Endoscopic resection is now considered a curative procedure for early gastric cancer. In Japan, it has increasingly replaced surgical resection for this indication, although in the West it has not been universally accepted as a first-line treatment. Recently, endoscopic submucosal dissection has been increasingly applied to colorectal disease, although it has not become a standard therapeutic procedure for early colorectal carcinoma because of its technical difficulty, the relatively long procedure time required, and the risk of complications, such as perforation and bleeding.
Key points
- •
Experience with endoscopic resection of early gastric cancer has been recorded in various studies.
- •
Expansion of the criteria for endoscopic mucosal resection will reduce the need for gastrectomy in early gastric cancer.
- •
Although endoscopic submucosal dissection has been increasingly applied to colorectal disease, it has not become a standard therapeutic procedure for early colorectal carcinoma because of its technical difficulty; the relatively long procedure time required; and the risk of complications, such as perforation and bleeding.
- •
To help overcome the learning curve of colorectal endoscopic submucosal dissection, Western and Eastern endoscopic societies should communicate with each other.
Upper gastrointestinal tract
Endoscopic resection is now considered a curative procedure for early gastric cancer. In Japan, it has increasingly replaced surgical resection for this indication, although in the West it has not been universally accepted as a first-line treatment. Endoscopic resection also has been used as a histologic staging technique for early gastric cancer by assessing the depth of tumor penetration, which is important in determining the best treatment.
Limited invasion of the tumor and absence of lymph node metastases are crucial for achieving a cure of gastric cancer with endoscopic therapy. The important endoscopic points to consider are as follows:
- 1.
The extent of the surface involvement and the morphology. Endoscopic mucosal resection (EMR) is not recommended for lesions greater than 2 cm in size. Endoscopic submucosal dissection (ESD) is recommend for lesions up to 3 cm if they are not ulcerated or scarred, or are less than 2 cm if they have ulcers or scars.
- 2.
The depth of invasion should not be deeper than the mucosa or the most superficial submucosal layer, the upper third of the submucosa (sm1). The depth may be determined by endoscopic ultrasound examination before endoscopic resection, and corroborated with histopathologic examination of the resected specimen.
- 3.
High-grade dysplasia is the earliest stage of malignancy. A high degree of cellular differentiation favors endoscopic treatment. Even though poorly differentiated lesions have a higher risk of distant spread, those less than 5 mm can be treated endoscopically.
- 4.
Multifocal early gastric cancer can be treated endoscopically provided the entire lesion is histologically consistent with intramucosal cancer. The degree of difficulty in performing endoscopic resection depends in part on the location of the tumor in the stomach; lesions in the posterior wall and lesser curvature are technically more difficult to remove.
- 5.
Successful endoscopic resection depends on having histologically clear lateral and vertical margins of the tumor at the time of resection, to be assessed by endoscopic biopsies in follow-up examinations.
- 6.
Assessment of the results of endoscopic resection is based on the rates of complete removal or destruction, en bloc resection, recurrence, and patient survival.
The Japanese Research Society for Gastric Cancer has established the following criteria for lesions suitable for endoscopic resection. Type I or IIa: well differentiated, less than 2 cm, limited to the mucosa, and without histologic ulceration. Type IIc: well differentiated, less than 1 cm, limited to the mucosa, and without ulceration. When these criteria are met the risk of lymph node involvement is only 1.7%.
Experience with endoscopic resection of early gastric cancer has been recorded in various studies. In a retrospective evaluation of 210 patients with early gastric cancer treated with EMR and followed for 15 years, the 5-year survival rate was 86% and the 10-year rate 56%; however, there were no cancer-related deaths. In another study, 106 patients with early gastric cancer up to 2 cm in diameter were treated with complete resection of the lesion in a single procedure, either by en bloc resection for lesions less than 10 mm (64%) or by piecemeal resection for larger lesions (36%); no recurrence after either technique was found in patients with tumor-negative margins, and the overall recurrence rate of cancer was 2.8%. All tumors that recurred were greater than 15 mm initially, and all were treated with piecemeal resection. Histologic reconstruction to confirm complete resection by piecemeal removal often is difficult, therefore patients treated with this method should be followed closely.
