Endoscopic Mucosal Resection and the Risk of Lymph-Node Metastases: Indications Revisited?

Four articles in the present issue of Endoscopy deal in different ways with Barrett’s esophagus and the management of cancer and its risk of development [1] [2] [3] [4]. A literature review found that endoscopic surveillance is unlikely to affect mortality, since surveillance covers so few of the actual Barrett’s patients; strategies therefore have to be developed to identify those at risk [1]. A method of submucosal dissection for neoplasia at the esophagogastric junction is reported by a Japanese group. In 30 such patients with esophagogastric junction tumors, at least four of the lesions were Barrett’s cancers [2]. The authors achieved a very high R0 en-bloc resection rate (97 %). This aim of this technique, which has been reported previously with variable success rates [5] [6] [7] [8] [9] [10], is to overcome the limitations of endoscopic mucosal resection (EMR). Using conventional techniques, EMR in most cases only allows piecemeal resection with a low primary R0 resection rate, or at least inherent uncertainty as to whether or not the tumor has been completely removed [11].

Missed lymph-node metastases are the most feared type of failure with EMR, however. It is now clear that the risk of lymph-node metastases is negligible in mucosal adenocarcinomas. There has been some debate recently on whether the risk might be similarly low in patients with minimal infiltration of the submucosa. An overview of surgical series reporting lymph-node metastases in early upper gastrointestinal adenocarcinomas is given in Table [1] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29]. With one exception reporting very high rates [21], this overview shows that the risk of lymph-node metastases is between 0 % and 8 %, mostly below 5 % in mucosal cancer, and ranges from 14 % to 45 % in submucosal cancer. In subtle submucosal infiltration (sm1 or < 500 μm), the rate of lymph-node metastasis was lower (0 - 14 %) than with deeper infiltration (21 - 46 %), so that additional prognostic factors come into play - e. g., lymphatic invasion in the resection specimen. In the huge series of early gastric cancers presented by Gotoda et al., several situations are defined that have a risk of lymph-node metastases that is close to zero [19] (Table [2]). We are only discussing adenocarcinomas in such cases. A high risk of early lymph-node metastases in squamous-cell carcinomas is well established; Koike et al., dividing mucosal infiltration into three levels (m1 - 3), already showed in 1998 that m3 squamous-cell carcinoma had an 8 % rate of lymph-node metastases, compared with a rate of 0 in m1 - m2 tumors [30]. Squamous-cell cancer is therefore a different story, and only superficial mucosal extension may be an accepted EMR indication.

Table 1 Rates of lymph-node metastases in surgical series reporting mucosal and submucosal early upper gastrointestinal tract adenocarcinoma. Only papers from 1999 to 2005 dealing with submucosal invasion and lymph nodes were retrieved from Medline (using the search terms ”early cancer”, ”lymph node”, and either ”stomach”, ”esophagus”, or ”Barrett”). Only selected earlier papers were taken into account. The results for mucosal cancers are also listed from these papers if reported First author, ref. n LN metastases System1 muc sm muc sm sm1/ < 500 μm sm2, 3/ > 500 μm Gastric cancer Kim 12 Only sm 377 371 4 % 16 % - - Pelz 13 Only sm 137 142 8 % 15 % - - Park 14 sm1 - 3 - 105 - 23 % 14 % 26 % < / ≥ 500 μm 9 % 27 % Abe 15 2 sm1/2 - 104 - 23 % 6 % 40 % Choi 16 Only sm - 88 - 19 %3 - - Folli 17 Only sm 296 288 4 % 25 % - - Yamada 18 sm1 - 3 - 104 - 14 % 0 21 % Gotoda 19 4 sm1/2 3016 2053 2 % 19 % 9 % 24 % Borie 20 5 Only sm 160 167 4 % 9 % Ishigami 21 6 sm1 - 3 - 170 - 45 % 29 % 57 % Yasuda 22 < /≥ 500 μm - 118 - 14 % 5 % 15 % Kurihara 23 sm1 - 3 - 245 - 14 % 2 % 18 % Esophagogastric junction cancer Stein 24 Only sm 57 87 0 18 % - - Westerterp 25* sm1 - 3 54** 65 2 %** 28 % 0 44 % Buskens 26* Only sm 33 42 0 29 % 0 46 % Rice 2001 27 Only sm 53 31 4 % 19 % . . Van Sandick 28 Only sm 12 20 0 30 % - - Ruol 29 Only sm 4 22 0 36 % - 1 The classification system used to measure the extent of submucosal infiltration (”only sm” means no further subclassification). 2 Lymphatic involvement was an important prognostic factor. Without sm1 and sm2, there were lymph-node metastasis rates of 0 % and 10 %, whereas positive lymphatic involvement yielded rates of 21 % (sm1) and 80 % (sm2). 3 The rate increased to 32 % when cytokeratin immunostaining for micrometastases was carried out in addition to hematoxylin-eosin staining. 4 A very large collection of cases. Lymphatic involvement was another important selection criterion: without lymphatic involvement (n = 670), the rate was 8 % (sm1 lesions 1 %; sm1 < 3 cm 0 %); with lymphatic involvement (n = 421), the rate was 32 %. See also Table 2. 5 Here again, the presence of lymphatic involvement led to further risk stratification: rate of lymph-node metastasis without lymphatic involvement 5 %, with lymphatic involvement 24 %. 6 The degree of lymphatic tumor invasion was a further prognostic factor; if it was minimal (grade 1 of three), the rate of lymph-node metastasis was 17 % for sm1 cancers.* Probably the same group of patients.** Patients with mucosal and sm1 submucosal infiltration when grouped together as ”muc”.

