile duct with T-tube drainage, gastro-jejunostomy and a feeding jejunostomy. It was similar to the procedures performed by Kim et al.[17] and Fatima et al.[3] However, Alfieri et al.[16] selectively used retroperitoneal laparostomy for retroperitoneal and an anterior approach for intraperitoneal fluid collections. The type of surgery did not have a significant influence on outcomes. A retrospective study[11] compared T-tube versus no T-tube duodenostomy added to biliary drainage and perforation repair showed an increased incidence of post-operative duodenal leaks in patients who had T-tube duodenostomies. Hence, it is still uncertain what constitutes the appropriate procedure for this condition.

In a study by Wu et al.[18] of 21 patients, re-exploration was required in 2 (10%) (n = 2) and Ercan et al. reexplored 4 (16%) (n = 4).[19] The most common indication for re-exploration was a duodenal leak. Whereas, in the present study, 20% (n = 4) underwent re-exploration, and all patients who underwent re-exploration died.

The mortality after-surgery patients ranges from 16.7% (n = 3)[15] to 46.6% (n = 7).[8] Studies with a larger cohort have reported mortality rates between 21% and 45%.[11],[14],[19] We had a mortality rate of 35% (n = 7) of patients, mainly those with delayed diagnosis of duodenal perforation.

The median duration of hospital stay in the present study was 21 (range 5–60 days), which was similar to one study[11] where there was a median stay of 16 days (range 4–74), but in another,[16] the mean hospital stay was 53.3 ± 28.4 days.

In the present study, there was no significant difference in outcomes with early or delayed surgical treatment in patients with duodenal perforation, similar to the study by Langerth et al.[14] However, in a study by Ercan et al.,[19] patients who underwent early surgery had a poorer outcome.

Retrospective studies[20] have shown that delay in diagnosis of perforation beyond 24 h is associated with high morbidity. Similarly, we also found that identification of perforation during the procedure and its management had better outcomes, especially in mortality.

Identifying ERCP-associated perforations during the procedure was accompanied by lower mortality, which was statistically significant (P= 0.015) in the present study.

The mortality rate was higher in patients with pre-operative organ failure but did not reach statistical significance (P = 0.06). Comorbidities did not influence the outcomes of the patients in this study.

Most endoscopic perforations are small in size and can be managed conservatively, based on the opinion of endoscopists. Similarly, when perforation occurs in a patient with malignancy,[5] the preferred choice is to manage them conservatively.

Several well-documented predictors of ERCP complications are ERCP-induced pancreatitis in females, previous post-ERCP pancreatitis (PEP), and previous pancreatitis, including risk factors for PEP.[21] For post-ERCP perforation and outcomes, a retrospective multivariate analysis of 11,497 patients reported that patients with previous surgeries like Billroth II had a higher incidence of bowel perforations. In the present study, we found that delay in diagnosis significantly affected the outcomes.[22],[23] In the absence of a specific predictor of outcomes, we used the Boey score[24] to predict mortality, and it showed a good association as depicted in [Figure 2].
Figure 2: ROC curve of boey score. ROC: Receiver operating characteristic

Click here to view

  Conclusion Top

Duodenal perforations after ERCP are now well recognised as severe adverse events associated with significant morbidity and mortality. A high index of suspicion, continuous observation, and early management may result in decreased morbidity and better survival. In addition, the Boey score may predict outcomes in patients undergoing surgery for duodenal perforation.

Financial support and sponsorship


Conflicts of interest

There are no conflicts of interest.

