- Email: [email protected]
Endoscopic ultrasonography and endoscopy for staging depth of invasion in early gastric cancer: a pilot study
Endoscopic ultrasonography and endoscopy for staging depth of invasion in early gastric cancer: a pilot study Hideo Yanai, MD, Yusuke Matsumoto, MD, Toshiya Harada, MD, Makoto Nishiaki, MD Hiroshi Tokiyama, MD, Toshinori Shigemitsu, MD, Masahiro Tada, MD, Kiwamu Okita, MD Ube, Japan
Background: We compared the accuracy of endoscopic ultrasonography (EUS) for staging depth of invasion of early gastric cancer with that of conventional endoscopy. Patients and methods: We assessed the depth of invasion of 108 lesions (104 patients) using EUS with a thin 20 MHz probe and compared the results with those of conventional endoscopy and of histologic examination of endoscopically or surgically resected specimens. Results: The overall accuracy rates for staging depth of invasion for conventional endoscopy and EUS were 72.2% and 64.8%, respectively. Lesions that were classified as limited to the mucosa on both conventional endoscopy and EUS were very likely (92.2%) to be limited to the mucosa on histologic examination. The rates for understaging and overstaging were 16.7% and 11.1%, respectively, for conventional endoscopy; and 7.4% and 24.1%, respectively, for EUS. The highest rate for understaging based on conventional endoscopy occurred for lesions in the gastric body (including cardia, 23.9%). Conclusions: EUS appears to be useful in combination with conventional endoscopy for staging depth of invasion of early gastric cancer. In particular, the two techniques in tandem may accurately predict that a lesion is limited to the mucosa, and EUS may be useful to overcome the potential for understaging by conventional endoscopy, particularly in the gastric body. (Gastrointest Endosc 1997;46:212-6.)
The depth of invasion of early gastric cancer (EGC) strongly correlates with lymph node metastasis, and therefore pretreatment depth staging is important for patient management. 1-3 Conventional endoscopy (CE) is the most useful diagnostic modality for EGC, and endoscopic ultrasonography (EUS) has been shown to be useful for staging. 4-9 We previously described the application of a 20 MHz thin ultrasound probe to differentiate mucosal can-
cer (m) from tumor invading more deeply into the submucosa (sin). 1°-12 However, the relationship between staging characteristics of CE and EUS has not been established, and the role of EUS in EGC depth staging has not been fully explored. To determine the usefulness of EUS for staging EGC, we retrospectively analyzed the nonblinded pilot data for the staging characteristics of both CE and EUS in patients with EGC.
Received September 26, 1996. For revision January 3, 1997. Accepted March 19, 1997. From the First Department of Internal Medicine, Yamaguchi University School of Medicine, Ube, Japan. Reprint requests: Hideo Yanai, MD, First Department of Internal Medicine, Yamaguchi University School of Medicine, 1144 Kogushi, Ube, Yamaguchi, 755 Japan. Copyright © 1997 by the American Society for Gastrointestinal Endoscopy 0016-5107/97/$5.00 + 0 37/1/81963
Subjects The study included 108 endoscopically diagnosed EGCs in 104 patients including 5 lesions for which the final histologic diagnosis was advanced gastric cancer (cancer invading into muscularis propria or deeper) (Table 1). The lesions clearly diagnosed as advanced lesions by endoscopy were excluded. There were 76 men and 28 women (mean age 67.2 years, range 31 to 87 years). Patients underwent EUS between January 1990 and August 1995.
PATIENTS AND METHODS
212
GASTROINTESTINAL ENDOSCOPY
VOLUME 46, NO. 3, 1997
EUS and endoscopy for staging depth of invasion in early gastric cancer
H Yanai, Y Matsumoto, T Harada, et al.
Table 1. Endoscopically diagnosed early gastric cancers staged by endoscopic ultrasonography (EUS) Macroscopic classification Type 0 I (protruded) I+IIa I+IIc Type 0 II (superficial) IIa (elevated) IIa+I Iia+IIc IIb+IIa IIc (depressed) IIc+IIa IIc+III Type 0 III(III) (excavated) III+IIa III+IIc
Lesions (n = 108) 10 1 1 23 2 12" 1 36* 3 13" 3 1 2t
*Includes one lesion found to be advanced gastric cancer. tIncludes two lesions found to be advanced gastric cancer.
