Clinical Outcomes of Gastrectomy after Incomplete EMR/ESD.

PURPOSE: Endoscopic resection is widely accepted as standard treatment for early gastric cancer (EGC) without lymph node metastasis. The procedure is minimally invasive, safe, and convenient. However, surgery is sometimes needed after endoscopic mucosal resection (EMR)/endoscopic submucosal dissection (ESD) due to perforation, bleeding, or incomplete resection [corrected]. We evaluated the role of surgery after incomplete resection.
MATERIALS AND METHODS: We retrospectively studied 29 patients with gastric cancer who underwent a gastrectomy after incomplete EMR/ESD from 2006 to 2010 at Korea University Hospital.
RESULTS: There were 13 incomplete resection cases, seven bleeding cases, three metachronous lesion cases, three recurrence cases, two perforation cases, and one lymphatic invasion case. Among the incomplete resection cases, a positive vertical margin was found in 10, a positive lateral margin in two, and a positive vertical and lateral margin in one case. Most cases (9/13) were diagnosed as mucosal tumors by endoscopic ultrasonography, but only three cases were confirmed as mucosal tumors on final pathology. The positive residual tumor rate was two of 13. The lymph node metastasis rate was three of 13. All lymph node metastasis cases were submucosal tumors with positive lymphatic invasion and no residual tumor in the gastrectomy specimen. No cases of recurrence were observed after curative resection.
CONCLUSIONS: A gastrectomy is required for patients with incomplete resection following EMR/ESD due to the risk of residual tumor and lymph node metastasis.

Introduction

Early gastric cancer (EGC) is defined as a neoplasm confined to the mucosa or submucosa, regardless of regional lymph node metastasis.(1) EGC without evidence of nodal metastasis is treated by endoscopic resection, either endoscopic mucosal resection (EMR) or endoscopic submucosal dissection (ESD). The endoscopic method is widely accepted because it is less invasive and less costly and requires a shorter hospital stay than surgical resection.(2-4)

Guidelines for EMR/ESD, established by the Japanese Gastric Cancer Association, are generally accepted and state that: 1) elevated EGCs less than 2 cm in diameter and 2) small (≤1 cm) depressed EGCs without ulceration are indications for EMR/ESD. At the same time, these lesions must be differentiated adenocarcinoma confined to the mucosa with no lymphatic or vascular involvement.(5)

The so-called extended criteria have been proposed to include: 1) mucosal cancers without ulceration regardless of lesion size, 2) ulcerated mucosal cancers with a size restriction of 30 mm, and 3) cancers with minute submucosal invasion (<500 µm) with a size restriction of 30 mm. Lesions with undifferentiated histology and lymphatic or vascular invasion are still excluded.(6,7)

Although endoscopic resection has the advantage of preserving most of the stomach and maintaining quality of life, some EMR/ESD patients have incomplete resections.(8) In addition, EMR/ESD is associated with a complication frequency, particularly perforation and post-operative bleeding (delayed bleeding).(9) The risk of synchronous and metachronous gastric cancer developing in the patient after endoscopic resection has also become a major problem. (10) Incomplete resection patients are treated empirically, by either gastrectomy, additional endoscopic treatment, or careful follow-up. (11)

We conducted this study to evaluate clinical outcomes and the role of surgery for patients with incomplete resection after EMR/ESD.

Materials and Methods

We retrospectively evaluated 510 gastric cancer patients who underwent EMR/ESD for gastric neoplasm at Korea University Hospital from 2006 to 2010. From these patients, 482 patients with diagnosed gastric cancer were identified. We studied 29 gastric cancer patients who underwent gastrectomy following EMR/ESD.

Incomplete resection was the most common cause of gastrectomy (13 of 29 cases) following EMR/ESD. Thirteen patients who underwent gastrectomy after incomplete EMR/ESD was enrolled in the study. Indications for EMR in our hospital encompassed the extended criteria including mucosal cancers without ulceration regardless of lesion size, ulcerated mucosal cancers with a size restriction of 30 mm, and cancers with minute submucosal invasion with a size restriction of 30 mm.

The EMR/ESD specimen was examined histopathologically. Resected specimens were systematically sectioned at 2-mm intervals, centered on the part of the lesion closest to the margin and the site of deepest invasion. Incomplete resection was classified as a positive lateral margin, a positive vertical margin and both positive lateral and vertical margin.

Results

1. Reasons for performing gastrectomy after EMR/ESD

The reasons for gastrectomy after EMR/ESD are shown in Table 1. There was bleeding in 7 cases, metachronous lesion in 3 cases, recurrence in 3 cases, perforation in 2 cases, and lymphatic invasion 1 in case. Curative resection was performed in all patients. The surgeries performed were subtotal gastrectomy, B-I & B-II and total gastrectomy with Roux-en-Y esophagojejunostomy.

Table 1

Reasons for performing gastrectomy after EMR/ESD

EMR = endoscopic mucosal resection; ESD = endoscopic submucosal dissection.

Characteristics of the patients and the gastric lesions in incomplete cases (n=13). Of the 13 patients, 7 were male and 6 were female. The mean age of the patients was 63.1±7.7 years. There were 8 cases with lesion size greater than 2 cm.

After surgery, the depth of invasion was at the mucosa level in 3 cases, at the submucosal level in 9 cases, and to the muscle in 1 case. Three cases had LN metastases and 1 case was stage N3.

In the final pathology, 8 cases (61.6%) were diagnosed as well or moderately differentiated adenocarcinoma. Four cases were poorly differentiated, and 1 case was diagnosed as signet ring cell carcinoma. The most common type of operation was subtotal gastrectomy with Billroth I anastomosis (Table 2).

