Sie sind auf Seite 1von 9

Update on the Paris Classification of Superficial Endoscopic Classification Review

Neoplastic Lesions in the Digestive Tract Group


Review

Background and Study Aims: Neoplastic lesions in the diges− pearance is that of a superficial lesion. Type 0 is divided into
tive−tract mucosa are termed “superficial” when the depth of in− three categories: protruding (0 ± I), nonprotruding and nonexca−
vasion is limited to the mucosa and submucosa. The endoscopic vated (0 ± II), and excavated (0 ± III). Type 0 ± II lesions are then
appearance has a predictive value for invasion into the submuco− subdivided into slightly elevated (IIa), flat (IIb), or depressed

Downloaded by: University of Connecticut. Copyrighted material.


sa, which is critical for the risk of nodal metastases. (IIc). Nonprotruding depressed lesions are associated with a
Materials and Methods: The endoscopic morphology of superfi− higher risk of submucosal invasion. After endoscopic resection,
cial lesions can be assessed with a standard video endoscope invasion into the submucosa is an important criterion for the ne−
after spraying of a dye ± an iodine−potassium iodide solution for cessity of additional surgical resection. Micrometer analysis of
the stratified squamous epithelium, or an indigo carmine solu− the depth of invasion in the specimen is more precise, and dis−
tion for the columnar epithelium. In 2002, a workshop was held tinct cut−off limits have been established in the esophagus,
in Paris to explore the relevance of the Japanese classification. stomach, and large bowel.
The conclusions were revised in 2003 in Osaka in relation to the Conclusions: The morphology of superficial and nonprotruding
definition of the subtypes used in endoscopy and the evaluation neoplastic lesions is relevant to the prognosis. Following endo−
of the depth of invasion into the submucosa. In Japan, the de− scopic detection, the lesions are analyzed using chromoendoscopy
scription of advanced cancer in the digestive−tract mucosa using and assigned a subtype of the type 0 classification. The choice be−
570 types 1 ± 4 is supplemented by a type 0 when the endoscopic ap− tween endoscopic or surgicaltreatment isbased on this description.

Introduction applied to the esophagus and large bowel as well. In Western


countries, specific attention was given during the same period
Tumors in the esophagus, stomach, and large bowel represent a to colorectal cancer. Many Western specialists have in the past
substantial proportion of the burden of cancer mortality. Sec− considered that nonprotruding precursors only play a small role
ondary prevention of cancer requires early detection at the stage in the large bowel and that the Japanese classification of “super−
when the lesion is still curable. Endoscopy provides the most ef− ficial” neoplasia into multiple subtypes is too complex for practi−
fective method of detecting premalignant and malignant precur− cal use. However, the eastern and western points of view have
sors of advanced cancer in which the tumor process is restricted now moved much closer together:
to the superficial layers of the digestive−tract wall. To ensure re− ± Nonprotruding precursors of advanced cancer play a major
liable detection, consensus guidelines on the morphological clas− role in Barrett’s esophagus.
sification of the endoscopic appearance of these precursors need ± The existence of potentially malignant, nonprotruding lesions
to be adopted throughout the world. in the large bowel is increasingly being accepted in the West.
± A consensus histopathological classification of epithelial neo−
In Japan, the prevention of gastric cancer has been a medical plastic lesions in the digestive−tract mucosa [1, 2] has been
priority for the last 40 years. The gross morphological classifica− adopted by many Asian, European, and American patholo−
tion used for superficial tumors in the stomach soon came to be gists.

Corresponding Author
R. Lambert, M. D., F.R.C.P. ´ International Agency for Research on Cancer ´ 150, cours Albert Thomas ´
Lyon 69372, Cedex 08 ´ France ´ Fax: +33−4−72 73 86 50 ´ E−mail: lambert@iarc.fr

Bibliography
Endoscopy 2005; 37 (6): 570±578  Georg Thieme Verlag KG Stuttgart ´ New York ´ ISSN 0013−726X
DOI 10.1055/s−2005−861352
± The clinical relevance of the Japanese endoscopic classifica− mucosal carcinoma” (subgroups 4−3 and 4−4). Low−grade and
tion of superficial neoplasia in the digestive tract was con− high−grade dysplasia and adenoma are equivalent terms to low−
firmed in a workshop held in Paris in 2002 by a group of spe− grade and high−grade noninvasive intraepithelial neoplasia.
cialists. The conclusions were recently published, including a High−grade intraepithelial lesions have the potential to progress
descriptive atlas of endoscopic images [3]. A further meeting after a short delay into submucosal or advanced carcinoma, as
was organized in Osaka in 2003, and a simplified and updated shown in the follow−up of untreated lesions in the gastric muco−
classification is now published here. sa in Japan [6]. In summary, all high−grade neoplastic lesions (in−
vasive and noninvasive) are now included in subdivisions of the
same group 4.
Premalignant Conditions and Premalignant Lesions
Two other terms apply to the description of superficial malignant
As a rule, discrete epithelial premalignant lesions in the diges− epithelial lesions: “de novo” cancer corresponds to small (often
tive−tract mucosa are the first step in a progression to superficial less than 5 mm), flat or depressed cancerous lesions when there
malignancy and then to advanced cancer. Premalignant lesions are no adenomatous glands in the specimen, suggesting that the

Review
often develop against a background of inflammation and diffuse carcinoma has not developed from an adenomatous precursor or
alterations in the digestive−tract mucosa; these are considered to that it has rapidly lost evidence of its adenomatous origin. “Early
be risk factors for cancer and are termed “premalignant condi− cancer” is a superficial tumor with a high potential for cure. Two
tions”. Premalignant conditions in the digestive−tract mucosa in− criteria are involved: depth of invasion and lymph−node status.
clude lesions in the squamous epithelium that occur in some re− In surgical specimens, using the pTNM classification, this corre−
gions of Asia, columnar metaplasia in the esophagus, chronic sponds to an N0 intramucosal tumor in the stratified squamous
gastritis associated with Helicobacter pylori infection, chronic in− epithelium of the esophagus and to an intramucosal or submu−

Downloaded by: University of Connecticut. Copyrighted material.


flammation in ulcerative colitis, and intestinal metaplasia in the cosal N0 or N1 tumor in the columnar epithelium of the stomach
esophagus or stomach. Distinct characteristics have been attrib− or large bowel. By extension, “early cancer” is used for the endo−
uted to specialized intestinal metaplasia in the esophagus and to scopic description when there is no evidence of lymph−node in−
intestinal metaplasia at the cardia and in the stomach [4]; how− vasion after an endoscopic ultrasound examination.
ever, the specificity of these has been challenged [5].
Hyperplastic, inflammatory, and juvenile polyps are nonneoplas−
tic lesions; their potential for progression to neoplasia is nil or
Superficial Epithelial Neoplastic Lesions very small, while serrated adenomas harbor a neoplastic compo−
in the Digestive−Tract Mucosa nent. Nonneoplastic polyps are frequent in the colon [7].

