Submucosal Dissection for T1 Colorectal Cancer

Total of colorectal ESD

(n = 1902)

Colorectal ESD for the T1b lesion

(n = 169)

R0 resection rate

1833 (96.4%)

125 (74.0%)

Curative resection rate

1665 (87.5%)

–

Regarding tumor morphology, laterally spreading tumors of non-granular type (LST-NG) >2 or 3 cm in diameter are considered lesions with formal indication of ESD due to a higher SM cancer risk and the difficulty to predict SM superficial invasion. In addition, when assessing the risk of SM invasion, a multifocal behavior has been previously reported [14, 15]. Furthermore, conventional EMR is often difficult for LST-NG >2 or 3 cm in diameter because of frequent non-lifting sign despite of intramucosal neoplasia.

Large LST granular, nodular mixed type (LST-G Mix) >3 or 4 cm in diameter is considered an adequate (suitable) lesion for en bloc ESD because of higher risk of SM invasion and lower sensitivity of pit pattern for SM invasion [15]. LST-G-type lesions harbor a polypoid growth (PG) cancer; therefore, pit pattern diagnosis for SM invasion is sometimes difficult from the surface pattern when no SM cancer is exposed to the tumor surface [14, 15].

Large sessile-type villous polyps are another indication for ESD; however, such lesions are technically challenging because of the higher incidence of SM fibrosis due to tumor weight and intestinal peristalsis [16].

10.3 The Possibility of Expansion for ESD Indication

Among unfavorable histological factors for lymph node metastasis, it has been reported that the invasion depth might be only a negligible risk for LNM compared to other risk factors [9, 11]. We should note, however, that this data corresponds to the risk for LNM in nonconsecutive surgically resected cases. There is no firm evidence of long-term outcome for T1b cancer (SM invasion deeper than 1000 μm from the muscularis mucosae) treated only by endoscopic resection and with no other risk factors. According to unpublished data from The Japanese Society for Cancer of the Colon and Rectum Project, the risk for LNM was 1.4% (12/881, 95%CI; 0–2.4%) for T1b cancers with no other risk factors but micrometastasis. Therefore, knowing the fact that those lesions correspond to surgically resected cases even when LNM was negative, we cannot consider such as over surgery cases.

10.4 Technical Feasibility of ESD for T1b Cancers

Until now, clinical T1b (cT1b) cancers had been surgically treated, and only pathological T1b (pT1b), those which were diagnosed as cTis (intramucosal) or T1a (SM superficial), had been treated with endoscopic resection. Occasionally, EMR or ESD was attempted as a diagnostic EMR/ESD when endoscopic diagnosis was difficult for T1a or T1b. Therefore it is still uncertain whether endoscopic R0 resection through ESD can be safely and certainly performed for pT1b cancers.

Unpublished data (from NCCH database) showed that R0 resection rate and curative resection rate in colorectal ESDs were 96.4% and 87.5%, respectively. R0 resection rate for pT1b was, however, lower (74.0%) and unsatisfactory (Table 10.1).

Considering this, ESD for cT1b should not be performed as an established procedure in a clinical setting but only in clinical trials with prior IRB approval. Thus, no muscle invasion and no LNM should be carefully confirmed with EUS and/or CT scan prior to ESD, which should be cautiously conducted by an ESD expert.

10.5 Clinicopathological Feasibility

When cancer invades the SM deep layer, the risk of lymphovascular invasion naturally increases. Diagnostic performance for lymphovascular invasion is, however, low and subjective even after immunochemical staining [17]. In such situation, endoscopic surveillance of T1b cancers with no additional surgery should be carefully considered even when there are no other unfavorable risk factors. Our multicenter case series [18] revealed that surgery would decrease the 5-year recurrence rate from 6.6% to 3.6% in patients with high-risk SM-CRC. In addition, 10 cases out of 15 recurrent patients showed no lymphovascular invasion.

Long-term outcomes of colorectal ESD also revealed a higher recurrence rate of 13% at 3 years in non-curative resection cases. The recurrence rate was better when patients underwent additional surgery [19].

Summing up, surgery with LN dissection would be the most appropriate treatment modality to date for pT1b cancers even without any other risk factor.

10.6 New Therapeutic Strategy for Rectal T1 Cancer

Our multicenter retrospective cohort study revealed that the risk for local recurrence was significantly higher in high-risk patients with submucosal rectal cancer compared to that of patients with submucosal colon cancer when treated only with endoscopic resection [20]. In addition, rectal surgery often results in lower QOL for the patient even if permanent stoma is not required [21, 22].

In this situation, a new treatment modality is necessary for rectal T1 cancer. Recently, a new treatment concept has been introduced for such cancers, which consists of adjuvant chemoradiotherapy similar to the approach for esophageal squamous cell cancer treatment. A meta-analysis [23] reported that the rate of local recurrence was 5% in patients who received radiotherapy or CRT after local excision, which was similar to that (4%) in those who underwent total mesorectal excision. A recent clinical trial from Japan also revealed an excellent result in which CRT for en bloc R0 resection for high-risk T1 rectal cancers showed similar recurrence control compared to the traditional surgery [24]. Concerning adverse effects of the radiation, such could be reduced to 45 gray for control of LNM recurrence after en bloc R0 resection for T1b cancer.

10.7 Technical Warnings Regarding ESD for T1 Cancer

The basic steps in ESD consist of injecting fluid into the submucosa to elevate the tumor, making an incision in the surrounding mucosa to gain access to the submucosa and dissecting the submucosa beneath the tumor to achieve tumor resection. Conventional ESD method comprises the following steps [25].