TRANSJEJUNAL ERCP (TJ-ERCP) THROUGH EUS-GUIDED ENTERO-ANASTOMOSES (EUS-EA) IN PATIENTS WITH ROUX-EN-Y HEPATICO-JEJUNOSTOMY (RYHJ): SHORT AND LONG-TERM OUTCOMES

BACKGROUND AND AIMS
Ampullary neoplasia (AN) is a rare disease, but its incidence is increasing. In the last 20 years, endoscopic papillectomy (EP) has become the gold standard treatment for ampullary adenomas and early stage adenocarcinomas, thereby replacing surgical resection, which is burdened by higher rates of morbidity and mortality. However, the data supporting safety and efficacy of EP derive from multiple retrospective studies, that included procedures mostly performed before 2015, when first guidelines on endoscopic management of AN were available. This had an impact on large variability in patient selection criteria and endoscopic techniques, resulting in heterogenous outcomes. Therefore, the aim of our study is to provide data on the efficacy and safety of this technique, by including consecutive patients treated after the standardization of this technique.

METHODS
All patients who underwent EP at 19 Italian centers between January 2016 and December 2021 were included. Clinical success was defined by the complete endoscopic management of the neoplasm and any eventual recurrence found in the follow-up period. EP-related adverse events and recurrences were recorded.

RESULTS
A total of 225 patients were included. The mean lesion’s size was 20 mm (5–80 mm). En bloc resection was possible in 72.5% of cases, with an overall R0 resection rate of 50.7%. During a mean follow-up period of 23.2 months, recurrences were diagnosed in 17.2% of patients, 61,3% of which were successfully treated with an additional endoscopic treatment. Thus, clinical success was achieved in 76.7% of the cases. In multivariate analysis, R1 resection, lesion size and histological diagnosis were predictors for recurrence. Intra-procedural bleeding occurred during 12,4% of EP. Post-EP adverse events (AE) occurred in 39,5% of patients, including delayed bleeding (20,9%), pancreatitis (13.3%) and perforation (2.2%). Complications were mild or moderate in 88,9%, while the 11.1% were severe, according the ASGE Lexicon. No EP-related deaths were recorded.

CONCLUSIONS
The results of our study confirm the efficacy of endoscopic papillectomy in the treatment of ampulla of Vater neoplasms in the current clinical practice. Most of recurrences were successfully endoscopically managed. However, even if performed by expert endoscopists, EP is a procedure associated with not negligible risk of complications.

Aim:

Endoscopic papillectomy is the recommended treatment for adenoma, in situ carcinoma and intramucosal adenocarcinoma of the papilla but positive margin resection and recurrence are common and led to complementary surgery with significant mortality rate (3-10%).
Radiofrequency ablation (RFA) as complementary treatment of papillectomy margin seems to be efficient and feasible.

Methods:

This single-center retrospective study included all patients who had a first biliary RFA following papillectomy in our institution, indicated when pathological margins were positive or in case of relapse.
Primary aim was the clinical success defined by the absence of recurrence at papillectomy site macroscopically and/or histologically 12 months after the first or the second RFA session. The second aims were the number of sessions needed for clinical success, the early (<30 days) and late (> 30 days) complications. Eradication failure was defined when more than 2 sessions were realized.

Results:
Thirty-four patients were treated with RFA as complementary treatment after papillectomy, 25 of them (mean age 71 years) had more than 12 month follow-up and were included, median follow-up was 36 months (12-80).
RFA were performed in cases of positive resection margin n=20 (Low-grade dysplasia (LGD) n=10, high-grade dysplasia n=5, in situ carcinoma (CIS) n=3, adenocarcinoma pT1a R1 refusing surgery n=1, neuroendocrine tumor grade 2 n=1) or relapse n=5 (LGD n=5).

Clinical success was 88% (22/25) with median number of RFA session needed of 1.1 (1-2).
Two patients had eradication failure, one after 12 RFA session (unfit for surgery) and one had a Whipple surgery (progression of LGD into the common bile duct inaccessible with RFA).
One patient died from a severe acute pancreatitis after a second RFA session (pancreatitis prophylactic stent introduction fail).
Our patient with adenocarcinoma is free from disease after 44 month of follow-up.
One patient with CIS had a lymph node relapse at the 33^th month, still treated by chemotherapy at the 55^th month.

We reported 3 early complications (12%): two acute pancreatitis (mild and fatal) and one bleeding requiring a new endoscopy. Twelve late complications occurred (48%) of a stenosis type, successfully managed endoscopically with a median follow-up of 17.6 months (0-56). Eight of these stenosis occurred in patient who had 1 or 2 additional resections sessions after the initial papillectomy.