Gross Morphology
Neoplastic lesions in the esophagus, stomach, and large bowel Methodology for Endoscopic Detection of Superficial 571
are called superficial when their endoscopic appearance sug− Neoplastic Lesions
gests that invasion is limited to the mucosa and submucosa. Su−
perficial, protruding lesions are also termed “polyps”; excavated Chromoendoscopy
lesions are called “ulcers”. Less conspicuous lesions (nonprotrud− The first step in the endoscopic exploration of the digestive−tract
ing and nonexcavated) are often called “flat”, but most are slight− mucosa is to identify slightly discolored areas and irregularities
ly elevated or slightly depressed. in the surface or in the microvascular network. Precise analysis
of the lesion with assessment of its boundaries and classification
Histopathology of its morphology requires chromoendoscopy, using an iodine±
Epithelial tumors developing in the digestive tract include in− potassium iodide solution (1.5 ± 2.0 %) [8] for the squamous stra−
traepithelial neoplasia (adenomas, dysplasia, premalignancy) in tified epithelium in the esophagus and an indigo carmine solu−
the mucosa and invasive lesions classified as intramucosal or tion (0.5 ± 1.0 %) [9,10] for the mucosa in the stomach and intes−
submucosal carcinomas. The progression of intraepithelial neo− tine. Other dyes include methylene blue [11 ± 13] (1 % solution),
plasia to invasive carcinoma is not invariable, and the delay can which stains intestinal metaplasia in the esophagus or stomach,
be short or long. The Vienna classification of epithelial neoplasia and acetic acid [14] (a fresh 3 % solution), which enhances the ar−
in the mucosa of the esophagus, stomach, and intestines (compa− chitecture of the columnar metaplasia in Barrett’s esophagus.
tible with the World Health Organization classification) now
provides a consensus terminology for Western and Asian specia− Magnifying Endoscopy
lists. The revised edition of the classification [2] features five dis− Recent video endoscopes with high−quality digital imaging meet
tinct groups, and neoplastic intramucosal lesions confirmed by all the requirements for the detection of superficial neoplastic le−
the pathologist are classified into groups 3 and 4, while submu− sions and for their classification into subtypes. If the endoscope
cosal carcinoma is in group 5. is equipped with an optical zoom ( ” 80), more or less reliable
prediction of the histological findings (optical biopsy) is possible
Group 3 includes premalignant lesions, termed “low−grade non− [15]. Magnification endoscopy with a contrast dye reveals the
invasive intraepithelial neoplasia”. Group 4 includes premalig− surface architecture of the epithelium. Magnification endoscopy
nant lesions, termed “high−grade noninvasive intraepithelial in transparency allows exploration of the microvascular network
neoplasia” (subgroups 4−1 and 4−2) and malignant lesions, [16 ± 19]. Two specific applications of magnification endoscopy
termed “invasive high−grade intraepithelial neoplasia” or “intra− are detection of intestinal metaplasia in Barrett’s esophagus

Lambert R et al. Superficial Neoplastic Lesions in the Digestive Tract ´ Endoscopy 2005; 37: 570 ± 578
[20 ± 22] and analysis of the pit pattern of the colonic epithelium
[23 ± 27]. The latter analysis method is also used for serrated ade− Table 1 The macroscopic classification of digestive−tract cancer
nomas [28, 29]. used in Japan

Superficial type 0 Superficial protruding or nonprotruding lesions


Endoscopic Classification in Type 0
Advanced type 1 Protruding carcinoma, attached on a wide base
The Japanese Gastric Cancer Association (Table 1) has published
Advanced type 2 Ulcerated carcinoma with sharp and raised margins
guidelines for the classification of stomach cancer [30]. Types 1 ± Advanced type 3 Ulcerated carcinoma without definite limits
5 characterize advanced cancer and are derived from Borrmann’s Advanced type 4 Nonulcerated, diffusely infiltrating carcinoma
description. Type 0 was added for superficial cancer. Type 0 is Advanced type 5 Unclassifiable advanced carcinoma
divided into three categories corresponding to protruding lesions
(0 ± I), nonprotruding and nonexcavated lesions (0 ± II), and exca−
vated lesions (0 ± III) (Table 2, Figures 1 ± 5). Type 0 ± I is subdivid−
ed into pedunculated (0 ± Ip) and sessile (0 ± Is) lesions (Figure 1).
Type 0 ± II is divided into three subtypes ± a, b, and c ± cor− Table 2 The macroscopic classification of type 0 digestive−tract
Review

responding to slightly elevated, flat, and depressed lesions (Fig− lesions, with a superficial appearance at endoscopy
ure 2). Type 0 ± III is an ulcer (Figure 3). Mixed patterns with ele−
vation and depression also occur and are classified into two Protruding
Pedunculated 0 ± Ip
groups: in 0 ± IIc + 0 ± IIa lesions, most of the surface is depressed; Sessile 0 ± Is
elevation is present in a segment of the lesion at the periphery; Nonprotruding and nonexcavated
in 0 ± IIa + 0 ± IIc lesions, there is a central depression in a globally Slightly elevated 0 ± IIa
Completely flat 0 ± IIb
elevated lesion (Figure 4). The combined patterns of excavation Slightly depressed 0 ± IIc

Downloaded by: University of Connecticut. Copyrighted material.


