The role of endoscopic ultrasound guidance for biliary and pancreatic duct access and drainage to overcome the limitations of ERCP: a retrospective evaluation.

BACKGROUND: Endoscopic ultrasound (EUS)-guided ductal access and drainage (EUS-DAD) of biliary/pancreatic ducts after failed endoscopic retrograde cholangiopancreatography (ERCP) is less invasive than percutaneous transhepatic biliary drainage (PTBD). The actual need for EUS-DAD remains unknown. We aimed to determine how often EUS-DAD is needed to overcome ERCP failure. METHODS: Consecutive duct access procedures (n = 2205; 95 % biliary) performed between June 2013 and November 2015 at a tertiary-care center were reviewed. ERCP was used first line, EUS-DAD as salvage after ERCP, and PTBD when both had failed. Procedures were defined as "index" in patients without prior endoscopic duct access and "combined" when EUS-DAD followed successful ERCP. The main outcomes were the EUS-DAD and PTBD rates. RESULTS: EUS-DAD was performed in 7.7 % (170/2205) of overall procedures: 9.1 % (116/1274) index and 5.8 % (54/931) follow-up. Most index EUS-DADs were performed following (46 %) or anticipating (39 %) ERCP failure, whereas 15 % followed successful ERCP (combined procedures). Among index procedures, the EUS-DAD rate was higher in surgically altered anatomy (58.2 % [39 /67)] vs. 6.4 % [77/1207]); PTBD was required in 0.2 % (3/1274). Among follow-up procedures, ERCP represented 85.7 %, cholangiopancreatography through mature transmural fistulas 8.5 %, and EUS-DAD 5.8 %; no patient required PTBD. The secondary PTBD rate was 0.1 % (3/2205). Six primary PTBDs were performed (overall PTBD rate 0.4 % [9/2205]). CONCLUSIONS: EUS-DAD was required in 7.7 % of ERCPs for benign and malignant biliary/pancreatic duct indications. Salvage PTBD was required in 0.1 %. This high EUS-DAD rate reflects disease complexity, a wide definition of ERCP failure, and restrictive PTBD use, not poor ERCP skills. EUS-DAD effectively overcomes the limitations of ERCP eliminating the need for primary and salvage PTBD in most cases.

Retrospective cohort study comparing endoscopic ultrasound-guided and percutaneous drainage of upper abdominal abscesses.

BACKGROUND AND AIM: Endoscopic ultrasonography (EUS)-guided drainage (EUS-D) has become the standard treatment for peripancreatic fluid collections. Its use in other intra-abdominal abscesses has been reported, although there is limited evidence. METHODS: We carried out a single-center retrospective cohort study comparing percutaneous drainage (PCD) and EUS-D of upper abdominal abscesses between January 2012 and June 2017. Pancreatic fluid collections and liver transplant recipients were excluded. Primary endpoints were technical and clinical success rates. RESULTS: We included 18 EUS-D (nine hepatic and nine intraperitoneal abscesses) and 62 PCD. There were no differences regarding age, gender and etiology. Size was larger in the PCD group (80 vs 65.5 mm, P = 0.04) and perivesicular location was more frequent in the PCD group (24.2% vs 11.1%, P = 0.003). In the EUS-D group, metal stents were deployed in 16 (88.9%) subjects (eight lumen-apposing metal stents and eight self-expandable metal stents), coaxial double-pigtail plastic stents in six (33.3%) and lavage/debridement was carried out in five (27.8%). There were no significant differences in technical success (EUS-D: 88.9%, PCD: 96.8%, P = 0.22) or clinical success (EUS-D: 88.9%, PCD: 82.3%, P = 0.50), with no relapses in the EUS-D group and 10 (16.1%) in the PCD group (P = 0.11). There were four (22.2%) adverse events in the EUS-D group, none of them severe, and 13 (21%) in the PCD group (P = 0.91). CONCLUSIONS: EUS-D is an alternative to PCD in the treatment of upper abdominal abscesses, reaching similar success, relapse and adverse events rates.

Cumulative risks of stent migration and gastrointestinal bleeding in patients with lumen-apposing metal stents.

BACKGROUND: Delayed gastrointestinal (GI) bleeding and stent migration are known adverse events which may occur following placement of lumen-apposing metal stents (LAMSs). METHODS: All consecutive patients who underwent LAMS placement between May 2011 and June 2017 at a single tertiary medical center were included. Demographics and procedural details were prospectively collected. Post-procedure follow-up and outcome measures were retrospectively collected. The cumulative risks of migration and LAMS-related GI bleeding were estimated using the life-table method. Risk predictors were assessed using Cox proportional hazards models. RESULTS: We analyzed 250 patients (64.8 % men; median age 71.6 [interquartile range (IQR) 57.9 - 83.6]). Median follow-up was 78.5 days (IQR 31 - 246.5 days). Thirty-four stent migrations (13.6 %) occurred (5 symptomatic). On multivariable analysis, associations with migration included nasocystic drains (hazard ratio [HR] 6.5, 95 % confidence interval [CI] 2.2 - 19.3), pancreatic fluid collections (PFCs; HR 4.2, 95 %CI 1.8 - 10.1), and double-pigtail stents (HR 2.4, 95 %CI 1.2 - 4.9). Migration risk at 12 months was 25.5 % (95 %CI 17.9 % - 35.7 %) and was higher for PFCs 48.9 % (33.4 % - 66.9 %) than other indications 8.4 % (4.9 % - 17.5 %; P < 0.001). LAMSs placed for longer durations (i. e. enteral anastomoses, biliary and gallbladder drainage) presented an 8.4 % cumulative risk at 2 years. There were 13 LAMS-related GI hemorrhages (5.2 %), two of them fatal, presenting a median of 3 days (IQR 1 - 9 days) after deployment. The cumulative risk of bleeding at 12 months was 6.9 % (3.6 % - 12.7 %). CONCLUSIONS: LAMS migration occurs in 1 out of 7 cases and is most common when treating PFCs. Bleeding related to LAMS placement occurs much less commonly but can be life-threatening.