Conclusion:
Our study is one of the largest series reporting biliary RFA following endoscopic papillectomy. RFA is an effective treatment to eradicate residual ampullary or recurrent lesions with 88% of clinical success. Late stenosis is the principal complication (48%) favored by the association of an additional resections sessions however, they are managed effectively by endoscopy. RFA seems to be a safe alternative to surgery in well-selected cases.

Introduction:
The endoscopic management of unresectable hilar cholangiocarcinoma is centered on the maintenance of biliary patency. While self-expandable metal stents have been shown to have higher stent patency, selection bias in comparative data limits interpretation and difficulty with re-intervention on metal stents make plastic stents (PS) a viable option for many patients. We aimed to review our single-center experience with the routine exchange of bilateral plastic stents in patients with unresectable hilar cholangiocarcinoma to identify factors associated with recurrent biliary obstruction.

Methods
Patients with unresectable cholangiocarcinoma undergoing ERCP for undergoing unilateral or bilateral plastic stent placement were eligible for inclusion. Patients receiving metal stents or percutaneous placement of external drains were excluded. Recurrent biliary obstruction (RBO) was defined according to the 2014 Tokyo guidelines as signs/symptoms consistent with obstruction or migration. Kaplan-Meier curves were used to evaluate RBO-free survival during the follow-up period. Cox regression was used to identify factors associated with RBO after plastic stent exchange, using a shared frailty model for recurrent events.

Results:
Consecutive patients treated for unresectable hilar cholangiocarcinoma between January of 2015 and June of 2022 were considered for eligibility. 62 patients were included in the analysis, encompassing a total of 199 ERCP’s. The average age of patients at the time of ERCP was 63.9 years (SD 14.7). 45% of patients were female. Patients received an average of 3.2 ERCP’s (SD 3.4) over a mean follow-up period of 40.5 months (SD 97.1). 85 ERCP’s (42.7%) were performed prior to a scheduled procedure (due to RBO). 58/85 (68%) emergent ERCP’s were performed for cholangitis. The Kaplan-Meier curve for RBO-free survival is shown in Figure 1. The median RBO-free survival was 98 days. The 12-week cumulative probability of RBO was 59% (95% CI 50%-69%). Results of univariable and multivariable Cox-proportional hazards regression are reported in Table 1; procedures performed ‘emergently’ (prior to scheduled stent exchange) were associated with greater risk of recurrent biliary obstruction. There was non-significant association between higher total bilirubin (p=0.055) and AST (p=0.062) two weeks following biliary drainage and early RBO.

Discussion:
Plastic stents are generally exchanged at 12-week intervals, currently based largely on expert opinion. We demonstrated that at 12-weeks, nearly 60% of patients have developed signs or symptoms of recurrent obstruction, and in nearly 70% of these cases this manifested as cholangitis. A shorter interval to routine plastic stent exchange, particularly in patients that have previously required emergent endoscopy, may be warranted to minimize the risk of recurrent biliary obstruction and its complications.

Background: Enteroscopy-assisted (EA-) ERCP can be difficult or impossible in patients with prior Roux-en-Y gastric bypass (RYGB). EUS-guided intervention (EUS-I), and EUS-directed transgastric ERCP (EDGE) are alternative methods for cholangiography and therapy.

Aim: To compare therapeutic success, complications, and procedure times of EA-ERCP, EUS-I, and EDGE in patients with RYGB.

Methods: Retrospective analysis was performed of all patients with RYGB receiving biliary intervention between April 2018 (when EDGE programmatically initiated) and Nov. 2022 at a single tertiary endoscopy center.

During EA-ERCP, various endoscope types were used to reach and cannulate the biliary orifice. For EUS-I, cholangiography was performed via EUS-guided transgastric needle puncture into a left intrahepatic bile duct. Direct anterograde intervention or rendezvous procedure would then be performed. For EDGE, gastro- or entero-gastric fistula created with lumen apposing metal stent (LAMS) provided a conduit for endoscope passage into the excluded stomach and duodenum to access the papilla for transgastric ERCP.