and depression are termed 0 ± III + IIc or 0 ± IIc + III, depending Elevated and depressed types 0 ± IIc + IIa
on the respective surface of the ulcer and of the depressed area 0 ± IIa + IIc
(Figure 5). Excavated
Ulcer 0 ± III
Excavated and depressed types 0 ± IIc + III
The subtype is decided only on the basis of the endoscopic ap− 0 ± III + IIc
pearance and should not be changed after the decision has been
taken. The classification into subtypes provides a controlled en−
doscopic description of superficial gastric cancer. A similar clas−
sification of type 0 lesions was later also adopted for the esopha− The pathology report will confirm the safety of a local excision or
gus [31] and the large bowel. recommend additional surgical resection. The decision is based
on qualitative criteria (tumor grade, vascular invasion, tumor
572 The classification helps predict the extent of invasion into the budding, cribriform pattern) and quantitative criteria (width
submucosa and thus the choice between endoscopic or surgical and depth of invasion into the submucosa).
treatment. A prediction of deep invasion is further confirmed if
the lesion fails to lift after injection of saline into the submucosa The depth of invasion into the submucosa correlates with the
(the nonlifting sign). The description of superficial neoplastic le− risk of lymph−node metastases [32 ± 36]. In surgically resected
sions in the digestive−tract mucosa requires an assessment of the specimens, the full thickness of the digestive−tract wall is avail−
size and the subtype. The diameter of the lesion is estimated as able, and invasion into the submucosa is often evaluated by the
precisely as possible, preferably using a graduated gauge. The pathologist using a semiquantitative method, with an artificial
elevation or depression of the suspect area is estimated in com− division into two layers (sm1 and sm2) or even three layers
parison with the adjacent mucosa. A biopsy forceps placed long− (sm1, sm2, sm3), ranging from superficial to deep. In specimens
itudinally next to the lesion is a helpful reference standard (the resected at endoscopy, semiquantitative evaluation of the depth
diameter with closed jaws is 2.5 mm, while the diameter of a sin− of invasion is not fully reliable. Quantitative micrometer (m)
gle jaw is 1.2 mm). The subtypes listed in Table 2 apply to the measurement from the lower limit of the muscularis mucosae is
esophagus, stomach, and intestines; the relative frequency of more precise when the position of the muscularis mucosae can
each subtype varies depending on the site, and an atlas of the en− be determined with precision in the area adjacent to the tumor.
doscopic appearances has been published [3].
Micrometer measurement should be generally used to allow
comparison of outcomes after either surgery or endoscopic mu−
Handling and Processing of Mucosectomy Specimens cosal resection (EMR). With regard to the depth of invasion into
the submucosa, the empirical cut−off limit adopted in Japan for
In the pathology laboratory, the analysis of the specimen obtain− safety after mucosectomy (EMR) varies depending on the site of
ed after “en−bloc” or “piecemeal” mucosectomy must be able to: the lesion. It is estimated as less than 200 m in the esophagus,
± Confirm the morphology of the lesion as type 0 and subtypes. 500 m in the stomach, and 1000 m in the large bowel (Fig−
± Confirm whether it is low−grade or high−grade neoplasia. ures 6 ± 8). In the large bowel, this cut−off value only applies to
± Assess the completeness of the resection at the margins of the sessile lesions.
specimen.
± Evaluate any invasion into the submucosa and its depth.

Lambert R et al. Superficial Neoplastic Lesions in the Digestive Tract ´ Endoscopy 2005; 37: 570 ± 578
0–II c + IIa

0–I p

0–II a + II c

0–I s

Figure 1 Endoscopic appearance of a superficial neoplastic lesion at 0–II a + II c


the surface of the digestive−tract mucosa: protruding type,
pedunculated (0 ± Ip) and sessile (0 ± Is). The distinction between a ses−
Figure 4 Endoscopic appearance of a superficial neoplastic lesion on
sile (protruding) lesion and a slightly elevated (nonprotruding) lesion is
the surface of the digestive−tract mucosa: elevation plus depression.
based on the extent of the elevation from the adjacent mucosa. The

Review
Type 0 ± IIc + IIa is a depressed lesion with an elevation in part of the
cut−off limit is 2.5 mm in the columnar epithelium and 1.2 mm in the
peripheral ring. Type 0 ± IIa + IIc is an elevated lesion with a central de−
stratified epithelium of the esophagus.
pression. The central depression is surrounded by an elevated ring.
When the level of the depression is higher than the mucosa adjacent
to the lesion, it is a relatively depressed lesion.

0–II a

Downloaded by: University of Connecticut. Copyrighted material.


0–II c + III
0–II b

0–II c 0–III + II c

Figure 2 Endoscopic appearance of a superficial neoplastic lesion on Figure 5 Endoscopic appearance of a superficial neoplastic lesion on
the surface of the digestive−tract mucosa: nonprotruding and nonexca− the surface of the digestive−tract mucosa; ulcer plus depression. Type
vated types, slightly elevated (0 ± IIa), completely flat (0 ± IIb), or slight− 0 ± IIc + III is a depressed lesion with a central ulcer. Type 0 ± III + IIc is an
ly depressed (0 ± IIc). The distinction between a slightly depressed le− ulcer with short depressed margins.
sion and an excavated lesion is based on the depth of the depression
from the adjacent mucosa. The cut−off limit is 1.2 mm in the columnar
epithelium and 0.5 mm in the stratified epithelium of the esophagus. 573
sed 0 ± IIc subtype is the most frequent (45 % of all lesions), and
excavated lesions (0 ± III) are uncommon (Table 3).

