EUS & ERCP Monthly

UKIEUS — Education & Media Series  ·  www.ukieus.co.uk

Vol. 1, Issue 6  ·  July 2026  ·  Editor: Dr Venkat Mahesh, MRCP, FEBGH

From the editor — July 2026

This edition turns on getting familiar devices and techniques right, rather than reaching for new ones. A Japanese multicentre trial (ESCAPE) shows that changing a retrieval basket’s geometry — not its size or material — reduces the need for mid-procedure device escalation for small bile duct stones, though whether basket or balloon should be first-line at all remains a separate, more finely balanced question. A large Helsinki registry finds that patients who undergo cholecystectomy before their PSC diagnosis carry markedly worse long-term outcomes than those diagnosed with an intact gallbladder — a timing signal with real implications for surveillance intensity. A multicentre Kuwait/UK cohort adds substantial real-world experience with a self-assembling peptide hydrogel for ERCP-related bleeding (PuraStat™), now backed by a genuine randomised trial rather than case series alone. A 13-centre Taiwanese registry identifies the precise anatomical view responsible for stent mis-deployment in EUS-guided gastroenterostomy, and shows that a simple avoidance strategy can drive the rate to zero. And a large 11,464-patient meta-analysis finally separates true recurrence from genuinely new disease in the pancreatic remnant after IPMN resection — a distinction that current guidelines already anticipate, but which has rarely been quantified this precisely. Six more papers worth your attention are listed below the Top 5.

— Dr Venkat Mahesh, Education & Media Lead, UKIEUS

RANKED #1  ·  SCORE 83/100  ·  MULTICENTRE RCT  ·  ERCP / BILIARY

Spiral basket geometry cuts device escalation for small bile duct stones — though the balloon-versus-basket question itself remains unsettled

Ogura T, Ohnishi K, Sato T, et al.  ·  200 patients, multicentre (Japan)  ·  Endoscopy 2026

DOI: 10.1055/a-2924-7286  ·  PMID: 42526878

The ESCAPE trial randomised 200 patients with common bile duct stones ≤8 mm 1:1 to a spiral-configured or standard 8-wire retrieval basket. Both study arms used the same 8-wire configuration, differing only in geometry; this was not a comparison against a “basic” basket. The primary outcome — first-line procedural success without device escalation within 10 minutes — was achieved in 97.9% of the spiral group versus 85.4% of the standard group (95% CI for the difference 4.9–20.1; p=0.004; number needed to treat 8). The spiral basket required fewer sweeps (2.71 vs 3.31; p=0.01) and less procedure time (3.96 vs 5.48 minutes; p=0.01), with lower device-related costs as reported directly by the trial authors in USD (mean difference −108, bootstrap 95% CI −185 to −42). Final stone clearance rates and adverse event rates did not differ significantly. The authors themselves are explicit that the spiral basket was not compared with balloon extraction and should not be interpreted as superior to it.

Two things temper how this should be read. First, the trial did not include a balloon-catheter arm, so it says nothing about whether basket or balloon should be first-line for small stones at all — a separate and more mixed literature, and one where trial-level differences in stone size cut-offs and device generation limit direct comparison. Sharma S, Sharma S, Singhal S, Sanaka MR. Endoscopy International Open 2022 (PMID 36397864, DOI 10.1055/a-1905-0251) pooled three RCTs (508 patients) and found a modest overall advantage for balloon catheters on complete clearance (89.8% vs 81.8%; RR 1.10, 95% CI 1.03–1.18; I²=0%), but the three component trials do not agree on direction: two favoured balloon (Ishiwatari et al., Endoscopy 2016, PMID 26760604, 92.3% vs 80.0%, p=0.037; Ekmektzoglou et al. 2020, PMID 33321014, 94.0% vs 84.1%) while a third found no difference (Ozawa et al., J Gastroenterol 2016, PMID 27631594, 83.9% vs 81.3%, p=0.70). This inconsistency, despite a statistically significant pooled result, is why the balloon-versus-basket question should be treated as unresolved rather than settled. Second, the spiral device evaluated is currently available primarily in Japan, which limits immediate UK applicability, though the underlying geometric principle may inform future device design more broadly.

Primary literature identified via PubMed systematic search. Broader clinical context, including identification of contemporaneous trials, verified through structured evidence synthesis. All references independently confirmed via PubMed prior to inclusion.