Table 2

Characteristics of the patients and the gastric lesions in incomplete cases (n=13)

LN = lymph node; RSG = radical subtotal gastrectomy; B-I = Biloth-I; B-II = Biloth-II.

2. Endoscopy, EMR/ESD and Operative findings

9 of 13 cases were diagnosed as mucosal tumors on endoscopic ultrasonography (EUS) but only 3 cases were confirmed as mucosal tumors on final pathology.

Three cases were diagnosed as tubular adenoma on endoscopy, with 1 case well differentiated and 2 cases moderately differentiated. Four cases were diagnosed as well differentiated on EMR/ESD with 1 case revised to moderately differentiated on final pathology (Table 3).

Table 3

Endoscopy, EMR/ESD and operative findings

EMR = endoscopic mucosal resection; ESD = endoscopic submucosal dissection; EUS = endoscopic ultrasonography; TA = tubular adenoma; WD = adenocarcinoma, well differentiated.

3. Relationship between positive margins with residual tumors and lymph node metastasis

We analyzed the relationship between positive margins with residual tumors and lymph node metastasis. Positive residual tumor was present in 2 of 13 positive margin cases.

There was 1 case with lymph node metastasis in the positive lateral margin and 2 cases in the positive vertical margin (Table 4).

Table 4

Relationship between positive margins with residual tumors and lymph node metastasis

LN = lymph node; LM = lateral margin; VM = vertical margin.

We also analyzed the relationship between depth of invasion with residual tumor and lymph node metastasis. One case with residual tumor had submucosal invasion and the other case had muscle invasion. All lymph node metastasis cases were submucosal tumors (Table 5).

Table 5

Relationship between depth of invasion with residual tumor and lymph node metastasis

LN = lymph node.

4. Cases of LN metastasis (n=3)

There were 3 cases of LN metastasis. The sizes were 1.5, 1.6, and 2.2 cm.

The histological results were moderately differentiated for case 1, and poorly differentiated in cases 2 and 3. All cases had submucosal invasion. The number of LN metastases were 7, 2, and 1.

Cases 1 and 3 had lymphatic invasions and no cases had vascular or neural invasions (Table 6).

Table 6

Cases of LN metastasis

LN = lymph node; MD = moderately differentiated; PD = poorly differentiated; SM = submucosa.

Discussion

The role of surgery after incomplete EMR/ESD was has been previously reported.

Song et al.(12) reported that gastrectomy with lymph node dissection should be performed in patients with submucosal invasion with or without margin involvement after EMR.

Nagano et al.(13) reported that patients with submucosal invasion or a positive vertical margin should undergo gastrectomy and lymph node dissection.

At first, we expected that if the margin was positive, there'll be residual tumor. But, the results were different. In our study, the positive residual tumor rate was only 15.4% (2 of 13) in positive lateral and/or vertical margin cases. 84.6% (11 of 13) had no residual tumor despite positive margins. In the case that had both the lateral and vertical margins positive, there was no lymph node metastasis.

In a recent study, Chang et al.(14) reported that in lesions invading the lateral margin, the possibility of residual cancer was low. This may be because the EMR/ESD specimen was not fully evaluated histologically. In addition, EMR/ESD is performed using a coagulating device which means that there might not be cells within the margin of the coagulated tissue.

With regard to the depth of invasion, the presence of residual tumor was associated with submucosal and muscle invasion. In particular, the muscle invasion case required EMR with piecemeal dissection.

All lymph node metastasis cases were associated with submucosal tumor. There was no lymph node metastasis in tumors confined to the mucosa.

These results suggest that the depth of invasion deserves greater consideration than positive margin status as an indication for surgery after incomplete EMR/ESD.

Recently there have been a few report regarding extended criteria for EMR/ESD, it is significant especially regarding the application of extended criteria its presence.

The indications for EMR/ESD in our hospital used the extended criteria. There are some concerns regarding the application of extended criteria for EMR/ESD.

Existing methods for preoperative detection of lymph node metastasis are not sufficiently accurate. The only reliable diagnostic method for lymph node metastasis is pathological examination after lymph node dissection.(15,16) Therefore, EMR/ESD may miss lymph node metastases.(17) Our study showed lymph node metastasis in 3 of 13 (23.1%) cases that met the extended indication criteria.

Ishikawa et al.(18) reported the relationship between lymph node metastasis and the tumor depth of wall invasion. When the depth of invasion was submucosal (<500 µm), 2/15 (13%) of EGC differentiated type without ulcer had lymph node metastasis. The authors suggested that indications for EMR/ESD should be limited to mucosal tumors. In our study, all lymph node metastasis cases were submucosal tumors. Extending the indications for EMR/ESD remains controversial, and the applicability of the extended criteria should be further evaluated.

The invasion of tumor can be assessed by performing EUS. The accuracy of EUS for gastric cancer from different authors ranges from 64.8%(19) to 92%(20) in T staging and 50%(21) to 90%(22) in N staging. Currently, EUS is the most reliable method in T and N staging of gastric cancer with high accuracy rates.(23)

In our results, 9 of 13 cases were diagnosed as mucosal tumors on EUS but only 3 cases were confirmed as mucosal tumors on final pathology. More accurate pretreatment staging by EUS should be undertaken for proper stage-dependent patient management.

In conclusion, the gastrectomy is needed for patients with incomplete resection after EMR/ESD, because of the risk of both residual tumor and lymph node metastasis. Precise T staging is essential to avoid unnecessary procedures.