Pathology Control
Invasion into the submucosa occurs more frequently in protrud−
0–III ing lesions (0 ± I) or excavated lesions (0 ± III); the lowest risk is
with completely flat (0 ± IIb) lesions (Table 4). In the same Japa−
Figure 3 Endoscopic appearance of a superficial neoplastic lesion on
the surface of the digestive−tract mucosa: excavated type (0 ± III). An nese series including 1562 patients [38], the risk of metastatic
ulcer is seen. lymph nodes was assessed in relation to the depth of invasion of
lesions classified into three groups: superficial in the mucosa
(m1 or intraepithelial + m2 or microinvasive); intermediate (m3
Neoplasia in the Squamous Epithelium of the Esophagus + sm1); deep in the submucosa (sm2 + sm3). The corresponding
proportions of lymph−node metastases in the three groups were:
Endoscopic Appearance 2 %, 19 %, and 44 %. Superficial invasion (m1 + m2), which is a safe
In the squamous stratified epithelium, the elevation or depres− indication for endoscopic treatment, was found in 69 % of 0 ± IIb
sion of a superficial neoplastic lesion is estimated with a specific lesions, 39 % of 0 ± IIc lesions, 20 % of 0 ± IIa lesions, and almost
millimeter scale [3, 37], using a single jaw of a biopsy forceps (1.2 never for 0 ± I and 0 ± III lesions.
mm) as a calibrating gauge. Protruding 0 ± I lesions rise more
than 1.2 mm above the adjacent surface. Lesions less elevated In conclusion, endoscopic treatment is recommended for nonde−
than the gauge are classified as 0 ± IIa. Lesions less deep than pressed neoplastic lesions up to a diameter of 20 mm and up to
0.5 mm are classified as slightly depressed (0 ± IIc). Lesions deep− 10 mm for depressed lesions. In the histopathological assess−
er than 0.5 mm are classified as excavated (0 ± III). In the esopha− ment, a cut−off limit for the depth of invasion into the submucosa
gus, protruding lesions are not frequent and have a sessile mor− is 200 m [32].
phology (0 ± Is). Most neoplastic lesions (75 %) are nonprotrud−
ing, as shown by the analysis of 1562 surgical specimens of “su−
perficial” cancer from 143 institutions in Japan [38]: the depres−

Lambert R et al. Superficial Neoplastic Lesions in the Digestive Tract ´ Endoscopy 2005; 37: 570 ± 578
1 2 3 4 5
Table 3 Relative proportions (%) of subtypes in type 0 lesions in Ja−
epithelium pan. Series based on endoscopic analysis and a pathology
specimen
lamina propria
musc. mucosae Subtype Esophagus 1 Stomach 2 Large bowel 3
(n = 1562) (n = 2098) (n = 3644)

200 µ submucosa 0±I * 16 % 3% 50 %


0 ± IIa 20 % 17 % 44 %
muscul. propria
0 ± IIb 14 % < 1% < 1%
0 ± IIc ** 45 % 78 % 5%
Figure 6 Esophagus: diagram of a pathology specimen, showing the
depth of invasion of superficial neoplastic lesions into the stratified epi− 0 ± III 5% < 1% < 1%
thelium. 1, intraepithelial neoplasia (or m1); 2, microinvasive cancer
1
with involvement of the basal membrane of the epithelium (or m2); Multicenter analysis by Kodama and Kakegawa [38]. 2 National Cancer Center,
Tokyo, 1990 ± 1999 [3]. 3 Niigata Hospital [3]. * All types Ip and s. ** Includes sim−
Review

3, intramucosal cancer (or m3); 4, cancer with superficial invasion of ple and combined IIc types.
the submucosa (sm1); 5, cancer with deep invasion of the submucosa
(sm2 or sm3). Endoscopic mucosectomy is fully justified in situations 1
or 2. The indication is debatable in situations 3 and 4, and inappropri−
ate in situation 5, when the depth of invasion into the submucosa is
200 m.
Table 4 Relative frequency (%) of submucosal invasion in subtypes
of type 0 lesions in Japan. Series based on endoscopic
analysis and a pathology specimen

Downloaded by: University of Connecticut. Copyrighted material.


1 2 3
Subtype Esophagus 1 Stomach 2 Large bowel 3
(n = 1562) (n = 2098) (n = 3644)
epithelium
% n

musc. mucosae 0±I * 79 % 57 %


0 ± Ip 5% 69/1303
0 ± Is 34 % 185/504
500 µ submucosa 0 ± IIa 48 % 29 % 4% 64/160
0 ± IIb 15 % 20 % 0 ±
muscul. propria
0 ± IIc ** 27 % 40 % 61 % 23/200
0 ± III 84 % ± ±
Figure 7 Stomach: diagram of a pathology specimen, showing the
depth of invasion of superficial neoplastic lesions into the columnar 1
Multicenter analysis by Kodama and Kakegawa [38]. 2 National Cancer Center,
574 epithelium. 1, intraepithelial or intramucosal neoplasia; 2, cancer with Tokyo, 1990 ± 1999 [3]. 3 Niigata Hospital [3]. * All types Ip and s. ** Includes sim−
superficial invasion of the submucosa (sm1); 3, cancer with deep inva− ple and combined IIc types.
sion of the submucosa (sm2). Endoscopic mucosectomy is justified
when the depth of invasion in the submucosa is 500 m or less (situa−
tion 1 or 2).
Neoplasia at the Esophagogastric Junction

1 2 3 Superficial neoplastic lesions that develop in the columnar epi−


thelium at the esophagogastric junction, including the cardia,
epithelium
have a similar morphology to those that develop in Barrett’s
musc. mucosae esophagus. The elevation or depression of the neoplastic lesion
is measured with a millimeter scale for all sites in the columnar
epithelium [3] using the closed jaws of a biopsy forceps (2.5 mm)
1000 µ submucosa as a calibrating gauge. This height is the cut−off limit between
protruding and nonprotruding elevated lesions. There are few
muscul. propria
data on the distribution of the subtypes of type 0, and further
morphological studies are needed. Protruding sessile and slight−
Figure 8 Large bowel: diagram of the pathology specimen, showing ly elevated lesions are more frequent than depressed (0 ± IIc) le−
the depth of invasion of superficial neoplastic lesions into the columnar
sions. In a series of 66 type 0 lesions in the esophagus [39], the
epithelium. 1, intraepithelial or intramucosal neoplasia; 2, cancer with
superficial invasion of the submucosa (sm1); 3, cancer with deep inva− respective proportions of protruding, nonprotruding, and exca−
sion of the submucosa (sm2 or sm3). Endoscopic mucosectomy is jus− vated lesions were found to be 21 %, 74 %, and 5 %. At the cardia,
tified when the depth of invasion into the submucosa is 1000 m or depressed lesions (0 ± IIc) were less frequent (13 %) than in the
less (situation 1 or 2). subcardial stomach [40]. It is assumed that endoscopic treat−
ment of these lesions is justified when the depth of invasion
into the submucosa is less than 500 m.

