Endoscopic retrograde cholangiopancreatography utilisation and outcomes in the first advanced endoscopy centre in Palestine at Al-Ahli Hospital: a retrospective cohort study

Introduction

Endoscopic retrograde cholangiopancreatography (ERCP) is an advanced endoscopic procedure that requires specialised training and experience. ERCP was first introduced in 1968, with the first sphincterotomy performed by Classen and Demling from Germany and Kawai from Japan in 1972.1 2 After that, the procedure gained widespread acceptance. ERCP is generally considered a safe and effective procedure. However, it has been found to carry the highest morbidity and complications rate of all endoscopic procedures (e.g., upper endoscopy and colonoscopy) despite technological advancement, adherence to safety standards and advanced training programmes.3 ERCP-guided interventions were very popular for the diagnosis and management of hepatopancreaticobiliary disorders. Recently, ERCP has evolved from a diagnostic procedure to being predominantly therapeutic, especially after the introduction of less to non-invasive diagnostic alternatives such as endoscopic ultrasound (EUS) and magnetic resonance cholangiopancreatography, respectively.4 Therefore, a careful review of procedure indications and alternatives is critical nowadays.

In Palestine, ERCP was started at Augusta Victoria Hospital in 1995 by a general surgeon, Dr Dibsi, for diagnostic purposes only. He was helped by a visiting team from the UK to carry out the initial few cases. However, in addition to EUS, advanced endoscopy and ERCP were not performed in Palestine until late 2017 when Dr Al Ashhab returned from the US and established a modern endoscopy centre at Al-Ahli Hospital. This, in turn, contributed to the introduction of advanced interventions such as balloon dilation, cholangioscopy, spyglass, radiofrequency ablation, electrohydraulic lithotripsy and EUS-guided rendezvous procedures.

During ERCP, the operator can perform stent placement, stone extraction, tissue biopsy, brushings and manometry.5 Despite mainly focusing on therapeutic indications, ERCP still has indispensable diagnostic indications in patients with a solitary dilated duct, cholangiocarcinoma, primary sclerosing cholangitis and autoimmune cholangitis.6 There are some complications associated with ERCP, such as acute pancreatitis (termed post-ERCP pancreatitis, PEP), bleeding, perforation, local or systemic infections and cardiopulmonary complications.7 8 PEP is the most common, if not the most serious, complication associated with ERCP developing within 24 hours. It occurs in about 5%–7% of patients and seems more frequent in therapeutic versus diagnostic cases.7–9 ERCP complications have varying degrees of severity, ranging from minor to life-threatening and vary significantly from case to case. Published trials have found that younger age, previous episodes of PEP and presence of sphincter of Oddi dysfunction (SOD) have also consistently shown to place the patient at higher risk for the development of another episode.10–12 The overall complications rates in published trials vary from 4% to as high as 15.9%, with procedure-related mortality ranging from 0% to 1%.10–14

This single centre-based study primarily focused on determining the outcomes of performing ERCP, the incidence of inpatient complication post ERCP, the risk factors for procedural-related complications and clinically valuable data for decision-making and development of quality standards. We retrospectively analysed the clinical records of patients undergoing ERCP in the first advanced ERCP centre in Palestine at Al-Ahli Hospital throughout the period from December 2017 to September 2022 and compared these with published figures. We described our experiences with ERCP’s indications, findings and technical success in a sample of the Palestinian population admitted to the centre.

Methods

Study design

A retrospective cohort study was conducted with the approval of the ethical review board of Al-Quds University and Al-Ahli Hospital. All patients provided consents for medical review. The data were collected retrospectively from Nebras data reporting system for all patients who underwent ERCP at our institution from December 2017 through September 2022 (n=1909). The majority of the procedures were conducted by Dr Al Ashhab personally, while the remainder were performed either under direct supervision of Dr Al Ashhab or independently by a junior faculty member. The database contains many potential variables, including demographics, clinical history and context, blood test results, procedural details, technical procedures, procedural findings, diagnoses and complications.

