Laparoscopic lymph node dissection after endoscopic submucosal dissection: a novel and minimally invasive approach to treating early-stage gastric cancer

Laparoscopic lymph node dissection after endoscopic submucosal dissection: a novel and minimally invasive approach to treating early-stage gastric cancer

The American Journal of Surgery 190 (2005) 496 –503 How I do it Laparoscopic lymph node dissection after endoscopic submucosal dissection: a novel a...

408KB Sizes 8 Downloads 68 Views

The American Journal of Surgery 190 (2005) 496 –503

How I do it

Laparoscopic lymph node dissection after endoscopic submucosal dissection: a novel and minimally invasive approach to treating early-stage gastric cancer Nobutsugu Abe, M.D., Ph.D.*, Toshiyuki Mori, M.D., Ph.D., Hirohisa Takeuchi, M.D., Takashi Yoshida, M.D., Atsuko Ohki, M.D., Hisayo Ueki, M.D., Osamu Yanagida, M.D., Ph.D., Tadahiko Masaki, M.D., Ph.D., Masanori Sugiyama, M.D., Ph.D., Yutaka Atomi, M.D., Ph.D. Department of Surgery, Kyorin University School of Medicine, 6-20-2 Shinkawa, Mitaka, Tokyo 181-8611, Japan Manuscript received November 22, 2004; revised manuscript May 2, 2005

Abstract Background and Objective: Endoscopic submucosal dissection (ESD), a newly developed endoscopic mucosal resection (EMR) technique, can completely cure a differentiated mucosal gastric cancer smaller than 2 cm. For early-stage gastric cancers (EGCs) deviating from the above-mentioned criterion, gastrectomy with lymph node dissection is performed for potential risk of lymph node metastasis (LNM). However, many of surgical EGC cases actually do not have LNM, indicating this surgery may not be necessary for many cases of EGC. To avoid this unnecessary surgery, we have introduced laparoscopic lymph node dissection (LLND) after ESD. Standard gastrectomy with extended lymph node dissection is indicated for patients if LLND reveals LNM. We present our novel approach and the preliminary results of EGC patients having potential risk of LNM. Methods: Five patients with EGC deviating from the EMR criterion underwent the combination of ESD and LLND. ESD was performed using a newly developed insulation-tipped diathermic knife. Lymph nodes, which were determined on the basis of the location of the primary tumor and lymphatic drainage of the stomach, were removed laparoscopically. The lymphatic drainage was visualized by submucosally injecting indocyanine green (ICG) around the post-ESD ulcerative scars during intraoperative gastroscopy. Results: The ESD enabled en bloc resection without any complications. The resected margins of all the lesions were free of cancer cells vertically and horizontally. LLND was successfully performed without any complications. The mean number of the dissected lymph nodes was 15 (range 6 to 22). In 4 of the 5 patients, the dissected lymph nodes were free of cancer cells, and therefore, the combination of ESD and LLND was considered a definitive treatment. The remaining patient was found to have LNM but chose not to undergo any surgery. During follow-ups, the patients’ previous quality of life was restored without any tumor recurrence. Conclusions: The combination of ESD and LLND enables the complete resection of the primary tumor and the histologic determination of lymph node status. This combination treatment is a potential, minimally invasive method, and may obviate unnecessary gastrectomy without compromising curability for EGC patients having the potential risk of LNM. © 2005 Excerpta Medica Inc. All rights reserved. Keywords: Early-stage gastric cancer; Endoscopic submucosal dissection; Laparoscopic lymph node dissection; Lymph node metastasis

Endoscopic submucosal dissection (ESD), a newly developed endoscopic mucosal resection (EMR) technique, is a valuable alternative to gastrectomy for treating early-stage gastric cancer (EGC) [1– 4]. The accurate assessment of the potential presence of lymph node metastasis (LNM) is re* Corresponding author. Tel.: ⫹81-422-47-5511; fax: ⫹81-422-47-9926. E-mail address: [email protected]

quired for the appropriate treatment (endoscopic treatment or gastrectomy with lymph node dissection). However, metastatic lymph nodes at an early stage of cancer are clinically undetectable [5,6] and therefore the definitive prediction of LNM at the pretreatment stage is often difficult. In Japan, ESD is indicated for a differentiated mucosal cancer smaller than 2 cm [1], because risk for LNM is negligible (0%) [7]. For other EGCs deviating from the

