Surgically altered anatomy remains a challenge when performing advanced endoscopic procedures such as endoscopic retrograde cholangiopancreatography (ERCP). Patients are frequently encountered in clinical practice with a history of the following surgical procedures that affect ERCP success: Billroth II, Roux-en-Y gastric bypass, Roux-en-Y hepaticojejunostomy and pancreaticoduodenectomy (PD; Whipple procedure). With respect to PD, immense surgical improvements have led to significantly decreased mortality. However, postoperative morbidity is still present in many cases, including complications such as biliary stricture, pancreaticocutaneous fistula, and delayed gastric emptying.^1–3 As the Whipple procedure has become more routine in the management of benign and malignant conditions, the need for endoscopic therapeutic interventions has grown in tandem.

Patients with post-PD anatomy can present with pancreaticobiliary symptoms stemming from changes in their anatomy, in addition to common pathologies including choledocholithiasis. Obstructive jaundice can occur from recurrent disease or surgical bilioenteric anastomotic strictures. Prior studies have demonstrated that 2.6% of patients develop bilioenteric strictures within 13 months, and 8.2% in 5 years.^3,4 Likewise, recurrent pancreatitis or abdominal pain can occur from a stricture of the pancreatic duct or pancreaticojejunal anastomosis. However, endoscopic management and evaluation with ERCP can be challenging. The difficulties in performing endoscopy are not only related to navigating through altered anatomy, but also to the lack of devices specifically designed for post-surgical anatomy.^5,6 Several techniques have been attempted to improve the success of pancreaticobiliary interventions. Most notably, single-balloon enteroscopy, double-balloon enteroscopy, and rotational overtubes have been utilized to improve post-Whipple ERCP technical success.^7–11 In addition, endoscopic ultrasound (EUS)-guided therapies including EUS-pancreatic duct drainage (EUS-PDD) and EUS-enteroenterostomy have been described as possible solutions to assist in ERCP.^12–16

Although ERCP maintains a success rate of 90% to 95% in patients with native gastric and pancreaticoduodenal anatomy, the data are less robust in patients with post-PD anatomy.^17,18 Moreover, the use of a rigidizing overtube in ERCP has not been well described as an adjunctive device in post-surgical anatomy.¹⁹ In this study, a retrospective analysis was conducted to evaluate the technical success rate, clinical success rate, reasons for failure, and adverse events in post-PD patients undergoing ERCP with and without a rigidizing overtube.

This was a retrospective study performed at MD Anderson Cancer Center between 2006 and 2021. The protocol (RCR05-0672) was approved by the institutional review board.

Patient and procedural characteristics

Patients who had undergone PD (Whipple’s procedure) and required ERCP were included. Electronic medical records were used to capture pertinent demographic information, oncologic history, indications for PD, biochemical data, and follow-up information. ERCP procedural reports were analyzed for the indication, selection of the endoscope used, primary anastomosis evaluated, stricture location, treatment for the stricture, and stent characteristics. Each procedure was included as a separate entry for patients who received multiple ERCPs. A Pathfinder (Neptune Medical) rigidizing overtube was used as an adjunctive device in a subset of cases.

Outcomes

The major outcomes were the technical and clinical success rates. The technical success rate was defined as the ability to perform an intended therapeutic intervention, including the evaluation of a stricture, stent placement, and stent removal. The clinical success rate was defined as a documented improvement of bilirubin to normal or 50% of the peak value within 4 weeks. The clinical success rate was not calculated for pancreatic indications. The additional outcomes evaluated included the reasons for technical failure and an analysis of the adverse events.

Statistics

Descriptive statistics were used to characterize the demographic and clinical characteristics of the study population. The statistical analysis was performed using Stata/SE version 16.0 statistical software (Stata Corp.).

Patient characteristics

In total, 102 ERCPs were performed in 47 patients with prior PD. Within this sample, there were 31 male (66.0%) and 16 female (34.0%) patients. The patients’ mean age was 61.5 years of age. The indication for PD in the majority of patients was pancreatic adenocarcinoma (n = 26; 55.3%). A comprehensive list of patient characteristics, including the primary malignancy and neoadjuvant therapy, can be found in Table 1.

Table 1 . Patient Characteristics.