Figure 1. Radially scanned 20 MHz EUS image of the normal gastric wall. In the nine-layered structure, a fine hypoechoic layer (2c) between the conventional second and third layers is considered to correspond to the muscularis mucosae. TM The hyperechoic 2b layer is considered the interface echo between the mucosa and musculafis mucosae. Endoscopic ultrasonography
For every lesion, EUS findings were compared to the endoscopic diagnosis and with the histologic diagnosis obtained from endoscopically or surgically resected specimens. The location of each lesion was endoscopically classified into three regions of the stomach: the gastric body including the cardia (body-cardia), the region of the angulus, or the antrum including the pylorus (antrumpylorus). Ia
Histology Specimens were fixed and stained with H&E, and the depth of invasion was classified as mucosal (m), submucosa] (sm), or advanced (the t u m o r invades the muscularis propria or deeper). 14
Endoscopy The endoscopes used in this study were the EVG-CT (Fujinon, Omiya, Japan), GIF-Q200 (Olympus, Tokyo, Japan), or GIF-2T200 (Olympus). Staging was performed following a s t a n d a r d protocol. 15 Small lesions t h a t p r o t r u d e d from t h e m u c o s a with a smooth surface w e r e classified as CE-M (mucosal). Lesions with small shallow depressions b u t w i t h o u t b a n k f o r m a t i o n or u n e v e n surfaces were also classified as CE-M. Lesions considered to be s u b m u c o s a l (CE-SM) were those t h a t showed a more rigid base w i t h an i r r e g u l a r l y s h a p e d nodule on the m a r g i n or those w i t h folds t h a t were i n t e r r u p t e d and enlarged. U l c e r a t i v e lesions surr o u n d e d by a t u m o r o u s bank, or those with folds t h a t were e l e v a t e d a n d merged, were considered a d v a n c e d (CE-AD). "VOLUME 46, NO. 3, 1997
The 20 MHz thin endoscopic u l t r a s o u n d probes used in this study were linear SP-101 (71 lesions) and radial-linear switchable SP-501 (37 lesions) interfaced to a Sonoprobe System (SPS, Fujinon, Omiya, Japan). The probes yielded high-quality cross-sectional images of the mucosa and submucosa and were easily directed to the EGC lesions u n d e r direct vision of the endoscope. 1°-12 Observing routine endoscopic studies, the EUS images were obtained by the same group of endoscopists. The EUS images were i n t e r p r e t e d with regard to t u m o r invasion according to the five layer architecture of the gastric wall, and lesions were classified as EUS-M (mucosa), EUS-SM (submucosa) or EUS-AD (advanced). On EUS, the normal gastric wall is visualized as the mucosa (combination of the first hyperechoic and second hypoechoic layers) and the submucosa (the third hyperechoic layer). The muscularis propria is visualized as the fourth hypoechoic layer, and the fifth hyperechoic layer is the serosa including the subserosa. 16, 17 According to our previous reports, the fine hypoechoic layer between the second and third layers is considered to correspond to the muscularis mucosae. 11 When the muscularis mucosae and i n t e r m u s c u l a r interface of the muscularis propria are visualized, the normal gastric wall is observed as a nine-layered structure (Fig. 1). Data were analyzed with the chi-square test.
RESULTS The accuracy rates of CE and EUS in staging t u m o r invasion depth were 72.2% and 64.8%, respecGASTROINTESTINAL ENDOSCOPY 213
H Yanai, Y Matsumoto, T Harada, et al.
EUS and endoscopy for staging depth of invasion in early gastric cancer
Table 2. Accuracy of invasion depth staging with conventional endoscopy and 20 MHz endoscopic ultrasonography (EUS) in endoscopically diagnosed early gastric cancers Conventional endoscopy CE-M SM
(n) Histopathology M (76) SM (27) AD (5) Total accuracy Understaged Overstaged
64 13 1
12 14 4 72.2% 16.7% 11.1%
Endoscopic ultrasonography SM AD Indeterminant
EUS-M 48 4 0
23 21 4
1 2 1
4 0 0
64.8% 7.4% 24.1%
Causes of staging errors (under or over)differed significantly between CE and EUS (p < .01). M, Mucosal; S M , submucosal; AD, advanced.