Table 2 Criteria from Gotoda et al. 19 for safe endoscopic mucosal resection of lesions with submucosal invasion Criteria Incidence 95 % CI Intramucosal cancerDifferentiated adenocarcinoma, no lymphatic vessel invasion, irrespective of ulcer findings, tumor < 3 cm in size 0/1230 (0 %) 0 - 0.3 % Intramucosal cancerDifferentiated adenocarcinoma, no lymphatic vessel invasion, without ulcer findings, irrespective of tumor size 0/929 (0 %) 0 - 0.4 % Undifferentiated intramucosal cancerNo lymphatic vessel invasion, without ulcer findings, tumor < 3 cm in size 0/141 (0 %) 0 - 2.6 % Minute submucosal penetration (sm1)Differentiated adenocarcinoma, no lymphatic vessel invasion, tumor < 3 cm in size 0/145 (0 %) 0 - 2.5 %

The papers by Bollschweiler et al. [3] and Shami et al. [4] in this issue also deal with lymph-node metastases in early cancers and ways of diagnosing them. The former paper, concerned with surgical resection specimens, shows that a 45 % rate of lymph-node metastases can be expected in submucosal esophageal carcinoma (squamous-cell carcinoma and adenocarcinoma). The rate increases with the depth of submucosal infiltration. In contrast, mucosal carcinoma showed no lymph-node metastases. Similarly, Shami et al. present a case series of 25 patients with suspected early cancers who were referred for EMR, during which lymph-node metastases were confirmed by endoscopic ultrasound-guided fine-needle aspiration in four cases [4].

Both of these papers add further emphasis to warnings against any uncritical use of EMR treatment in the absence of appropriate long-term studies, since lymph-node metastases may occur even in mucosal cancers, and particularly in cases with minimal submucosal infiltration, recently under discussion as an expanded indication for EMR. Two questions arise:

Is minimal submucosal infiltration in differentiated tumors an acceptable indication for definitive EMR? Can lymph-node metastases be reliably recognized before EMR in order to overcome this problem?

One should recall here the EMR criteria that are still regarded as being standard, mainly on the basis of Japanese experience with gastric cancer [31]. High-risk factors in early carcinomas that exclude EMR as the sole treatment option are therefore:

Poor differentiation. Infiltration into the submucosal layer. Size exceeding 2 - 3 cm. Ulcerated tumors. Involvement of lymphatic vessels (Table 1). Venous invasion is not a contraindication against local endoscopic treatment, since distant metastases due to venous invasion cannot be prevented with surgery either.

What do we need to assess these criteria before EMR?

Endoscopy provides information about the size and shape of lesions (ulcers). Biopsy provides information about grading (although this may occasionally change when the full specimen is examined) and lymphatic vessel involvement. However, the diagnosis of submucosal infiltration (and all the more so if further subclassification is required) and lymph-node involvement remains problematic.