The variables documented for each case include the characteristics of the patients and the specifics of each procedure, amounting to a total of 51 variables. The database contains numerous potential variables, encompassing demographics (patient’s identification number, age, age group, gender, referral site and year of admission), clinical history and context (a history of pancreatitis, gallstones, cholecystitis, hypertension, diabetes, coronary artery disease, malignancy, other chronic diseases, smoking and family history), blood test results, procedural details (previous ERCP, cannulation result, type of procedure, indication, outcome, technique, type of stent, causes of failure and management), technical procedures (pancreatic cannulation, pancreatic injection and stenting, sphincterotomy, clearance and drainage, radiofrequency ablation, cholangioscopy), procedural findings (strictures, ERCP brush cytology and EUS biopsy), complications and hospital stay. Our analysis’s primary outcomes of interest were success rates, procedural outcomes, overall complications, pancreatitis and bleeding. We also discussed the possible risk factors for postprocedural complications and the management of failed procedures. The sociodemographic indications, outcomes and complications were represented as frequency and percentage, while mean and SD were used to represent the continuous variables.

Protocol

ERCP was conducted in a well-designed, isolated room specially equipped to perform this procedure under sedation or general anaesthesia. The decision for either sedation or general anaesthesia was made after the evaluation of the overall status of the patients by the anaesthesiologist and the endoscopist. In healthy patients (American Society of Anesthesiologists, ASA I), the procedures were done under deep sedation with midazolam and fentanyl in addition to ketamine or propofol. On the other hand, for patients with systemic diseases (ASA II or more), the choice of drugs depended on the patient’s health status, comorbidities, procedure complexity and the duration of the procedure.15 General anaesthesia with endotracheal intubation was performed for patients with gastric outlet obstruction due to aspiration risk; thus, the procedure was done under rapid sequence induction. It was also done in case of complicated and prolonged procedures (i.e., >2 hours). In those patients, anaesthesia induction was done by midazolam, fentanyl, muscle relaxant (succinylcholine chloride or rocuronium bromide) as well as ketamine or propofol. Maintenance of anaesthesia was achieved either by inhalation or intravenous anaesthesia, depending on the patient’s health status. At the end of the procedure, patients who failed to be extubated or had procedural-related complications were transferred to the intensive care unit (ICU) for close monitoring and follow-up.

Endoscopes were PENTAX medical ED34-i10T2 along with a PENTAX EPK-i7010 video processor. The long-wire system was being used in most of the procedures. Accessories used during ERCPs were mostly of Boston Scientific and Cook medical brands, and to a lesser extent, Olympus accessories were used in a few cases.

All patients received prophylactic non-steroidal anti-inflammatory drugs (NSAIDs) suppositories to decrease the risk of PEP. Prophylactic antibiotics were given to all patients undergoing ERCP to reduce the risk of post-ERCP cholangitis because single-use equipment had to be used multiple times after proper re-sterilisation and reprocessing due to limited resources and availability in our country.

In terms of complications, PEP, similar to other subtypes of pancreatitis, typically presents with severe epigastric abdominal pain radiating to the back, elevated pancreatic enzymes (three times the upper limit) and certain abnormal imaging findings such as signs of pancreas inflammation or fluid collection. The severity of PEP was categorised as mild, moderate or severe depending on the clinical features, severity of symptoms, length of hospital stays, need for ICU admission and respiratory support and response to treatment. Post-ERCP bleeding has been considered in patients who had overt bleeding during the procedure that requires haemostatic interventions or had clinically significant bleeding from the site of the procedure within the 30 days following the procedure. All the patients who presented with symptoms such as hematemesis, melena, hematochezia, abdominal pain or signs of haemodynamic instability following an ERCP procedure were examined to rule out postprocedural haemorrhage.

Bowel perforation following the procedure is quite rare, and the diagnosis was made either by overt perforation during the procedure or by clinical and radiological evidence of perforation. Patients who had suffered from postprocedural hepatobiliary pyogenic infections (e.g., ascending cholangitis and liver abscess) or sepsis were considered to have a post-ERCP infection. All the complications have been addressed during the 30 days following the procedure.

Patient and public involvement

None.

Statistical analysis

To identify significant independent correlates of overall and specific complications, clinically relevant risk factors were obtained from previous studies and considered in a univariate analysis (χ2). Factors with a significant p value were considered in multivariate logistic regression models using stepwise forward selection for each of the two categories of complications. The first one determined the association between the probability of post-ERCP complications and possible risk factors like difficult cannulation. On the other hand, the second one studied the correlation between PEP and possible risk factors like pancreatic duct cannulation and prior ERCP procedure. Age and sex were empirically included in each model. Adjusted ORs and their 95% CI were calculated. All analyses were performed using RStudio Statistical Software (V.4.1.3; R Core Team, 2022).

Results

Demographic characteristics

A total of 1909 ERCP procedures, performed on 1303 patients between December 2017 and September 2022, were selected to participate in the study. Among them, 28.6% were ≤40 years, while 35.3% and 36.1% were between 41 and 64 years and ≥65 years, respectively.