Amano and colleagues have retrospectively evaluated endoscopic therapy in patients with early gastric cancer that does not meet the Japanese Research Society morphologic criteria for lesions suitable for EMR. Endoscopic therapy consisted of EMR, thermal therapy, or both. Poorly differentiated and well-differentiated tumors from 1 to 3 cm were included. Some patients with submucosal invasion limited to sm1 also were included. Curative resection is defined as where the lateral and vertical margins of the specimens were free of cancer and there was no submucosal invasion deeper than 500 μm from the muscularis mucosae, lymphatic invasion, or vascular involvement, and it was achieved in 95%. The rate of cure in this group was statistically similar to that of cancers that fulfilled the standard morphologic criteria for EMR resection (98%).
Adequacy of EMR should be assessed by measuring the distance from the margin of the cancer to the edge of the resected specimen. In one study, no cancer recurred when this distance was more than 2 mm, whereas 16% recurred when the distance was less than 2 mm ; presence of cancer at the edge of the specimen was associated with a 45.8% recurrence rate. In another study, no recurrence was observed if the distance from the margin of the cancer to the edge of the resected specimen was more than 7 mm. Thus, it seems that adequate distance (preferably at least 2 mm) between the cancer and the edge of the specimen needs to be achieved to ensure a complete resection. Margin-negative resections were more likely (81.2%) in cancers that are less than 1 cm in diameter than in cancers that are more than 2 cm in diameter.
A prospective analysis of 479 early gastric cancers in 405 patients treated with EMR over an 11-year period at Tokyo National Cancer Center has been reported. The selection criteria were as follows: well- or moderately well-differentiated gastric cancer; morphologic type I, IIa, or IIc; no histologic evidence of ulceration; diameter less than 3 cm; histologic confirmation of intramucosal carcinoma; no lymphovascular invasion; and clean margins. Complete resection was achieved in 69% (278 of 405). The recurrence rate was only 2% after complete resection, and all recurrences were treated successfully with a modified combination therapy of EMR and laser. Two hundred seventy-eight lesions were followed after endoscopic treatment for a median period of 38 months (range, 3–120 months); no cancer recurred and no cancer-related deaths were reported.
Recently, EMR and ESD have become established alternatives to surgical therapy for early gastric cancer in Korea and Japan. The traditional indication criteria for endoscopic resection are elevated-type intramucosal cancer (0–IIa) less than 20 mm, depressed-type mucosal cancer without ulceration (0–IIb, 0–IIc) less than 10 mm, and well-differentiated or moderately differentiated intestinal-type adenocarcinoma.
In a report by Abe and colleagues, depressed cancers were associated with lymph node metastases (86%) when there was submucosal infiltration, a size of 20 mm or more, or lymphatic vessel involvement.
Some authors have proposed extended indications for EMR in early gastric cancer: well-differentiated lesions up to 30 mm, without an ulcer or ulcer scar; mucosal cancers less than 20 mm, with an ulcer or ulcer scar; sm1 lesions less than 20 mm, without an ulcer or ulcer scar; and poorly differentiated lesions less than 10 mm.
The risk for nodal metastasis for differentiated early or mucosal cancers is approximately 0.4%. Undifferentiated mucosal cancers are not recommended to be treated by EMR, because the higher risk for nodal metastasis is approximately 4%. However, according to one study, poorly differentiated and signet-ring cell carcinomas less than 5 mm in size can be treated with EMR. The proposed extended criteria for endoscopic resection in the ESD era are summarized in Fig. 1 .
Expansion of the criteria for EMR will reduce the need for gastrectomy in early gastric cancer. However, because resection of large or ulcerated lesions by conventional EMR is difficult, ESD has been developed. Over the years, substantial experience in the use of this technique has been gained. In Japan, ESD performed with the insulation-tip knife or others has become standard treatment of early gastric cancer with estimated minimum metastatic risk.
