EMR vs ESD for Large Colon Polyps in 2026: When En Bloc Is Worth the Knife

Clinical Bottom Line

Lesion scenario	2026 practical answer
Adenomatous LNPCP >=20 mm, no invasion signs	Conventional hot EMR is the ESGE strong-recommendation default [1]. After piecemeal hot EMR, STSC the margin. Clip a complete right-colon defect.
Same lesion, ESD available in a high-volume center	ESD is a weak alternative (rec 7, low-quality evidence), not a mandate. Jacques: 6-month recurrence 0.6% vs 5.1% after STSC-EMR, more adverse events, 47 vs 14.5 minutes [7].
Suspected limited SM invasion that a snare cannot take en bloc	En bloc: ESD, en bloc EMR if it will fit, EID, EFTR, or surgery (ESGE rec 11). Do not piecemeal a possible T1.
High-risk rectal morphology (large nodular / covert-cancer lane)	Strongest Western case for ESD rather than piecemeal EMR. Bahin: universal ESD was only cost-effective among higher-risk rectal lesions [16].
Fibrotic residual after piecemeal EMR	Conventional snare EMR often fails. UEMR salvage or ESD/referral. See the UEMR article.
SSL without suspected dysplasia	Cold snare or cold EMR. Not this article. See cold vs hot EMR.
Obvious deep invasion	Surgery. A longer ESD does not convert non-lifting cancer into an endoscopic cure.

What do I do tonight?

1. Photograph and classify first (Paris, LST type, NICE/JNET, pit pattern). The tool follows optical diagnosis, not lesion diameter alone.
2. If it is an adenomatous LNPCP without invasion signs and you do EMR well: resect by conventional hot EMR tonight.
3. If invasion is suspected and you cannot take it en bloc with a snare: stop adding pieces. Tattoo, photograph, and refer for ESD or surgery.
4. After piecemeal hot EMR: STSC the margin (Klein) and clip a complete right-colon defect when you can (ESGE rec 10). Details live on the clip-closure article.
5. If the lesion is fibrotic residual after piecemeal EMR: do not keep injecting and snaring air-EMR into scar. That is UEMR or ESD.
6. Site check at about 6 months after piecemeal EMR. En bloc R0 ESD does not need the same early scar hunt.

What do the societies say?

Is ESD the default for large polyps?

No. ESGE 2024 rec 5 still lists conventional hot EMR as the strong default for adenomatous LNPCPs. Rec 7 is a weak, low-quality alternative in selected cases and high-volume rooms. The 2022 ESD guideline says the same thing in different clothes: polypectomy or EMR for most colorectal lesions, ESD when you need a true en-bloc specimen [2].

Bahin's cost model found selective ESD for high SMIC suspicion was the least expensive strategy. Universal ESD cost about $210,000 per extra surgery avoided and was only cost-effective among higher-risk rectal lesions [16]. ESGE quotes that analysis: universal ESD cannot be justified beyond high-risk rectal lesions.

What did RESECT-COLON actually show?

Jacques randomized 360 patients with large nonpedunculated colonic adenomas (>=25 mm) to ESD or piecemeal EMR with snare-tip soft coagulation of the margin [7]. Primary analysis: 318 lesions. Six-month recurrence 1/161 ESD (0.6%) vs 8/157 EMR (5.1%). No recurrence among R0 ESD cases (90% of the ESD arm). Adverse events 35.6% vs 24.5%. ESGE extracted median procedure time from the same trial: 14.5 minutes for EMR vs 47 minutes for ESD.

Read that as a recurrence win for ESD in expert French hands, not as a reason to ESD every 25 mm adenoma tonight. The EMR arm already used STSC, so 5.1% is modern EMR, not Belderbos-era 20% piecemeal recurrence. The recurrences after EMR were still an endoscopic problem. This is a colonic adenoma trial, not a rectal T1 cancer trial.

How bad is piecemeal EMR recurrence in 2026?

Belderbos (2014, pre-STSC): pooled local recurrence 15%, 20% after piecemeal vs 3% after en bloc, most found by 6 months [9]. Klein SCAR changed the EMR story: first-surveillance recurrence 5.2% with thermal margin ablation vs 21.0% without [8]. That is why ESGE rec 8 is strong.

Moss ACE (1000 successful wide-field EMRs): early residual 16%, usually diminutive; 93% of recurrences treated endoscopically; 98.1% adenoma-free and surgery-free at 16 months when index EMR succeeded and there was no SM invasion needing an operation [10].

What do you trade for en bloc?

Fujiya: ESD had higher en bloc and curative resection and lower recurrence (OR 0.08), with longer procedures, more additional surgery (OR 2.16), and more perforation (OR 4.96) [11]. Arezzo (>20 mm called noninvasive): en bloc 89.9% vs 34.9%, perforation 4.9% vs 0.9%, further surgery 7.8% vs 3.0% [12].