Both of the papers dealing with the lymph-node problem in this issue deserve further detailed comments. Bollschweiler et al. show a 25 % rate of lymph-node metastases (three of 12) even in sm1 esophageal cancers. A detailed analysis shows that these were often distant lymph nodes (M1a/b), but also that they mainly occurred in squamous-cell cancers. The numbers of patients in the subgroups concerned are quite small - one of three patients with sm1 squamous-cell cancer and one of seven with sm1 adenocarcinomas had lymph-node metastases. Shami et al. found four patients with lymph-node metastases among 25 with early Barrett’s cancer who were referred for EMR. Only those patients in whom endoscopic ultrasonography (EUS) detected lymph nodes, some of them quite large, underwent aspiration - as shown in Fig. [1 b] in the paper [4]. Unfortunately, the findings were not confirmed at surgery and histopathological analysis in any of the four patients; the local tumor infiltration was submucosal in three patients and mucosal in one. In addition, in the FNA-positive patients, the lymph nodes were 3 mm (!), 9 mm, 22 mm, and 30 mm in size [4], so that it is difficult to draw conclusions regarding selection criteria for which lymph nodes should be aspirated. Nor is it known how many lymph nodes might have been missed on EUS.

Returning to the first question: what about adenocarcinoma and subtle submucosal infiltration? First of all, the methods used to determine and classify minimal submucosal invasion have to be critically examined. The depth of submucosal infiltration cannot be accurately determined by a subdivision into sm1 - 2 or sm1 - 3 in the EMR specimen, which often does not contain the full thickness of submucosa and may have been damaged during injection and cutting. Reliable assessment of each third of the submucosa is only possible in surgical specimens that include the muscularis propria. An invasion depth of 500 μm has therefore been set as an auxiliary parameter for maximum infiltration into the submucosa in upper gastrointestinal tract adenocarcinoma, which would then define the limit for curative endoscopic therapy. The situation might be similar in Barrett’s esophagus, but the data are limited (Table [1]). Thus, if EMR shows subtle submucosal infiltration, further histopathological characterization is needed in accordance with the criteria listed above. In any case, further research is needed in Barrett’s cancer in particular in order to determine the outcome during the long-term follow-up in patients who undergo surgery and those who do not.

Can the depth of infiltration be diagnosed before EMR? The most precise technique available - EUS with high-frequency miniprobes, usually in early gastric cancer - has unfortunately not been consistently shown to provide the necessary information with a high level of accuracy, at least in Barrett’s esophagus and gastric adenocarcinoma, and different studies have reported quite variable results [32] [33] [34] [35] [36] [37]. Even in the most recent publication on early Barrett’s cancer, only two of 14 patients with submucosal tumor invasion diagnosed after EMR had been correctly diagnosed with EUS beforehand; in this study, the endoscopic assessment was as good (or poor as far as specificity was concerned) as EUS [37]. It is not yet clear whether further improvements in EUS (e. g., the use of 30-MHz probes) or the application of other methods (such as optical coherence tomography or confocal laser microscopy) will be able to solve this problem.

The second question was whether lymph-node metastases can be reliably diagnosed before EMR. It might also be possible to use EUS for this purpose, and the study by Shami et al. may support this hypothesis [4]. However, previous evidence from a few series on early gastric adenocarcinoma has shown disappointingly low sensitivity rates for identifying the few cases of metastatic lymph nodes that are present [33] [38] [39]. Quite often, in the authors’ experience, smaller lymph nodes are seen in Barrett’s esophagus irrespective of the presence of neoplasia, and it is not clear which of these should be aspirated before EMR (e. g., with selection according to size, shape, or pattern). Possible methods of improving lymph-node diagnosis (e. g., with contrast EUS or elastography) are at present only speculative.

Fig. [1] and [2] present two clinical cases that highlight the way in which these decision-making problems can arise in everyday practice. In the first case, endoscopy and EUS should not have been used for decision-making (surgery could have been avoided); while in the second, the information provided by EUS should have been used as a basis for not carrying out EMR.

Figure 1 A 53-year-old man with long-segment Barrett’s esophagus and a diagnosis of early cancer from routine biopsies. a Endoscopy revealed a large suspicious area measuring about 3 × 4 cm in the long Barrett’s esophagus segment. b A hypoechoic lymph node 2 cm in size was seen on endoscopic ultrasonography and was also suspected to be positive. Surgery was therefore recommended. The postoperative histopathological grading of the lesion was T1mN0.

