Introduction

Optical evaluation differentiating between adenomatous and serrated histopathology is the first step in management of large non-pedunculated colorectal polyps (LNPCP) as it informs appropriate endoscopic resection technique selection. Performance of optical evaluation is modest for diminutive polyps but has not been reported for LNPCP.¹

Aims & Methods

We sought to evaluate the performance of optical evaluation to differentiate between adenomatous and serrated histopathology for LNPCP. Consecutive patients > 18 years of age who underwent endoscopic resection for a LNPCP were enrolled in a prospective multi-center observation cohort study (ClinicalTrials.gov NCT05402696). Pre-resection optical evaluation was performed using high-definition white-light and narrow-band imaging (NBI) with or without near-focus. The Japanese NBI Expert Team (JNET) classification was used to differentiate between serrated (JNET I) vs. adenomatous (JNET IIA, IIB) LNPCPs. Traditional serrated adenomas (TSA) and cancers were excluded from analysis. Sensitivity, specificity, and accuracy were used to evaluate optical evaluation performance.

Results

Between June 2022 to April 2024, 440 patients underwent 476 procedures for 558 LNPCP. Median LNPCP size was 30 mm (IQR 20-40) and 404 (72.4%) LNPCP were proximal to the splenic flexure. Histopathology of all lesions included 365 (65.4%) adenomas, 137 (24.6%) serrated lesions, 34 (6.1%) cancers, and 22 (3.9%) other lesions, which included 6 (1.1%) TSA. Adenomatous histopathology was found predominantly in LNPCP with Paris 0-IIA+IS (107/136, 78.7%) and Paris 0-IS (70/82, 85.4%) morphology. Paris 0-IIA lesions were both adenomatous (184/328, 56.1%) and serrated (125/328, 38.1%). Of 137 serrated LNPCP, 133 (97.1%) were predicted as serrated, with sensitivity, specificity, and accuracy of 97.08% (95% CI 92.69-99.20), 99.45% (95% CI 98.03-99.93), and 98.80% (95% CI 97.42-99.56), respectively. Of 365 adenomatous LNPCP, 350 (96.0%) were predicted as adenomatous, with sensitivity, specificity, and accuracy of 95.89% (95% CI 93.31- 97.68), 97.08% (95% CI 92.69- 99.20), and 96.22% (95% CI 94.15- 97.71), respectively.

Table 1: Optical Performance

Conclusion

For LNPCP, optical evaluation had greater than 95% accuracy to differentiate between serrated and adenomatous lesions, forming the basis of a selective resection algorithm.

References

1. Yamada M, Sakamoto T, Otake Y, Nakajima T, Kuchiba A, Taniguchi H, et al. Investigating endoscopic features of sessile serrated adenomas/polyps by using narrow-band imaging with optical magnification. Gastrointestinal Endoscopy. 2015 Jul 1;82(1):108–17.

Introduction

Endoscopic resection is the primary treatment strategy for most large (≥20 mm) non-pedunculated colonic polyps (LNPCP). A selective resection algorithm with endoscopic mucosal resection (EMR) and endoscopic submucosal dissection (ESD) has been shown to be cost effective and optimizes oncologic outcomes in the rectum.^1,2 However, an algorithm for the colon which incorporates cold snare resection (CSR) has not been described.

Aims & Methods

We sought to evaluate the performance of a colon-specific selective resection algorithm for LNPCP. Consecutive patients > 18 years of age who underwent endoscopic resection for a LNPCP were enrolled in a prospective multi-center observation cohort study (ClinicalTrials.gov NCT05402696). Modality selection was determined by optical evaluation using the Paris and the Japan NBI Expert Team classifications: 1) JNET I – CSR; 2) JNET IIA – EMR; 3) JNET IIB or Paris 0-IIC morphology – en bloc resection (EMR/ESD); 4) JNET III – referral to surgery/multi-disciplinary team review. Algorithm performance was evaluated by technical success (all neoplastic tissue removed at index procedure), procedure-related adverse events, referral to surgery, and recurrence at first surveillance colonoscopy (SC1).