0002-9610/05/$ – see front matter © 2005 Excerpta Medica Inc. All rights reserved. doi:10.1016/j.amjsurg.2005.05.042

N. Abe et al. / The American Journal of Surgery 190 (2005) 496 –503

497

Table 1 Clinical data on five patients Patient

Age/gender

Location

Macroscopic type

Pre-ESD diagnosis

Pathological diagnosis of ESD specimens

Time to LLND from ESD

1 2 3 4 5

66/M 69/F 56/M 69/M 53/M

U/less L/ant U/less U/less U/gre

IIb ⫹ IIc IIc ⫹ IIb IIa IIa IIc

m, 20 mm, por m, 30 mm, sig m, 22 mm, well sm, 30 mm, well m, 15 mm, por

m, 33 mm, por, ly(⫺), lm(⫺), vm(⫺) m, 35 mm, sig, ly(⫺), lm(⫺), vm(⫺) sm, 21 mm, well, ly(⫹), lm(⫺), vm(⫺) sm, 34 mm, well, ly(⫹), lm(⫺), vm(⫺) m, 22 mm, por, ly(⫺), lm(⫺), vm(⫺)

5 3 2 15 4

weeks weeks weeks months weeks

These factors were evaluated out according to the Japanese Classification of Gastric Carcinoma established by the Japanese Research Society for Gastric Cancer. U ⫽ upper third; L ⫽ lower third of the stomach; less ⫽ lesser curvature; gre ⫽ greater curvature; ant ⫽ anterior wall; IIb ⫽ superficial flat type; IIc ⫽ superficial depressed type; IIa ⫽ superficial elevated type; m ⫽ intramucosal cancer; sm ⫽ submucosal cancer; por ⫽ poorly differentiated adenocarcinoma; sig ⫽ signet-ring cell carcinoma; well ⫽ well-differentiated adenocarcinoma; ly ⫽ lymphatic involvement; lm ⫽ lateral margin; vm ⫽ vertical margin.

above-mentioned criterion, major surgery such as gastrectomy with lymph node dissection is usually performed for potential risk (8.8%) [7] for LNM even though ESD may remove the primary gastric lesion. However, this surgery may eventually be unnecessary for many EGC patients, since approximately 91% of surgically treated EGC patients actually do not have LNM [7]. If the gastric lesions can be completely removed endoscopically and lymph node status can be histologically determined before gastrectomy, this unnecessary surgery could be obviated. We introduced the combination of ESD using the newly developed insulation-tipped diathermic knife, which is one of the EMR procedures promising the complete removal of the primary tumor as a single fragment (en bloc resection) [1– 4], and laparoscopic lymph node dissection (LLND) without performing gastrectomy [4]. The combination of ESD and LLND enables the complete resection of the primary tumor and the histologic determination of the lymph node status. If LLND reveals LNM, standard gastrectomy with extended lymph node dissection is indicated. We have employed this combination method for EGC patients having the potential risk of LNM, deviating from the above-mentioned ESD criterion. We present our novel approach and the preliminary results of five EGC patients.

