EUS & ERCP Monthly
UKIEUS — Education & Media Series · www.ukieus.co.uk
August’s literature has a clear direction of travel: EUS is moving from rescue tool to first choice. A 29-centre propensity-matched study suggests that when standard cannulation fails in distal malignant obstruction with a dilated duct, going straight to EUS-guided drainage halves adverse events compared with escalating to precut. A network meta-analysis ranks EUS-gastroenterostomy above both duodenal stenting and surgical bypass, and a randomised trial challenges percutaneous liver biopsy on specimen adequacy. Closer to everyday ERCP, UK data from Cambridge, Oxford and King’s show biodegradable pancreatic stents removing the need for follow-up endoscopy, and an RCT-only network meta-analysis asks whether we choose biliary SEMS covering for the right reasons. In September, ESGE published its new Biliary Stricture Management guideline, which I reference throughout. From 358 papers harvested and 175 screened, five are featured and twelve listed.
— Dr Venkat Mahesh, MRCP, FEBGH · Education & Media Lead, UKIEUS
Top 5 Featured Papers
Early EUS-guided biliary drainage halves adverse events compared with advanced cannulation when ERCP access is difficult in distal malignant obstruction
Spadaccini M, … Repici A et al. · 29 centres (USA, Canada, Europe) · 313 vs 626 matched · Am J Gastroenterol 2026
DOI: 10.14309/ajg.0000000000004182 · PMID: 42572882
Standard cannulation failed in 2,066 of 10,738 ERCPs (19.2%) in patients with distal malignant obstruction and a CBD over 12 mm. After 1:2 propensity matching, early EUS-BD carried fewer adverse events than advanced ERCP cannulation (9.3% vs 18.2%) and higher technical success (95.9% vs 82.0%), with comparable clinical success. Post-ERCP pancreatitis accounted for most of the excess, occurring after 12.0% of advanced-cannulation attempts. Regression and IPTW sensitivity analyses supported the main result.
The direction fits the randomised evidence for primary EUS-BD: lower pancreatitis and higher technical success in the 2018 Korean RCT (Paik, Am J Gastroenterol), DRA-MBO and ELEMENT (both Gastroenterology 2023), and a 2026 RCT in ducts ≥15 mm showing pancreatitis of 1.8% vs 7.3%. A 2025 RCT of precut vs EUS-guided rendezvous also found less pancreatitis with the EUS approach (8.7% vs 1.9%, P=0.06). Two cautions apply. This is retrospective, and the choice of early EUS-BD reflects which centres had the expertise in the room. And the corroborating 2025 Endoscopy cohort and the 2026 Gastroenterology RCT come from the same research group, so they are not independent replication.
Primary literature identified via PubMed systematic search. Broader clinical context, including identification of contemporaneous trials, meta-analyses, and guideline positions, verified through structured evidence synthesis. All references independently confirmed via PubMed prior to inclusion.
Biodegradable pancreatic stents match plastic stents for post-ERCP pancreatitis prophylaxis — and remove the follow-up endoscopy
Galanakis V, … Corbett G · Cambridge, Oxford, King’s College Hospital · n=355 · Endoscopy 2026
DOI: 10.1055/a-2936-0508 · PMID: 42628939
Across three UK centres (2020–2025), 178 patients received biodegradable and 177 plastic pancreatic stents. After inverse probability weighting, post-ERCP pancreatitis was 7.3% vs 6.2%, an adjusted risk difference of +3.59% (95% CI −3.27 to +11.37). Repeat endoscopy within three months fell from 45.2% to zero. The upper confidence limit of about 11% means equivalence is plausible but non-inferiority is not proven. The authors acknowledge residual imbalance by procedure year, because biodegradable stents arrived later in the study period.
A UK NHS cost model from Wolverhampton (Pancreatology 2025) found biodegradable stents cost-effective at £9,316.70 per QALY. A 2026 Korean RCT (Clin Gastroenterol Hepatol) adds an important counterpoint: checking plastic-stent passage at 4 rather than 2 weeks cut endoscopic removal from 18.8% to 6.6% with no rise in pancreatitis. Part of the plastic arm’s follow-up burden may therefore be a pathway problem rather than a device problem.
Primary literature identified via PubMed systematic search. Broader clinical context, including identification of contemporaneous trials, meta-analyses, and guideline positions, verified through structured evidence synthesis. All references independently confirmed via PubMed prior to inclusion.