Table 3. Comparison of invasion depth staging between endoscopy and endoscopic ultrasonography
Table 4. Accuracy of invasion depth staging according to anatomic location Gastric body Antrum including Angular including cardia region pylorus (46 lesions) (22 lesions) (40 lesions)
Endoscopic ultrasonography EUS-M EUS-SM EUS-AD Indeterminant Conventional endoscopy CE-M CE-SM
m, 47 sm, 4
m, 12 sm, 8
m, 1
m, 11 sin, 13 ad, 4
m, 1 sm, 1 ad, 1
m, 4
sm, 1
M, Mucosal; S M , submucosal; AD, advanced.
tively (Tables 2 to 4). No statistically significant difference existed b e t w e e n the a c c u r a c y of t h e two modalities. H o w e v e r , t h e r e w e r e significant differences between the two modalities with r e g a r d to understaging and overstaging (p < .01). For CE, the rates of u n d e r s t a g i n g and overstaging were 16.7% and 11.1%, respectively. In contrast, for E U S the rates of u n d e r s t a g i n g and overstaging were 7.4% and 24.1%, respectively. T h e r e were several (3.7%) i n d e t e r m i n a n t lesions on EUS, which resulted in a lower overall accuracy t h a n t h a t for CE (Table 2, Fig. 2). There was no significant difference in the accuracy rates between the two types of thin u l t r a s o u n d probes. Overall, 92.2% (47 of 51) of all lesions classified as CE-M a n d E U S - M were histologically found to h a v e only mucosal (m) i n v a s i o n (Table 3). In the gastric body-cardia, t h e a c c u r a c y r a t e s for i n v a s i o n d e p t h staging for CE a n d E U S w e r e 67.4% a n d 67.4%, r e s p e c t i v e l y (no significant difference). The r a t e o f u n d e r s t a g i n g for CE was 23.9% of lesions a n d 10.9% for EUS. In the region of the angulus, the a c c u r a c y r a t e s for CE a n d E U S w e r e 63.6% a n d 45.5%, respectively. In the a n t r u m - p y l o r u s , the ac214 GASTROINTESTINAL ENDOSCOPY
Endoscopy, n (%) Accurately staged Understaged Overstaged Endoscopic ultrasonography,* n (%) Accurately staged Understaged Overstaged Indeterminant
31 (67.4) 11 (23.9) 4 (8.7)
i4 (63.6) 4 (18.2) 4 (18.2)
33 (82.5) 3 (7.5) 4 (10.0)
31 (67.4) 5 (10.9) 9 (19.6) 1 (2.2)
10 (45.5) 3 (13.6) 7 (31.8) 2 (9.1)
29 (72.5) 0 (0.0) 10 (25.0) 1 (2.5)
*20 MHz. c u r a c y r a t e s for CE a n d E U S w e r e 82.5% a n d 72.5%, r e s p e c t i v e l y (Table 4).
DISCUSSION R e c e n t r e p o r t s h a v e s u g g e s t e d t h a t the distinction b e t w e e n i n v a s i o n to m u c o s a or s u b m u c o s a is especially i m p o r t a n t in EGC b e c a u s e it directly influences the choice of endoscopic t r e a t m e n t v e r s u s surgical resection. 1-3 E U S p r o m i s e s to be useful for this distinction; however, t h e r e h a s b e e n no r e p o r t c o n c e r n i n g a d e t a i l e d c o m p a r i s o n of E U S w i t h the e s t a b l i s h e d diagnostic m e t h o d of CE. We p r e v i o u s l y r e p o r t e d t h a t a h i g h - r e s o l u t i o n 20 M H z E U S probe is suitable for staging t h e d e p t h of i n v a s i o n of EGC.11, 12 To assess t h e u s e f u l n e s s of the E U S , we c o m p a r e d t h e e s t i m a t e s of d e p t h of i n v a s i o n b a s e d on CE a n d on E U S w i t h a 20 M H z probe. Overall, E U S a n d CE yielded similar a c c u r a c y r a t e s for staging of d e p t h of invasion. H o w e v e r , t h e staging c h a r a c t e r i s t i c s for CE a n d E U S w e r e differVOLUME 46, NO. 3, 1997
EUS and endoscopy for staging depth of invasion in early gastric cancer
H Yanai, Y Matsumoto, T Harada, et al.