Differences between childhood- and adulthood-onset eosinophilic esophagitis: An analysis from the EoE connect registry.

BACKGROUND: Direct comparisons of childhood- and adulthood-onset eosinophilic esophagitis (EoE) are scarce. AIM: To compare disease characteristics, endoscopic and histological features, allergic concomitances and therapeutic choices across ages. METHODS: Cross-sectional analysis of the EoE CONNECT registry. RESULTS: The adulthood-onset cohort (those diagnosed at ≥18y) comprised 1044 patients and the childhood-onset cohort (patients diagnosed at <18 y), 254. Vomiting, nausea, chest and abdominal pain, weight loss, slow eating and food aversion were significantly more frequent in children; dysphagia, food bolus impaction and heartburn predominated in adults. A family history of EoE was present in 16% of pediatric and 8.2% of adult patients (p<0.001). Concomitant atopic diseases did not vary across ages. Median±IQR diagnostic delay (years) from symptom onset was higher in adults (2.7 ± 6.1) than in children (1 ± 2.1; p<0.001). Esophageal strictures and rings predominated in adults (p<0.001), who underwent esophageal dilation more commonly (p = 0.011). Inflammatory EoE phenotypes were more common in children (p = 0.001), who also presented higher eosinophil counts in biopsies (p = 0.015) and EREFS scores (p = 0.017). Despite PPI predominating as initial therapy in all cohorts, dietary therapy and swallowed topical corticosteroids were more frequently prescribed in children (p<0.001). CONCLUSIONS: Childhood-onset EoE has differential characteristics compared with adulthood-onset, but similar response to treatment.

[Comparison between insufflation with air or carbon dioxide during the colonoscopy in sedated patients with propofol]. / Comparación entre la insuflación con aire ambiente o con dióxido de carbono durante la colonoscopia en pacientes sedados con propofol.

OBJECTIVES: compare the intensity of pain experienced after colonoscopy with air or with CO2 and evaluate the safety of CO2 in colonoscopies performed with moderate/deep sedation. MATERIALS AND METHODS: individuals undergoing ambulatory colonoscopy without exclusion criteria (severe respiratory disease, morbid obesity) were randomized in air or CO2 group. We recorded different variables prior to, during and upon completion of the colonoscopy, performing monitoring using pulse oximetry and capnography. Each patient rated, using a visual numeric scale, the intensity of post-colonoscopy pain at different moments. RESULTS: 141 individuals in the air group (sex M/F 63/78, age 24-83) and the CO2 group (sex M/F 59/70, age 24-82). No significant differences existed in the recorded variables in both groups except for the greater number of explorations performed by an endoscopist in training (TE) in the air group compared to those by a more experienced endoscopist (SE). CO2 in expired air, episodes of oxygen desaturation and of apnoea and dose of propofol, of midazolam were similar in both groups. No episodes of hypercapnea or any complication requiring cardiopulmonary resuscitation measures were recorded. The pain in the air group was significantly higher at 15 minutes and at 1, 3 and 6 hours after the endoscopy, equalising at 24 hours. After multivariant adjustment for type of doctor (TE vs. SE) the differences observed in pain intensity for each group were maintained. CONCLUSIONS: a) the use of CO2 in colonoscopy causes significantly less pain in the first 6 hours after the procedure; b) its use in patients with moderate/deep sedation is safe; and c) performance of the endoscopic technique is not modified, nor are times reduced.

EUS-guided single-incision needle-knife biopsy: description and results of a new method for tissue sampling of subepithelial GI tumors (with video).

BACKGROUND: The diagnostic efficacy of current tissue sampling techniques for upper GI subepithelial tumors (SETs) appears to be limited. Better tissue acquisition techniques are needed to improve the diagnostic yield in this setting. OBJECTIVE: Our purpose was to determine the safety and diagnostic yield of EUS-guided needle-knife incision and forceps biopsy (SINK biopsy) of upper GI SETs. DESIGN: Retrospective database review. SETTING: Academic tertiary-care referral center. PATIENTS: This study involved 14 consecutive patients referred for EUS evaluation of upper GI SETs with previous unsuccessful attempts at tissue diagnosis by conventional forceps biopsy. INTERVENTION: EUS-guided needle-knife incision and forceps biopsy. MAIN OUTCOME MEASUREMENTS: The safety and diagnostic yield of this method, compared with EUS-guided fine-needle aspiration (EUS-FNA), when possible. RESULTS: SINK biopsy provided tissue samples that were sufficient for definite histologic diagnosis in 13 of 14 cases (diagnostic yield 92.8%). There were 8 gastric GI stromal tumors. In 7 of 8, the size of SINK specimens allowed immunohistochemical analysis, and the evaluation of malignant potential was carried out by means of mitotic index determination in 5 cases (71.42%). SINK biopsies determined the pathological diagnosis of all (4 of 4) nonmesenchymal lesions. Eight patients underwent both EUS-FNA and SINK, with final histologic diagnosis determined in 6 of 8 cases (75%) by SINK versus 1 of 8 cases (12.5%) by EUS-FNA (Fisher exact test, P = .023). There were no procedure-related complications. LIMITATIONS: A single-center, retrospective analysis with small sample size. CONCLUSION: SINK biopsy appears to be an easy, safe, and effective technique for determining the definitive pathological diagnosis, evaluation of the malignant potential, and planning management of SETs. It could be a reliable alternative to conventional FNA, providing larger samples that improve the histologic yield.