  References Top

Rustagi T, Jamidar PA. Endoscopic retrograde cholangiopancreatography- related adverse events: General overview. Gastrointest Endosc Clin N Am 2015;25:97-106.  Back to cited text no. 1
Christensen M, Matzen P, Schulze S, Rosenberg J. Complications of ERCP: A prospective study. Gastrointest Endosc 2004;60:721-31.  Back to cited text no. 2
Fatima J, Baron TH, Topazian MD, Houghton SG, Iqbal CW, Ott BJ, et al. Pancreaticobiliary and duodenal perforations after periampullary endoscopic procedures: Diagnosis and management. Arch Surg 2007;142:448-54.  Back to cited text no. 3
Andriulli A, Loperfido S, Napolitano G, Niro G, Valvano MR, Spirito F, et al. Incidence rates of post-ERCP complications: A systematic survey of prospective studies. Am J Gastroenterol 2007;102:1781-8.  Back to cited text no. 4
Chaudhary A, Aranya RC. Surgery in perforation after endoscopic sphincterotomy: Sooner, later or not at all? Ann R Coll Surg Engl 1996;78:206-8.  Back to cited text no. 5
Avgerinos DV, Llaguna OH, Lo AY, Voli J, Leitman IM. Management of endoscopic retrograde cholangiopancreatography: Related duodenal perforations. Surg Endosc 2009;23:833-8.  Back to cited text no. 6
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;232:191-8.  Back to cited text no. 7
Tavusbay C, Alper E, Gökova M, Kamer E, Kar H, Atahan K, et al. Management of perforation after endoscopic retrograde cholangiopancreatography. Ulus Travma Acil Cerrahi Derg 2016;22:441-8.  Back to cited text no. 8
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;206:180-6.  Back to cited text no. 9
Cotton PB, Lehman G, Vennes J, Geenen JE, Russell RC, Meyers WC, et al. Endoscopic sphincterotomy complications and their management: An attempt at consensus. Gastrointest Endosc 1991;37:383-93.  Back to cited text no. 10
Patil NS, Solanki N, Mishra PK, Sharma BC, Saluja SS. ERCP-related perforation: An analysis of operative outcomes in a large series over 12 years. Surg Endosc 2020;34:77-87.  Back to cited text no. 11
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:E706-9.  Back to cited text no. 12
Kim BS, Kim IG, Ryu BY, Kim JH, Yoo KS, Baik GH, et al. Management of endoscopic retrograde cholangiopancreatography-related perforations. J Korean Surg Soc 2011;81:195-204.  Back to cited text no. 13
Langerth A, Isaksson B, Karlson BM, Urdzik J, Linder S. ERCP-related perforations: A population-based study of incidence, mortality, and risk factors. Surg Endosc 2020;34:1939-47.  Back to cited text no. 14
Preetha M, Chung YF, Chan WH, Ong HS, Chow PK, Wong WK, et al. Surgical management of endoscopic retrograde cholangiopancreatography-related perforations. ANZ J Surg 2003;73:1011-4.  Back to cited text no. 15
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;27:2005-12.  Back to cited text no. 16
Kim JH, Yoo BM, Kim JH, Kim MW, Kim WH. Management of ERCP-related perforations: Outcomes of single institution in Korea. J Gastrointest Surg 2009;13:728-34.  Back to cited text no. 17
Wu HM, Dixon E, May GR, Sutherland FR. Management of perforation after endoscopic retrograde cholangiopancreatography (ERCP): A population-based review. HPB (Oxford) 2006;8:393-9.  Back to cited text no. 18
Ercan M, Bostanci EB, Dalgic T, Karaman K, Ozogul YB, Ozer I, et al. Surgical outcome of patients with perforation after endoscopic retrograde cholangiopancreatography. J Laparoendosc Adv Surg Tech A 2012;22:371-7.  Back to cited text no. 19
Bell RC, Van Stiegmann G, Goff J, Reveille M, Norton L, Pearlman NW. Decision for surgical management of perforation following endoscopic sphincterotomy. Am Surg 1991;57:237-40.  Back to cited text no. 20
Chen JJ, Wang XM, Liu XQ, Li W, Dong M, Suo ZW, et al. Risk factors for post-ERCP pancreatitis: A systematic review of clinical trials with a large sample size in the past 10 years. Eur J Med Res 2014;19:26.  Back to cited text no. 21
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;70:80-8.  Back to cited text no. 22
Freeman ML. Adverse outcomes of ERCP. Gastrointest Endosc 2002;56:S273-82.  Back to cited text no. 23
Boey J, Choi SK, Poon A, Alagaratnam TT. Risk stratification in perforated duodenal ulcers. A prospective validation of predictive factors. Ann Surg 1987;205:22-6.  Back to cited text no. 24


  [Figure 1], [Figure 2]

  [Table 1], [Table 2]


Similar in PUBMED
   Search Pubmed for
   Search in Google Scholar for
 Related articles
Access Statistics
Email Alert *
Add to My List *
* Registration required (free)

  In this article
Materials and Me...
Article Figures
Article Tables

 Article Access Statistics
    PDF Downloaded94    
    Comments [Add]    

Recommend this journal