Lambert R et al. Superficial Neoplastic Lesions in the Digestive Tract ´ Endoscopy 2005; 37: 570 ± 578
Neoplasia in the Stomach (Cardia Excluded) Neoplasia in the Large Bowel

Endoscopic Appearance Endoscopic Appearance


The elevation or depression of superficial lesions is evaluated The elevation or depression of superficial lesions is assessed with
with a millimeter scale for the columnar epithelium: protruding a calibrating gauge placed along the adjacent surface, using the
neoplastic lesions (0 ± I) rise more than 2.5 mm above the adja− scale of the columnar epithelium. Some nonprotruding elevated
cent surface; lesions lower than this limit are classified as 0 ± IIa. colonic lesions have been called “laterally spreading” lesions,
To estimate the depth of depression, the reference scale is a sin− when their diameter is more than 20 mm. Slightly depressed le−
gle jaw of the biopsy forceps (approximately 1.2 mm). Lesions sions (0 ± IIc) in the colon also have specific characteristics. They
less deep than 1.2 mm are classified as depressed (0 ± IIc). Lesions are termed “absolutely depressed” when the surface is at a lower
deeper than this value are classified as excavated (0 ± III). level than that of the adjacent mucosa and the entire thickness of
the mucosa is reduced. They are called “relatively depressed”
In the distal stomach, protruding lesions are sessile but rare (0 ± when the shallow depression at the top of an elevation is still
Is); less than 10 % of polypoid adenomas show malignant foci above the surface of the adjacent mucosa. Combined types are

Review
[41]. Lesions combining elevation and depression are classified classified either as 0 ± IIc + IIa (depressed lesions with elevated
into two groups: depressed lesions in which most of the surface margins) or 0 ± IIa + IIc (elevated lesions with a central depres−
is depressed and there is elevation in a portion of the peripheral sion). Relatively depressed lesions are a variant of the IIa + IIc le−
ring are classified as 0 ± IIc + 0 ± IIa, while elevated lesions with a sions.
central depression encircled by the elevated ring at the periphery
are called 0 ± IIa + 0 ± II c (Figure 4). The combined patterns of ex− In the published series [3] from Niigata Hospital, protruding
cavation and depression are called 0 ± III + IIc or 0 ± IIc + III, de− polyps represented approximately 50 % of all lesions (Table 3);

Downloaded by: University of Connecticut. Copyrighted material.


pending on the relative surface area of the ulcer and of the de− they can be pedunculated (0 ± Ip) or sessile (0 ± Is). The other
pressed area (Figure 5). 50 % of the lesions are nonprotruding (Table 3); most of these
(44 % of all cases) are elevated (0 ± IIa). Completely flat lesions
In a previously reported series of 2098 superficial cancers at the (0 ± IIb) are very rare, and depressed lesions (0 ± IIc) are not fre−
National Cancer Center Hospital in Tokyo [3], most lesions (95 %) quent (5 %). Excavated lesions (0 ± III) almost never occur. Similar
were found to be nonprotruding, and the depressed subtype (0 ± figures were reported in the series (9533 cases) in the Akita Red
IIc) was the most frequent (78 % of cases; Table 3). Cross Hospital [9], in which the proportion of nonprotruding le−
sions was 43 %.
Pathology Control
Invasion of the submucosa occurred in 38 % of cases in the same Pathology Control
series from the National Cancer Center Hospital in Tokyo. The Invasion of the submucosa. In sessile protruding lesions, the
highest risk of invasion is seen in protruding (0 ± I) or depressed frequency of submucosal invasion is high (34 %), as seen in Ta− 575
(0 ± II c) lesions (Table 4). A similar distribution was reported for ble 4. In nonprotruding lesions, the highest risk occurs with de−
3223 type 0 gastric cancers in the files of the National Mass pressed (0 ± IIc) lesions; it reached 61 % in the series of pathology
Screening program in Japan [42]. The relation between the depth specimens from Niigata Hospital (Table 4). The risk is still high in
of invasion into the submucosa and lymph−node metastases was depressed lesions of small size, as shown in the previously re−
analyzed in 1091 cases at the National Cancer Center Hospital in ported series from the Akita and Yokoyama Hospitals [3]: inva−
Tokyo. When the invasion was less than the cut−off limit of sion into the submucosa occurred in 44 % of small (6 ± 10 mm)
500 m, the proportion was only 6 %; beyond this limit it in− 0 ± IIc lesions.
creased to 21 %.
Lymph−node metastases. Lymph−node metastases correlate
In conclusion, endoscopic treatment is recommended for super− with the depth of invasion into the submucosa. In the same se−
ficial and well−differentiated neoplastic lesions up to a diameter ries from the Akita and Yokoyama Hospitals [3], the frequency
of 20 mm for nondepressed lesions and up to 10 mm for depres− of nodal metastases was less than 1 % in superficial submucosal
sed lesions. In the histopathological assessment, in the absence invasion (sm1), 6 % (seven of 105 patients) in medium invasion
of adverse qualitative criteria, the cut−off limit for the depth of (sm2), and reached 14 % (10 of 71) in deep invasion (sm3).
invasion into the submucosa is 500 m. Lymph−node metastases are rare if the depth of invasion into
the submucosa is less than 1000 m based on a precise microme−
ter measurement [32, 33]. In protruding lesions, the risk of nodal
Neoplasia in the Small Intestine metastases is low for the pedunculated subtype, even if there is
invasion of the submucosa. For the sessile protruding subtype,
Lesions readily accessible to endoscopic exploration and treat− the depth of submucosal invasion has better prognostic value:
ment are located in the duodenum; data on the morphological depth of invasion measured at more than 1000 m or classified
classification are scarce, and there have been no large series. Ele− as sm3 is an argument for additional surgery.
vation or depression in neoplastic lesions with a superficial ap−
pearance is assessed using the same criteria as in the distal stom− In conclusion, endoscopic treatment of superficial and nonpro−
ach. Most lesions of the protruding (0 ± Is) or nonprotruding (0 ± truding neoplastic lesions is recommended up to a diameter of
IIa) types are sessile; depressed lesions exist, but are less fre− 20 mm for well−differentiated nondepressed lesions. Safe resec−
quent. tion of large, laterally spreading lesions is also possible using the