Figure 2 A 72-year-old man with two elevated early cancers at the gastric cardia on endoscopy. a The endoscopic impression of the lesion at the 12-o’clock position was borderline. b The endoscopic ultrasound findings were also doubtful with regard to submucosal infiltration. However, diagnostic endoscopic mucosal resection had to be stopped due to deeper tumor infiltration.

At present, many endoscopists argue that they are not concerned with details about infiltration depth, as they just go on to EMR when they can guess from the endoscopic appearance that the lesion is probably completely resectable. It is claimed that cap EMR is easy and safe and that the histopathologist can provide the missing information about the infiltration depth on the basis of which the lymph-node risk can be calculated (”test EMR” or ”diagnostic EMR”). We agree in part, but would like to issue several warnings:

It may be true that EMR is associated with a low rate of complications rate in expert hands, but there is a lack of large, prospective multicenter series to confirm this. The conventional techniques of cap and banding EMR lead to piecemeal resection in the majority of cases and thus a very low R0 resection rate, with a need for further interventions 7 8 11. It is not fully clear whether this type of approach, contradicting oncological principles, is really safe, and it requires at least a careful follow-up regimen. Newer and more elaborate and time-consuming techniques 2 5 6 7 8 9 10 11 that might ultimately meet the oncological resection criteria and lead to en-bloc and R0 resection in a high percentage of cases are at present too time-consuming to be carried out during diagnostic EMR - so that once again we return to the need for precise preprocedural assessment.

In summary, the diagnostic tools available for differentiating between mucosal and submucosal infiltration and for recognizing lymph-node metastases are at present inadequate. Endoscopic experience in assessing the lesions, together with endoscopic biopsies, will have to suffice for the time being, although efforts should be made to develop improvements. With regard to the risk of metastatic lymph nodes, larger studies on EUS and lymph-node detection should be conducted, but would require a careful and long-term EUS follow-up regimen, since most patients will undergo EMR and potential lymph-node metastases are not easily detected during only short-term follow-up periods. In addition, Bollschweiler et al. show that patients may also have distant lymph-node metastases, which can escape even surgical lymphadenectomy if it is not radical enough. Accurate exclusion of lymph-node metastases is therefore currently a diagnostic gap. Since the likelihood of lymph-node metastases is quite low per se (< 5 %), the specificity of any given diagnostic technique has to be very high (> 95 %) for it to be of clinical benefit.

It should not be forgotten that for a gastroenterologist who has to take decisions about further patient management after EMR, the most crucial issue is precise work-up of the neoplastic lesion from the EMR specimen or specimens. The pathologist’s report should contain the following information [40] [41]:

Site of the lesion (taken from the endoscopy report). Type of neoplasia. Grading of the neoplasia. Size of the neoplastic lesion. Presence/absence of vascular invasion. Description of the invasion front (single cells present/absent). Resection status (complete /incomplete). If incomplete: lateral or basal. Exact depth of infiltration. TNM classification.

Do the papers mentioned expand the range of evidence available? Yes and no. There is always a referral bias in most of these studies. In the paper by Bollschweiler et al. [3], the patients who underwent surgery might have been judged by the referring gastroenterologist to have unresectable lesions, but this is not reported in detail. There may also be a referral bias with regard to lymph nodes, since patients were also sent for EUS-FNA in the series reported by Shami et al. [4]. Similarly, however, series reporting endoscopic mucosal resection may only concentrate on smaller and more suitable lesions. Gastroenterologists and surgeons should therefore realize that series from each discipline are likely to provide complementary rather than contradictory information. There have not as yet been sufficient multicenter studies reporting long-term data with careful follow-up and large numbers of patients to allow a conclusive assessment of the limitations of EMR in Barrett’s esophagus. Extending the boundaries of EMR treatment on the basis of the information currently available - which mainly implies accepting minimal submucosal infiltration in certain conditions (Table [2]) - should therefore only be undertaken in strict study conditions and possibly in patients who are at higher surgical risk. The main aim should be to develop risk profiles to allow careful selection of appropriate treatment options in a specialized clinical setting - supporting an interdisciplinary approach providing modern and individualized treatment for patients with gastrointestinal neoplasia.