Figure 2. A, Endoscopic image of a superficial depressed (0 IIc) type early gastric cancer in the gastric body (arrows). The tumor depression is shallow and smooth, and the lesion is thought to be confined to the mucosa (CE-M). B, Radially scanned image of a representative gastric tumor. The hypoechoic 2c layer (muscularis mucosae) and the hyperechoic third layer (submucosa) are disrupted (arrow heads) and hypoechoic tumor (T) has penetrated into the third layer. The tumor invasion depth is thought to be limited to the submucosa (EUS-SM). C, Linearly scanned image of the tumor. The 2c layer is disrupted and the third layer is irregularly narrowed. The tumor was thought to penetrate to the submucosa (EUS-SM). D, Histologic examination (H&E) of the surgically resected specimen corresponded with the EUS findings. This is a CE-understaged case.
ent. CE was more likely to result in understaging for invasion depth. In contrast, EUS showed a lower (7.4%) understaging rate. When both CE and EUS classified a lesion as CE-M and EUS-M, the accuracy was improved to 92.2%. This suggests t h a t EUS can correct the understaging of ES. The main causes for overstaging in EUS were benign ulcerous change, fibrosis, benign cystic glands in the submucosal layer, inflammatory change, or an anomaly of the muscularis mucosae. Sano et al. 15 reported t h a t CE had an accuracy of 71.9% in 206 cases in the differentiation of mucosal from submucesal cancer. Shirao et al. ~s also reported t h a t CE was accurate in 73% (182 of 250) of EGC cases. Our CE accuracy rate of 72.2% was almost the same as in these reports. There are very few data on the differentiation between M and SM VOLUME 46, NO. 3, 1997
cancers using the high-frequency thin ultrasound probes. Yasuda 19 reported t h a t the accuracy rate for depth of gastric cancer invasion as determined by ultrasound probes was 83.7% in 43 cases. Our accuracy rate of 64.8% using a 20 MHz endoscopic ultrasound probe was lower. The accuracy for staging of tumor invasion depth by CE and EUS varied according to the location within the stomach. In the antrum-pylorus, CE showed a higher accuracy t h a n EUS and had low rates for both overstaging and understaging. Therefore, EUS m a y not be required for staging of lesions in this region t h a t appear at endoscopy to be early stage tumors. In contrast, in the gastric body-cardia, the accuracy rates of CE and EUS were similar, and CE showed a 23.9% understaging rate. Therefore, EUS may be useful to compensate for understaging GASTROINTESTINAL ENDOSCOPY 215
H Yanai, Y Matsumoto, T Harada, et al.
EUS and endoscopy for staging depth of invasion in early gastric cancer
by CE in the gastric body-cardia. This difference in CE accuracy between the antrum-pylorus and gastric body-cardia may be derived from the fact that the endoscopes used in this study are forward viewing. In the gastric body, it is difficult to achieve an en face view of the lesion with a forward viewing instrument. In contrast, lesions in this location are easily examined by EUS because the ultrasound beam is directed laterally. Furthermore CE can observe only the surface findings. When EGC lesions have a smooth surface and no swelling of fold tips, cross-sectional visualization using EUS may be useful. Our results suggest that the combination of CE and EUS is useful in evaluating depth of invasion in EGC. EUS may compensate for the understaging of lesions by CE, especially in the gastric body-cardia. When the same endoscopist performs both CE and E U S consecutively, each modality may result in similar staging data concerning invasion depth due to unconscious integration of both modalities. A prospective blinded study of EUS in EGC appears to be warranted. REFERENCES 1. Lambert R. Endoscopic treatment of esophagogastric tumors. Endoscopy 1996;28:27-37. 2. Yamao T, Shirao K, Ono H, Kondo H, Saito D, Yamaguchi H, et al. Risk factors for lymph node metastasis from intramucosal gastric carcinoma. Cancer 1996;77:602-6. 3. Oguro Y. Recent advances in endoscopic treatment. In: Oguro Y, Takagi K, editors. Endoscopic approaches to cancer diagnosis and treatment. Gann monograph on cancer research No. 37. London: Taylor & Francis; 1990. p. 89-99. 4. Longo WE, Zucker KA, Zdon MJ, Ballantyne GH, Cambria RP, Modlin IM. Role of endoscopy in the diagnosis of early gastric cancer. Arch Surg 1987;122:292-5. 5. Tio TL, Schouwink MH, Cikot RJLM, Tytgat GNJ. Preoperative TNM classification of gastric carcinoma by endosonography in comparison with the pathological TNM system: a prospective study of 72 cases. Hepatogastroenterology 1989; 36:51-6.
