Society: ASGE

BACKGROUND: Endoscopic Snare Papillectomy (ESP) is a minimally invasive option for the management of noninvasive ampullary adenomas. The primary aim of this study is to determine risk factors for incomplete resection of ampullary adenomas following ESP, as well as the impact of rectal indomethacin and pancreatic duct (PD) stenting on post-papillectomy pancreatitis (PPP).

METHODS: This retrospective multicenter study included consecutive patients undergoing index ESP for ampullary adenomas from 1/2012-9/2020 at two referral centers. Patient demographics, procedural indications, details, adverse events, pathology, and subsequent interventions were collected. Wilcoxon rank-sum test and Fisher’s exact test were used to compare continuous and categorical variables. Stepwise logistic regression was used to identify predictors for residual disease and PPP.

RESULTS: Of 136 included patients, 24 (18%) carried a diagnosis of Familial Adenomatous Polyposis. Overall median adenoma size was 20mm (IQR, 15mm, 25mm) and 47 (35%) adenomas were laterally spreading. Common bile duct extension was seen on pre-procedure EUS or index ERCP in 25 (19%) patients. En bloc resection was used in 77 (57%) patients. Advanced neoplasia was seen in 60% of patients (Table 1). Median duration of surveillance was 899 days (IQR 382, 2041 days). Residual neoplasia was seen in 66 (49%) patients on first surveillance endoscopy, of whom 6 (9%) had biopsy-confirmed residual malignancy. Factors not meeting inclusion threshold during stepwise regression (P<0.20) were adenoma size, location, and presence of periampullary diverticulum. Factors associated with increased risk of residual adenoma were biliary extension (OR 6.87, 95% CI 2.10-22.48) and piecemeal resection (OR 2.32, 95% CI 1.09-4.92) (Table 2). Median number of endoscopic sessions required for successful ampullary adenoma eradication was 1 (IQR 1 session, 2 sessions). Surgical pancreaticoduodenectomy was ultimately required in 13 (10%) patients.

PPP occurred in 20 (15%) patients. Prophylactic PD stent was placed in 122 (90%) patients, of whom 62 (51%) underwent PD stenting alone and 60 (49%) received combination PD stenting and rectal indomethacin. Decreased rates of PPP were seen with PD stenting (OR 0.22, 95% CI 0.06-0.78, P=0.020), but not with rectal indomethacin alone (OR 0.82, 95% CI 0.32-2.12) or in combination with PD stenting (OR 0.49, 95% CI 0.18-1.37).

CONCLUSIONS: Residual adenoma was seen in nearly half of patients with ampullary adenomas after ESP. CBD extension and piecemeal resection were predictive of residual disease. ESP was associated with a 15% rate of PPP, which was lowered by PD stenting. Reduction of PPP was not demonstrated by the addition of rectal indomethacin or indomethacin alone.

Background and aims: Adverse events (AEs) after endoscopic retrograde cholangiopancreatography (ERCP) are not uncommon. Post-ERCP pancreatitis (PEP) is the main AE. Albeit its pathophysiology is not fully understood, thermal injury from biliary sphincterotomy may play an important role and affect other AEs. Therefore, this study evaluates the outcomes of two electric current modes used during biliary sphincterotomy.

Methods: From October-2019 to August-2021, consecutive patients submitted for ERCP with native papilla undergoing biliary sphincterotomy were randomized to either the pure cut (30-50W - WEM S200 and ICC 200 electrosurgical units) or endocut (effect 2 - VIO 300 and VIO 300 electrosurgical units) after transpapillary cannulation in two terciary reference centers. The study was conducted according to CONSORT statements and analysis were performed under intention-to-treat basis. The primary outcome was PEP incidence. Secondary outcomes included intraprocedural and delayed bleeding, infection, and perforation.

Results: 550 patients were randomized (272 for pure cut and 278 for endocut – figure 1). The overall PEP rate was 4.0%, and it was significantly higher in the endocut arm (5.8% x 2.2%, p=0.034 – table 1). Univariate analysis revealed >5 attempts (p=0.004) and endocut mode (p=0.034) as risk factors for PEP. Multivariate analysis revealed >5 attempts (p=0.005) and a trend for endocut mode as a risk factor for PEP (p=0.052). Intraprocedural bleeding occurred more often with pure cut (p=0.018). Delayed bleeding was more frequent with endocut (p=0.047). There was no difference in terms of perforation (p=1.0) or infection (p=0.499) among the groups.

