Colorectal Polyps
A colorectal polyp is a protrusion of tissue into the lumen above the surrounding mucosa. Usually asymptomatic, but may bleed, cause obstructive symptoms when large, or be a cancer precursor. General content below is from a textbook reference chapter (Sabiston Ch52 - Colorectal Neoplasia & Colon Cancer Surgery); the Malignant polyps section and everything below it has been substantially expanded with a second, more detailed textbook chapter specifically on malignant-polyp management (ASCRS Ch23 - Management of Malignant Polyps) — neither is primary literature.
Classified endoscopically as pedunculated (with a stalk) or sessile (flat), and histologically (after excision/biopsy) as adenomas, hamartomas, inflammatory, serrated, etc. Main clinical importance: neoplastic polyps with CRC potential should be removed to reduce cancer risk.
Nonneoplastic polyps
- Hyperplastic polyps — small, sessile, usually <5 mm; elongated colonic crypts with a papillary epithelial configuration, no atypia. Very common, frequently grossly indistinguishable from small adenomas. No malignant potential (small, distal ones — see "serrated polyps" below for the malignant-potential subset).
- Exception: small (<10 mm) hyperplastic polyps in the rectum or sigmoid specifically are considered low-risk with a 10-year surveillance interval (see surveillance table below) — consistent with their lack of malignant potential.
- Inflammatory polyps (pseudopolyps) — found in areas of healing inflammation (e.g. IBD — see Inflammatory Bowel Disease), formed after full-thickness epithelial ulceration in which new mucosa regenerates in an irregular polypoid configuration. No intrinsic neoplastic potential, but can be large and mimic a neoplasm; must be differentiated from true neoplastic lesions, particularly in diseased colons otherwise at risk for cancer (e.g. longstanding UC/Crohn's).
- Hamartomas — uncommon GI polyps; can be sporadic or associated with a genetic syndrome (Peutz-Jeghers syndrome, juvenile polyposis syndrome, PTEN hamartoma syndrome — see Hamartomatous Polyposis Syndromes). No intrinsic malignant potential. Removal indicated for obstructive symptoms or bleeding.
Serrated polyps
Divided into three types: hyperplastic polyps (not precancerous — see above), sessile serrated polyps, and traditional serrated adenomas. The latter two combine adenomatous and hyperplastic features — colonic crypts with a saw-tooth serrated configuration plus nuclear atypia. Patients with sessile serrated polyps or traditional serrated adenomas have an increased CRC risk; CRC arising in these patients usually follows the serrated neoplasia pathway rather than the classic adenoma-carcinoma sequence (see Colorectal Cancer Molecular Pathways — this pathway corresponds to CIMP). These polyps should be removed, with patients followed by serial endoscopy.
Serrated polyposis syndrome (SPS)
Per ASCRS Ch22 - Sporadic and Inherited Colorectal Cancer: SPS is the most common polyposis syndrome currently known, found in as many as 1:111 individuals in screening cohorts (though overall population incidence is estimated at <1%), predominantly in patients of European lineage. Previously called hyperplastic polyposis syndrome; the definition has broadened to include other serrated lesions (sessile serrated polyps, serrated adenomas), not just hyperplastic polyps. WHO 2019 diagnostic criteria (either one, cumulative lifetime polyp count): (1) at least 5 serrated polyps proximal to the rectum, with 2 of these >10mm in size, or (2) over 20 serrated polyps of any size distributed throughout the large bowel, with at least 5 proximal to the rectum.
Unlike the syndromes above, SPS does not have a fully elucidated genetic basis — diagnosis relies exclusively on clinical criteria; BRAF mutations are commonly observed, and mutations in the ubiquitin ligase gene RNF43 have been identified in some SPS families. Considerable phenotypic overlap exists with MUTYH-associated polyposis (see Familial Adenomatous Polyposis (FAP)'s MAP section) — testing for a MUTYH mutation is reasonable in the setting of concurrent adenomas.
Lifetime risk for advanced neoplasia may be as high as 50%, though precise estimates are unclear; limited data suggest intensive colonoscopic surveillance may reduce CRC risk in SPS patients. SPS is strongly associated with smoking, and smoking cessation as a modifiable risk factor should be reinforced in patients meeting SPS clinical criteria.
Treatment: complete clearance of all polyps should be performed during colonoscopy, with a surveillance interval of 1–3 years based on polyp burden. Patients who develop cancer should undergo segmental or total colectomy after informed discussion regarding risks and anticipated bowel function.
Neoplastic polyps
All adenomas have malignant potential.
- Tubular adenomas — branched tubular glands; the most common type, 65–80% of polyps removed; frequently pedunculated.
