Background and Aim:
Piecemeal endoscopic mucosal resection (EMR) is established as the preferred method for treatment of large (≥20mm) non-pedunculated colorectal polyps (LNPCP); however, it may not provide cure if low risk cancer (superficial submucosal invasion [SM1], low tumour grade and no lymphovascular invasion [LVI] or high grade tumour budding [TB]) is present. Consequently, endoscopic submucosal dissection (ESD) for en bloc excision has been advocated by some as the preferred treatment strategy for all LNPCP. In the right colon, the rates of cancer are lower, and surgery is associated with less morbidity compared with the left colon and rectum. The effect of a universal ESD strategy on oncological outcomes for the treatment of LNPCP in the right colon is unknown and we sought to evaluate this in a large Western population

Methods:
We retrospectively analysed prospectively collected data from major tertiary centres where patients were referred for resection of LNPCPs between April 2008 and June 2023 (182 months). The right colon was defined as proximal to the splenic flexure. In patients where more than one cancer was detected in the right colon, the more advanced cancer was included. In patients who were found to have cancer after EMR and those referred directly to surgery based on endoscopic suspicion of, or biopsy-proven cancer, the histopathology report of the EMR and/or surgical specimen was obtained and SM depth, LVI, TB and tumour grade were recorded. The primary outcome was to determine the proportion of right colon LNPCPs with low risk cancer which could have been cured if resected by ESD

Results:
4350 sporadic LNPCP were referred for ER at 7 sites; 2842 were located in the right colon (median size 30mm [IQR 15])(Figure 1). 67 (2.4%) were referred directly to surgery and cancer was proven in 58 (86.7%). 39 (1.4%) LNPCP were referred to surgery when EMR was impeded by technical factors. 2736 (96.2%) of LNPCP underwent EMR (median size 30mm [IQR 15], n=904/2736 [33%] ≥40mm) of which 69 (2.5%) were cancers. The overall prevalence of cancer in the right colon was 4.5% (127/2842). Detailed histopathological analysis was possible in 107 cancers (65 after EMR, 42 direct to surgery, Table 1). 24/2822 (0.85%) sporadic R colon LNPCP were low risk cancers which would have been cured by ESD. Of all cancers potentially cured by ESD, 5/24 (20.8%) were cured by en bloc EMR.

Conclusions:
The proportion of right colon LNPCP referred for EMR containing low risk early cancer amenable to cure by ESD was extremely low at <1%, in a high volume, multicentre Western cohort. More than 95% of right colon LNPCP (median size 30mm) did not contain cancer and 99.3% were cured by EMR. Given its significant resource demands and higher risk of adverse events, a universal ESD strategy cannot be justified for the management of right colon LNPCP.

Figure 1: STROBE Diagram

<b>Table 1</b>: Presence of histopathologic parameters from EMR and/or surgical specimen predicting high risk for lymph node or distant tumour spread amongst right colon LNPCP referred for ER<br /> <br /> *Missing data; n=5

Table 1: Presence of histopathologic parameters from EMR and/or surgical specimen predicting high risk for lymph node or distant tumour spread amongst right colon LNPCP referred for ER

*Missing data; n=5

Simon Zanati

Appearances