UEMR vs Conventional EMR in 2026: When Skipping the Lift Is the Better Capture

Clinical Bottom Line

Lesion scenario	2026 practical answer
10-20 mm sessile adenoma	UEMR is the better capture technique. Yamashina [7]: R0 69% vs 50%, en bloc 89% vs 75%. Deng 2024 replicated the R0 gain. R0 is still not 100%.
20-30 mm adenomatous LNPCP	Reasonable UEMR-first alternative to conventional hot EMR. Spanish RCT: overall recurrence similar, 20-30 mm recurrence 3.4% vs 13.1%.
>30-40 mm LNPCP	En bloc remains limited (Nagl [8] 33% vs 18%). Recurrence benefit sat in a subgroup, not the primary endpoint. Suspected invasion is ESD or surgery, not more water.
Residual or fibrotic lesion after piecemeal EMR	Strongest indication. Kim: [6] endoscopic complete removal 89% vs 32%. Injection into scar often worsens geometry.
Appendiceal orifice adenoma	Feasible (Binmoeller 2016: 89% endoscopic success). If you cannot exclude extension into the appendix, operate.
SSL without suspected dysplasia	Not this article. ESGE 2024 [1] default is cold snare or cold EMR. See the cold vs hot EMR post.
6-9 mm nonpedunculated	Not a UEMR indication. Zhang: UEMR was not better than CEMR. Cold snare is the current default.

What do I do tonight?

1. If the lesion is a 10-20 mm sessile adenoma, UEMR is the better capture technique (Yamashina R0 69% vs 50%).
2. If it is a 20-30 mm adenomatous LNPCP, UEMR is a reasonable alternative to conventional hot EMR, not a mandate.
3. If it is larger than 30-40 mm or invasion is suspected, do not use more water as an en bloc cancer operation. That is ESD or surgery.
4. If this is residual or fibrotic tissue after piecemeal EMR, UEMR is the strongest indication (Kim 89% vs 32% complete removal).
5. If the lesion floats, skip injection. If it will not float, convert to a lift (Siau needed a lift in 30% of >=10 mm cases).
6. If the defect is a large right-colon EMR bed, clip it anyway. Underwater resection does not cancel delayed-bleed prophylaxis.

What do the societies say?

Do not use UEMR as an en bloc cancer operation. Do not skip clipping a right-colon large defect because the resection was underwater. SSL without dysplasia stays on the cold pathway.

Is UEMR the default, or still an alternative?

This page is underwater EMR versus conventional gas-insufflation EMR with a lift. It is not the cold-versus-hot debate.

ESGE 2024 rec 5: conventional diathermy EMR remains the strong-recommendation default for adenomatous large nonpedunculated colorectal polyps. Rec 6: UEMR can be considered an alternative (weak recommendation, moderate evidence). USMSTF 2020 describes the physics without replacing CEMR. ASGE's EMR curriculum flags UEMR for difficult-to-lift and recurrent lesions.

The mechanism is real. With water filling, mucosa and submucosa involute and float off a still-circular muscularis propria. Gas distention flattens and thins the wall. Skip injection when the lesion floats. Do not treat injection as banned. Siau needed a lift in 29.9% of lesions >=10 mm, mostly when size hit 30 mm. Convert rather than force a dry snare.

When is UEMR better than conventional EMR?

Yamashina (five Japanese centers, 10-20 mm sessile, 108 UEMR vs 102 CEMR): R0 69% vs 50%. En bloc 89% vs 75%. Adverse events 2.8% vs 2.0%. Deng 2024 (10-20 mm, unsedated, 99 per arm): R0 73.3% vs 56.3%. En bloc 91.1% vs 80.6%. That is the highest-quality reason to pick UEMR for an intermediate sessile adenoma.

Do not launder Le 2025 (Vietnamese tertiary, 10-30 mm, R0 98.4% vs 90.3%) as the expected Western R0. Yamashina and Nagl are the realistic numbers.

Does UEMR reduce recurrence?

Nagl (single German center, 20-40 mm): 6-month overall recurrence 15.1% vs 24.6%, not significant. The >30-40 mm subgroup was 6.3% vs 42.9%. En bloc 33.3% vs 18.4%. Rodriguez-Sanchez (11 Spanish hospitals, 311 lesions): overall recurrence 9.5% vs 11.7%, not significant. The 20-30 mm band showed 3.4% vs 13.1%. Lenz (10-40 mm): 6-month recurrence 2% vs 15%, driven by 21-40 mm lesions.

RCT-only metas do not agree on recurrence. Chandan [12] 2023: recurrence RR 0.62. Chowdhury 2023: no significant recurrence difference. They do agree that en bloc is higher with UEMR. Write the disagreement into the consent conversation. Do not sell UEMR as an across-the-board recurrence killer.