Conclusion: This RCT revealed that endocut is associated with higher rates of PEP and delayed bleeding, whereas pure cut is associated with increased intraprocedural bleeding. As all intraprocedural bleeding were controlled during ERCP, pure cut should be the preferred electrocautery mode for biliary sphincterotomy. Additionally, no more than five cannulation attempts should be performed, as this is also associated with PEP.

Introduction:
Over 500,000 ERCPs are performed annually in the United States. Currently, most endoscopists perform an ERCP with only a side-viewing endoscope (i.e., duodenoscope). We hypothesized that given the non-forward-viewing design of a duodenoscope, there is a significant miss rate of important findings. Therefore, the primary aim of this study was to evaluate differences in clinical findings between a forward-viewing and side-viewing examination during ERCP.

Methods:
We conducted a blinded, tandem, single-center prospective clinical trial at a tertiary academic center (NCT: NCT05627882). Patients undergoing ERCP for standardized indications were recruited. Six credentialed attending gastroenterologists were outcome assessors. Included patients underwent a standard forward-viewing endoscopic examination with one attending. Findings were deemed to be clinically significant (leading to management change) or classified as non-management altering. Results were recorded by a nurse not involved with the study so the answers could not be modified. A second attending, blinded to the forward viewing results, subsequently performed the ERCP with a duodenoscope. Results were recorded in a similar fashion. Continuous data were compared using the two-sample t-test or Wilcoxon rank-sum test and categorical data were compared using the Chi-square or Fisher’s exact test, as appropriate.

Results:
A preliminary trial analysis of 44 included participants was performed. Baseline demographics including co-morbidity history and anticoagulation are shown in Table 1. Forward-viewing endoscopic examination noted more endoscopic abnormalities when compared to side-viewing (56.81% vs 18.18%; P<0.001) – total miss rate of 38.63%. Discordant results were identified in 52.27% of patients (23/44). Clinically significant discrepancies (i.e., practice changing) were noted in 38.64% of patients (17/44), with non-significant differences among 22.73% of patients (10/44) [P=0.109]. No abnormalities were found with duodenoscope examination that were not identified on forward-viewing examination. Of the findings missed on side-viewing examination, 16 were in the esophagus, 8 in the stomach, and 3 in the duodenum. 65% of the practice changing findings were found in outpatient cases. Examples of missed findings included: ulcerated gastric sarcoma, deep ulceration requiring endoscopic clipping, Barrett’s esophagus, as well as esophageal and gastric varices.

Discussion:
Current clinical practice for ERCP traditionally relies solely on examination with a side-viewing endoscope. However, this blinded, tandem prospective trial found a significant 38.63% miss rate when patients underwent forward-viewing examination. These findings suggests that forward-viewing examination may be valuable in addition to use of a duodenoscope during routine ERCP procedures.

Introduction
ERCP, like all fluoroscopic procedures, poses the risk of ionizing radiation exposure (RE) to patients and staff. RE increases the risk of adverse biological effects such as skin burns, cataracts, sterility, cancer, and reduced lifespan. Newly developed fluoroscopic imaging equipment (C-Arm) provides options to limit radiation delivery in the form of lower radiation dose and frame rate or time-limited “pulsed” settings. However, the impact of lower radiation settings on image quality in ERCP has not been assessed, and currently, there is no standard protocol for fluoroscopy settings used during ERCP.

Aim
To determine whether low-dose time-limited "pulsed" fluoroscopy provides adequate imaging compared with standard-dose time-limited “pulsed” fluoroscopy.

Methods
This study was a double-blind, randomized, prospective trial of patients undergoing standard-of-care ERCP. Patients were randomized into two groups: 1) standard-dose time-limited “pulsed” fluoroscopy (n = 41); 2) low-dose time-limited “pulsed” fluoroscopy (n = 37). Standard-dose time-limited pulse was defined as x-ray in the standard-dose setting limited to 3 seconds each time the foot switch is depressed, whereas low-dose time-limited pulse fluoroscopy was defined as x-ray in the low-dose setting limited to 3 seconds each time the foot switch is depressed. The blinded endoscopist rated image quality using a Likert scale (1=excellent to 5=unacceptable).

