Selection bias complicates comparisons of endoscopic submucosal dissection vs. endoscopic mucosal resection in the treatment of colorectal neoplasms

We have read with great interest the paper by Rolny entitled “The need for surgery after endoscopic treatment of colorectal neoplasms is the most important outcome criterion” [1]. We agree with the author that endoscopic submucosal dissection (ESD) achieves higher rates of en bloc and R0 resections as well as a lower incidence of recurrence when removing large colorectal neoplasms. Nonetheless, the author argues that there is no clear advantage of colorectal ESD compared with endoscopic mucosal resection (EMR) in terms of reducing the need for additional surgery after endoscopic treatment. Although this opinion is interesting and challenging, we would like to point out serious concerns about that conclusion.

Rolny cites a prospective multicenter study reporting that additional surgery was required in 7.5 % of patients after ESD and in 4.0 % after EMR [2]. However, in that study, the proportion of submucosal cancers in the ESD group was more than double that in the EMR group (150 /816 [18 %] vs. 70 /1029 [7 %]) [2]. Moreover, the size of the lesions was significantly larger in the ESD group (42 % > 40 mm) compared with the EMR group (10 % > 40 mm) [2]. Similar discrepancies were also reported in the meta-analyses cited by Rolny, showing that the lesions were significantly larger in patients undergoing ESD compared with those undergoing EMR [3] [4]. These differences demonstrate that ESD and EMR were used for different types of target lesions, and such comparisons between ESD and EMR are not meaningful because of selection bias. In addition, Rolny states that half of the patients with superficial submucosal invasive carcinoma required surgery after ESD based on a large multicenter study from Japan [5]. This is not surprising considering that half of the submucosal cancers in that study were histologically diagnosed as SM2 or deeper cancers and therefore underwent additional surgery after ESD in order to reduce the risk of lymph node metastasis [5].

We conclude that as long as there are no randomized trials, comparisons between ESD and EMR for removal of colorectal lesions are complicated. Such comparisons might not be meaningful when the target lesions and indications differ significantly between ESD and EMR, and are prone to end up being affected by selection bias.

• References

• 1 Rolny P. The need for surgery after endoscopic treatment of colorectal neoplasms is the most important outcome criterion. Endoscopy 2017; 49: 80-82
• 2 Nakajima T, Saito Y, Tanaka S. et al. Current status of endoscopic resection strategy for large, early colorectal neoplasia in Japan. Surg Endosc 2013; 27: 3262-3270
• 3 Fujiya M, Tanaka K, Dokoshi T. et al. Efficacy and adverse events of EMR and endoscopic submucosal dissection for the treatment of colon neoplasms: a meta-analysis of studies comparing EMR and endoscopic submucosal dissection. Gastrointest Endosc 2015; 81: 583-595
• 4 Arezzo A, Passera R, Marchese N. et al. Systematic review and meta-analysis of endoscopic submucosal dissection vs endoscopic mucosal resection for colorectal lesions. United European Gastroenterol J 2016; 4: 18-29
• 5 Saito Y, Uraoka T, Yamaguchi Y. et al. A prospective, multicenter study of 1111 colorectal endoscopic submucosal dissections (with video). Gastrointest Endosc 2010; 72: 1217-1225