Do not quote Japanese R0 as your R0. Fuccio: standard colorectal ESD R0 85.6% in Asia vs 71.3% outside Asia; en bloc 93% vs 81.2%; surgery for ESD adverse events 3.1% vs 0.8% [13]. Hybrid ESD (knife then snare) R0 60.6%. That is not a shortcut to Japanese numbers. Ohata's Japanese prospective cohort (1814 neoplasms >=20 mm) is what expert ESD looks like: local recurrence 0.5%, 5-year intestinal preservation 88.6% [15]. Piecemeal ESD and a positive margin predicted recurrence. North American mixed-organ ESD (Draganov) reached en bloc 91.5% and R0 84.2% in selected centers, with perforation 2.9% across organs, not colon-only [17].

Zhao's meta found higher ESD perforation outside Japan (8% vs 4%) and linked right-colon polyps to perforation risk [14]. The same paper concluded ESD should be first-line without a 20 mm restriction. ESGE 2024 rec 5 contradicts that conclusion. Use Zhao for geography of harm, not as an indication statement.

When is the knife the better medicine?

When the pathologist needs intact lateral and vertical margins, or when a snare cannot get the lesion off. Limited SM invasion signs: a demarcated depressed area with an irregular surface, or a large protruding/bulky component, especially over 20 mm, particularly in the rectum [2]. Fibrosis and residual after piecemeal EMR belong here or with UEMR, not with another lift-and-hope EMR. Deep invasion signs still go to surgery.

JGES 2020 pushes ESD earlier (LST-NG, fibrosis, recurrence) [4]. AGA's US ESD update is about who should be doing it, not about replacing EMR in every suite [5]. If you do not do colorectal ESD weekly, tonight's job is optical diagnosis, a clean EMR when that is enough, and a clean referral when it is not.

Pitfalls

Pitfall	Better move
Piecemeal EMR of a lesion with invasion signs	You destroy the T1 staging specimen. Refer for en bloc or operate.
Quoting Belderbos 20% as 2026 EMR recurrence	That is pre-STSC. Klein 5.2% and Jacques EMR 5.1% are the current numbers when the margin is ablated.
Selling ESD as safer or faster	It is neither. More perforation in every meta. Jacques: more AEs and 47 vs 14.5 minutes.
Laundering Japanese R0 or Ohata 0.5% recurrence as US community ESD	Fuccio Western R0 71.3%. Hybrid ESD is worse.
Universal ESD because "en bloc is always better"	Bahin: selective ESD. Universal ESD only pencils out in higher-risk rectum.
Skipping STSC or right-colon clips because ESD exists in the building	If you chose EMR, finish EMR: STSC the margin, clip the right-colon defect.
ESD for SSL without dysplasia, or for obvious deep invasion	Cold pathway for SSL. Surgery for deep invasion.

Selected references

1. Ferlitsch M, et al. ESGE Guideline Update: colorectal polypectomy and EMR. Endoscopy. 2024.
2. Pimentel-Nunes P, et al. ESGE Guideline Update: ESD for superficial GI lesions. Endoscopy. 2022.
3. Kaltenbach T, et al. USMSTF: endoscopic removal of colorectal lesions. Am J Gastroenterol. 2020.
4. Tanaka S, et al. JGES colorectal ESD/EMR guidelines. Dig Endosc. 2020.
5. Draganov PV, et al. AGA CPU: ESD in the United States. Clin Gastroenterol Hepatol. 2019.
6. Copland AP, et al. AGA CPU: appropriate and tailored polypectomy. Clin Gastroenterol Hepatol. 2024.
7. Jacques J, et al. En bloc ESD vs piecemeal EMR of large colonic adenomas (RESECT-COLON). Ann Intern Med. 2024.
8. Klein A, et al. Thermal ablation of EMR margins (SCAR). Gastroenterology. 2019.
9. Belderbos TD, et al. Local recurrence after EMR of nonpedunculated lesions. Endoscopy. 2014.
10. Moss A, et al. ACE study: long-term recurrence after wide-field EMR. Gut. 2015.
11. Fujiya M, et al. EMR vs ESD for colon neoplasms: comparative meta-analysis. Gastrointest Endosc. 2015.
12. Arezzo A, et al. ESD vs EMR for colorectal lesions: meta-analysis. United European Gastroenterol J. 2016.
13. Fuccio L, et al. Colorectal ESD outcomes, Asia vs West. Gastrointest Endosc. 2017.
14. Zhao S, et al. ESD vs EMR for colorectal polyps: meta-regression. World J Gastroenterol. 2021.
15. Ohata K, et al. Long-term outcomes after colorectal ESD in Japan. Gastroenterology. 2022.
16. Bahin FF, et al. Cost-effectiveness of selective vs universal ESD. Gut. 2018.
17. Draganov PV, et al. ESD in North America: prospective multicenter study. Gastroenterology. 2021.