216 GASTROINTESTINAL ENDOSCOPY
6. Akahoshi K, Misawa T, Fujishima H, Chijiiwa Y, Maruoka A, Ohkubo A, et al. Preoperative evaluation of gastric cancer by endoscopic ultrasound. Gut 1991;32:479-82. 7. Lightdale CJ. Endoscopic ultrasonography in the diagnosis, staging and follow-up of esophageal and gastric cancer. Endoscopy 1992;24 (Suppl 1):297-303. 8. Grimm H, Binmoeller KF, Hamper K, Koch J, Henne-Bruns D, Soehendra N. Endosonography for preoperative locoregional staging of esophageal and gastric cancer. Endoscopy 1993;25:224-30. 9. Colin-Jones DG, RSsch T, Dittler HJ. Staging of gastric cancer by endoscopy. Endoscopy 1993;25:34-8. 10. Takemoto T, Yanai H, Tada M, Aibe T, Fujimura H, Murata N, et al. Application of ultrasonic probes prior to endoscopic resection of early gastric cancer. Endoscopy 1992;24 (Suppl 1):329-33. 11. Yanai H, Fujimura H, Suzumi M, Matsuura S, Awaya N, Noguchi T, et al. Delineation of the gastric muscularis mucosae and assessment of depth of invasion of early gastric cancer using a 20-megahertz endoscopic ultrasound probe. Gastrointest Endosc 1993;39:505-12. 12. Yanai H, Tada M, Karita M, Okita K. Diagnostic utility of 2O-megahertz linear endoscopic ultrasonography in early gastric cancer. Gastrointest Endosc 1996;44:29-33. 13. Ma~atka Z. Gastroscopy. In: Ma~atka Z, editor. Terminology, definitions and diagnostic criteria in digestive endoscopy. 2nd edition. Bad Homburg, Germany: Normed Verlag; 1989. p. 47-59. 14. Japanese Research Society for Gastric Cancer. Japanese Classification of Gastric Carcinoma. 1st English edition. Tokyo: Kanehara & Co., Ltd.; 1995. p. 37-71. 15. Sano T, Okuyama Y, Kobori O, Shimizu T, Morioka Y. Early gastric cancer: endoscopic diagnosis of depth of invasion. Dig Dis Sci 1990;35:1340-4. 16. Aibe T, Fuji T, Okita K, Takemoto T. A fundamental study of normal layer structure of the gastrointestinal wall visualized by endoscopic ultrasonography. Scand J Gastroenterol 1986;21 (Suppl 123):6-15. 17. Kimmey MB, Martin RW,Haggitt RC, Wang KY, Franklin DW, Silverstein FE. Histologic correlates of gastrointestinal ultrasound images. Gastroenterology 1989;96:433-41. 18. Shirao K, Saito D, Yamaguchi H, Kido M, Fukuda H, Yoshida S. Accuracy of preoperative diagnosis of early gastric cancer on submucosal invasion: a study for indications of endoscopic treatment. Stomach and Intestine 1992;27:1175-84. 19. Yasuda K. Endoscopic ultrasonic probes and mucosectomy for early gastric carcinoma. Gastrointest Endosc 1996;43:$29-31.