Results
A total of 78 patients undergoing ERCP were enrolled and randomized. No difference in age, gender, and BMI were found between the two groups (Table 1). Using a mixed multivariable linear model, no significant difference was found in image quality between those who did vs. did not receive low-dose “pulsed” fluoroscopy (p=0.925), even when controlling for BMI, radiation dosing, or fluoroscopy time (Table 2). The low-dose group was exposed to significantly less radiation when compared to standard-dose pulsed (p=0.00007). Fluoroscopy time (minutes) was similar in both groups (median: group 1 = 2.0 vs. group 2= 1.9), further suggesting that group assignment had no impact on image quality or procedure time. A total of 5 cases were switched from protocol to standard-dose continuous fluoroscopy due to poor visualization. Four patients in group 2, versus 1 patient in group 1 (mean image quality rating: Fair - 3.2 vs. 3, respectively).

Conclusion
Our study shows that low-dose pulsed fluoroscopy is a reliable method that substantially reduces radiation dose without compromising image quality, nor affecting procedural or fluoroscopy times. This study highlights the need for a standardized protocol regarding fluoroscopy settings in ERCP and greater emphasis on limiting procedural radiation exposure in a stepwise approach.

Table 1: Baseline Characteristics<br /> Group 1: Standard<br /> Group 2: Low-Dose

Table 1: Baseline Characteristics
Group 1: Standard
Group 2: Low-Dose

Table 2: Difference in Image Quality<br /> Group 1: Standard Dose<br /> Group 2: Low Dose

Table 2: Difference in Image Quality
Group 1: Standard Dose
Group 2: Low Dose

BACKGROUND AND AIM: Malignant hilar obstruction usually presents in advanced stage disease with a poor prognosis. Effective biliary drainage is essential for the beginning of palliative chemotherapy. Some recent studies suggest that effective drainage of significant liver parenchyma is necessary to achieve clinical success. This study aimed to correlate clinical success rate with the volume of liver drained. METHODS: We conducted a retrospective study including patients with malignant hilar biliary obstruction who underwent retrograde endoscopic cholangiography for biliary drainage, from January 2010 to December 2017 in Intitudo do Cancer do Estado de Sao Paulo. Assessment of liver volume drained was made based on the relationship between the liver segment(s) drained and the preoperative volume of each segment. Main outcome was correlation of clinical success rate with hepatic volume drained. Secondary outcome was correlation of clinical success rate with the hepatic sectorial ducts drained and laterality. RESULTS: 82 patients met inclusion criteria, 58.5% female, with a mean age of 60±13 years. The main cause of hilar obstruction was cholangiocarcinoma (32.9%). Technical success was achieved in 75 patients (91.5%), clinical success in 45 patients (60%). We found significant correlation between clinical success with drained hepatic volume, when at least 50% of viable parenchyma was drained (p=0.016). Using a ROC curve for sensitivity and specificity analysis, the area under the curve was 72.07 [95%CI 59.7 to 84.4], p<0.01. A cut-off of 67% of drained liver volume parenchyma (Figure 2) was established to achieve best relation between sensitivity and specificity towards clinical success. Considering sectorial ducts, higher clinical success rates were found when at least 2 sectorial ducts were drained (p<0.001). Correlation between unilateral versus bilateral drainage and clinical success on the other hand, was not statistically significant. DISCUSSION: The determination of the hepatic segments related to each sectorial duct by the analysis of pre-procedure CT-scan and cholangiography findings is possible and can facilitate and improve clinical results when CT volumetry is not available. It is reasonable to assume that the drainage of more than one liver sector will drain at least 50% of hepatic parenchyma. In our study, drainage of at least 2 sectoral ducts accounted for significant improvement in clinical success rates (p<0.001). The same result was not achieved when comparing unilateral and bilateral stenting, compatible with current literature that shows dissonant results in this subject. CONCLUSION: Drainage of at least 50% of volume hepatic parenchyma was associated with better outcomes as well as drainage of at least 2 hepatic sectors, regardless its laterality.

† TABLE 3. Correlation of clinical success with percentage of drained liver parenchyma, sectoral ducts and laterality.<br /> *Considered significant p values >0.05.

† TABLE 3. Correlation of clinical success with percentage of drained liver parenchyma, sectoral ducts and laterality.
*Considered significant p values >0.05.

FIGURE 2. Roc curve analysis: blue line is generated by analyzing the data, between sensitivity and specificity, and the area under the curve was 72.07 [95%CI 59.7 to 84.4], p<0.01.

FIGURE 2. Roc curve analysis: blue line is generated by analyzing the data, between sensitivity and specificity, and the area under the curve was 72.07 [95%CI 59.7 to 84.4], p<0.01.

Bruno Martins

Appearances