Patients and Methods Patients EGC is defined as a lesion confined to the mucosa or submucosa regardless of the presence or absence of LNM according to the Japanese Classification of Gastric Carcinoma established by the Japanese Research Society for Gastric Cancer [8]. Five EGC patients (4 men and 1 woman) underwent the combination of ESD and LLND between February 2002 and May 2004. The ages of the patients ranged from 58 to 67 years (mean 63). Table 1 shows the pre-ESD diagnostic findings and pathologic assessments of the primary lesion obtained by ESD for each patient. At the pre-ESD stages, the depth of

cancer invasion was diagnosed based on endoscopic examination and/or endoscopic ultrasonography. The criterion for local treatment (EMR or ESD) generally accepted in Japan are a differentiated mucosal cancer smaller than 2 cm [1]; such cancers rarely metastasize to the lymph nodes [7]. Therefore, local treatment alone was not considered to be a curative treatment for the present 5 patients because of the potential LNM (Table 1). At the post-ESD stages, the predicted probabilities of LNM were presumed to be 4.1%, 4.1%, 29.6%, 39.4%, and 2.7% in patients 1, 2, 3, 4, and 5, respectively, according to a previous report on the incidence of LNM in a large number of EGC cases [7]. These patients were accordingly informed of the need for standard surgery such as gastrectomy with lymph node dissection at the preand/or post-ESD stages, but they expressed preference for the combination of ESD and LLND without performing gastrectomy. Patient 4 had undergone ESD in another hospital. The computed tomography and endoscopic sonography 14 months after the ESD disclosed an enlarged lymph node in the perigastric region on the lesser curvature of the stomach. Due to the concern of a high possibility of the presence of LNM, the patient was informed of the need for gastrectomy with lymphadenectomy. However, the patient chose to undergo LLND alone. This patient was referred to our hospital for LLND. Before applying ESD or LLND, informed consent was obtained from each patient; the procedure was conducted in accordance with the ethical standards of our institution’s Committee on Human Experimentation. In addition, when LNM was detected after LLND, we informed the patients of the need for additional gastrectomy with extended lymph node dissection. The lymph node status was evaluated by a conventional method using formalin-fixed sections, not frozen sections of specimens obtained intraoperatively. ESD and pathologic evaluation ESD was performed with the newly developed insulation-tipped diathermic knife (IT knife, type KD-IL; Olympus Optical Co, Tokyo, Japan) as previously described (Fig. 1A–E) [1– 4]. The ESD enabled en bloc resection without

498

N. Abe et al. / The American Journal of Surgery 190 (2005) 496 –503

Fig. 1. ESD procedures (patient 2). (A) Endoscopic view of slightly depressed type cancer (arrowheads) on the anterior wall of the antrum of the stomach. (B) The tumor was separated from the surrounding normal mucosa by a circumferential incision using the IT knife. (C) The IT knife (arrow) was used to dissect the submucosal layer under the circumferentially incised area. The IT knife is a needle knife, which has a ceramic ball at its tip. (D) The resection site after ESD. (E) The resected specimen, measuring 55 mm in diameter, showed signet ring cell carcinoma (arrowheads, 33 mm in diameter) with a clear lateral margin.

major complications, such as significant bleeding and perforation of all the lesions. The formalin-fixed specimens were cut into multiple slices at an interval of 3 to 5 mm. All microscopic sections were stained with hematoxylin and eosin (H&E), and were examined histologically by pathologists. The specimens were evaluated basically according to the Japanese Classification of Gastric Carcinoma established by the Japanese Research Society for Gastric Cancer [8]. Pathologic examinations showed that the resected margins of all the lesions were free of cancer cells vertically and horizontally (Table 1). Therefore, all the lesions were completely removed locally. Determination of dissection area in LLND LLND was performed under general endotracheal anesthesia. A 12-mm initial access site was established at the supraumbilicus by the Hasson open technique. After a forward-viewing laparoscope was introduced through the supraumbilical trocar, a pneumoperitoneum was created with

carbon dioxide at a pressure of 10 mm Hg and 4 additional trocars were placed. Two 11-mm trocars were placed, 1 in the left lateral abdominal wall and the other in the left subcostal area. Two 5-mm trocars were then added, 1 in the right subcostal area and the other in the right lateral abdominal wall. The left lobe of the liver was elevated with a snake retractor. The anatomic distribution and numbering of the regional lymph nodes are based on the classification system of the Japanese Research Society for Gastric Cancer (Table 2) [8]. The area for lymph node dissection was determined on the basis of the location of the primary tumor and the lymphatic drainage of the stomach [9,10]. The lymphatic drainage was visualized with a laparoscope 5 to 10 minutes after the submucosal injection 8 mL of indocyanine green (ICG) around the post-ESD ulcerative scars during intraoperative gastroscopy [11]. In patients 1, 3, and 4, the ICG injection revealed lymphatic flows around branches from the left gastric artery (LGA) in the perigastric region on the lesser curvature of the stomach. Accordingly, lymph nodes, in-