VOLUME 46, NO. 3, 1997
Background: We compared the accuracy of endoscopic ultrasonography (EUS) for staging depth of invasion of early gastric cancer with that of conventional endoscopy. Patients and methods: We assessed the depth of invasion of 108 lesions (104 patients) using EUS with a thin 20 MHz probe and compared the results with those of conventional endoscopy and of histologic examination of endoscopically or surgically resected specimens. Results: The overall accuracy rates for staging depth of invasion for conventional endoscopy and EUS were 72.2% and 64.8%, respectively. Lesions that were classified as limited to the mucosa on both conventional endoscopy and EUS were very likely (92.2%) to be limited to the mucosa on histologic examination. The rates for understaging and overstaging were 16.7% and 11.1%, respectively, for conventional endoscopy; and 7.4% and 24.1%, respectively, for EUS. The highest rate for understaging based on conventional endoscopy occurred for lesions in the gastric body (including cardia, 23.9%). Conclusions: EUS appears to be useful in combination with conventional endoscopy for staging depth of invasion of early gastric cancer. In particular, the two techniques in tandem may accurately predict that a lesion is limited to the mucosa, and EUS may be useful to overcome the potential for understaging by conventional endoscopy, particularly in the gastric body. (Gastrointest Endosc 1997;46:212-6.)
The depth of invasion of early gastric cancer (EGC) strongly correlates with lymph node metastasis, and therefore pretreatment depth staging is important for patient management. 1-3 Conventional endoscopy (CE) is the most useful diagnostic modality for EGC, and endoscopic ultrasonography (EUS) has been shown to be useful for staging. 4-9 We previously described the application of a 20 MHz thin ultrasound probe to differentiate mucosal can-
cer (m) from tumor invading more deeply into the submucosa (sin). 1°-12 However, the relationship between staging characteristics of CE and EUS has not been established, and the role of EUS in EGC depth staging has not been fully explored. To determine the usefulness of EUS for staging EGC, we retrospectively analyzed the nonblinded pilot data for the staging characteristics of both CE and EUS in patients with EGC.
Received September 26, 1996. For revision January 3, 1997. Accepted March 19, 1997. From the First Department of Internal Medicine, Yamaguchi University School of Medicine, Ube, Japan. Reprint requests: Hideo Yanai, MD, First Department of Internal Medicine, Yamaguchi University School of Medicine, 1144 Kogushi, Ube, Yamaguchi, 755 Japan. Copyright © 1997 by the American Society for Gastrointestinal Endoscopy 0016-5107/97/$5.00 + 0 37/1/81963
Subjects The study included 108 endoscopically diagnosed EGCs in 104 patients including 5 lesions for which the final histologic diagnosis was advanced gastric cancer (cancer invading into muscularis propria or deeper) (Table 1). The lesions clearly diagnosed as advanced lesions by endoscopy were excluded. There were 76 men and 28 women (mean age 67.2 years, range 31 to 87 years). Patients underwent EUS between January 1990 and August 1995.
PATIENTS AND METHODS
212
GASTROINTESTINAL ENDOSCOPY
VOLUME 46, NO. 3, 1997
EUS and endoscopy for staging depth of invasion in early gastric cancer
H Yanai, Y Matsumoto, T Harada, et al.
Table 1. Endoscopically diagnosed early gastric cancers staged by endoscopic ultrasonography (EUS) Macroscopic classification Type 0 I (protruded) I+IIa I+IIc Type 0 II (superficial) IIa (elevated) IIa+I Iia+IIc IIb+IIa IIc (depressed) IIc+IIa IIc+III Type 0 III(III) (excavated) III+IIa III+IIc
Lesions (n = 108) 10 1 1 23 2 12" 1 36* 3 13" 3 1 2t
*Includes one lesion found to be advanced gastric cancer. tIncludes two lesions found to be advanced gastric cancer.