Lambert R et al. Superficial Neoplastic Lesions in the Digestive Tract ´ Endoscopy 2005; 37: 570 ± 578
endoscopic technique of submucosal dissection with a needle− (0 ± IIc) lesions has an impact on the efficacy of cancer preven−
knife. The size limit for safe resection of depressed lesions is es− tion, as these lesions have a high potential for progression to in−
timated at 10 mm. At histopathological assessment, in the ab− vasive malignancy. The precursors of colorectal cancer do not
sence of adverse qualitative characteristics, a cut−off limit of consist exclusively of protruding lesions [48, 49]; a recent study
1000 m for the depth of invasion into the submucosa is a safe from Sweden suggests that more than 40 % of advanced colorec−
guideline. This requires precise identification of the position of tal cancers develop from a nonprotruding precursor [50].
the muscularis mucosae adjacent to the lesion [43]. The cut−off
value applies to sessile protruding lesions and nonprotruding le− For the endoscopist, training and experience depend on constant
sions; more flexible limits are used for polypoid precursors. interaction with the pathologist. In specimens resected at endos−
copy, the depth of invasion into the submucosa is checked and
measured with a micrometer in order to confirm the justification
Discussion for endoscopic treatment.

Quality Assurance in Diagnostic Endoscopy In the upper gastrointestinal tract, most precursors of advanced
Review

Recent video endoscope models with high resolution are suitable cancer have a nonprotruding morphology. In the absence of ad−
for early detection of the nonprotruding precursors of advanced verse qualitative criteria, the risk of lymphatic metastases is ex−
cancer. However, the proportion of false−negative procedures is tremely small when the depth of invasion into the submucosa is
far from zero. Underestimation of the role of nonprotruding le− less than 200 m in the esophagus or 500 m in the stomach.
sions as precursors of advanced colorectal cancer in Western These are the cut−off values for the safety of endoscopic muco−
populations is the result of technical differences rather than eth− sectomy.
nic ones. Nonprotruding depressed 0 ± IIc neoplastic lesions are

Downloaded by: University of Connecticut. Copyrighted material.


invariably neglected if endoscopists do not bear in mind the dis− In the large bowel, the proportion of protruding precursors is
colored spots that correspond to these lesions and if chromoen− around 50 %, and pedunculated polyps account for approximate−
doscopy is not routinely used. False−negative findings in gastro− ly one−third of the precursors of advanced cancer. In the absence
scopy for early gastric cancer have recently been estimated at of adverse qualitative criteria, endoscopic mucosectomy is safe
19 % in Japan [43]. The miss rate for small adenomas (< 1 cm) dur− in nonprotruding lesions when the depth of invasion into the
ing colonoscopy is high in the USA [44] and other countries. It is submucosa is less than 1000 m. For nonprotruding and depres−
currently a widely−held view that missed cancers occur rarely sed lesions, the risk of lymphatic metastases is high, while it is
during the follow−up period after a negative colonoscopy [45]. extremely small in nonprotruding and elevated lesions. The cut−
However, more realistic figures are obtained when the history off value of 1000 m has to be adjusted to some extent for sessile
of patients with a confirmed cancer are investigated. In a cohort protruding lesions with a broad implantation in the submucosa
of 557 patients, 5.2 % were found to have had one or more nega− and for pedunculated lesions with a stalk [51]. In these cases,
576 tive colonoscopies during the 5 years before the diagnosis of can− the extent of the width is taken into account well as the depth,
cer [46]. A recent population study confirmed a 4 % rate of false− and more flexible criteria have been proposed combining two
negative findings for cancer; in 2654 individuals, 105 had had at cut−off values: 2000 m for the depth [35, 36] and 4000 m for
least one colonoscopy within 3 years of their admission for surgi− total submucosal width of the carcinoma [36].
cal resection of a right−sided colon cancer [47].

When a potentially resectable premalignant or malignant lesion Acknowledgments


is detected, two potential erroneous decisions are: unnecessary
resection of a nonneoplastic lesion; and inappropriate excision This collaborative study was supported by an unrestricted edu−
of a carcinoma with deep submucosal invasion. Quality assur− cational grant from the Olympus Medical Systems Corporation.
ance in diagnostic endoscopy can aid in decision−making regard− It is based on a workshop held in Paris in November 2002 and a
ing the appropriate treatment; rigorous analysis of the lesion can workshop held during the Digestive Disease Week meeting in Ja−
serve as a safeguard against erroneous decisions. pan in Osaka in October 2003, and it has received the approval of
the Organisation Mondiale d’Endoscopie Digestive (OMED, World
Classification of Superficial Neoplastic Lesions Organization of Digestive Endoscopy) and Organisation Mondiale
The classification of type 0 superficial neoplastic lesions into de Gastro−EntØrologie (OMGE, World Gastroenterology Organiza−
subtypes provides a safeguard in diagnostic endoscopy against tion).
false−negative procedures and erroneous treatment decisions.
The consensus classification established in Paris [3] for the
esophagus, stomach, and intestines was confirmed a year later
in Osaka, where further points were debated and integrated into
an updated version. This particularly affected the definitions of
the subtypes combining elevation and depression and the specif−
ic characteristics of neoplastic lesions in the large bowel.

The frequent misclassification between small 0 ± IIa and 0 ± Is le−


sions is of minimal relevance for the prevention of digestive−
tract cancer. On the other hand, underdetection of depressed

Lambert R et al. Superficial Neoplastic Lesions in the Digestive Tract ´ Endoscopy 2005; 37: 570 ± 578
6
Members of the Endoscopic Classification Review Group Tsukuma H, Oshima A, Narahara H et al. Natural history of early gas−
tric cancer: a non−concurrent long term follow up study. Gut 2000; 47:
618 ± 621
A. Axon, Leeds Teaching Hospitals National Health Service Trust, 7
Winawer SJ, Zauber AG, O’Brien MJ et al. The National Polyp Study: de−
Leeds, United Kingdom sign, methods, and characteristics of patients with newly diagnosed
M. D. Diebold, Centre Hospitalier Universitaire Robert DebrØ, polyps. The National Polyp Study Workgroup. Cancer 1992; 70:
Reims, France 1236 ± 1245
8
Inoue H, Rey JF, Lightdale C. Lugol chromoendoscopy for esophageal
M. Fujino, Yokohama Seamen’s Insurance Hospital, Yokohama, squamous cell cancer. Endoscopy 2001; 33: 75 ± 79
Japan 9
Kudo S, Kashida H, Tamura S et al. The problem of “flat” colonic adeno−
R. Fujita, Yokohama Seamen’s Insurance Hospital, Yokohama, ma. Gastrointest Endosc Clin N Am 1997; 7: 87 ± 98
10
Japan, Cancer Institute Hospital, Tokyo, Japan Jaramillo E, Watanabe M, Slezak P et al. Flat neoplastic lesions of the
colon and rectum detected by high resolution video−endoscopy and
R. M. Genta, Hôpitaux Universitaires de Gen›ve, Geneva, chromoscopy. Gastrointest Endosc 1995; 42: 114 ± 122
Switzerland 11
Canto MI, Setrakian S, Willis JE et al. Methylene blue staining of dys−
J. J. Gonvers, Polyclinique MØdicale Universitaire, Lausanne, plastic and nondysplastic Barrett’s esophagus: an in vivo and ex vivo
Switzerland study. Endoscopy 2001; 33: 391 ± 400