RANKED #2  ·  SCORE 74/100  ·  PROSPECTIVE REGISTRY  ·  HEPATOBILIARY / PSC

Cholecystectomy performed before PSC diagnosis linked to a fivefold higher risk of death, malignancy, or transplantation

Färkkilä MA, Arola JT, Tenca A, et al.  ·  977 patients, Helsinki PSC registry  ·  JHEP Reports 2026

DOI: 10.1016/j.jhepr.2026.101967  ·  PMID: 42508616

Among 977 patients with PSC enrolled in the prospective Helsinki registry, 163 (16.7%) underwent cholecystectomy — 100 before and 63 after their PSC diagnosis. Patients cholecystectomised before diagnosis had more advanced disease at baseline (higher ERCP scores, Mayo risk scores, FIB-4 index, and more advanced fibrosis on liver histology) and reached a composite endpoint of liver-related death, hepatobiliary malignancy, or transplantation in 37.0% of cases, versus 6.5% with an intact gallbladder and 5.9% cholecystectomised after diagnosis. After adjustment for timing, age, sex, and disease severity, pre-diagnosis cholecystectomy carried a fivefold higher risk of that composite endpoint (HR 5.16, 95% CI 3.00–8.88, p<0.001; the wide interval means the true effect could be as modest as a tripling of risk) and a restricted mean survival time 5 years shorter than patients with an intact gallbladder. Because the endpoint includes malignancy rather than measuring transplant-free survival specifically, the finding should not be read as a transplant-specific outcome.

A population-based Ontario PSC-IBD cohort (476 patients, linked administrative data, multistate modelling; Leung KK et al., Hepatology 2026, PMID 41637315, DOI 10.1097/HEP.0000000000001692) offers an important counterweight: mortality was higher after cholecystectomy in that cohort too, but there was no detected difference in subsequent cancer incidence or transplantation, and the authors attributed the excess mortality to the disease processes that prompted surgery in the first place — dysplasia, malignancy, or severe biliary disease — rather than to the surgery itself. This tempers how far the Helsinki finding should be read as causal. Neither dataset establishes causation, and no current guideline (AASLD 2023, PMID 36083140; EASL-aligned data in Altenmüller et al. 2025, PMID 40856347) ties the cholecystectomy decision to timing relative to PSC diagnosis at all — both trigger surgery on gallbladder polyp size (>8 mm, or growth) found on routine surveillance imaging, not on disease chronology.

Primary literature identified via PubMed systematic search. Broader clinical context, including a contemporaneous population-based cohort and current AASLD/EASL guidance, verified through structured evidence synthesis. All references independently confirmed via PubMed prior to inclusion.

RANKED #3  ·  SCORE 71/100  ·  MULTICENTRE PROSPECTIVE COHORT  ·  ERCP (UK AUTHOR +5)

Self-assembling peptide hydrogel achieves 97.7% immediate haemostasis for ERCP-related bleeding in a 138-patient real-world cohort

Alali AA, Alkandari AA, Hashim A, Alzafiri R, Bhandari P.  ·  138 patients, multicentre (Kuwait/UK)  ·  Surgical Endoscopy 2026

DOI: 10.1007/s00464-026-13197-3  ·  PMID: 42509506

This multicentre prospective cohort followed 138 consecutive patients (mean age 51.3, 65.9% female) in whom a self-assembling peptide hydrogel was used to manage ERCP-related bleeding — predominantly intraprocedural oozing (82.6%), with smaller numbers for delayed bleeding (10.9%) and prophylactic use (6.5%). Technical success was achieved in all patients; immediate haemostasis was achieved in 97.7% of those with active bleeding. Thirty-day rebleeding occurred in 5.1%; other adverse events (3.6%) were post-ERCP pancreatitis, with no procedure-related deaths.