The gender distribution was 42.1% and 57.9% between males and females, respectively. Most of the participants were mainly from West Bank (94%), while the other 6% were from Gaza Strip. The distribution of the procedures per year between 2017 and 2022 was 0.8%, 13.4%, 22.2%, 19.5%, 20.0% and 24.1% respectively. The relative drop in cases in 2020 and 2021 was due to a 14-month COVID-19 pandemic.

Procedural characteristics

The primary indications observed in our patient population were primarily related to obstructive jaundice, with a significant proportion (35.3%) attributed to known stones. Other indications included stent removal or exchange (20.3%), obstructive jaundice with known malignancy (13.0%) and obstructive jaundice with unknown biliary stricture suspected for malignancy (7.5%). The remaining 14% encompassed a variety of conditions, such as dilated common bile duct on imaging without jaundice, postoperative complications, acute cholangitis, pancreatic disease other than malignancy and other types of obstructive jaundice and periampullary lesion (table 1).

Table 1

Indications of the procedure

As a procedural technique, sphincterotomy was done in 35.8%, double wire cannulation in 10.7% and previous sphincterotomy was found to be performed in 41.4% of the patients. This was due to the fact that the procedure has been repeated more than once in some cases and due to multiple referrals after failed attempts. Cannulation without sphincterotomy, cannulation with pancreatic papillotomy, cannulation with altered anatomy/obstruction and precut papillotomy were distributed over the remaining 4.1% of the cases. Failure to attempt cannulation was encountered in 1.7% of the procedures.

The predominant outcomes observed from the ERCP procedures were related to stones/sludge (36.6%), followed by stent removal (16.2%) and strictures (10.7%). Stent exchange (8.1%) and stent insertion (6.1%) were also relatively common outcomes. Conversely, normal ERCP was observed in a minority of cases (5.0%). The remaining 17.3% of procedure outcomes were distributed among various other conditions, including failure of cannulation, leaking, periampullary diverticulum with biliary stones, ampullary mass/lesion and others (table 2).

Table 2

Outcomes of the procedure

During a follow-up period of 30 days after the procedure, complications were noted in only 5% of the total cases. Within this group, cardiopulmonary instability was observed in 2.1% of them, while bleeding was noted in 9.5%. Infection was ocurred in 21.1% of cases, while PEP was the most prevalent complication in 45.3% of cases. Other complications, such as perforation, death and others, accounted for 7.4%, 11.6% and 3.2%, respectively. The complication rate in patients with native papilla of Vater was 8.7%, with pancreatitis and death representing 54% and 12.7%, respectively.

Compared with the international published studies, all complication rates in our centre were found to be on the average rates except for the mortality rate, which was found to be slightly higher (0.5%) with a total of 11 cases (table 3). Reviewing the databases, the causes of death were found to be as follows: six cases were debilitated due to advanced malignancy with unresectable cancer, two cases died from cardiogenic shock and multiorgan failure, one case died from decompensated advanced liver cirrhosis and another case died from postprocedure myocardial infarction. The last case was an ICU patient due to septic shock post-COVID-19 pneumonia, and during the stay, the patient developed jaundice requiring ERCP. After that, ERCP was performed without any intervention and was normal. The patient was returned to the ICU, and later, the patient’s status deteriorated and died from COVID’s complications, within 30 days of the procedure, so the case was included. Therefore, none of the cases were directly related to the procedure.

Table 3

Comparison with international published studies in terms of sample volume, success and complication rates

Native papilla of Vater was present in 55.9% (n=1067) of the procedures. The native papilla of Vater’s successful cannulation rate was 94.3% (n=1006), and the overall successful cannulation rate of the native and non-native papilla of Vater was 95.4% (n=1821) of the procedures. The overall failure prevalence was merely 4.6% (n=88). This failure of the ERCP procedure was distributed as 37.5% (n=33) for obstructing tumour, 15.9% (n=14) for altered anatomy, 35.2% (n=32) for failure of cannulation, 6.5% (n=5) for failure due to complications and 6.5% (n=5) for others. The second attempt was performed successfully in 45.5% (n=40) of the failed cases, percutaneous transhepatic drainage in 22.7% (n=20), percutaneous transhepatic cholangiogram with rendezvous in 9.1% (n=8), surgical intervention in 3.4% (n=3), other interventions in 4.6% (n=4) and 14.7% (n=13) left in their palliative management without any further interventions.