Competing interests: None

References

• 1 Quera R, O’Sullivan K, Quigley E A. Surveillance in Barrett’s esophagus: will a strategy focused on a high-risk group reduce mortality from esophageal adenocarcinoma?.  Endoscopy. 2006;  38 162-169
• 2 Kakushima N, Yahagi N, Fujishoro M. et al . The efficacy and safety of endoscopic submucosal dissection for esophagogastric junctional tumors.  Endoscopy. 2006;  38 170-174
• 3 Bollschweiler E, Baldus S E, Schröder W. et al . High rate of lymph-node metastasis in submucosal esophageal squamous-cell carcinoma and adenocarcinoma.  Endoscopy. 2006;  38 149-156
• 4 Shami V, Villaverde A, Stearns L. et al . Impact of EUS and EUS-FNA in patients with Barrett’s esophagus and high-grade dysplasia or intramucosal carcinoma referred for endoscopic therapy.  Endoscopy. 2006;  38 157-161
• 5 Muto M, Miyamoto S, Hosokawa A. et al . Endoscopic mucosal resection in the stomach using the insulated-tip needle-knife.  Endoscopy. 2005;  37 178-182
• 6 Makuuchi H, Yoshida T, Ell C. Four-step endoscopic esophageal mucosal resection (EEMR) tube method of resection for early esophageal cancer.  Endoscopy. 2004;  36 1013-1018
• 7 Rösch T, Sarbia M, Schumacher B. et al . Attempted endoscopic en bloc resection of mucosal and submucosal tumors using insulated-tip knives: a pilot series.  Endoscopy. 2004;  36 788-801
• 8 Ono H, Kondo H, Gotoda T. et al . Endoscopic mucosal resection for treatment of early gastric cancer.  Gut. 2001;  48 225-229
• 9 Giovannini M, Bories E, Pesenti C. et al . Circumferential endoscopic mucosal resection in Barrett’s esophagus with high-grade intraepithelial neoplasia or mucosal cancer: preliminary results in 21 patients.  Endoscopy. 2004;  36 782-787.
• 10 Seewald S, Akaraviputh T, Seitz U. et al . Circumferential EMR and complete removal of Barrett’s epithelium: a new approach to management of Barrett’s esophagus containing high-grade intraepithelial neoplasia and intramucosal carcinoma.  Gastrointest Endosc. 2003;  57 854-859
• 11 Vieth M, Ell C, Gossner L. et al . Histological analysis of endoscopic resection specimens from 326 patients with Barrett's esophagus and early neoplasia.  Endoscopy. 2004;  36 776-781
• 12 Kim D Y, Joo J K, Ryu S Y. et al . Factors related to lymph node metastasis and surgical strategy used to treat early gastric carcinoma.  World J Gastroenterol. 2004;  10 737-740
• 13 Pelz J, Merkel S, Horbach T. et al . Determination of nodal status and treatment in early gastric cancer.  Eur J Surg Oncol. 2004;  30 935-941
• 14 Park D J, Lee H K, Lee H J. et al . Lymph node metastasis in early gastric cancer with submucosal invasion: feasibility of minimally invasive surgery.  World J Gastroenterol. 2004;  10 3549-3552
• 15 Abe N, Sugiyama M, Masaki T. et al . Predictive factors for lymph node metastasis of differentiated submucosally invasive gastric cancer.  Gastrointest Endosc. 2004;  60 242-245
• 16 Choi H J, Kim Y K, Kim Y H. et al . Occurrence and prognostic implications of micrometastases in lymph nodes from patients with submucosal gastric carcinoma.  Ann Surg Oncol. 2002;  9 13-19
• 17 Folli S, Morgagni P, Roviello F. et al . Risk factors for lymph node metastases and their prognostic significance in early gastric cancer (EGC) for the Italian Research Group for Gastric Cancer (IRGGC).  Jpn J Clin Oncol. 2001;  31 495-499
• 18 Yamada H, Nihei Z, Yamashita T. et al . Is lymphadenectomy needed for all submucosal gastric cancers?.  Eur J Surg. 2001;  167 199-203
• 19 Gotoda T, Yanagisawa A, Sasako M. et al . Incidence of lymph node metastasis from early gastric cancer: estimation with a large number of cases at two large centers.  Gastric Cancer. 2000;  3 219-225
• 20 Borie F, Millat B, Fingerhut A. et al . Lymphatic involvement in early gastric cancer: prevalence and prognosis in France.  Arch Surg. 2000;  135 1218-1223