Results

From June 2022 to April 2024, 382 patients underwent 402 procedures for 480 LNPCP. Median age was 67 years (IQR 62-74 years) and 174 (45.5%) were female. Median lesion size was 30 mm (IQR 20-40 mm) and 386 (80.4%) were located in the proximal colon. Cancer was present in 16 (3.3%) LNPCP. Based on optical evaluation, LNPCP were resected by CSR (25.8%), EMR (72.5%), and ESD (1.7%). Out of protocol deviations occurred in 23 (4.8%) LNPCP. Technical success was 97.0%, 100.0% and 100.0% for EMR, CSR, and ESD, respectively (p=0.172). Procedure duration was significantly shorter for CSR (median 10 min; IQR 5-15 min) compared to EMR (15 min; IQR 10-30 min) and ESD (55 min; IQR 38-90 min) (p<0.001). Perforation was seen in 11 (3.3%), 1 (14.3%) and 0 (0%) patients undergoing EMR, ESD and CSR, respectively (p=0.030). There were no significant differences between resection modalities in their frequencies of clinically significant post-endoscopic resection bleeding (5.2%) or delayed perforation (0.0%). 26 (5.4%) LNPCP were referred to surgery. Of 194 lesions completing SC1, there were 4 recurrences (0.8%), comprised of 1 (0.8%) and 3 (0.9%) in the CSR and EMR group, respectively (p=1.000); all of which were treated endoscopically.

Table 1: LNPCP Characteristics and Outcomes

Conclusion

A colon-specific selective resection algorithm for LNPCP using optical evaluation to select between CSR, EMR, and ESD has high efficacy, efficiency, and safety.

References

1. Bahin, Farzan F., et al. Wide-field endoscopic mucosal resection versus endoscopic submucosal dissection for laterally spreading colorectal lesions: a cost-effectiveness analysis. Gut 67.11 (2018): 1965-1973.
2. Shahidi N, Vosko S, Gupta S, Whitfield A, Cronin O, O’Sullivan T, et al. A Rectum-Specific Selective Resection Algorithm Optimizes Oncologic Outcomes for Large Nonpedunculated Rectal Polyps. Clin Gastroenterol Hepatol. 2023 Jan;21(1):72-80.e2.

Introduction

Studies report that almost 50% of Large Non-Pedunculated Colorectal Polyps (LNPCP) have been subjected to heavy manipulation in form of prior attempts at resection or extensive biopsy sampling prior to referral to an expert centre. This significantly impacts several critical outcomes of endoscopic resection (ER). However, understanding of standardised classifications of LNPCP, endoscopic diagnosis and indications for ER has increased in general endoscopy practice which may have improved decision making prior to referral. We aimed to compare trends in patterns of prior injudicious manipulation and the significance of biopsy sampling over sequential time periods in a large cohort of patients referred to a tertiary referral centre for ER.

Aims & Methods

Referral documentation, including clinic and referral letters, endoscopy reports, and histopathology result, were reviewed for evidence of prior resection attempts or biopsy sampling. Data included the number of biopsy samples taken and corresponding histological results. Patients referred between January 2011 and August 2024 were categorised into five time periods. Heavy manipulation was defined as either a prior resection attempt or ≥6 biopsy samples. Analyses included comparisons of resection attempts, heavy manipulation, biopsy sampling, and the detection of high-grade dysplasia on biopsies (as a marker of improved lesion assessment and targeted sampling). Trends in practice changes were assessed using the Chi-square test for trend.

Results

1407 LNPCP were treated by ER. There was a significant trend toward fewer prior failed attempts at resection over sequential time periods with 30% in Period 1 and only 11% in Period 5 (p<0.001); fewer LNPCP subjected to heavy manipulation with 61% in Period 1 and 27% in Period 5 (p<0.001) and fewer LNPCP with prior biopsy sampling with 77% in Period 1 and 47% in Period 5 (p<0.001). However, there was no trend found for an improvement in biopsy sampling yields of HGD with 18% in Period 1 and 25% in Period 5 (p=0.1).

Conclusion

In a large cohort of ER of LNPCP, this study demonstrates a significant reduction over time in failed resection attempts, heavy manipulation, and biopsy sampling of LNPCP before referral to a tertiary centre for endoscopic resection (ER). This suggests an improvement in endoscopic lesion assessment and decision making by general endoscopists. However, heavy manipulation remains a significant problem. Although fewer LNPCP are biopsied prior to referral, the yield of significant findings did not change, suggesting biopsies are not effectively targeted, and highlighting the need for further improvements in lesion assessment.