This is an uncontrolled cohort, there is no comparator arm, and 68.2% of bleeding treated was mild oozing — a category that frequently resolves without any intervention, which inflates apparent efficacy relative to a randomised comparison. The source study itself supports this concern directly: its own subgroup analysis found significantly better outcomes for mild (oozing) bleeding than for moderate bleeding, and found equal efficacy whether the hydrogel was used as primary or rescue (secondary) therapy, confirming it was not used exclusively as sole treatment throughout the cohort. This sits alongside genuine randomised evidence for the first time: PROTECT-EST (Ogura T, Ikeura T, Takenaka M, et al., Gastrointestinal Endoscopy 2026, PMID 42250776, DOI 10.1016/j.gie.2026.05.039), a 130-patient Japanese multicentre non-inferiority trial, found the same class of hydrogel non-inferior to balloon tamponade for oozing-type post-sphincterotomy bleeding; in a prespecified exploratory superiority analysis (the trial’s own designation, not an inference of ours), initial haemostasis favoured the hydrogel (92.3% vs 76.9%; rescue device use 7.7% vs 23.1%, p=0.033). One trial of 138 patients is a meaningful step from case series to randomised evidence, but it is not yet a settled evidence base. The 2020 ESGE guideline on ERCP-related adverse events makes only one formal recommendation for post-sphincterotomy bleeding — temporary covered metal stent placement for bleeding refractory to standard measures — and does not yet endorse topical hydrogels as first-line, a position echoed in the 2025 ESGE technical review, which describes these agents as promising adjuncts still awaiting further randomised and cost-effectiveness data.

UK author: Pradeep Bhandari, Portsmouth Hospitals University NHS Trust and Mayo Clinic London — whose primary published expertise is luminal EMR/ESD rather than ERCP; his co-authorship here does not itself signal ERCP-specific procedural authority.

Primary literature identified via PubMed systematic search. Broader clinical context, including a contemporaneous randomised controlled trial and current ESGE guidance, verified through structured evidence synthesis. All references independently confirmed via PubMed prior to inclusion.

RANKED #4  ·  SCORE 69/100  ·  MULTICENTRE REGISTRY  ·  THERAPEUTIC EUS

Cross-sectional bowel view identifies the anatomy behind stent misdeployment in EUS-guided gastroenterostomy, and avoiding it drove the rate to zero

Kuo YT, Yang CY, Chen JH, et al.  ·  165 patients, 13 centres (Taiwan)  ·  Digestive Endoscopy 2026;38(7):e70235

DOI: 10.1111/den.70235  ·  PMID: 42466906

This registry across 13 Taiwanese tertiary centres enrolled 165 patients undergoing EUS-guided gastroenterostomy for malignant gastric outlet obstruction between July 2020 and June 2025. Technical and clinical success were 98.2% and 95.2% respectively; overall stent misdeployment occurred in 4.8% of procedures, with same-session endoscopic salvage successful in 87.5% of those cases. A cross-sectional (short-axis) EUS view of the target bowel loop — the registry’s own “EM type C” classification — was the sole independent predictor of misdeployment on multivariable analysis (adjusted OR 16.9, p=0.001), carrying a 33% misdeployment risk versus the favourable longitudinal or oblique views (EM types A/B). Fluoroscopic morphology added complementary risk stratification. From 2023, the centres adopted a stepwise strategy prioritising favourable EUS views and minimising stomach-to-bowel distance; misdeployment fell to zero thereafter in this registry.

This is a before-and-after observation within a single registry, not a randomised comparison; the authors themselves caution that the zero-misdeployment period in 2023–2025 likely reflects proactive avoidance of high-risk anatomy rather than improved management of it, and note the later cohort was significantly older and had worse baseline performance status, yet still achieved zero adverse events of any kind. The adjusted OR of 16.9 for EM type C (95% CI 3.17–89.71, p=0.001) carries a very wide interval, reflecting that only 8 misdeployment events occurred across the whole cohort; the authors describe their own findings as hypothesis-generating pending larger validation, not confirmatory. Fluoroscopic morphology type II was associated with misdeployment on univariable analysis but did not reach significance after adjustment (aOR 7.13, 95% CI 0.82–61.99, p=0.075); the authors position FM as a qualitative “macroscopic roadmap” alongside EM, not as an independently validated predictor. Separately, misdeployment was independently associated with worse survival (adjusted HR 2.57, 95% CI 1.22–5.41, p=0.013) even with an 87.5% same-session salvage rate — a signal that these events carry prognostic weight beyond their technical management. Misdeployment rates in the wider literature range from roughly 4.6% (Giri S et al., pooled meta-analysis, PMID 38747173) to 9.85% (Ghandour B et al., Gastrointestinal Endoscopy 2022, PMID 34352256, DOI 10.1016/j.gie.2021.07.023).