In a report from Europe, ESD was performed with a new double-channel endoscope in 10 patients, nine early gastric cancers and one adenoma with a median diameter of 22 mm (R-scope; Olympus, Tokyo, Japan). ESD was successful in six patients. Perforation occurred in two, who were then treated with surgery. In another study, ESD was performed in 19 patients with superficial gastric lesions (15–30 mm) that had high-grade (N = 15) or low-grade (N = 4) noninvasive epithelial neoplasia. R0 resection was performed in 89% and en bloc resection in 79%. Major bleeding occurred in one patient (5%); there were no perforations. In a median follow-up of 10 months, one cancer (5%) recurred.
Endoscopic resection has been performed also for undifferentiated intramucosal cancer. In 38 such patients with 42 undifferentiated intramucosal cancers, who had declined surgical therapy, ESD was performed with dedicated devices by experienced expert endoscopists. The en bloc resection rate was 83.3% and complete resection rate 80.9%. Clinical remission was achieved in 92.8%, with recurrence in only 7.14%, during follow-up of 15 months. In undifferentiated gastric cancer, grossly normal gastric mucosa surrounding the resected lesion can contain cancer cells beneath the epithelium ; this characteristic may explain why the complete resection rate of ESD for undifferentiated cancer is lower than that reported for well-differentiated cancer.
The en bloc resection rate is better with ESD than with conventional EMR. However, the procedure time is longer for ESD, a disadvantage that might be improved with experience.
Upper gastrointestinal tract
Endoscopic resection is now considered a curative procedure for early gastric cancer. In Japan, it has increasingly replaced surgical resection for this indication, although in the West it has not been universally accepted as a first-line treatment. Endoscopic resection also has been used as a histologic staging technique for early gastric cancer by assessing the depth of tumor penetration, which is important in determining the best treatment.
Limited invasion of the tumor and absence of lymph node metastases are crucial for achieving a cure of gastric cancer with endoscopic therapy. The important endoscopic points to consider are as follows:
- 1.
The extent of the surface involvement and the morphology. Endoscopic mucosal resection (EMR) is not recommended for lesions greater than 2 cm in size. Endoscopic submucosal dissection (ESD) is recommend for lesions up to 3 cm if they are not ulcerated or scarred, or are less than 2 cm if they have ulcers or scars.
- 2.
The depth of invasion should not be deeper than the mucosa or the most superficial submucosal layer, the upper third of the submucosa (sm1). The depth may be determined by endoscopic ultrasound examination before endoscopic resection, and corroborated with histopathologic examination of the resected specimen.
- 3.
High-grade dysplasia is the earliest stage of malignancy. A high degree of cellular differentiation favors endoscopic treatment. Even though poorly differentiated lesions have a higher risk of distant spread, those less than 5 mm can be treated endoscopically.
- 4.
Multifocal early gastric cancer can be treated endoscopically provided the entire lesion is histologically consistent with intramucosal cancer. The degree of difficulty in performing endoscopic resection depends in part on the location of the tumor in the stomach; lesions in the posterior wall and lesser curvature are technically more difficult to remove.
- 5.
Successful endoscopic resection depends on having histologically clear lateral and vertical margins of the tumor at the time of resection, to be assessed by endoscopic biopsies in follow-up examinations.
- 6.
Assessment of the results of endoscopic resection is based on the rates of complete removal or destruction, en bloc resection, recurrence, and patient survival.
The Japanese Research Society for Gastric Cancer has established the following criteria for lesions suitable for endoscopic resection. Type I or IIa: well differentiated, less than 2 cm, limited to the mucosa, and without histologic ulceration. Type IIc: well differentiated, less than 1 cm, limited to the mucosa, and without ulceration. When these criteria are met the risk of lymph node involvement is only 1.7%.