Introduction

Endoscopic resection (ER) of large non-pedunculated colorectal polyps (LNPCP) has advanced significantly, pushing boundaries to treat lesions of any size. However, a dearth of published data exists on the extremes of ER practice, particularly for large lesions creating circumferential or near-circumferential mucosal defects. While Japanese studies, focusing on ESD-resected LNPCP, show conflicting outcomes, two prominent Western series lean towards EMR-resected lesions. The Western landscape is evolving, with a notable rise in ESD for LNPCP. Our study delivers fresh insights, presenting outcomes from ER with ≥75% circumferential mucosal defects in a leading Western centre that predominantly favours ESD for such lesions. This signifies a transformative trend in ER methodology, underlining the efficacy of ESD in addressing intricate colorectal lesions, thereby shaping evolving standards.

Aims & Methods

1477 ERs of LNPCP ≥20mm were performed and grouped according to the extent of the mucosal defect and comparisons made between those with <75% and ≥75% defects. Independent predictors of stenosis were identified.

Results

137 patients had ER leaving ≥75% circumference defect, most located at or distal to the rectosigmoid, with ≥90% defects in 29 and 100% in 14. 31 (22.6%) were resected using EMR and 106 (77.3%) using ESD. En bloc resection was achieved in 64%, increasing to 88% in the last 4 years of the study where ESD was increasingly employed routinely. Overall complications were more likely in these patients (11.7% versus 3.4%, p<0.001), although there was no difference in the incidence of clinically significant perforation (p=0.41) or post-procedure bleeding (p=0.45). 101 patients had undergone surveillance. Stenosis developed in only 2 of 71 (2.8%) patients with defects ≥75% circumference but 10 of 18 (55.6%) patients with defects ≥90% and 10 of 12 patients (83.3%) with 100% circumferential defects. ≥90% circumference defect was an independent risk factor for stenosis (p<0.001). All patients with stenosis were managed successfully with either expectant management if asymptomatic or balloon dilatation. None have required surgical intervention for stenosis. Recurrence was more likely in those with ≥75% defect (OR 7.9, 95% CI 3.8-16.4, p<0.001) but was acceptable at an incidence of 15% at SC1

Conclusion

This is the largest western series to date reporting on outcomes of ER of LNPCP leaving near or fully circumferential defects, and the first to include significant numbers of ER performed by ESD. ER of LNPCP resulting in defects ≥75% of the luminal circumference is challenging but highly effective in avoiding surgery when performed in an expert centre. En bloc resection can be achieved in most cases, reducing recurrence. The only independent predictor of stenosis is ≥90% circumference defect but some patients improve with expectant management therefore pre-emptive intervention may not be warranted.

Introduction

The safety and efficacy of endoscopic resection (ER) as a first-line treatment for large non-pedunculated colorectal polyps (LNPCP) is well established. Significant complications are infrequent. Risk factors for immediate perforation have been described although it is frequently amenable to endoscopic closure and rarely results in significant deleterious sequelae. Delayed perforation or sepsis is a rare complication with potentially devastating consequences. Emergent surgery may be necessary with its attendant major complications. Risk factors for delayed perforation/sepsis after ER of LNPCP are not well established.

Aims & Methods

We analysed a large series of advanced ER of LNPCP to identify factors associated with delayed perforation.
Consecutive patients with LNPCP treated by ER at a tertiary referral centre between September 2011 and September 2023 were included. Lesions were resected using EMR, piecemeal EMR and ESD. Univariate analysis was used to examine clinicopathological factors potentially associated with delayed perforation.

Results

1883 ER for LNPCP were performed over the study period with a mean size of 46.4mm +/- 26.4mm (median 40mm, range 20-210mm). ESD was used for 890. Delayed perforation or sepsis occurred in 24 (1.3%) patients of which 8 (0.4%) required emergent surgery and 1 required CT guided drainage of an abscess. 3 patients developed sepsis after ESD of rectal lesions without evidence of perforation on CT with luminal contrast and required intensive care unit (ICU) management. There was no mortality. In patients who suffered a delayed perforation or sepsis, 54% of LNPCP were in the right colon, 8% in the left colon and 38% in the rectum. Delayed perforation/sepsis was significantly associated with extreme size (OR 3.02, 95% CI 1.35-6.76, p<0.001); difficult access (OR 2.63, 95% CI 1.16-5.97, p=0.02); ESD (OR 4.08, 95% CI 1.52-10.98, p=0.003) and long procedure time over 240 min (OR 5.20, 95% CI 2.02-13.36, p<0.001) but not location, F2 fibrosis, or age.