- Villous adenomas — long, fingerlike projections of surface epithelium; 5–10% of adenomas; commonly sessile.
- Tubulovillous adenomas — elements of both; 10–25% of adenomas.
Malignancy risk increases with size (large), gross shape (sessile), histologic type (villous), and grade of dysplasia. An advanced adenoma — defined as ≥1 cm in size, high-grade dysplasia, or tubulovillous/villous histology — carries a significantly increased CRC risk. For example: <5% incidence of carcinoma in a tubular adenoma <1 cm, vs. a 50% chance a villous adenoma >2 cm contains a cancer.
Removal: adenomatous polyps found at colonoscopy should be excised. Pedunculated polyps are commonly removed with cold or hot snare polypectomy. Sessile polyps are frequently elevated from the underlying muscularis by saline injection, then excised with various techniques. Sessile polyps with a central depression that does not elevate adequately with saline injection (the "nonlifting sign") are at increased risk both for perforation with endoscopic removal and for harboring neoplasia — commonly referred for surgical removal via segmental colectomy. Large polyps not removable endoscopically are also referred for surgery, though larger polyps can sometimes still be removed endoscopically with techniques such as endoscopic mucosal resection (EMR) and endoscopic submucosal dissection (ESD).
Malignant polyps
A malignant polyp is one where histologic examination after removal reveals a focus of carcinoma that has invaded through the muscularis mucosa into the submucosa (pT1 on the TNM classification). The key question: is complete endoscopic removal sufficient, or is further surgery (completion/ segmental colectomy) needed? Carcinomas that do not breach the muscularis mucosa carry negligible metastatic risk (no lymphatics in the mucosa/lamina propria layer) and are considered "benign" despite historically confusing terminology (intramucosal adenocarcinoma, carcinoma in situ, high-grade dysplasia have all been used loosely for these lesions) — adequately treated with complete endoscopic resection alone, no further workup needed. Carcinomas that do invade the submucosa carry a real risk of local recurrence and lymph node metastasis, with depth of submucosal invasion (SMI) as the single most important risk factor.
Malignant polyps account for ~12% of all polyps, with incidence possibly rising as screening programs improve.
Depth of invasion and lymph node metastasis risk
Haggitt classification (for pedunculated polyps — level 0 corresponds to carcinoma limited to the mucosa, i.e. not actually a malignant polyp by the pT1 definition above):
| Level | Description |
|---|---|
| 0 | Carcinoma in situ, limited to the mucosa (not a true malignant polyp) |
| 1 | Invading the submucosa, limited to the head of the polyp |
| 2 | Invading to the neck (junction of head and stalk) |
| 3 | Invading any part of the stalk |
| 4 | Invading the submucosa of the colon wall, below the level of the stalk but above the muscularis propria |
Sessile polyps with submucosal invasion are, by definition, Haggitt level 4 (no head/neck/stalk anatomy to stage against) — the Kikuchi (Sm) classification is used instead, based on the fraction of submucosal thickness invaded: Sm1 = upper third, Sm2 = middle third, Sm3 = lower third. Kitajima et al. standardized histopathological SMI evaluation and found: for pedunculated (Haggitt) lesions, lymph node metastasis rate was 0% at Haggitt level 1, and 0% at level 2/3 when SM depth was <3000μm. For non-pedunculated (sessile) lesions, lymph node spread risk was 0% when SMI was <1000μm, 3.9% when SMI was <2000μm, and 17.1% when SMI was ≥2000μm — i.e. risk scales continuously with measured invasion depth, not just the three-tier Sm1/2/3 bucket. Select malignant polyps with SMI <1000μm and no other unfavorable histopathological factors (see below) can be adequately treated with endoscopic resection alone.
Histopathological risk factors for lymph node metastasis
Beyond depth of invasion, several histopathological features independently predict lymph node metastasis. A systematic review and meta-analysis (23 studies, 4510 patients) found: unfavorable tumor grade — poorly differentiated adenocarcinoma or mucinous adenocarcinoma vs. well/moderately differentiated (OR 5.60), lymphovascular invasion (OR 4.81), tumor budding — single or cluster cancer cells at the invasive front (OR 7.74), and depth of SMI >1000μm (OR 3.87) were all significantly associated with lymph node spread. A malignant polyp's endoscopic resection is considered curative only if all of the following are met: all resection margins negative, SMI <1000μm, well-to-moderately differentiated tumor grade, and absence of lymphovascular invasion and/ or tumor budding. Completion/segmental colectomy is commonly recommended when any of these criteria are not met (e.g., pedunculated Haggitt level 4, sessile Kikuchi Sm2/Sm3, poor differentiation, lymphovascular invasion, tumor budding, or incomplete removal/close resection margins <1mm) — the systematic review above found these features carry an additive risk (see the ACPGBI risk-stratification tool below).