Figure 1. Radially scanned 20 MHz EUS image of the normal gastric wall. In the nine-layered structure, a fine hypoechoic layer (2c) between the conventional second and third layers is considered to correspond to the muscularis mucosae. TM The hyperechoic 2b layer is considered the interface echo between the mucosa and musculafis mucosae. Endoscopic ultrasonography
For every lesion, EUS findings were compared to the endoscopic diagnosis and with the histologic diagnosis obtained from endoscopically or surgically resected specimens. The location of each lesion was endoscopically classified into three regions of the stomach: the gastric body including the cardia (body-cardia), the region of the angulus, or the antrum including the pylorus (antrumpylorus). Ia
Histology Specimens were fixed and stained with H&E, and the depth of invasion was classified as mucosal (m), submucosa] (sm), or advanced (the t u m o r invades the muscularis propria or deeper). 14
Endoscopy The endoscopes used in this study were the EVG-CT (Fujinon, Omiya, Japan), GIF-Q200 (Olympus, Tokyo, Japan), or GIF-2T200 (Olympus). Staging was performed following a s t a n d a r d protocol. 15 Small lesions t h a t p r o t r u d e d from t h e m u c o s a with a smooth surface w e r e classified as CE-M (mucosal). Lesions with small shallow depressions b u t w i t h o u t b a n k f o r m a t i o n or u n e v e n surfaces were also classified as CE-M. Lesions considered to be s u b m u c o s a l (CE-SM) were those t h a t showed a more rigid base w i t h an i r r e g u l a r l y s h a p e d nodule on the m a r g i n or those w i t h folds t h a t were i n t e r r u p t e d and enlarged. U l c e r a t i v e lesions surr o u n d e d by a t u m o r o u s bank, or those with folds t h a t were e l e v a t e d a n d merged, were considered a d v a n c e d (CE-AD). "VOLUME 46, NO. 3, 1997
The 20 MHz thin endoscopic u l t r a s o u n d probes used in this study were linear SP-101 (71 lesions) and radial-linear switchable SP-501 (37 lesions) interfaced to a Sonoprobe System (SPS, Fujinon, Omiya, Japan). The probes yielded high-quality cross-sectional images of the mucosa and submucosa and were easily directed to the EGC lesions u n d e r direct vision of the endoscope. 1°-12 Observing routine endoscopic studies, the EUS images were obtained by the same group of endoscopists. The EUS images were i n t e r p r e t e d with regard to t u m o r invasion according to the five layer architecture of the gastric wall, and lesions were classified as EUS-M (mucosa), EUS-SM (submucosa) or EUS-AD (advanced). On EUS, the normal gastric wall is visualized as the mucosa (combination of the first hyperechoic and second hypoechoic layers) and the submucosa (the third hyperechoic layer). The muscularis propria is visualized as the fourth hypoechoic layer, and the fifth hyperechoic layer is the serosa including the subserosa. 16, 17 According to our previous reports, the fine hypoechoic layer between the second and third layers is considered to correspond to the muscularis mucosae. 11 When the muscularis mucosae and i n t e r m u s c u l a r interface of the muscularis propria are visualized, the normal gastric wall is observed as a nine-layered structure (Fig. 1). Data were analyzed with the chi-square test.
RESULTS The accuracy rates of CE and EUS in staging t u m o r invasion depth were 72.2% and 64.8%, respecGASTROINTESTINAL ENDOSCOPY 213
H Yanai, Y Matsumoto, T Harada, et al.
EUS and endoscopy for staging depth of invasion in early gastric cancer
Table 2. Accuracy of invasion depth staging with conventional endoscopy and 20 MHz endoscopic ultrasonography (EUS) in endoscopically diagnosed early gastric cancers Conventional endoscopy CE-M SM
(n) Histopathology M (76) SM (27) AD (5) Total accuracy Understaged Overstaged
64 13 1
12 14 4 72.2% 16.7% 11.1%
Endoscopic ultrasonography SM AD Indeterminant
EUS-M 48 4 0
23 21 4
1 2 1
4 0 0
64.8% 7.4% 24.1%
Causes of staging errors (under or over)differed significantly between CE and EUS (p < .01). M, Mucosal; S M , submucosal; AD, advanced.