N. Abe et al. / The American Journal of Surgery 190 (2005) 496 –503 Table 2 Regional lymph nodes dissected Numbering No. No. No. No. No. No. No. No. No.

1 2 3 4sa 4sb 4d 5 6 7

Right paracardial LN Left paracardial LN LN along the lesser curvature LN along the short gastric vessels LN along the left gastroepiploic vessels LN along the right gastroepiploic vessels Suprapyloric LN Infrapyloric LN LN along the left gastric artery

The anatomic distribution and numbering of the regional lymph nodes are based on the classification system of the Japanese Research Society for Gastric Cancer. LN ⫽ lymph node.

499

cluding those firstly stained with the dye, that is, sentinel nodes, located in the perigastric region near the resected tumor (no. 1), no. 3, and 7 lymph nodes were removed in these patients (Fig. 2A). In patient 2, ICG injection revealed lymphatic flow around branches from the left and right gastric arteries in the perigastric region on the lesser curvature of the stomach as well as that around branches from the right gastroepiploic artery in the perigastric region on the greater curvature of the stomach. Thus, no. 3, 4d, 5, 6, and 7 lymph nodes were accordingly removed (Fig. 2B). In patient 5, the ICG injection revealed no lymphatic flow. On the basis of the location of the primary tumor (greater curvature of the upper stomach), no. 1, 2, 3, 4sa, 4sb, and 7 lymph nodes were removed (Fig. 2C).

Fig. 2. Schematic of the area for lymph node dissection. (A) No. 1, 3, and 7 lymph nodes were removed in patients 1, 3, and 4. (B) No. 3, 4d, 5, 6, and 7 lymph nodes were removed in patient 2. (C) No. 1, 2, 3, 4sa, 4sb, and 7 lymph nodes were removed in patient 5.

500

N. Abe et al. / The American Journal of Surgery 190 (2005) 496 –503

Fig. 3. Resection specimen in patients 4 (A), 2 (B and C), and 5 (D).

The lymph nodes obtained from resected specimens were fixed in 10% formalin solution. A single section was prepared from the plane of the largest dimension of each node, and stained with H&E for histologic examination. LLND technique in patients 1, 3, and 4 An incision was made in the lesser omentum along the lesser curvature of the stomach at the level of the gastric angle. The incision in the lesser omentum was extended laterally near to its attachment to the left lobe of the liver using a Harmonic Scalpel (Ethicon Endo-Surgery, Cincinnati, OH) or a Vessel Sealing System (LigaSure Atras; Valleylab, Boulder, CO). The incision was then extended proximally toward the abdominal esophagus. During the

skeletonization of the upper third of the lesser curvature (dissection for no. 1 and 3 lymph nodes), the LGA was exposed and divided at its origin with triple clips, allowing the lymph node dissection along this vessel (no. 7). The coronary vein was also exposed and divided. The specimen, including no. 1, 3, and 7 lymph nodes, was finally obtained. The hepatic branch of the vagal nerve was identified and left intact throughout the surgery. The removed specimen is shown in Fig. 3A. LLND technique in patient 2 The gastrocolic ligament was divided 4 cm distal to the epiploic arcade at the level of the junction of the right and left gastroepiploic arteries. The division of the gastrocolic

N. Abe et al. / The American Journal of Surgery 190 (2005) 496 –503

501

Table 3 Operative and postoperative data on five patients Patient

Dissected area (LNs number)

Blood loss (ml)

Time (min)

Number of dissected lymph nodes

Histological lymph node status

Postoperative hospital stay (days)