More granular LVI/differentiation-combination data, per ASCRS Ch27 - Rectal Cancer Local Excision (Chang et al., 943-patient single-institution cohort of pT1–2 rectal cancers): lymphovascular invasion was the variable most strongly associated with lymph node metastasis, with a reported odds ratio of 11.5.
Flag — discrepant LVI and differentiation effect-size estimates, not reconciled: Chang et al.'s OR 11.5 for LVI is substantially higher than the OR 4.81 reported above from the 23-study/ 4510-patient systematic review; similarly, a 17-study meta-analysis (Bosch et al., also per ASCRS Ch27) reports RR 4.9 (95% CI 3.3–6.9) for poor differentiation, vs. this page's OR 5.60 from the 23-study review. Both discrepant pairs point the same direction (LVI and poor differentiation are each strong, independent LNM risk factors) but differ in magnitude — plausibly single-institution vs. pooled-meta-analysis variation, but not adjudicated here; reported side by side per this wiki's practice rather than silently merged.
Chang et al.'s combination-level data (their Table 27.1, reused with permission):
| Cancer | LVI | Poor differentiation | LNM, n | LNM, % |
|---|---|---|---|---|
| pT1 | − | − | 18/241 | 7.5% |
| pT1 | − | + | 1/5 | 20% |
| pT1 | + | − | 9/15 | 60% |
| pT1 | + | + | 3/3 | 100% |
| pT2 | − | − | 87/569 | 15.3% |
| pT2 | − | + | 5/16 | 31.3% |
| pT2 | + | − | 59/88 | 67.0% |
| pT2 | + | + | 6/6 | 100% |
LVI and poor differentiation combine roughly additively (each present factor increases risk substantially), with LVI+/poor-differentiation+ tumors carrying essentially universal nodal involvement in this cohort at both pT1 and pT2.
LVI and recurrence pattern (rather than just nodal metastasis), per Hogan et al. (also cited in ASCRS Ch27, their Table 27.2, reused with permission) — a Cox multivariate model of independent association between LVI-positivity and locoregional (LR) vs. systemic (SR) recurrence in rectal cancer specifically:
| Parameter | LR: p | LR: HR (95% CI) | SR: p | SR: HR (95% CI) |
|---|---|---|---|---|
| Age | 0.37 | 1.02 (0.98–1.07) | 0.49 | 0.99 (0.95–1.03) |
| Male | 0.88 | 1.09 (0.37–3.21) | 0.42 | 0.68 (0.24–1.80) |
| Stage III | 0.05 | 2.57 (1.01–6.93) | 0.05 | 1.66 (0.99–6.42) |
| Total LN | 0.09 | 0.92 (0.83–1.01) | 0.06 | 0.99 (0.91–1.07) |
| LVI+ | 0.40 | 1.57 (0.55–4.49) | 0.04 | 2.57 (1.04–6.39) |
In rectal cancer, LVI+ was an independent predictor of systemic recurrence (p=0.04) but not of locoregional recurrence (p=0.40) — the opposite pattern from colon cancer in the same study, where LVI+ independently predicted locoregional recurrence (p=0.02) but not systemic recurrence (p=0.31). Lymphovascular invasion has also been separately reported to be associated with reduced overall survival in rectal cancer patients.
Tumor budding — definition and the ITBCC 2016 consensus
Per ASCRS Ch27 - Rectal Cancer Local Excision: tumor budding was first described by Hase in 1993, defined as small clusters of undifferentiated cancer cells at the invasive front of a lesion. Initial reports of its prognostic significance came primarily from the Japanese literature; more recent Western-center studies have supported it as an independent adverse prognostic factor in colorectal carcinoma, potentially offering risk stratification beyond TNM staging alone — particularly relevant in early (T1) rectal cancer, where it factors into the endoscopic-resection-curative-criteria and ACPGBI tool above.