Age (yr)	61.5 ± 11.6
Sample size (n)	47
Male	31 (66.0)
Female	16 (34.0)
Primary malignancy
Pancreatic adenocarcinoma	26 (55.3)
Pancreatic neuroendocrine tumor	5 (10.6)
Ampullary adenocarcinoma	4 (8.5)
Ampullary adenoma	1 (2.1)
Ampullary neuroendocrine tumor	1 (2.1)
Duodenal adenocarcinoma	2 (4.3)
Duodenal adenoma with high grade dysplasia	1 (2.1)
Duodenal neuroendocrine tumor	1 (2.1)
Intraductal papillary mucinous neoplasm	3 (6.4)
Mucinous cystic neoplasm	1 (2.1)
Metastatic renal cell carcinoma	1 (2.1)
Cholangiocarcinoma	1 (2.1)
Neoadjuvant therapy
Chemoradiation	11 (23.4)
Chemotherapy	9 (19.1)
Radiation	2 (4.3)
No neoadjuvant therapy	25 (53.2)

Procedure details

Among the 102 procedures, there were 98 biliary and 4 pancreatic indications for ERCP, the majority of which included obstructive jaundice (n = 22; 21.6%), cholangitis (n = 35; 34.3%), and stent evaluation, including removal or exchange (n = 30, 29.4%). A complete list of procedural indications can be found in Table 2. Choledochojejunal or pancreaticojejunal anastomoses were identified in 90 procedures (89 choledochojejunal and 1 pancreaticojejunal). The anastomosis was reached mainly with a therapeutic upper endoscope (n = 31; 34.4%), adult colonoscope (n = 20; 22.2%), and pediatric colonoscope (n = 17; 18.9%). The anastomosis was reached using a duodenoscope (n = 1; 1.1%) and upper endoscope (n = 2; 2.2%) in a minority of patients. Of note, the endoscope used to reach the anastomosis was not specified in 19 (21.1%) procedures.

Table 2 . Procedure Characteristics.

	n (%)
Total procedures (n)	102
Indications
Biliary	98 (96.1)
Obstructive jaundice	22 (21.6)
Cholangitis	35 (34.3)
Choledocholithiasis	1 (1.0)
Ductal dilation	4 (3.9)
Evaluation of stricture	5 (4.9)
Stent evaluation	30 (29.4)
Elevated liver enzymes	1 (1.0)
Pancreatic	4 (3.9)
Pancreatitis	4 (3.9)
Endoscope used for successful interventions
Adult colonoscope	20 (23.8)
Duodenoscope	1 (1.2)
Pediatric colonoscope	15 (17.9)
Therapeutic upper endoscope	30 (35.7)
Upper endoscope	2 (2.4)
Not specified	16 (19.0)
Indicated cannulations	101
Biliary	97 (96.0)
Successful	86 (88.7)
Unsuccessful	11 (11.3)
Pancreatic	4 (4.0)
Successful	0 (0.0)
Unsuccessful	4 (100.0)
Indicated cannulations (adjusted after anastomosis identification)	89 (87.3)
Biliary	88 (98.9)
Successful	86 (97.7)
Unsuccessful	2 (2.3)
Pancreatic	1 (1.1)
Successful	0 (0.0)
Unsuccessful	1 (100.0)
Cholangiograms and pancreaticograms conducted	90 (88.2)
Biliary	89 (98.9)
Choledochojejunal stricture	51 (57.3)
Common hepatic stricture	2 (2.2)
Hilar stricture	2 (2.2)
Intrahepatic stricture	6 (6.7)
No stricture	28 (31.5)
Pancreatic	1 (1.1)
No stricture	1 (100.0)
Stent characteristics	55
Plastic	49 (89.1)
Uncovered metal	4 (7.3)
Covered metal	2 (3.6)

Biliary (n = 97) or pancreatic (n = 4) duct cannulation was indicated in 101 procedures. One patient underwent ERCP and stent removal without the need for cannulation or additional biliary intervention. Successful cannulation occurred in 86 (88.7%) patients requiring biliary intervention, and 0 (0.0%) patients requiring pancreatic intervention. When the anastomosis of interest was identified, cannulation was successful in 86 (97.7%) patients requiring biliary interventions and 0 (0.0%) patients requiring pancreatic interventions.

Among all patients, 90 (88.2%) patients had cholangiograms (n = 89; 98.9%) or pancreaticograms (n = 1; 1.1%) conducted. In the patients with cholangiograms, most strictures were choledochojejunal in origin (n = 51; 57.3%). Conversely, 28 patients did not have strictures (31.5%). The other strictures identified on cholangiograms included intrahepatic (n = 6; 6.7%), hilar (n = 2; 2.2%), and common hepatic duct (n = 2; 2.2) strictures. The single pancreaticogram conducted showed no evidence of stricture or additional pathology. Among the patients requiring stent placement, 49 required plastic stent placement (n = 49; 89.1%). Two patients had fully covered metal stents deployed (3.6%), and four patients had uncovered metal stents inserted (7.3%). All patients who underwent uncovered stent deployment had choledochojejunal strictures. Of the patients receiving fully covered metal stent placement, one patient had a choledochojejunal stricture and the other did not have evidence of stricture. All other patients had plastic stents deployed regardless of stricture location (Table 3).