Follow-up period (months)

Outcome

1 2 3 4 5

1, 3, 1, 1, 1,

517 217 157 10 12

208 288 234 254 214

28 21 11 7 6

Negative Negative Negative Positive (1/7) Negative

9 7 7 7 7

37 27 20 14 11

No No No No No

3, 7 4d, 5, 6, 7 3, 7 3, 7 2, 3, 4sa, 4sb, 7

ligament was extended toward the pylorus using the Harmonic Scalpel. The root of the right gastroepiploic artery was exposed and skeletonized, and divided with double clips at about 2 cm distal its origin, allowing the excision of all adjacent lymph nodes. This procedure permitted the complete dissection of no. 6 lymph node. The 4d lymph nodes were dissected along with the skeletonization of the lower third of the greater curvature of the stomach. The removed specimen is shown in Fig. 3B. After the incision of an avascular area along the right side of the lesser omentum, the right gastric artery was exposed and divided at its origin with double clips. The no. 5 lymph node was dissected from the lesser curvature using the Harmonic scalpel. The LGA was then exposed and divided at its origin with triple clips, allowing the dissection of the no. 7 lymph node. The No. 3 lymph node was dissected along with the skeletonization of the middle third of the lesser curvature of the stomach (Fig. 3C). LLND technique in patient 5 The gastrocolic ligament was divided 4 cm distal to the epiploic arcade at the level of the junction of the right and left gastroepiploic arteries. The division of the gastrocolic ligament was advanced toward the spleen. After the roots of the right gastroepiploic and short gastric arteries were exposed and divided at their origins in the hilum of the spleen, the skeletonization of the upper third of the greater curvature of the stomach was performed. These procedures permitted a complete dissection of lymph nodes no. 4sa, 4sb, and 2. The removed specimens are shown in Fig. 3D.

Results The entire laparoscopic procedures were successfully performed without any complications. The mean operating time and blood loss were 240 (range 208 to 288) minutes and 183 (range 10 to 517) mL, respectively (Table 3). In none of patients did the stomach showed ischemia at the end of LLND. Postoperatively, all of the patients did very well, resuming oral intake on postoperative day 3 and being discharged from hospital on postoperative day 7 to 10. The mean number of dissected lymph nodes was 15

reccurrence reccurrence reccurrence reccurrence reccurrence

(range 6 to 22). Histologic examination revealed that all the nodes were free of cancer cells in patients 1, 2, 3, and 5 (Table 3). On the basis of these findings, the combination of ESD and LLND was considered the definitive treatment in these patients. In patient 4, histologic examination confirmed matastasis to the enlarged lymph node disclosed by imaging modalities. However, the other 6 lymph nodes dissected were free of cancer cells. Patient 4 was informed of the need for additional conventional gastrectomy with extended lymph node dissection but chose not to undergo any surgery. All of the patients have been monitored periodically by physical examination, serum carcinoembryonic antigen monitoring, chest radiography, abdominal computed tomography, and upper gastrointestinal endoscopy every 3 to 6 months. During the mean follow-up of 22 months, they were almost asymptomatic without any tumor recurrence (Table 3). Their previous quality of life was restored with no dietary restrictions and body weight loss. Although no patients showed severe gastritis, gastric ulcer, or reflux esophagitis, a slight gastric residue was detected in 2 patients. All of the patients were satisfied with the outcome of the combination of ESD and LLND.

Comments Gastrectomy with lymph node dissection has been generally accepted as a treatment of choice in such patients as the present 5 patients with respect to the potential presence of LNM. However, the present procedures applied, the combination of ESD and LLND, could eliminate unnecessary gastrectomy, particularly in patients 1, 2, 3, and 5. ESD allowed the complete local resection of the primary tumor, and LLND enabled the confirmation of negative LNM. LNM was fortunately absent in patients 1, 2, 3, and 5, and therefore, this combination was considered the definitive treatment for these patients. At present, conventional gastrectomy with extended lymph node dissection is considered to be required in order to eliminate the potential presence of downstream LNM if LLND reveals LNM, as in patient 4. This is because it remains unclear whether LLND is sufficient treatment for EGC with LNM. A long-term follow-up study in selected patients such as patient 4 is required to