The International Tumor Budding Consensus Conference (ITBCC) 2016 issued eleven GRADE-based consensus statements formalizing tumor budding's definition, assessment, and reporting role:
| # | Statement | Vote (consensus) | Evidence |
|---|---|---|---|
| 1 | Tumor budding is defined as a single tumor cell or a cell cluster of four tumor cells or fewer | Strong, 22/22 (100%) | High |
| 2 | Tumor budding is an independent predictor of lymph node metastasis in pT1 colorectal cancer | Strong, 23/23 (100%) | High |
| 3 | Tumor budding is an independent predictor of survival in stage II colorectal cancer | Strong, 23/23 (100%) | High |
| 4 | Tumor budding should be taken into account along with other clinicopathological features in a multidisciplinary setting | Strong, 23/23 (100%) | High |
| 5 | Tumor budding is counted on H&E | Strong, 19/22 | Moderate |
| 6 | Intramural budding exits in colorectal cancer and has been shown to be related to lymph node metastasis | Strong, 22/22 (100%) | Low |
| 7 | Tumor budding is assessed in one hotspot (a field measuring 0.785 mm²) at the invasive front | Strong, 22/22 (100%) | Moderate |
| 8 | For tumor budding assessment in colorectal cancer, the hotspot method is recommended | Strong, 22/22 (100%) | Moderate |
| 9 | A three-tier system should be used along with the budding count in order to facilitate risk stratification in colorectal cancer | Strong, 23/23 (100%) | Moderate |
| 10 | Tumor budding should be included in guidelines/protocols for colorectal cancer reporting | Strong, 23/23 (100%) | High |
| 11 | Tumor budding and tumor grade are not the same | Strong, 23/23 (100%) | High |
The consensus explicitly clarifies that tumor budding and tumor grade (differentiation) are distinct features (statement 11) — both should be assessed and reported independently, consistent with the combination-level data above (Chang et al.'s table) treating them as separate variables rather than a single composite.
Endoscopic resection of malignant polyps
Endoscopic resection is the preferred treatment for benign precursor lesions and for malignant polyps without features suggesting deep SMI, given it spares the cost/morbidity/mortality of surgery. Endoscopic features suggestive of deep SMI (which should prompt biopsy and surgical referral rather than resection attempt): Paris 0-IIc (depressed) morphology, Paris 0-III (ulcerated/excavated) morphology, and Kudo type V/Vn or NICE type III pit/vascular pattern — see Endoscopic Polypectomy, EMR & ESD for the full Paris/Kudo/NICE classification detail and the non-lifting sign (this page does not repeat that detail). En bloc resection is mandatory for malignant polyps — assessment of SMI depth and resection margin status cannot be reliably obtained from a fragmented (piecemeal) specimen, and piecemeal EMR of a malignant polyp is considered non-curative regardless of favorable histology, since it inevitably requires surgery anyway due to the high risk of understaging from compromised pathologic interpretation. Malignant polyps ≤2cm can often be removed en bloc with EMR; larger lesions usually require ESD (see Endoscopic Polypectomy, EMR & ESD for full EMR/ESD technique — not repeated here). The resected specimen should be pinned (cork board or similar) before formalin fixation to prevent edge-curling, which can compromise margin measurement and depth-of-SMI accuracy.
Malignant polyps and early rectal cancers are also well suited to transanal endoscopic surgery (TES) (see Local Excision of Rectal Neoplasms for TEM/TAMIS technique) — TES can remove lesions en bloc in either the submucosal or full-thickness plane, with low specimen-fragmentation rates (<10%) and polyp recurrence rates of 5–11% reported; the chapter's cited margin-status range for TES (66–93%) is reported here exactly as stated, though its placement in a sentence about "low rates" reads as though it should describe negative rather than positive margins — flagged as an apparent internal inconsistency in the source rather than silently resolved. A North American multicenter series of rectal ESD (n=171) found an en-bloc resection rate of 82.5%, R0 resection rate of 74.9%, and was curative in 81.8% of malignant polyps with favorable histologic features.
Predicting risk of residual disease: the ACPGBI risk-stratification tool
Assessing occult lymph node metastasis risk in a resected T1 CRC is an imperfect science — no single histologic feature perfectly predicts it. The ACPGBI (Association of Coloproctology of Great Britain and Ireland) position statement provides a weighted risk-scoring tool, summing points for each unfavorable feature present:
| Histologic criterion | Points |
|---|---|
| Resection margin <1mm | 4 |
| Resection margin 1–2mm | 1 |
| Pedunculated, Haggitt level 4 | 4 |
| Sessile, Kikuchi Sm2 | 2 |
| Sessile, Kikuchi Sm3 | 4 |
| Poor differentiation | 3 |
| Mucinous tumor | 1 |
| Tumor budding | 1 |
| Lymphovascular invasion | 2 |
| Total score | Risk grade | Estimated residual-cancer risk | Recommended action |
|---|---|---|---|
| 0 | Very low | <3% | Routine follow-up |
| 1 | Low | <5% | Assess other factors, close follow-up |
| 2 | Medium | 5–10% | Discuss risk/benefit of surgery vs. follow-up |
| 3 | High | 8–15% | Discuss risks, err toward surgery |
| ≥4 | Very high | >20% | Recommend surgery unless patient unfit |
In a 3-year UK regional-cancer-network study using this tool to guide multidisciplinary teams across 173 patients after endoscopic resection of malignant colorectal polyps, 37 patients (21.4%) underwent primary surgical resection with a residual disease rate of 43%, while 136 patients managed with surveillance had only a 4.4% recurrence rate — suggesting the tool meaningfully separated higher- from lower-risk patients in practice.
