Endoscopic Submucosal Dissection of Colorectal Lesions

Abstract

Diagnosis of superficial colorectal neoplasms [1] increases as a result of the effort and adoption of screening and surveillance programs. Most of superficial neoplasms of the gastrointestinal tract with limited risk of lymph node (LN) metastasis are locally excised by endoscopic resection (ER), with 5-year survival rates similar to those of surgery, lower mortality and morbidity, and greater cost-effective ratio [2–4]. Multicenter studies show that mortality of surgery is 1% and 2% for low anterior resection and right hemicolectomy, respectively [5, 6], and these data are confirmed by a National Surgical Quality Improvement Program [4]. However, major postoperative adverse events rate occur in 15% of the cases with 8% readmissions, 4% redo surgeries, and 2% colostomies or ileostomies. Conversely, ER of large neoplasms has a much lower (0.1%) mortality rate both in a review of 6440 patients [7] and a prospective study of 1050 advanced lesions. Moreover, ER has a very low morbidity rate [8], and age is not an independent risk fact [9, 10], although severe comorbidities should be cautiously considered [11]. ER and surgery in the treatment algorithm of colorectal neoplasms are complementary according to the risk of LN metastasis, difficulty gradients of resection, and quality of life.

Appendix

Patients with cecal neoplasms involving the appendix have been considered candidates for laparoscopic surgery. Actually, neoplasms at the appendix should be classified as: Type 0, in proximity to but not in contact with the appendiceal orifice; Type 1, at the border but not entering into orifice; Type 2, entering into the orifice with visible margins on inspection of the appendiceal lumen; Type 3, entering the orifice deeply with no visible margins [133]. Type 0 neoplasms should be considered as cecal neoplasms. After case reports, Jacob et al. [133] retrospectively reported their ESD experience on a large cohort of patients (Table 6) with an en bloc resection rate of 83% for Type 2 lesions. Another retrospective center study found similar ESD outcomes [134].

Table 6 ESD involving the orifice and eventually the lumen of the appendix

Difficulties for ER at the appendix are: highly prevalent SM fibrosis; the very narrow appendiceal lumen that may preclude the visualization of neoplasm margins (a prerequisite to perform a complete resection); difficulty in maintaining an adequate luminal distension; potential development of a delayed appendicitis. Type 3 lesions should not undergo ER unless the patient had a previous appendectomy or the endoscopist believes there is a reasonable chance that a complete resection can be carried out. Perforations are more frequent in these cases leading to surgery in almost a quarter of cases [136]. The technical strategy has not been standardized, and different methods have been proposed: a tapered transparent distal attachment can be useful to partially intubate the appendix [134, 135, 137], traction methods can be useful [136, 138]; water immersion may be more effective in maintaining the distension of the appendiceal lumen [137, 139].

Ano-Rectal Junction

The perineal rectum is characterized by: a narrow lumen at the anal sphincter that makes visualization and operability difficult; abundant blood vessels and hemorrhoids; squamous epithelium rich in sensory nerves. Holt et al. [140] showed that EMR at the ano-rectal junction (ARJ) is almost always piecemeal with a high 22% recurrence rate. The disadvantages of piecemeal EMR at the ARJ are even more relevant that in other locations: incomplete resections and scarring recurrences becomes even more difficult to be visualized and retreated; inaccurate T1 cancer microstaging leads to the unnecessary demolitive surgery. Due to these limitations, tumors at the ARJ have been mainly resected by transanal surgical approach with suboptimal outcomes: high recurrence rates ranging from ranging from 23% to 31% [141], and complications, such as temporary ileostomy in some cases [142]. Recent data indicate that ESD at the ARJ is successful (Table 7), although operation times are longer [143,144,145], and curative resection rates are lower due to burning artifacts at the anal side of the specimen since mucosal incision is often conducted below the dentate line.

Table 7 ESD at the ano-rectal junction

Technical points for ESD at the ARJ are: a tapered transparent hood may be more effective than straight ones to maintain a good visual field into the anal canal; injection of 1% lidocaine (100 mg/10 mL) on the anal side reduces pain while dissecting at the ARJ; SM dissection should be conducted just above the muscularis propria to shut off vessels that penetrate vertically and supply hemorrhoids. Specific adverse events of ER at the ARJ are a postoperative anal pain in 16–18% of patients that can be managed by oral non-steroidal anti-inflammatory drugs, and a proctostenosis if resection is circumferential [150]. The risk of delayed bleeding of ESD at the ARJ is similar to that of rectal ESD, and not increased by hemorrhoids .