502

N. Abe et al. / The American Journal of Surgery 190 (2005) 496 –503

demonstrate the direct evidence of the treatment efficacy of LLND. This combination is applicable in any EGC patients in whom the primary lesion can be completely removed by ESD despite their potential risk of LNM. In treating EGC by this combination, a complete local resection of the primary tumor is an important initial step to achieve. To obtain this, we have employed the newly developed ESD technique as an EMR procedure [4]. Conventional methods such as strip biopsy [12,13] are presumed to be inadequate for en bloc resection of large lesions. Fragmented resection prevents histopathologic evaluation of the completeness of resection, which may lead to imcomplete resection and a high rate of local recurrence [4]. By contrast, ESD using an IT knife enables en bloc resection and complete histopathologic examination, resulting in a low incidence of local recurrence [1–3]. Therefore, we consider that ESD technique is desirable for local resection as a part of the present combined treatment. LLND has some problems to be solved. One is the delineation of the dissection area. LNM of relatively large (2 to 4 cm in diameter) EGCs, particularly for mucosal cancers, is reported to be confined to the perigastric area [10,14]. Based on these data, as well as on the successful use of ICG, we believe that the procedure used in the present cases enabled the complete removal of regional lymph nodes. In this study, ICG injection was used simply for the visualization of lymphatic flow from the stomach, but not mainly for the detection of sentinel nodes. A radiopharmaceutical-guided sentinel node detection has been reported to demonstrate a high accuracy in detecting sentinel nodes but does not enable the visualization of lymphatic flow [6]. Moreover, it remains unclear whether sentinel node detection should be performed using dyes such as ICG or radiopharmaceutical or both [15]. For these reasons, we have chosen the ICG injection method for determining the dissection area. A second problem is the detection method of lymph node micrometastasis. Recently, several investigators have demonstrated that molecular techniques using highly sensitive reverse trancriptase-polymerase chain reaction may be useful in the detection of lymph node micrometastasis. Immunohistochemistry can also enhance the accuracy of the detection of lymph node micrometastasis. However, these diagnostic techniques are time-consuming and relatively laborious compared with conventional H&E staining, and have been considered to be of limited practical value. Although relatively simple and easy to perform, H&E staining remains as the primary standard in lymph node diagnosis in a clinical setting even today. On the basis of the above considerations, we determined the lymph node status by conventional H&E staining in this study. A third potential problem is whether the stomach has useful motility postoperatively. Although the hepatic branch (and the pyloric branch) of the vagal nerve was preserved, the anterior and posterior vagal trunks innervating the lesser curvature of the stomach were resected

for lymph node dissection. A slight gastric residue was detected in 2 patients by postoperative gastroscopy. However, none of our patients presented with postprandial static symptoms, such as epigastric fullness, nausea,a and vomiting, observed in contrast to patients after pylorus-preserving gastrectomy [16]. Quality of life was actually restored, with no dietary restrictions in our patients. Gastric peristalsis may be maintained when the pacemaker zone in the proximal corpus along the greater curvature is preserved [10,17]. A fourth potential problem is whether ischemic change may occur in the preserved stomach. During the surgical procedures, the extramural vessels and their branches were removed. However, LLND can be performed safely without gastric ischemia. A previous experiment showed the whole stomach continuously receiving blood supply through the intramural network with the preservation of 2 of the 5 major blood vessels [18]. In patient 2, although the right and left gastoepiploic arteries were divided at their origin (the short and left gastroepiploic arteries were preserved), the whole stomach did not become ischemic. In the follow-up periods, no patient developed gastric ulcer. Although laparoscopic local gastric resection with regional lymphadenectomy for EGC is the treatment of choice with respect to providing good quality of life, complications such as gastric stenosis have been reported [14]. The use of this procedure for a lesion located on the lesser curvature of the cardia, as in the present cases (patients 1, 3, and 4) is technically difficult. The combination of ESD and LLND, by contrast, is applicable to any areas in the stomach with fewer complications, and it preserves quality of life without compromising curability. In conclusion, the combination of ESD and LLND enables the complete resection of the primary tumor and histologic determination of lymph node status. This combination can be a potential, minimally invasive treatment for EGC patients having the potential risk of LNM. This combination may obviate the need for gastrectomy without compromising curability.