Binmoeller's 2015 2-4 cm LST series killed the stub's "huge 30 mm lesions in a single snare" line: endoscopic en bloc 55% with a 33 mm snare. Li's single-arm pool: en bloc 83% at 10-19 mm vs 36% at >=20 mm. Water does not make a 30 mm LST an en bloc cancer operation.

Is UEMR safer than conventional EMR?

Chandan's RCT-only meta: delayed bleed 1.14% vs 2.05%, perforation 0.2% vs 0.26%, both not significant. Garg's mixed meta: delayed bleed 2.2% vs 1.8%. USMSTF quotes 2-5% delayed bleed for UEMR of lesions >=10 mm. Ponugoti reported a perforation. Kawamura's 30 mm ascending lesion would not float, then got injected, then perforated. Thin right colon is still thin.

Clip large right-colon defects after UEMR the same way you would after conventional EMR. ESGE rec 10 is about the defect, not the insufflation gas. Prophylactic closure details live on the clip-closure post.

After piecemeal hot UEMR of an LNPCP, still ablate the margin (ESGE rec 8). Klein SCAR is CEMR data. The logic still applies: water does not replace snare-tip soft coagulation.

When is UEMR the salvage tool?

Kim 2014 is the comparative fibrotic-residual paper: UEMR vs conventional salvage after piecemeal EMR of LSTs >=2 cm. En bloc 47% vs 16%. Endoscopic complete removal 89% vs 32%. Later recurrence 10% vs 39%. Retrospective, single center, and still the cleanest salvage signal. This is where UEMR earns its keep.

Appendiceal orifice: Binmoeller 2016, n=27, technical success 89%. Three index failures because adenoma into the appendix could not be excluded. Uchima's cap-suction trick is an adjunct when a flat orifice lesion will not capture. Success is not 100%.

Learning curve data (Wang, Nagl, Pimentel 2026) describe CEMR-skilled operators. Pimentel: no en bloc difference after 10 or 15 cases in 10-35 mm lesions. That is not a fellow's first EMR. Yen mixed cold snare for 6-9 mm polyps into both arms; do not cite it as a pure hot-UEMR vs hot-CEMR trial for small polyps.

Pitfalls

Pitfall	Better move
Treating UEMR as "no injection ever"	Skip the lift when the lesion floats. Convert when it stays tethered (Siau: 30% still needed a lift).
Selling UEMR as safer perforation-wise	RCT metas show comparable perforation. Case reports exist.
Using overall recurrence from Nagl or Rodriguez-Sanchez as the headline	Both overall endpoints missed. Benefits sat in size subgroups and in R0/en bloc.
Skipping right-colon clips because there was no lift	Clip the defect, not the insufflation method.
UEMR for 6-9 mm lesions or SSL without dysplasia	Cold resection owns those lanes.
Calling 30 mm LSTs routinely en bloc	Binmoeller 2015: 55%. Nagl: 33%.
UEMR as ESD-lite for NICE 3 / Kudo V	Those lesions still need true en bloc or surgery.

Selected references

1. Ferlitsch M, et al. ESGE Guideline Update: colorectal polypectomy and EMR. Endoscopy. 2024.
2. Kaltenbach T, et al. USMSTF: endoscopic removal of colorectal lesions. Am J Gastroenterol. 2020.
3. Binmoeller KF, et al. Underwater EMR without submucosal injection. Gastrointest Endosc. 2012.
4. Binmoeller KF, et al. Attempted underwater en bloc resection of 2-4 cm LSTs. Gastrointest Endosc. 2015.
5. Binmoeller KF, et al. Underwater EMR of appendiceal orifice adenomas. Gastrointest Endosc. 2016.
6. Kim HG, et al. Underwater EMR for recurrences after piecemeal resection. Gastrointest Endosc. 2014.
7. Yamashina T, et al. Underwater vs conventional EMR of intermediate-size polyps. Gastroenterology. 2019.
8. Nagl S, et al. Underwater vs conventional EMR of large sessile polyps. Gastroenterology. 2021.
9. Rodriguez Sanchez J, et al. Underwater versus conventional EMR of LNPCLs. Gastrointest Endosc. 2023.
10. Lenz L, et al. Underwater versus conventional EMR for nonpedunculated lesions. Gastrointest Endosc. 2023.
11. Deng Q, et al. UEMR vs CEMR for medium-sized sessile polyps. Sci Rep. 2024.
12. Chandan S, et al. RCT meta-analysis of UEMR vs CEMR. Endosc Int Open. 2023.
13. Chowdhury AR, et al. Updated RCT meta-analysis of UEMR vs CEMR. Endosc Int Open. 2023.
14. Siau K, et al. Feasibility of UEMR for >=10 mm polyps. Surg Endosc. 2018.
15. Ponugoti PL, Rex DK. Perforation during underwater EMR. Gastrointest Endosc. 2016.
16. Pimentel-Nunes P, et al. UEMR learning curve. GE Port J Gastroenterol. 2026.