Nonlifting Lesions

Neoplasms may have a positive non-lifting sign due to: (1) carcinomatous SM invasion; (2) scars of previous biopsies and/or resection attempts; and (3) specific morphologies. Biopsies should never be performed when ER is possible. Difficult neoplasms should be referred to high-volume referral centers: the first attempt at ER should be associated with the highest rate of success; an attempt to see if ER might be possible should be avoided. If SM invasive neoplasms with a positive nonlifting sign should not undergo ER, nonlifting intramucosal neoplasms are a specific ESD indication: EMR has a high failure rate in these cases since the snare slips over. However, severe and/or diffuse SM fibrosis is the most important prognostic of difficulty for ESD [69, 151], and its presence need to be assessed preoperatively: residual neoplasms have visible scars; naïve non-granular and bulky morphologies have a high prevalence of SM fibrosis. Matsumoto et al. [152] classified SM fibrosis in: F0, no fibrosis, blue transparent layer; F1, mild fibrosis, white web-like structures in the blue submucosal layers; F2, severe fibrosis, white muscular structure without blue transparent layer. F0–1 and F2 neoplasms showed significantly different en bloc resection rates (84% vs. 33%) and perforation rates (17% vs. 11%). Subsequent studies confirmed the negative impact of severe fibrosis on ESD outcomes: longer operating time, lower en bloc resection and complete resection rates, and higher discontinuance and perforation rates (Table 8).

Table 8 ESD of residual scarred neoplasms

A unique technique is required for ESD of nonlifting neoplasms, the main issue being the facilitation of the SM access. The PCM was found to be advantageous for ESD of F2 neoplasms with higher en bloc and complete resection rates, and lower operating time [156]. Mucosal incision should be performed at a larger distance so that dissection can be initiated into nonfibrotic SM tissue in order to create the lesion flap and a facilitated SM access. The colonic localization represents an addition difficulty and increases the risk for perforation. Previous histological results are determinant before planning ER of residual or local recurrent lesions.

Inflammatory Bowel Disease

Patients with inflammatory bowel disease (IBD) have an increased risk of colorectal dysplasia and cancer. Endoscopic resection is appropriate when complete, synchronous invisible dysplasia is absent in the adjacent mucosa and in the whole colitic mucosa. Although this assumption is evidence-based for patients with small (<10 mm) polypoid neoplasms, it remains to be established if large (>20 mm), nonpolypoid IBD neoplasms can undergo ER without a high risk of metachronous cancer. Piecemeal EMR is associated with high incomplete resection rates and the need of salvage surgery [158, 159], which often means total colectomy due to the chronic inflammation cancer field effect. ESD is the only possibility, as indicated in all guidelines, but it has been evaluated only recently in few small experiences. The first study by Iacopini et al. [160] comprised 10 patients underwent ESD in Italy and Japan. ESD resulted highly successful, and no metachronous cancer and/or invisible dysplasia developed during the short follow-up (median 24 months). Similar results have been reported by other studies (Table 9).

Table 9 ESD for superficial neoplasms within colitic mucosa of patients with inflammatory bowel diseases

ESD in IBD is difficult. SM fibrosis is highly prevalent due to long-standing chronic inflammation, and also because most neoplasms are submitted to multiple biopsies and/or attempts of resection before being referred for ESD in expert centers. Delineation of neoplasm margins may be extremely difficult and requires prolonged times, the best high-definition scope, chromoendoscopy by narrow band imaging and dye application with acetic acid and indigo carmine. Magnification may be helpful. IBD neoplasms may show the coexistence of adenomatous and nonadenomatous patterns (hyperplastic and inflammatory), so that there could be overlap features of lesion margins and the background mucosa. The ESD strategy should be modified to include peripheral dot markings 5 mm beyond the lesion margins as for gastric ESD; mucosal incision be performed at a wider distance from neoplasm edges to guarantee a wide rim of normal mucosa and exclude adjacent invisible dysplasia that is pivotal to confirm a curative resection. Finally, endoscopists and patients should strictly adhere to strict long-term follow-ups .