Results: 52 patients underwent EDGE, 26 EUS-I, and 17 EA-ERCP. Technical therapeutic success for EDGE (98%) was superior to EUS-I (85%, p=0.0399) and EA-ERCP (76%, p=0.0116). No significant difference in complication rates occurred – EDGE = 8% (3 gastrogastric stent migrations salvaged without clinical consequence; 1 cautery-induced perforation of contralateral excluded gastric wall closed with single clip), EUS-I = 8%,p=1 (1 biloma, 1 pancreatitis+pneumonia), EA-ERCP = 0%,p=0.5654 2 patients who had staged EDGE re-presented while awaiting enterogastrostomy tract maturation for delayed ERCP (abdominal pain, gallstone pancreatitis). EDGE was faster than ERCP (58 vs 78 min, p=0.0029) but slower than EUS-I (58 vs 42 min, p=0.0094).

32 patients had single session EDGE (ss-EDGE, same session LAMS placement + transgastric ERCP). ss-EDGE therapeutic efficacy (30/32) was similar to that of EUS-I (p=0.3926) and EA-ERCP (p=0.0782). Complication rates were similar between ss-EDGE (4 EDGE complications noted above) and EUS-I (p=0.6814) or EA-ERCP (p=0.2837). ss-EDGE (52 min) was also slower than EUS-I (p=0.0019) but faster than EA-ERCP (p<0.0001). When initial endoscopic approach failed, pursuit of an alternative salvage endoscopic approach, when necessary, achieved composite therapeutic success of 92%.

Conclusion: In patients with RYGB undergoing EA-ERCP, EUS-I or EDGE, EDGE achieved in higher technical therapeutic success with no difference in complications when allowing for multi-session EDGE. ss-EDGE, however, was therapeutically comparable to EA-ERCP and EUS-I, while EUS-I was fastest. EUS-I may be considered in appropriate cases at specialized centers to achieve successful therapy up front, avoiding multiple procedures and potential interim clinical biliary events.

Background: EUS-Directed trans-Gastric ERCP (EDGE) is emerging for Roux-en-Y gastric bypass (RYGB). Promising short-term outcomes of related approaches using lumen-apposing metal stents (LAMS) as a shortcut for retrograde biliary access have been reported in heterogeneous post-surgical anatomy scenarios. Roux-en-Y Hepatico-Jejunostomy (RYHJ) patients with biliary obstruction (BO) continue to be managed primarily by percutaneous IR, particularly when complex underlying disease (strictures/hepatolithiasis) dictates the need for iterative revisions and/or adhesions preclude enteroscopy-assisted ERCP. TJ-ERCP via EUS-EA through a LAMS warrants further study in patients with RYHJ.
Aims: To assess feasibility, safety and efficacy of EUS-EA-assisted TJ-ERCP in patients with RYHJ.
Methods: Consecutive adult patients from two tertiary centers with BO and RYHJ who underwent EUS-EA-assisted TJ-ERCP using LAMS (Fig.1) between 2014-2022 were retrieved from prospective databases and reviewed under IRB approval. Baseline, procedural and clinical outcome data were pooled. Technical success: TJ-ERCP completion. Short-term clinical success: BO resolution without relapses within 30 days. Long-term clinical success: removal of indwelling percutaneous biliary drainage catheters with no recurrences or unscheduled re-interventions within 365 days +/- stricture resolution/complete stone clearance.
Results: 73 patients [48 (65,75%) male; 64 (11.71) years] (Table 1 and 2) underwent attempted TJ-ERCP. Benign BO was present in 64 (88%) patients. Median time (IQR) between index surgical RYHJ and EUS-EA was 7.4 (12.03) years. 54 patients (74%) had previous biliary drainage attempts, either failed or only temporarily successful. EUS-EA was technically successful in 69/73 (94.5%). 15-mm LAMS was used in 59/73 (83.1 %) and 20-mm LAMS in 12/73 (16.9%). Type of EUS-EA: 29/73 (39.7%) gastro-jejunostomy, 30/73 (41,1%) duodeno-jejunostomy (10 bulb; 20 second duodenal portion) and 14/73 (19.2%) jejuno-jejunostomy. Same-session ERCP was performed in 28/43 (40.6%). There were 6/73 (8.5%) AEs. There were 7 LAMS dislodgements (all managed endoscopically). Long-term clinical success was achieved in 59/73 (80.8%).
Conclusions: EUS-EA assisted TJ-ERCP using LAMS appears safe and effective. Technical success appears higher than for enteroscopy-ERCP in this setting, and patient tolerability higher than for IR-based approaches. TJ-ERCP might be considered the preferred approach to BO in RYHJ patients at centers with the required interventional EUS expertise. Underlying biliary disease is typically more complex than in RYGB patients undergoing EDGE. Iterative ERCP is required in 34% and easily carried out through indwelling LAMS. Definitive BO resolution is achieved in 80.8%. The main predictor of definitive therapeutic success is short-term clinical success following index ERCP.