Recurrence and surveillance after endoscopic resection
Recurrence/metastatic disease after endoscopic resection of a malignant polyp is mainly reported within 3–5 years. Local recurrence is rare after en-bloc resection — in a retrospective study of 310 colorectal ESD cases (53 of them T1 CRC), disease-free survival was 100% and all local recurrences (2% of the full cohort) occurred in patients who had piecemeal resection. There is no consensus on the specific surveillance method or timing; experts suggest periodic CEA and chest/ abdominal-pelvic CT, though no definitive interval is agreed. Two specific guideline positions are cited: the European Society of Gastrointestinal Endoscopy (ESGE) recommends surveillance colonoscopy at 6 months following piecemeal endoscopic resection of any colorectal polyp >10mm; since local recurrence is rare after en-bloc resection specifically, the Japan Gastroenterological Endoscopy Society (JGES) suggests follow-up colonoscopy within 3 years after en-bloc resection of a malignant polyp.
Postpolypectomy surveillance
Finding an adenoma at colonoscopy is a risk factor for additional polyps. (This table covers benign precursor lesions generally; see the malignant-polyp-specific surveillance guidance immediately above for T1 CRC managed endoscopically.)
| Index colonoscopy findings | Repeat colonoscopy |
|---|---|
| Small (<10 mm) hyperplastic polyps in rectum or sigmoid | 10 years |
| Low risk: 1–2 small tubular adenomas <10 mm, low-grade dysplasia | 5–10 years (AGA) / 10 years (ESGE) |
| High risk: villous histology, high-grade dysplasia, size ≥10 mm, or ≥3 polyps | 3 years |
| Piecemeal removal of a polyp | 6 months |
| Sessile serrated polyp <10 mm | 5 years |
| Sessile serrated polyp >10 mm, with dysplasia, or traditional serrated adenoma | 3 years |
Related pages
- Endoscopic Polypectomy, EMR & ESD — full EMR/ESD technique, Paris/Kudo/NICE classification detail, and non-lifting sign; this page's malignant-polyp content builds directly on that technique detail without repeating it, and this update resolves that page's previously-flagged open item about cross-checking EMR/ESD malignancy-risk figures against this page's content.
- Local Excision of Rectal Neoplasms — TAMIS/TEM technique for TES; this page adds malignant-polyp-specific TES outcome data. That page's T1N0 criteria, oncologic results, and adjuvant/neoadjuvant-therapy content build directly on this page's depth-of-invasion and nodal-risk-factor data.
- Colorectal Cancer Molecular Pathways — the adenoma-carcinoma and serrated-neoplasia pathway background behind this page's Neoplastic/Serrated polyps sections above.
- Colorectal Cancer Staging (TNM) — pT1 is the TNM classification underlying the malignant-polyp definition used throughout this section.
- Hereditary Colorectal Cancer Syndromes — hereditary polyposis syndromes (which change surveillance intervals dramatically) are covered separately.
Open items / gaps
- No primary trial data on EMR/ESD outcomes specifically, the ACPGBI tool's original validation cohort, or the comparative accuracy of the nonlifting sign — see Endoscopic Polypectomy, EMR & ESD for that page's own flagging of the same limitation. Chang et al., Hogan et al., and the ITBCC 2016 consensus (added above via ASCRS Ch27 - Rectal Cancer Local Excision) are likewise not independently ingested from primary publications.
- The TES margin-status figure (66–93% "positive margins," in a sentence otherwise describing low complication rates) is reported here exactly as extracted from the chapter — flagged as a likely transcription/printing anomaly in the source rather than silently corrected.
- Hereditary polyposis syndromes (which change surveillance intervals dramatically) are covered separately — see Hereditary Colorectal Cancer Syndromes.
- Two unresolved effect-size discrepancies added this ingest: LVI's odds ratio for lymph node metastasis (11.5 per Chang et al. vs. 4.81 per the existing 23-study systematic review) and poor differentiation's risk ratio (4.9 per Bosch et al.'s 17-study meta-analysis vs. OR 5.60 per the existing 23-study review) — both flagged in the Histopathological risk factors section above rather than merged into a single figure.