Table 3 . Stricture Types and Stents Utilized.

Stricture type	Plastic	Covered metal	Uncovered metal
Choledochojejunal	33	1	4
Common hepatic	2	0	0
Hilar	2	0	0
Intrahepatic	6	0	0
No stricture	6	1	0

Major outcomes

The technical success rate for ERCP in post-PD patients was 82.4% (84/102). Biliary procedures had a technical success rate of 84.7% (83/98), whereas pancreatic procedures had a technical success rate of 25.0% (1/4). The clinical success rate was 75.5% (74/98). Failure of biliary interventions mostly occurred secondary to the inability to reach or visualize the anastomosis (n = 10). Three additional procedures were unsuccessful given the degree of stricture at the level of the choledochojejunal anastomosis and the inability to pass instruments across the stricture. One of these procedures included an EUS-guided rendezvous procedure, in which the guidewire could not be passed antegrade across the stricture after obtaining biliary access. Moreover, two procedures were unsuccessful secondary to stent-related issues: one failure of stent deployment, and one in which there was no appropriate stent size available. With regard to pancreatic evaluation, technical success was achieved in one case for the evaluation of a suspected pancreaticojejunal stricture. The anastomosis was visualized, but the pancreatic duct could not be cannulated. However, a pancreaticogram was conducted and did not demonstrate any stricture or abnormality. All three unsuccessful cases were related to difficulty in identifying the pancreaticojejunal anastomosis and the inability to perform a pancreaticogram (Table 4).

Table 4 . Overall Technical and Clinical Success.

	n (%)
Overall technical success
Biliary and pancreatic procedures	102
Successful	84 (82.4)
Unsuccessful	18 (17.6)
Overall biliary technical success
Biliary procedures	98
Successful	83 (84.7)
Unsuccessful	15 (15.3)
Overall pancreatic technical success
Pancreatic procedures	4
Successful	1 (25.0)
Unsuccessful	3 (75.0)
Overall clinical success
Biliary procedures	98
Successful	74 (75.5)
Unsuccessful	24 (24.5)
Pathfinder technical success
Pathfinder procedures	11
Successful	10 (90.9)
Unsuccessful	1 (9.1)
Pathfinder clinical success
Pathfinder procedures	11
Successful	10 (90.9)
Unsuccessful	1 (9.1)

Rigidizing overtubes

Pathfinder rigidizing overtubes were utilized in 11 ERCPs and 5 patients. The indications included stent evaluation, including removal, insertion, or exchange (n = 5; 45.5%), obstructive jaundice (n = 3; 27.3%), evaluation of stricture (n = 2; 18.2%) and cholangitis (n = 1; 9.1%). The choledochojejunal anastomosis was reached in all cases (n = 11; 100%). The endoscopes used with rigidizing overtubes included therapeutic endoscopes (n = 6; 54.5%) and adult colonoscopes (n = 3; 27.3%). The remaining endoscopes used with rigidizing overtubes were unspecified (n = 2; 18.2%). A cholangiogram was obtained in all cases. The majority of strictures were choledochojejunal in origin (n = 9; 81.8%). There was one additional intrahepatic stricture (n = 1; 9.1%). The last patient did not have a stricture identified on a cholangiogram. All stents deployed for intervention were plastic (n = 9; 100%) (Table 5). Technical success was achieved in 10 of 11 procedures (90.9%). The unsuccessful procedure (described above) was secondary to a severe choledochojejunal stricture, in which a guidewire could not be passed antegrade under an EUS-rendezvous approach. Clinical success was achieved in 10 of 11 procedures (90.9%).

