Rectal ESD Guidelines
Rectal endoscopic submucosal dissection (ESD) should be performed for large (>20 mm) rectal lesions with suspected limited submucosal invasion (depressed morphology, irregular surface pattern, or bulky component) or lesions that cannot be completely removed by snare-based techniques, as it achieves superior en bloc resection rates (96%) and dramatically lower recurrence rates (2.3%) compared to EMR (18.8% recurrence). 1, 2
Primary Indications for Rectal ESD
The US Multi-Society Task Force and ESGE guidelines establish clear criteria 1, 2:
Large lesions >20 mm with features suggesting limited submucosal invasion:
- Demarcated depressed areas (Paris 0-IIc)
- Irregular or non-granular surface pattern
- Large protruding or bulky components
- LST non-granular type (particularly pseudodepressed variants)
Lesions with fibrosis from prior biopsy or attempted resection where snare capture is difficult
Local residual early carcinoma after previous endoscopic resection
Non-polypoid colorectal dysplasia in inflammatory bowel disease patients
Any rectal lesion where piecemeal EMR would be required but en bloc resection is preferred for accurate histologic staging
Why ESD Over EMR for Rectal Lesions
The most recent meta-analysis (2026) definitively shows ESD superiority for rectal tumors 3:
- En bloc resection: 96.1% (ESD) vs 9.7% (EMR)
- Local recurrence: 2.3% (ESD) vs 18.8% (EMR)
- Complete resection rates: equivalent (79-82%)
- Adverse events: no significant differences in bleeding, perforation, or stricture rates
This evidence strongly supports ESD as first-line treatment for complex rectal lesions where en bloc resection matters for oncologic outcomes.
Lesions NOT Requiring ESD
Standard polypectomy or EMR remains appropriate for 1:
- Small polypoid lesions <20 mm without concerning features
- LST granular homogeneous type without nodularity
- Lesions with clear lifting sign and no fibrosis
- Sessile serrated lesions (cold EMR highly effective)
Technical Approach
The 2020 US guidelines describe the standard technique 1:
- Submucosal injection with lifting solution
- Circumferential incision 3-4 mm beyond lesion margins
- Submucosal dissection using ESD knife through expanded submucosal plane under direct visualization with cap assistance
- En bloc resection in one piece
Hybrid ESD (partial submucosal dissection followed by snare resection) provides a safety bridge for lesions with severe fibrosis or when full ESD expertise is limited 1.
Recent innovations (2026) include tunneling techniques, traction systems, and advanced closure methods for large rectal lesions 4, 5.
Quality Benchmarks
Expert centers should achieve 6, 2:
- R0 resection rate: >90%
- Perforation rate: <1%
- Bleeding rate: <1%
Critical Caveats
Avoid practices that induce submucosal fibrosis 7:
- Never tattoo beneath or immediately adjacent to lesions
- Avoid partial snare resection for "diagnostic biopsies" of large lesions
- These practices can make subsequent ESD technically impossible
Referral to high-volume centers is essential 7, 5. Rectal ESD is technically demanding with a steep learning curve. The 2019 AGA update emphasizes that endoscopists should complete structured training (didactic courses, animal models, expert observation) before attempting patient cases, starting with lowest-complexity lesions 7.
Curative Resection Criteria
After rectal ESD, resection is considered curative (no additional treatment needed) when 2:
- En bloc R0 resection achieved
- Well to moderately differentiated histology
- No lymphovascular invasion
- Superficial submucosal invasion (sm1) if present
- No tumor budding grade 2-3
- Lesion of any size (colorectal-specific criterion)
High-risk features requiring multidisciplinary discussion include lymphovascular invasion, deeper than sm1 invasion, positive vertical margins, undifferentiated tumor, or budding grade 2-3 2.
Surveillance
Scheduled endoscopic surveillance with high-definition white-light and chromoendoscopy (virtual or dye-based) is recommended after curative ESD, with biopsies only of suspicious areas 2.