• 21 Ishigami S, Natsugoe S, Hokita S. et al . Carcinomatous lymphatic invasion in early gastric cancer invading into the submucosa.  Ann Surg Oncol. 1999;  6 286-289
• 22 Yasuda K, Shiraishi N, Suematsu T. et al . Rate of detection of lymph node metastasis is correlated with the depth of submucosal invasion in early stage gastric carcinoma.  Cancer. 1999;  85 2119-2123
• 23 Kurihara N, Kubota T, Otani Y. et al . Lymph node metastasis of early gastric cancer with submucosal invasion.  Br J Surg. 1998;  85 835-839
• 24 Stein H J, Feith M, Bruecher B L. et al . Early esophageal cancer: pattern of lymphatic spread and prognostic factors for long-term survival after surgical resection.  Ann Surg. 2005;  242 566-575
• 25 Westerterp M, Koppert L B, Buskens C J. et al . Outcome of surgical treatment for early adenocarcinoma of the esophagus or gastro-esophageal junction.  Virchows Arch. 2005;  446 497-504
• 26 Buskens C J, Westerterp M, Lagarde S M. et al . Prediction of appropriateness of local endoscopic treatment for high-grade dysplasia and early adenocarcinoma by EUS and histopathologic features.  Gastrointest Endosc. 2004;  60 703-710
• 27 Rice T W, Blackstone E H, Goldblum J R. et al . Superficial adenocarcinoma of the esophagus.  J Thorac Cardiovasc Surg. 2001;  122 1077-1090
• 28 Van Sandick J W, van Lanschot J J, ten Kate F J. et al . Pathology of early invasive adenocarcinoma of the esophagus or esophagogastric junction: implications for therapeutic decision making.  Cancer. 2000;  88 2429-2437
• 29 Ruol A, Merigliano S, Baldan N. et al . Prevalence, management and outcome of early adenocarcinoma (pT1) of the esophago-gastric junction: comparison between early cancer in Barrett’s esophagus (type I) and early cancer of the cardia (type II).  Dis Esophagus. 1997;  10 190-195
• 30 Yoshida M, Hanashi T, Momma K. et al . Endoscopic mucosal resection for radical treatment of esophageal cancer. [In Japanese.  ] Gan To Kagaku Ryoho. 1995;  22 847-854
• 31 Adachi Y, Shiraishi N, Kitano S. Modern treatment of early gastric cancer: review of the Japanese experience.  Dig Surg. 2002;  19 333-339
• 32 Yanai H, Noguchi T, Mizumachi S. et al . A blind comparison of the effectiveness of endoscopic ultrasonography and endoscopy in staging early gastric cancer.  Gut. 1999;  44 361-365
• 33 Akahoshi K, Chijiwa Y, Hamada S. et al . Pretreatment staging of endoscopically early gastric cancer with a 15 MHz ultrasound catheter probe.  Gastrointest Endosc. 1998;  48 470-476
• 34 Akahoshi K, Chijiiwa Y, Sasaki I. et al . Pre-operative TN staging of gastric cancer using a 15 MHz ultrasound miniprobe.  Br J Radiol. 1997;  70 703-707
• 35 Hizawa K, Iwai K, Esaki M. et al . Is endoscopic ultrasonography indispensable in assessing the appropriateness of endoscopic resection for gastric cancer?.  Endoscopy. 2002;  34 973-978
• 36 Yoshida S, Tanaka S, Kunihiro K. et al . Diagnostic ability of high-frequency ultrasound probe sonography in staging early gastric cancer, especially for submucosal invasion.  Abdom Imaging. 2005;  30 518-523
• 37 May A, Gunter E, Roth F. et al . Accuracy of staging in early oesophageal cancer using high resolution endoscopy and high resolution endosonography: a comparative, prospective, and blinded trial.  Gut. 2004;  53 634-640
• 38 Hamada S, Akahoshi K, Chijiwa Y. et al . Relationship between histological type and endosonographic detection of regional lymph node metastases in gastric cancer.  Br J Radiol. 1997;  70 697-702
• 39 Nakamura K, Morisaki T, Noshiro H. et al . Morphometric analysis of regional lymph nodes with and without metastasis from early gastric carcinoma.  Cancer. 2000;  88 2438-2442
• 40 Vieth M, Stolte M. Pathology of early gastrointestinal neoplasia.  Best Pract Res Clin Gastroenterol. 2005;  19 857-869
• 41 Schmumacher B, Vieth M. How can the pathological examination of a mucosectomy specimen be improved?.  Acta Endosc. 2006;  [in press]