Experience with endoscopic resection of early gastric cancer has been recorded in various studies. In a retrospective evaluation of 210 patients with early gastric cancer treated with EMR and followed for 15 years, the 5-year survival rate was 86% and the 10-year rate 56%; however, there were no cancer-related deaths. In another study, 106 patients with early gastric cancer up to 2 cm in diameter were treated with complete resection of the lesion in a single procedure, either by en bloc resection for lesions less than 10 mm (64%) or by piecemeal resection for larger lesions (36%); no recurrence after either technique was found in patients with tumor-negative margins, and the overall recurrence rate of cancer was 2.8%. All tumors that recurred were greater than 15 mm initially, and all were treated with piecemeal resection. Histologic reconstruction to confirm complete resection by piecemeal removal often is difficult, therefore patients treated with this method should be followed closely.
Amano and colleagues have retrospectively evaluated endoscopic therapy in patients with early gastric cancer that does not meet the Japanese Research Society morphologic criteria for lesions suitable for EMR. Endoscopic therapy consisted of EMR, thermal therapy, or both. Poorly differentiated and well-differentiated tumors from 1 to 3 cm were included. Some patients with submucosal invasion limited to sm1 also were included. Curative resection is defined as where the lateral and vertical margins of the specimens were free of cancer and there was no submucosal invasion deeper than 500 μm from the muscularis mucosae, lymphatic invasion, or vascular involvement, and it was achieved in 95%. The rate of cure in this group was statistically similar to that of cancers that fulfilled the standard morphologic criteria for EMR resection (98%).
Adequacy of EMR should be assessed by measuring the distance from the margin of the cancer to the edge of the resected specimen. In one study, no cancer recurred when this distance was more than 2 mm, whereas 16% recurred when the distance was less than 2 mm ; presence of cancer at the edge of the specimen was associated with a 45.8% recurrence rate. In another study, no recurrence was observed if the distance from the margin of the cancer to the edge of the resected specimen was more than 7 mm. Thus, it seems that adequate distance (preferably at least 2 mm) between the cancer and the edge of the specimen needs to be achieved to ensure a complete resection. Margin-negative resections were more likely (81.2%) in cancers that are less than 1 cm in diameter than in cancers that are more than 2 cm in diameter.
A prospective analysis of 479 early gastric cancers in 405 patients treated with EMR over an 11-year period at Tokyo National Cancer Center has been reported. The selection criteria were as follows: well- or moderately well-differentiated gastric cancer; morphologic type I, IIa, or IIc; no histologic evidence of ulceration; diameter less than 3 cm; histologic confirmation of intramucosal carcinoma; no lymphovascular invasion; and clean margins. Complete resection was achieved in 69% (278 of 405). The recurrence rate was only 2% after complete resection, and all recurrences were treated successfully with a modified combination therapy of EMR and laser. Two hundred seventy-eight lesions were followed after endoscopic treatment for a median period of 38 months (range, 3–120 months); no cancer recurred and no cancer-related deaths were reported.
Recently, EMR and ESD have become established alternatives to surgical therapy for early gastric cancer in Korea and Japan. The traditional indication criteria for endoscopic resection are elevated-type intramucosal cancer (0–IIa) less than 20 mm, depressed-type mucosal cancer without ulceration (0–IIb, 0–IIc) less than 10 mm, and well-differentiated or moderately differentiated intestinal-type adenocarcinoma.
In a report by Abe and colleagues, depressed cancers were associated with lymph node metastases (86%) when there was submucosal infiltration, a size of 20 mm or more, or lymphatic vessel involvement.
Some authors have proposed extended indications for EMR in early gastric cancer: well-differentiated lesions up to 30 mm, without an ulcer or ulcer scar; mucosal cancers less than 20 mm, with an ulcer or ulcer scar; sm1 lesions less than 20 mm, without an ulcer or ulcer scar; and poorly differentiated lesions less than 10 mm.
The risk for nodal metastasis for differentiated early or mucosal cancers is approximately 0.4%. Undifferentiated mucosal cancers are not recommended to be treated by EMR, because the higher risk for nodal metastasis is approximately 4%. However, according to one study, poorly differentiated and signet-ring cell carcinomas less than 5 mm in size can be treated with EMR. The proposed extended criteria for endoscopic resection in the ESD era are summarized in Fig. 1 .