Conclusion

Delayed perforation/sepsis after ER of LNPCP is rare, and a requirement for surgery extremely rare, but varies according to lesion and procedure characteristics. The consequences can be severe and include surgery, guided drainage or ICU admission. Delayed perforation/sepsis can affect patients with LNPCP in any location including the rectum. Given the risk factors identified in this study, these data could help realistically counsel patients with LNPCP of extreme size, difficult access and requiring complex prolonged procedures regarding the risks of delayed perforation and sepsis.

Introduction

Large non-pedunculated colorectal polyps (LNPCP's) are significant precursors of colorectal cancer, commonly treated by piecemeal endoscopic mucosal resection (pEMR). When histology reveals a focus of adenocarcinoma invading the sub-mucosa, the resection is often considered non curative and salvage surgery is performed. Recent evidence shows that the risk of residual cancer in the bowel wall or lymph nodes of the surgical specimen are uncommon and that in some cases surgery may be avoided even when a focus of invasive cancer is found in the EMR specimen.

Aims & Methods

This is a retrospective observational study aimed to evaluate the prevalence, characteristics, and outcomes of covert carcinoma in LNPCPs resected by EMR in a two academic centers.
Data was collected from a prospective cohort of LNPCPs at two academic centers in Israel. We collected patient, lesion and procedural characteristics prospectively.
EMR was performed by 3 experienced endoscopist with dedicated EMR training. Piecemeal EMR was performed after endoscopic assessment for polyps that were deemed not to have sub-mucosal invasion according to accepted classification systems. Lesions suspected to containing invasive cancer were referred to surgery without endoscopic resection were excluded from the analysis.
For lesions containing invasive carcinoma, data was collected on Kikuchi classification, presence of tumor budding and lymphovascular invasion, grade of tumor differentiation, and status of the deep margin. Surgical data included presence of residual carcinoma in the surgical specimen, grading, margin involvement, and lymph node involvement.

Results

We analyzed 546 LNPCPs referred for resection at the two centers. We identified 27 cases of covert carcinoma, accounting for 4.9% of all LNPCPS resected by piecemeal EMR. A multidisciplinary tumor board deemed resection curative in 7 cases, and 18 patients were refferd for surgical resection. One patient refused surgery, and One patient was lost to follow up.
None of the surgical specimens had lymph node metastasis. In one case residual carcinoma was found in the bowel wall of the surgical specimen. No recurrence of colorectal cancer, presence of metastasis on imaging or elevation of CEA was documented for any of the patients during follow up of two year.

Conclusion

In conclusion, our study shows that covert carcinoma is relatively rare when EMR is performed in a high-volume center by expert endoscopists. In addition, the data suggests that even in these cases, the risk of metastatic disease is probably very low.

References

1. Kudo S ei, Lambert R, Allen JI, Fujii H, Fujii T, Kashida H, et al. Nonpolypoid neoplastic lesions of the colorectal mucosa. Gastrointest Endosc. 2008 Oct;68(4 SUPPL.).
2. Ferlitsch M, Moss A, Hassan C, Bhandari P, Dumonceau JM, Paspatis G, et al. Colorectal polypectomy and endoscopic mucosal resection (EMR): European Society of Gastrointestinal Endoscopy (ESGE) Clinical Guideline. Vol. 49, Endoscopy. Georg Thieme Verlag; 2017. p. 270–97.
3. Lopez A, Bouvier AM, Jooste V, Cottet V, Romain G, Faivre J, et al. Outcomes following polypectomy for malignant colorectal polyps are similar to those following surgery in the general population. Gut. 2019 Jan 1;68(1):111–7.
4. Choi YS, Kim WS, Hwang SW, Park SH, Yang DH, Ye BD, et al. Clinical outcomes of submucosal colorectal cancer diagnosed after endoscopic resection: A focus on the need for surgery. Intest Res. 2020 Jan 1;18(1):96–106.
5. Gibson DJ, Sidhu M, Zanati S, Tate DJ, Mangira D, Moss A, et al. Oncological outcomes after piecemeal endoscopic mucosal resection of large non-pedunculated colorectal polyps with covert submucosal invasive cancer. Gut. 2022 Mar 7;

Disclosure

Conflicts of interest: None. Financial disclosures: None.