EUS liver biopsy gives longer, more adequate cores than percutaneous biopsy — at the cost of more minor adverse events in this trial
Johri I, … Daniel J · Institute of Liver and Biliary Sciences, New Delhi · n=90 · Gastrointest Endosc 2026
DOI: 10.1016/j.gie.2026.07.046 · PMID: 42580638
Ninety patients were randomised to EUS biopsy (19 G Franseen needle, modified wet-heparin suction, both lobes) or percutaneous biopsy (18 G full-core needle, right lobe). Adequacy (length ≥15 mm and ≥8 complete portal tracts) was 97.8% vs 64.4% (risk difference 33.3%, 95% CI 17.9–48.1), meeting non-inferiority and superiority. Median specimen length was 6.9 vs 2.0 cm and portal tracts 42 vs 9, but EUS cores were more fragmented and narrower. Adverse events were 26.7% vs 4.4%, all but one minor.
The safety signal runs the opposite way to the 2026 Japanese RCT (Okuwaki, J Gastroenterol Hepatol, n=80), which reported 2.4% vs 25.6% largely because of transient hypertension in the percutaneous arm. A 2025 meta-analysis of four RCTs (258 patients, Endoscopy) found no difference in adverse events (RR 0.54, 0.20–1.46) but less pain with EUS. The trials disagree because they count different minor events. For a UK audience, note that the percutaneous comparator here was an 18 G needle, thinner than the 16 G the BSG recommends, so the control arm was disadvantaged on adequacy.
Primary literature identified via PubMed systematic search. Broader clinical context, including identification of contemporaneous trials, meta-analyses, and guideline positions, verified through structured evidence synthesis. All references independently confirmed via PubMed prior to inclusion.
No metal stent type wins on patency or survival in malignant distal biliary obstruction — but fully covered stents cost more gallbladders
Simadibrata DM, … Chandrasekhara V · Mayo Clinic / San Raffaele · 10 RCTs, n=1,659 · Endoscopy 2026
DOI: 10.1055/a-2948-9408 · PMID: 42667933
Restricting the analysis to randomised trials removes the observational bias of earlier meta-analyses. There was no difference between fully covered, partially covered and uncovered SEMS in time to recurrent obstruction, any recurrent obstruction, or overall survival. Each type failed differently: uncovered stents had the least migration, tumour overgrowth and sludge, while fully covered stents best prevented tumour ingrowth. Adverse events were otherwise similar, but fully covered stents — not partially covered — carried significantly more acute cholecystitis than uncovered stents in patients with an intact gallbladder.
The finding agrees with a 2025 meta-analysis of 21 studies (5,753 patients, GIE): covered stents had more cholecystitis (RR 1.71) and migration (RR 7.53) but less ingrowth (RR 0.13). A large single-centre cohort (GIE 2017) reported 7.8% vs 1.2% cholecystitis with an intact gallbladder. Older evidence suggests tumour involvement of the cystic duct origin may matter more than the covering itself, and this network cannot test that.
Primary literature identified via PubMed systematic search. Broader clinical context, including identification of contemporaneous trials, meta-analyses, and guideline positions, verified through structured evidence synthesis. All references independently confirmed via PubMed prior to inclusion.
EUS-gastroenterostomy outranks duodenal stenting and surgical bypass for malignant gastric outlet obstruction on almost every outcome
Peng YN, … Chen YI · McGill University · 6 RCTs + 10 causally adjusted cohorts, n=1,585 · Gastrointest Endosc 2026
DOI: 10.1016/j.gie.2026.07.038 · PMID: 42546781
EUS-GE had lower re-intervention than enteral stenting (RR 0.17) and surgical gastrojejunostomy (RR 0.30), and higher clinical success than both (RR 1.20 and 1.17). Compared with surgery it shortened hospital stay by about six days and time to oral intake by about two. It ranked first on most outcomes. Surgery retained the highest technical success in the network estimate. Most of the evidence comes from expert centres, and the authors flag that outcomes elsewhere are unknown.
The randomised anchors are DRA-GOO (Lancet Gastroenterol Hepatol 2025), where re-intervention by six months was 4% with EUS-GE vs 29% with duodenal stenting with no difference in 30-day mortality or adverse events, and a 2025 RCT against surgery (Gut) with a composite failure rate of 7.9% vs 38.9% and lower cost. A Markov analysis (Int J Surg 2025) suggests stenting remains competitive when expected survival is around a month.
† Editorial override: moved into the Top 5 in place of Paik et al. (score 77), because the latter’s NSAID-free trial population limits direct relevance to UK practice. Override logged in the Ranking & Selection tab.
Primary literature identified via PubMed systematic search. Broader clinical context, including identification of contemporaneous trials, meta-analyses, and guideline positions, verified through structured evidence synthesis. All references independently confirmed via PubMed prior to inclusion.
Also Published This Month
1. [ERCP] † Acetate is no better than lactated Ringer’s for preventing post-ERCP pancreatitis without NSAIDs
Multicentre double-blind RCT, n=813: PEP 11.5% vs 12.4% (RR 0.93). Two-thirds of patients needed only a 4-hour symptom-guided hydration protocol.