References [1] Ono H, Kondo H, Gotoda T, et al. Endoscopic mucosal resection for treatment of early gastric cancer. Gut 2001;48:225–9. [2] Kondo H, Gotoda T, Ono H, et al. Early gastric cancer: endoscopic mucosal resection. Ann Ital Chir 2001;72:27–31. [3] Miyamoto S, Muto M, Hamamoto Y, et al. A new technique for endoscopic mucosal resection with an insulation-tip electrosurgical knife improves the completeness of resection of intramucosal gastric neoplasms. Gastrointest Endosc 2002;55:576 – 81. [4] Abe N, Mori T, Izumisato Y, et al. Successful treatment of an undifferentiated early gastric cancer by combined en bloc endoscopic mucosal resection and laparoscopic regional lymphadenectomy. Gastrointest Endosc 2003;57:972–5. [5] Ziegler K, Sanft C, Zimmer T, et al. Comparison of computed tomography, endosonography, and intraoperative assessment in TN staging of gastric carcinoma. Gut 1993;34:604 –10.

N. Abe et al. / The American Journal of Surgery 190 (2005) 496 –503 [6] Miwa K, Kinami S, Taniguchi K, et al. Mapping sentinel nodes in patients with early-stage gastric carcinoma. Br J Surg 2003;90: 178 – 82. [7] Gotoda K, Yanagisawa A, Sasako M, et al. Incidence of lymph node metastasis from early gastric cancer. The estimation using a large number of cases in two large centers. Gastric Cancer 2000;3:219 –25. [8] Japanese Gastric Cancer Association. Japanese Classification of Gastric Carcinoma. 2nd English edition. Gastric Cancer 1998;1:10 –24. [9] Sarrazin R, Pissas A, Dyon JF, et al. Lymphatic drainage of the stomach. Anat Clin 1980;2:95–110. [10] Seto Y, Nagawa H, Muto Y, et al. Preliminary report on local resection with lymphadenectomy for early gastric cancer. Br J Surg 1999;86:526 – 8. [11] Hiratsuka M, Miyashiro I, Ishikawa O, et al. Application of sentinel node biopsy to gastric cancer surgery. Surgery 2001;129:335– 40. [12] Karita M, Tada M, Okita K. The successive strip biopsy partial resection technique for large early gastric and colon cancers. Gastrointest Endosc 1992;38:174 – 8.

503

[13] Tada M, Murakami A, Karita M, et al. Endoscopic resection of early gastric cancer. Endoscopy 1993;25:445–50. [14] Seto Y, Yamaguchi H, Shimoyama S, et al. Results of local resection with regional lymphadenectomy for early gastric cancer. Am J Surg 2001;182:498 –501. [15] Aikou T, Higashi H, Natsugoe S, et al. Can sentinel node navigation surgery reduce the extent of lymph node dissection in gastric cancer? Ann Surg Oncol 2001;8:90 –3. [16] Nakane Y, Akehira K, Inoue K, et al. Postoperative evaluation of pyloru-preserving gastrectomy for early gastric cancer. Hepatogastroenterology 2000;47:590 –5. [17] Grundy D, Scratcherd T. The role of the vagus and sympathetic nerves in the control of gastric motility. In: Akkermans LMK, Johnson AG, Read NW, eds. Gastric and Gastroduodenal Motility. New York, NY: Praeger, 1984:21–33. [18] Seto Y, Nakayama H, Ishigami H, et al. An experimental study of intramural blood supply network of the stomach wall. Hepatogastroenterology 1999;46:2673– 6.