Table 5 . Procedure Characteristics (Pathfinder).

	n (%)
Total procedures	11 (10.8)
Indications
Biliary	11 (100.0)
Obstructive jaundice	3 (27.3)
Cholangitis	1 (9.1)
Choledocholithiasis	0 (0.0)
Ductal dilation	0 (0.0)
Evaluation of stricture	2 (18.2)
Stent evaluation	5 (45.5)
Elevated liver enzymes	0 (0.0)
Endoscope used for procedure	11
Adult colonoscope	3 (27.3)
Duodenoscope	0 (0.0)
Pediatric colonoscope	0 (0.0)
Therapeutic upper endoscope	6 (54.5)
Upper endoscope	0 (0.0)
Not specified	2 (18.2)
Indicated cannulations*	11
Biliary	11 (100.0)
Successful	11 (100.0)
Unsuccessful	0 (0.0)
Cholangiograms conducted	11
Choledochojejunal stricture	9 (81.8)
Common hepatic stricture	0 (0.0)
Hilar stricture	0 (0.0)
Intrahepatic stricture	1 (9.1)
No stricture	1 (9.1)
Stent characteristics	9
Plastic	9 (100.0)
Uncovered metal	0 (0.0)
Covered metal	0 (0.0)

Adverse events

There were 5 (4.9%) adverse events in the 102 procedures. There were three cases of fever, and one case of cholangitis, all of which resolved either with supportive care or antibiotics. One patient had post-procedure nausea and abdominal pain, which resolved with supportive management. No additional procedure-related adverse events occurred, including bleeding, perforation, pancreatitis, the need for surgical intervention, or death following ERCP. In addition, there were no procedure-related adverse events in patients in whom rigidizing overtubes were utilized.

A retrospective analysis of 47 patients undergoing 102 ERCPs post-PD demonstrated an overall technical success rate of 82.4% and clinical success rate of 75.5%. Rigidizing overtubes were utilized in a subgroup of patients with a technical success rate of 90.9% and clinical success rate of 90.9%. The overall adverse event rate was 4.9%, and most adverse events were mild (self-resolving fevers). Therefore, ERCP is a safe and effective therapeutic procedure in post-PD anatomy for biliary intervention. The use of a rigidizing overtube may enhance technical and clinical success in ERCP; however, additional studies need to be undertaken for further evaluation.

In a review of the literature, our biliary technical success rate was similar to that reported by Chahal et al.¹⁷ In comparison to the study of Chahal et al,¹⁷ in which most procedures were conducted with duodenoscopes, our procedures were conducted mainly with therapeutic upper endoscopes, adult colonoscopes, and pediatric colonoscopes with similar success. Chahal et al¹⁷ reported two adverse events: one medically managed retroperitoneal perforation and one self-resolving Mallory Weiss tear. The increased rate of perforation is a well-described adverse event of side-viewing endoscopes, and as such, the utilization of duodenoscopes was limited in our sample. Despite the use of forward-viewing endoscopes, similar technical success was achieved.

In our study, biliary cannulation was highly successful (97.7%) if the choledochojejunostomy was identified. Most failed biliary ERCPs were related to an inability to visualize or navigate the endoscope to the level of the anastomosis. Advancement of the endoscope to the choledochojejunostomy can be challenging. It requires accurate identification of the afferent limb and traversal to the anastomotic site. Post-PD anatomy encourages acute bowel angulation and loop formation, making it difficult to advance to the choledochojejunostomy site (Fig. 1, 2). Moreover, the limited visibility of a side-viewing scope, decreased flexibility, and larger caliber duodenoscope hinders the traversal of the endoscope to the anastomosis and heightens the risk for perforation. To mitigate some of the anatomic challenges of post-PD anatomy, a novel finding in this study related to the use of Pathfinder rigidizing overtubes in a subgroup of patients requiring ERCP. Rigidizing overtubes can dynamically rigidize and transform between a flexible and rigid state, which can straighten acute angulations, prevent loop formation, and stabilize endoscopes for therapeutic intervention (Fig. 3–5). Based on this limited sample, rigidizing overtubes can be a safe and effective adjunctive tool in reaching the choledochojejunostomy site and facilitating successful cannulation in patients with post-PD anatomy. Most studies have cited the successful use of rigidizing overtubes in difficult colonoscopies, but the literature has been sparse in demonstrating its use in ERCP, especially in patients with altered anatomy.¹⁹ Prior studies have demonstrated the use of rotational overtubes without dynamic rigidization in facilitating ERCP in patients with surgically altered anatomy.⁷ The overall success rate of ERCP performed with rotational overtubes reported in the literature was lower (65%) than that observed for dynamic rigidizing overtubes (90.9%) in our sample. However, additional studies assessing the efficacy of rigidizing overtubes in ERCP are required before further conclusions can be drawn given the small size of our sample.

Figure 1. Endoscopic configuration (A) to reach choledochojejunostomy. Low-grade biliary stricture (B, C) with an uncovered metal stent placed across choledochojejunostomy stricture (D).

Figure 2. Choledochojejunostomy stricture (yellow arrow in A) undergoing balloon dilation (B, C) and double pigtail stent placement (D).