Paik WH et al. · United European Gastroenterol J 2026 · DOI: 10.1002/ueg2.70281 · PMID: 42640715
2. [EUS-GBD] Transduodenal access and larger LAMS linked to better outcomes in EUS-gallbladder drainage — 18 centres, n=496
Transduodenal access lowered LAMS-related adverse events (SHR 0.42); LAMS over 10 mm was associated with less recurrent cholecystitis and mortality. Conversion to open surgery at interval cholecystectomy was 6.5%.
De Wispelaere L et al. · United European Gastroenterol J 2026 · DOI: 10.1002/ueg2.70275 · PMID: 42592708
3. [EUS / endohepatology] EUS-measured portal pressure gradient predicts decompensation better than HVPG
Prospective exploratory cohort, n=90: no decompensation when EUS-PPG was below 10 mmHg; C-index 0.78 vs 0.56. Agreement with HVPG was poor.
Santopaolo F et al. · Endoscopy 2026 · DOI: 10.1055/a-2948-9632 · PMID: 42674000
4. [EUS-GE] International consensus recommends EUS-GE over duodenal stenting for malignant gastric outlet obstruction
66 experts, modified Delphi with GRADE: 20 of 25 statements reached consensus, including technique, training and institutional requirements.
Vanella G et al. · Endoscopy 2026 · DOI: 10.1055/a-2948-2876 · PMID: 42674001
5. [ERCP] Disposable elevator-cap duodenoscopes did not reduce MDR post-ERCP cholangitis — n=1,039
Multidrug-resistant cholangitis was 2.5% in both groups, although surveillance cultures grew organisms only from conventional scopes.
Kim CH et al. · Gastrointest Endosc 2026 · DOI: 10.1016/j.gie.2026.08.006 · PMID: 42575379
6. [ERCP / EUS] EUS before ERCP avoided unnecessary ERCP in 79% of high-likelihood CBD-stone patients with negative imaging
Prospective multicentre, n=48: accuracy 95.8%, negative predictive value 100%. Small, but directly relevant to UK pathways.
Sugiura R et al. · J Hepatobiliary Pancreat Sci 2026 · DOI: 10.1002/jhbp.70163 · PMID: 42544755
7. [ERCP / devices] Evidence does not support wholesale replacement of reusable endoscopes with single-use models [UK: Tyldesley-Marshall N and colleagues — Birmingham, Warwick, King’s, Darlington, Hull, Coventry]
Systematic review and meta-analyses of 50 studies: single-use duodenoscopes had a 7% crossover rate to reusable scopes, and infection-risk estimates are limited by reporting bias.
Tyldesley-Marshall N et al. · J Gastroenterol Hepatol 2026 · DOI: 10.1111/jgh.70637 · PMID: 42568228
8. [EUS / AI] AI assistance improved detection of solid pancreatic tumours on EUS video, mainly for small, transient lesions
Randomised tandem video study, 57 endosonographers: accuracy 70.8% → 77.1%.
Montanelli J et al. · Endoscopy 2026 · DOI: 10.1055/a-2902-2838 · PMID: 42617651
9. [EUS] EUS findings excluded 10% of cholangiocarcinoma patients from curative-intent surgery or transplant
Meta-analysis of 9 studies, n=831; the incremental benefit over cross-sectional imaging was 3%.
Sabrie N et al. · Endoscopy 2026 · DOI: 10.1055/a-2944-4167 · PMID: 42648707
10. [EUS / varices] Adding conventional injection after EUS-guided therapy reduced gastric variceal rebleeding
Propensity-matched, three centres, n=141: rebleeding 7.1% vs 25.0%, re-intervention 12.5% vs 35.7%.
Yu F et al. · Ther Adv Gastroenterol 2026 · DOI: 10.1177/17562848261483777 · PMID: 42668851
11. [Cholangioscopy] Antibiotic prophylaxis may lower post-cholangioscopy cholangitis, but the evidence is very low certainty
Meta-analysis of 6 studies, 2,531 patients: 3.0% vs 7.8% (RR 0.46, 0.22–0.95), not robust to removing individual studies.
Alsaleh T et al. · Gastrointest Endosc 2026 · DOI: 10.1016/j.gie.2026.08.018 · PMID: 42624299
12. [ERCP / sedation] Remimazolam-esketamine halved sedation-related adverse events compared with propofol-fentanyl in older ERCP patients
Single-centre RCT, n=160: 17.5% vs 35.0% (RR 0.50).
Yu Y et al. · Drug Des Devel Ther 2026 · DOI: 10.2147/DDDT.S633603 · PMID: 42668967