Review
12
Kiesslich R, Hahn M, Herrmann G et al. Screening for specialized co−
M. Guelrud, Tufts±New England Medical Center, Boston, lumnar epithelium with methylene blue: chromoendoscopy in pa−
Massachusetts, USA tients with Barrett’s esophagus and a normal control group. Gastroin−
H. Inoue, Showa University Northern Yokohama Hospital, test Endosc 2001; 53: 47 ± 52
13
Yokohama City, Kanagawa, Japan Sharma P, Topalovski M, Mayo MS et al. Methylene blue chromoen−
doscopy for detection of short−segment Barrett’s esophagus. Gastroin−
M. Jung, St. Hildegardis−Krankenhaus, Mainz, Germany test Endosc 2001; 54: 289 ± 293
H. Kashida, , Showa University Northern Yokohama Hospital, 14
Lambert R, Rey JF, Sankaranarayanan R. Magnification and chromosco−
Yokohama City, Kanagawa, Japan py with the acetic acid test. Endoscopy 2003; 35: 437 ± 445

Downloaded by: University of Connecticut. Copyrighted material.


15
S. Kudo, Showa University Northern Yokohama Hospital, Hurlstone DP, Cross SS, Adam I et al. Efficacy of high magnification
chromoscopic colonoscopy for the diagnosis of neoplasia in flat and
Yokohama City, Kanagawa, Japan depressed lesions of the colorectum: a prospective analysis. Gut.
R. Lambert, International Agency for Research on Cancer, 2004; 53: 284 ± 290
16
Lyons, France Inoue H, Kumagai Y, Yoshida T et al. High magnification endoscopic di−
C. Lightdale, Columbia±Presbyterian Medical Center, New York, agnosis of the superficial esophageal cancer. Dig Endosc 2000; 12
(Suppl): S32 ± S35
New York, USA 17
Yao K, Oishi T. Microgastroscopic findings of mucosal microvascular
T. Nakamura, Teikyo University School of Medicine, Tokyo, Japan architecture as visualized by magnifying endoscopy. Dig Endosc
H. Neuhaus, Evangelisches Krankenhaus, Düsseldorf, Germany 2001; 13 (Suppl): S27 ± S33
18
H. Niwa, St. Marianna University, Kawasaki, Japan Tobita K. Study on minute surface structures of the depressed−type
early gastric cancer with magnifying endoscopy. Dig Endosc 2001;
K. Ogoshi, Niigata Cancer Center Hospital, Niigata, Japan 13: 121 ± 126
J. F. Rey, Institut Arnault Tzanck, Saint Laurent du Var, France 19
Kumagai Y, Inoue H, Nagai K et al. Magnifying endoscopy, stereoscopic
R. Riddell, Mount Sinai Hospital, Toronto, Ontario, Canada microscopy, and the microvascular architecture of superficial esopha− 577
M. Sasako, National Cancer Center Hospital, Tokyo, Japan geal carcinoma. Endoscopy 2002; 34: 369 ± 375
20
Guelrud M, Herrera I, Essenfeld H et al. Enhanced magnification en−
T. Shimoda, National Cancer Center Hospital, Tokyo, Japan doscopy: a new technique to identify specialized metaplasia in Bar−
H. Suzuki, University of Tokyo, Japan rett’s esophagus. Gastrointest Endosc 2001; 53: 559 ± 565
21
G. N. J. Tytgat, Academic Medical Center, Amsterdam, Endo T, Awakawa T, Takahashi H et al. Classification of Barrett’s epithe−
The Netherlands lium by magnifying endoscopy. Gastrointest Endosc 2002; 55: 641 ±
647
K. Wang, Mayo Clinic, Rochester, Minnesota, USA 22
Sharma P, Weston AP, Topalovski M et al. Magnification chromoen−
H. Watanabe, Graduate School of Medical and Dental Sciences, doscopy for the detection of intestinal metaplasia and dysplasia in
Niigata, Japan Barrett’s oesophagus. Gut 2003; 52: 24 ± 27
23
T. Yamakawa, Teikyo University Mizonokuchi Hospital, Kudo S, Kashida H, Tamura T et al. Colonoscopic diagnosis and man−
agement of non polypoid early colorectal cancer. World J Surg 2000;
Kawasaki, Kanagawa, Japan 24: 1081 ± 1090
S. Yoshida, National Cancer Center Hospital East, Chiba, Japan 24
Hurlstone DP, Fujii T, Lobo AJ. Early detection of colorectal cancer
using high−magnification chromoscopic colonoscopy. Br J Surg 2002;
89: 272 ± 282
25
Kato S, Fujii T, Koba I et al. Assessment of colorectal lesions using mag−
References nifying colonoscopy and mucosal dye spraying: can significant lesions
be distinguished? Endoscopy 2001; 33: 306 ± 310
1
Schlemper RJ, Riddell RH, Kato Y et al. The Vienna classification of gas− 26
Konishi K, Kaneko K, Kurahashi T. A comparison of magnifying and
trointestinal epithelial neoplasia. Gut 2000; 47: 251 ± 255 nonmagnifying colonoscopy for diagnosis of colorectal polyps: a pro−
2
Dixon MF. Gastrointestinal epithelial neoplasia: Vienna revisited. Gut spective study. Gastrointest Endosc 2003; 57: 48 ± 53
2002; 51: 130 ± 131 27
Kudo S, Rubio CA, Teixeira CR et al. Pit pattern in colorectal neoplasia:
3
The Paris Endoscopic Classification of Superficial Neoplastic Lesions. endoscopic magnifying view. Endoscopy 2001; 33: 367 ± 373
Esophagus, stomach, and colon. Gastrointest Endosc 2003; 58 (Suppl 28
Morita T, Tamura S, Miyasaki J et al. Evaluation of endoscopic and his−
6): S3 ± S43 topathological features of serrated adenoma of the colon. Endoscopy
4
Ormsby AH, Vaezi MF, Richter JE et al. Cytokeratin immunoreactivity 2001; 33: 761 ± 765
patterns in the diagnosis of short segment Barrett’s esophagus. Gas− 29
Matsumoto T, Mizuno M, Shimizu M et al. Serrated adenoma of the
troenterology 2000; 119: 683 ± 690 colorectum: colonoscopic and histologic features. Gastrointest Endosc
5
Mohammed IA, Streutker CJ, Riddell RH et al. Utilization of cytokera− 1999; 49: 736 ± 742
tins 7 and 20 does not differentiate between Barrett’s esophagus and 30
Japanese Gastric Cancer Association. Japanese classification of gastric
gastric cardiac intestinal metaplasia. Mod Pathol 2002; 15: 611 ± 616 carcinoma: 2nd English edition. Gastric Cancer 1998; 1: 10 ± 24