Figure 3. The colonoscope (not visualized) and rigidizing overtube were advanced to the choledochojejunal anastomosis. The colonoscope (A) was withdrawn and a therapeutic endoscope (B) was placed through the rigidizing overtube. The preexisting stents were removed (C) and wire-guided cannulation/cholangiography was conducted (C) with subsequent stent insertion (D).

Figure 4. Dynamic rigidizing overtube in its flexible conformation.

Figure 5. Dynamic rigidizing overtube in its rigidized conformation.

Device-assisted enteroscopy has emerged as the leading modality for reaching the enterobiliary anastomosis and facilitating ERCP in patients with surgically altered anatomy.^6,8–10 However, both single-balloon and double-balloon enteroscopes have longer working lengths and smaller working channels, which preclude the utilization of standard ERCP devices and larger caliber stents. Contrastingly, colonoscopes and short double-balloon enteroscopes can provide additional length while maintaining compatibility with some devices and allow larger caliber stent placement, but lack an elevator for therapeutic intervention. More recently, EUS-guided enteroenterostomy has emerged as a possible solution to reaching the anastomotic site that may enable appropriate device utilization with a standard duodenoscope and facilitate cannulation with an elevator.^12,15,20 In addition, the use of rigidizing overtubes, as demonstrated in this study, may allow adequate access to the anastomotic site for biliary intervention.¹⁹

Our sample furnished limited data on the endoscopic evaluation and management of the pancreas in post-PD anatomy. However, prior literature has demonstrated limited technical success in post-PD patients.^17,18 Similar technical limitations in traversing the afferent limb are heightened in pancreatic evaluations given the longer length of the afferent limb that is required to reach the pancreaticojejunostomy. This has led to the development of EUS-guided pancreatic interventions, including enteroenterostomy, anterograde stenting, and the rendezvous technique. A recent study compared the safety and efficacy of EUS-PDD and enteroscopy-assisted endoscopic retrograde pancreatography (e-ERP) for the management of pancreatic-related adverse events in patients with post-PD anatomy.¹⁶ EUS-PDD had a higher technical success rate than e-ERP (92.5% vs. 20%). However, there were significantly more adverse events in patients who received EUS-PDD than in those who received e-ERP (35% vs. 2.9%). While improvements are necessary to reduce the occurrence of procedure-related adverse events when performing EUS-PDD, it is emerging as a promising alternative to the traditional methods of pancreatic management in the post-PD population.

A limitation of this study is that it was a retrospective analysis conducted at a single tertiary care center. The study was also limited by a relatively small number of patients. Although there were a larger number of total ERCPs than in prior studies, some patients received multiple ERCPs. Patients with prior interventions including stent placement may serve as a roadmap for identification of choledochojejunal anastomosis, facilitating a subsequent successful intervention. As a result, the number of repeated ERCPs may have biased our technical success rate. In addition, as a large comprehensive cancer center, post-surgical anatomy may be more commonly encountered at our center than at other centers. As such, the reported technical and clinical success rates at our center may not be generalizable at a population level. Moreover, although the utilization of a rigidizing overtube demonstrated high technical success, the device was used in a limited number of patients. Finally, the number of patients in our sample referred for pancreatic evaluation was low. As a result, it is difficult to draw any conclusions regarding ERCP in post-PD patients for pancreatic indications from our study. Larger studies are required to further assess the utility of rigidizing overtubes in patients with surgically altered anatomy. Additional studies will also be required to evaluate conventional ERCP for pancreatic indications and to assess the generalizability of our results across samples in post-surgical patients.

In conclusion, ERCP is an effective therapeutic biliary procedure in post-PD patients. This study demonstrates the successful use of a rigidizing overtube in facilitating post-PD ERCP. However, post-PD ERCP remains a technically challenging procedure. Altered anatomy may preclude successful navigation of the afferent limb, given the additional length that needs to be traversed along with sharp angulation and looping in the gastrointestinal tract. Moreover, the lack of devices specifically designed for post-surgical anatomy poses an additional challenge. However, additional devices and techniques, including balloon-assisted enteroscopy, EUS-guided techniques, and rigidizing overtubes may be potential solutions to assist in post-PD patients requiring ERCP. In particular, the use of a rigidizing overtube may assist in straightening the angulation, preventing loop formation, and enhancing stability to reach the anastomotic site and facilitate the intervention. However, further studies will need to be conducted to evaluate the use of rigidizing overtubes in post-PD ERCP patients.

The data that support the findings of this study are available from the corresponding author upon reasonable request.

None.

No potential conflict of interest relevant to this article was reported.