Our study’s multivariate regression analysis of PEP-related risk factors revealed significant associations with age and pancreatic cannulation. Specifically, being young (≤40 compared with ≥65) was associated with a significant risk for PEP, with an OR of 2.84 (95% CI: 1.20 to 7.46, p=0.023). Furthermore, patients who underwent pancreatic cannulation exhibited a significant increase in the folds of PEP by 3.64 compared with those who did not (OR=3.64 (95% CI: 1.93 to 6.92, p<0.001)). Although pancreatic stenting was not significantly associated with PEP at the multivariate level, it showed significant associations with PEP at the univariate level. Even though multiple studies showed an increased risk of PEP in females, our cohort concluded that there was no gender variation when it came to PEP, with a p value of 0.719. Patients with native papilla also experienced a higher risk for PEP. Our study showed that patients who did not have native papilla at the time of the procedure had a 54% decrease in the folds of PEP compared with the patients with native papilla, with an OR of 0.46 (95% CI: 0.2 to 0.97, p=0.05) (table 4).

Table 4

The multivariate regression analysis of PEP

In the multivariate regression analysis model examining the risk factors for post-ERCP complications, we found that procedural technique variation was the only variable significantly associated with an increased risk of complications. Specifically, double-wire cannulation was shown to be associated with a higher risk of postprocedural complications compared with other techniques. Our study reported a p value of 0.033 and an OR of 2.29 (95% CI: 1.08 to 4.95) when comparing the risk of post-ERCP complications between double wire cannulation and sphincterotomy. On the other hand, other techniques showed no significant increase in this risk compared with sphincterotomy. Other variables like age, gender variation, first ERCP procedure, presence of native papilla and pancreatic cannulation did not show any clear relationship with the risk of postprocedural complications. These findings suggest that the presence of two wires increases the complexity of the procedure, potentially leading to a higher risk of tissue trauma or injury. Therefore, endoscopists should consider the use of alternative cannulation techniques to reduce the risk of complications in patients undergoing ERCP or the use of prophylactic pancreatic duct stenting to mitigate the potential for postprocedural complications, especially PEP (table 5).

Table 5

The multivariate regression analysis of postprocedural complications

Discussion

Since its invention in 1968, ERCP has become a crucial endoscopic method for treating biliary and pancreatic pathologies. The clinical need for ERCP has grown dramatically over time; however, use patterns indicate that more therapeutic ERCPs are being performed than diagnostic ERCPs. This is most often attributed to the advancement of gastrointestinal tract imaging modalities. The less invasive ERCP ensures enhanced survival in patients with malignancies such as cholangiocarcinoma and also plays a role in dramatic cures for life-threatening illnesses.16 17 Compared with more often performed routine gastrointestinal procedures like esophagogastroduodenoscopy and colonoscopy, ERCP is technically far more challenging, necessitating greater physician competence and taking longer to master. Research has indicated that ERCP carries a higher risk of complications, which may reach as high as 15.7%, with a mortality rate of 0.7% in the general population.10 11 18 The most frequent side effects are PEP, post-ERCP cholangitis, bleeding and visceral perforations.

It is possible that development in therapeutic technology, such as non-invasive radiologic imaging techniques, sphincterotomes for sphincterotomy, inflatable balloons or stents to dilate strictures, electrocautery to stem hemobilia and baskets or inflatable balloons to retrieve choledocholithiasis, have caused the shift in the role of ERCP from diagnostic to therapeutic over the past 10 years.19 In the diagnosis and treatment of duodenal and pancreaticobiliary disorders, ERCP is superior to traditional interventions in terms of limiting potential trauma to internal structures, streamlining the procedure and shortening recovery times; however, like any intervention, it comes with its own set of complications.20

ERCP provides the ability of stent placement in the bile and pancreatic ducts, which play an essential role in treating pancreaticobiliary obstruction and leakages and preventing PEP. Various plastic stents with different models, shapes, diameters, lengths and materials are available in the market. Self-expandable metal stents (SEMS) are another type available as uncovered, partially or fully covered. The stent covering material, type of wire and design determine the mechanical characteristics of the stent.20 A cholangiogram should be performed before stent placement to evaluate the length and location of the obstruction or leak.21 A final radiograph image should be obtained to ensure proper stent placement and drainage. Of all our cases, 903 have undergone biliary stent placement, 80.5% received plastic stents and 19.5% received SEMS.