Lambert R et al. Superficial Neoplastic Lesions in the Digestive Tract ´ Endoscopy 2005; 37: 570 ± 578
31
Japanese Society for Esophageal Diseases. Guidelines for the clinical nocarcinoma of the distal stomach and esophagus. Oncology 2001;
and pathologic studies on carcinoma of the esophagus. 9th ed. Tokyo: 61: 1 ± 9
41
Kanehara, 1999 Park DI, Rhee PL, Kim JE et al. Factors suggesting malignant transfor−
32
Watanabe H, Komukai S, Ajioka Y et al. Histopathology of m3 and sm1 mation of gastric adenoma: univariate and multivariate analysis. En−
invasive squamous cell carcinoma of the esophagus, with special re− doscopy 2001; 33: 501 ± 506
42
ference to endoscopic resection [in Japanese]. Stomach Intest 1998; National Report of the Group Medical examination for digestive can−
33: 1001 ± 1009 cer in 1999 [in Japanese]. J Gastroenterol Mass Surv 2002; 40: 57 ± 76
33 43
Kobayashi M, Watanabe H, Maeo S et al. Correlation of histological Hosokawa O, Tsuda S, Kidani E et al. Diagnosis of gastric cancer up to
atypia and cancer−sprouting with vascular permeation and lymph three years after negative upper gastrointestinal endoscopy. Endosco−
nodal metastasis by our new histological classification of submucosal py 1998; 30: 669 ± 674
44
invasion by colorectal carcinomas [in Japanese]. Stomach Intest 1994; Rex DK, Cummings OW, Helper DJ et al. 5−year incidence of adenomas
29: 1151 ± 1160 after negative colonoscopy in asymptomatic average−risk persons.
34
Yokoyama J, Ajioka Y, Watanabe H et al. Lymph node metastasis and Gastroenterology 1996; 11: 1178 ± 1181
45
micrometastasis of submucosal invasive colorectal carcinoma: an in− Winawer SJ, Zauber AG, Ho MN et al. Prevention of colorectal cancer
dicator of the curative potential of endoscopic treatment. Acta Med by colonoscopic polypectomy. The National Polyp Study Workgroup.
Biol 2002; 50: 1 ± 8 N Engl J Med 1993; 329: 1977 ± 1981
35 46
Egashira Y, Yoshida T, Hirata I et al. Analysis of pathological risk factors Gorski TF, Rosen L, Riether R et al. Colorectal cancer after surveillance
Review

for lymph node metastasis of submucosal invasive colon cancer. Mod colonoscopy: false−negative examination or fast growth? Dis Colon
Pathol 2004; 17: 503 ± 511 Rectum 1999; 42: 877 ± 880
36 47
Ueno H, Mochizuki H, Hashiguchi Y et al. Risk factors for an adverse Bressler B, Paszat LF, Vinden C et al. Colonoscopic miss rates for right−
outcome in early invasive colorectal carcinoma. Gastroenterology sided colon cancer: a population−based analysis. Gastroenterology
2004; 127: 385 ± 394 2004; 127: 452 ± 456
37 48
Watanabe H, Tada T, Iwafuchi M et al. New definition and macroscopic Burgart L. Colorectal polyps and other precursor lesions: need for an
characteristics of early carcinoma of the esophagus. Stomach Intest expanded view. Gastroenterol Clin North Am 2002; 31: 959 ± 970
49
1990; 25: 1075 ± 1086 Jass JR. Histopathology of early colorectal cancer. World J Surg 2000;
38
Kodama M, Kakegawa T. Treatment of superficial cancer of the esoph− 24: 1016 ± 1021

Downloaded by: University of Connecticut. Copyrighted material.


50
agus: a summary of responses to a questionnaire on superficial cancer George SM, Makinen MJ, Jernvall P et al. Classification of advanced
of the esophagus in Japan. Surgery 1998; 123: 432 ± 439 colorectal carcinomas by tumor edge morphology: evidence for differ−
39
Ell C, May A, Gossner L et al. Endoscopic mucosal resection of early ent pathogenesis and significance of polypoid and nonpolypoid tu−
cancer and high−grade dysplasia in Barrett’s esophagus. Gastroenter− mors. Cancer 2000; 89: 1901 ± 1909
51
ology 2000; 18: 670 ± 677 Tanaka S, Yoshida S, Chayama K. Clinical usefulness of high frequency
40
Tajima Y, Nakanishi Y, Yoshino T et al. Clinicopathological study of ear− ultrasound probes for new invasion depth diagnosis in submucosal
ly adenocarcinoma of the gastric cardia: comparison with early ade− colorectal carcinoma. Dig Endosc 2004; 16 (Suppl 2): S161 ± S164

578

Lambert R et al. Superficial Neoplastic Lesions in the Digestive Tract ´ Endoscopy 2005; 37: 570 ± 578

Das könnte Ihnen auch gefallen