Primary literature identified via PubMed systematic search. Broader clinical context, including guideline positioning and a late-complication case series, verified through structured evidence synthesis. All references independently confirmed via PubMed prior to inclusion.

RANKED #5  ·  SCORE 64†/100  ·  SYSTEMATIC REVIEW / META-ANALYSIS  ·  PANCREATIC CYSTS

Meta-analysis of 11,464 patients separates true recurrence from new disease in the pancreatic remnant after IPMN resection

van Bodegraven AJ, Marstaller-Walz KM, Padoan V, et al.  ·  66 studies, 11,464 patients  ·  BJS Open 2026;10(4)

DOI: 10.1093/bjsopen/zrag087  ·  PMID: 42441942

This PRISMA/Cochrane-guided meta-analysis (PROSPERO 1009399) set out to separate three things that post-resection IPMN literature has historically lumped together as “recurrence”: progression of residual disease, de novo development of a genuinely new lesion, and true recurrence of an invasive component. Two headline figures matter, and they apply to different populations. Among patients who had invasive IPMN-derived pancreatic cancer at resection, recurrence of that invasive cancer in the remnant occurred in 41.9% (1,646/3,925), with a 5-year pooled recurrence-free survival of only 46.6%; lymph node involvement and tubular subtype were the strongest independent predictors. Separately, among patients resected for non-invasive IPMN, overall recurrence in the remnant occurred in 11.2% (831/7,446); within that smaller group, 49.1% of recurrences were reclassified as de novo lesions in the remnant pancreas, 20.9% as progression of a cyst already present at the index operation, and the remaining 30.0% could not be classified from available data — a substantial fraction that should temper how much weight the de novo finding carries. The meta-analysis does not distinguish whether “progression of a cyst already present” reflects a lesion intentionally left in situ or genuinely incomplete resection; these carry different clinical implications and the source data do not allow the two to be separated. More fundamentally, across 66 pooled studies this de novo versus progression distinction almost certainly rests on radiological and temporal criteria (new location, distance from the resection margin, time since surgery) rather than molecular clonal lineage tracing; the pooled 49.1% de novo figure should be read as a radiological classification, not a proven biological one.

The de novo finding matters for surveillance policy, not for relaxing it. Current guidelines already anticipate this: the Fukuoka/IAP international consensus (Tanaka M et al., Pancreatology 2017, PMID 28735806) explicitly frames the remnant pancreas as a “field defect” at risk of new IPMN or concomitant PDAC independent of the original lesion, and requires lifelong surveillance for all resected IPMN regardless of index grade, explicitly rejecting the 2015 AGA guideline’s five-year stop rule for low/intermediate-grade lesions. In line with this guidance, the case for continued lifelong surveillance of the whole gland is strengthened, not weakened, by the de novo finding. A West of Scotland cohort (McGuigan MC, Cameron AJ, Dickson EJ, et al., Annals of Surgery 2024, PMID 39480643, DOI 10.1097/SLA.0000000000006580 — e-published October 2024, so older than this month’s other featured evidence) offers a complementary UK data point: 746 Glasgow patients aged 65 or above with cysts stable below 15 mm for five or more years had a standardised incidence ratio of zero for malignant progression, suggesting a defensible basis for discontinuing surveillance in a narrowly defined low-risk subgroup — a more targeted approach than a blanket time-based cut-off.

† Editorial tie-break: this paper scored equally (64/100) with two EUS-GBD studies, including GALLBLADEUS-2 (Noteworthy Mention #1, below) and a second EUS-GBD paper on contained gallbladder perforation not featured in this edition. The editor selected van Bodegraven for the Top 5 given its direct bearing on active surveillance-policy debate.

Primary literature identified via PubMed systematic search. Broader clinical context, including current international consensus guidelines and UK cohort data, verified through structured evidence synthesis. All references independently confirmed via PubMed prior to inclusion.

1. Predictors of biliary drainage dysfunction after EUS-guided gallbladder drainage for malignant distal biliary obstruction (GALLBLADEUS-2)

Ascites was the sole independent predictor of drainage dysfunction (HR 2.45, 95% CI 1.17–5.13 — a wide interval, with the lower bound close to 1.0) across 166 patients at 17 centres; 15 mm stents were associated with fewer high-grade adverse events (14.3% vs 32.9%) without affecting patency, though statistical significance of this specific comparison was not confirmed by this newsletter beyond the editor’s carried-forward summary.