From 2012 to 2015, the total incidence of ERCP-related adverse events steadily increased. The primary cause behind that is assumed to be the rise in PEP. A recent cohort research from the USA similarly showed an increase in PEP-related admission rates and deaths.22 Recent efforts to handle more complex patients will likely bring this trend of growing PEP. One systematic review and meta-analysis of 28 studies involving over 54 000 patients identified several patient and procedural factors associated with an increased risk of PEP. These factors included female gender, suspected SOD, a history of pancreatitis, previous PEP, difficult cannulation, main pancreatic duct injection, intraductal papillary mucinous neoplasm, endoscopic sphincterotomy and use of precut sphincterotomy. On the other hand, this study also found that the risk of PEP did not vary depending on the indication for the procedure, with no increase in the risk of PEP in patients undergoing ERCP for therapeutic purposes compared with diagnostic purposes. It also concluded that endoscopic biliary stenting and pancreatography have no obvious correlation with the risk of PEP.23 In the current study, only 5% of the cases had postprocedural consequences. Cardiopulmonary instability was seen in 2.1% of this group, and bleeding was seen in 9.5%. PEP was the most common consequence, occurring in 45.3% of the cases, while infection was observed in 21.1% of instances as well. Perforation, mortality and other problems made for 7.4%, 11.6% and 3.2% of all complications, respectively. Except for PEP, which was seen more frequently in younger age groups, complications were found to be evenly distributed throughout all age categories.

In addition to identifying risk factors, several studies have investigated interventions that may reduce the risk of PEP. One meta-analysis of seven randomised controlled trials found that the use of prophylactic rectal NSAIDs significantly reduced the risk of PEP. The study also found that the protective effect of NSAIDs was more pronounced in high-risk patients, such as those with a SOD or biliary sphincterotomy. However, this study showed rectal NSAIDs did not significantly decrease the risk of PEP when it comes to patient-related risk factors such as female gender and young age.24

Different techniques for performing a biopsy during ERCP include brush cytology, forceps biopsy, needle aspiration biopsy and snare polypectomy. The sensitivity and specificity of these biopsy techniques vary depending on the location and type of tissue being sampled. The diagnostic sensitivity of brush cytology is variable among different studies. In general, the sensitivity for brush cytology remains low. Some studies suggested that the sensitivity for this procedure varies between 8% and 70%. In a study published in the Journal of Gastrointestinal Endoscopy, the sensitivity of brush cytology in detecting malignant pancreaticobiliary malignancies was found to be 56.2%%, with a specificity of 100%.25 A systematic review and meta-analysis published by Navaneethan et al found that the pooled sensitivity of brush cytology in detecting malignant bile duct strictures was found to be 45%, with a specificity of 99%.26 In our study, we conducted brush cytology in 103 patients presenting with suspected malignant pancreaticobiliary strictures. Among these patients, 45 underwent further diagnostic techniques involving EUS to confirm the brush cytology results. Of the 45 patients who underwent EUS, only 29 were ultimately confirmed to have malignant strictures based on EUS, follow-up imaging, biopsy or surgical pathology. Brush cytology correctly identified 19 of these confirmed cases, while 10 cases were missed, resulting in a sensitivity of 65.5% for detecting malignant strictures.

Moreover, all cases positively identified by brush cytology were truly positive, yielding a specificity of 100% in detecting malignant strictures. The positive predictive value was 100%, and the negative predictive value was 61.5%. The study concluded that brush cytology is a valuable tool for the diagnosis of malignant biliary strictures, with high specificity and moderate sensitivity, and may be used in conjunction with other diagnostic modalities to improve overall accuracy.

The complexity of the ERCP procedure, akin to other endoscopic procedures, contributes to a failure rate that may be notably higher than that of other interventions. Multiple studies conducted from 2010 to 2018 found to carry an average failure rate of 7%.27–30 Besides this, as a country with limited resources, there were many obstacles and challenges. One of the limitations was that we had to use single-use accessories multiple times after proper sterilisation and reprocessing due to a lack of accessories. Additionally, we had to train staff from scratch to assist in performing this complex procedure, which affects the success rate and postprocedural complications. Despite all the limitations and lack of resources at our centre, our complication and failure rates were fairly parallel to the internationally published results.

In conclusion, pancreatic duct cannulation, double wire cannulation and diagnostic ERCP were the procedures that led to the greatest ERCP-related adverse events. Young patients getting diagnostic ERCP should be treated with extra attention due to an increased risk of PEP. The findings of this cross-country study can be used as the foundation for practical policy requirements for quality assurance and direction of ERCP practice in the future.

This post was originally published on https://bmjopen.bmj.com