Goudot M, Lorenzo D, Caillo L, et al.  ·  Digestive Endoscopy 2026;38(7):e70211  ·  DOI: 10.1111/den.70211  ·  PMID: 42446476

2. Transcystic and choledochotomy laparoscopic bile duct exploration show distinct outcomes against ERCP + laparoscopic cholecystectomy [UK authors: Ahmed Z, Thompson-Reil C, D’Souza P, D’Souza R, Davidson BR — Imperial, Colchester, Royal Free]

Meta-analysis of 11 RCTs (1,472 patients): choledochotomy achieved the highest ductal clearance (93.5%) but significantly higher bile leak than ERCP+LC (9.8% vs 0.8%; RR 7.00); transcystic had the lowest overall morbidity but higher conversion to open surgery (3.5%).

Ahmed Z, Thompson-Reil C, D’Souza P, et al.  ·  Updates in Surgery 2026  ·  DOI: 10.1007/s13304-026-02694-8  ·  PMID: 42484834

3. Conscious sedation reduces adverse cardiovascular events versus general anaesthesia for emergency ERCP in frail elderly patients

Single-centre RCT, 108 patients aged 70–90: adverse cardiovascular events occurred in 17% vs 40% (p<0.05, exact value not confirmed) favouring conscious sedation, with fewer new-onset arrhythmias (5.7% vs 18.2%). Single-centre and modestly sized; published in an anaesthesia rather than gastroenterology journal. Whether patients were formally frailty-scored, versus simply falling in this age band, could not be confirmed by this newsletter.

Li C, Zeng N, Wang F, et al.  ·  Saudi Journal of Anaesthesia 2026;20(3):567–576  ·  DOI: 10.4103/sja.sja_94_26  ·  PMID: 42553904

4. Long-term fistula persistence and weight regain after EUS-directed transgastric ERCP (EDGE) in Roux-en-Y gastric bypass patients

73 patients, median 4.9-year follow-up after LAMS removal: persistent or recurrent fistula in only 4%; modest long-term weight regain (+3.1 kg). Whether this sits within an expected post-bariatric trajectory is this newsletter’s characterisation, not a comparator directly cited from the source study.

Ichkhanian Y, Ali AH, Easler JJ, et al.  ·  Surgical Endoscopy 2026  ·  DOI: 10.1007/s00464-026-13109-5  ·  PMID: 42414776

5. Ultra-low-pass whole-genome sequencing of bile cell-free DNA detects malignancy a median 38 days before routine diagnosis in biliary strictures

95 patients: bile cfDNA (cell-free DNA; not tumour-specific ctDNA — the assay measures total cell-free DNA burden, not a confirmed tumour-derived fraction) was detected in 51% of initially indeterminate strictures later confirmed malignant, in some cases over 100 days before diagnosis; specificity in confirmed benign strictures was 96%. Note that 51% detection in this subset functions as a sensitivity figure and is modest — a negative bile cfDNA result should not be read as excluding malignancy.

Rojo C, Rullan M, Sogbe M, et al.  ·  eGastroenterology 2026;4(3):e100425  ·  DOI: 10.1136/egastro-2026-100425  ·  PMID: 42602860

6. First JGES International consensus statement on EUS-guided biliary drainage terminology and indications

14 experts from 6 countries, structured Delphi process: EUS-BD recognised as a safe and effective alternative to ERCP-BD but not yet standard first-line; unified terminology proposed for interventional EUS procedures.

Iwashita T, Mukai S, Nakai Y, et al.  ·  Digestive Endoscopy 2026;38(7):e70194  ·  DOI: 10.1111/den.70194  ·  PMID: 42429199

Mid-month quiz — coming soon

Two clinical questions based on this month’s featured papers — with scenario-based reasoning and full literature-linked explanations — will be available on the UKIEUS website once the new Education section is live. Visit www.ukieus.co.uk and navigate to Education & Guidelines.

© UKIEUS 2026  ·  Editorial commentary by Dr Venkat Mahesh, MRCP, FEBGH

Paper titles and abstracts cited from PubMed with DOI attribution; no copyrighted full text reproduced.

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.

The evidence score is a relative editorial ranking tool, not a validated instrument. For educational use only.  ·  www.ukieus.co.uk