Table 3. Comparison of invasion depth staging between endoscopy and endoscopic ultrasonography
Table 4. Accuracy of invasion depth staging according to anatomic location Gastric body Antrum including Angular including cardia region pylorus (46 lesions) (22 lesions) (40 lesions)
Endoscopic ultrasonography EUS-M EUS-SM EUS-AD Indeterminant Conventional endoscopy CE-M CE-SM
m, 47 sm, 4
m, 12 sm, 8
m, 1
m, 11 sin, 13 ad, 4
m, 1 sm, 1 ad, 1
m, 4
sm, 1
M, Mucosal; S M , submucosal; AD, advanced.
tively (Tables 2 to 4). No statistically significant difference existed b e t w e e n the a c c u r a c y of t h e two modalities. H o w e v e r , t h e r e w e r e significant differences between the two modalities with r e g a r d to understaging and overstaging (p < .01). For CE, the rates of u n d e r s t a g i n g and overstaging were 16.7% and 11.1%, respectively. In contrast, for E U S the rates of u n d e r s t a g i n g and overstaging were 7.4% and 24.1%, respectively. T h e r e were several (3.7%) i n d e t e r m i n a n t lesions on EUS, which resulted in a lower overall accuracy t h a n t h a t for CE (Table 2, Fig. 2). There was no significant difference in the accuracy rates between the two types of thin u l t r a s o u n d probes. Overall, 92.2% (47 of 51) of all lesions classified as CE-M a n d E U S - M were histologically found to h a v e only mucosal (m) i n v a s i o n (Table 3). In the gastric body-cardia, t h e a c c u r a c y r a t e s for i n v a s i o n d e p t h staging for CE a n d E U S w e r e 67.4% a n d 67.4%, r e s p e c t i v e l y (no significant difference). The r a t e o f u n d e r s t a g i n g for CE was 23.9% of lesions a n d 10.9% for EUS. In the region of the angulus, the a c c u r a c y r a t e s for CE a n d E U S w e r e 63.6% a n d 45.5%, respectively. In the a n t r u m - p y l o r u s , the ac214 GASTROINTESTINAL ENDOSCOPY
Endoscopy, n (%) Accurately staged Understaged Overstaged Endoscopic ultrasonography,* n (%) Accurately staged Understaged Overstaged Indeterminant
31 (67.4) 11 (23.9) 4 (8.7)
i4 (63.6) 4 (18.2) 4 (18.2)
33 (82.5) 3 (7.5) 4 (10.0)
31 (67.4) 5 (10.9) 9 (19.6) 1 (2.2)
10 (45.5) 3 (13.6) 7 (31.8) 2 (9.1)
29 (72.5) 0 (0.0) 10 (25.0) 1 (2.5)
*20 MHz. c u r a c y r a t e s for CE a n d E U S w e r e 82.5% a n d 72.5%, r e s p e c t i v e l y (Table 4).
DISCUSSION R e c e n t r e p o r t s h a v e s u g g e s t e d t h a t the distinction b e t w e e n i n v a s i o n to m u c o s a or s u b m u c o s a is especially i m p o r t a n t in EGC b e c a u s e it directly influences the choice of endoscopic t r e a t m e n t v e r s u s surgical resection. 1-3 E U S p r o m i s e s to be useful for this distinction; however, t h e r e h a s b e e n no r e p o r t c o n c e r n i n g a d e t a i l e d c o m p a r i s o n of E U S w i t h the e s t a b l i s h e d diagnostic m e t h o d of CE. We p r e v i o u s l y r e p o r t e d t h a t a h i g h - r e s o l u t i o n 20 M H z E U S probe is suitable for staging t h e d e p t h of i n v a s i o n of EGC.11, 12 To assess t h e u s e f u l n e s s of the E U S , we c o m p a r e d t h e e s t i m a t e s of d e p t h of i n v a s i o n b a s e d on CE a n d on E U S w i t h a 20 M H z probe. Overall, E U S a n d CE yielded similar a c c u r a c y r a t e s for staging of d e p t h of invasion. H o w e v e r , t h e staging c h a r a c t e r i s t i c s for CE a n d E U S w e r e differVOLUME 46, NO. 3, 1997
EUS and endoscopy for staging depth of invasion in early gastric cancer
H Yanai, Y Matsumoto, T Harada, et al.
Figure 2. A, Endoscopic image of a superficial depressed (0 IIc) type early gastric cancer in the gastric body (arrows). The tumor depression is shallow and smooth, and the lesion is thought to be confined to the mucosa (CE-M). B, Radially scanned image of a representative gastric tumor. The hypoechoic 2c layer (muscularis mucosae) and the hyperechoic third layer (submucosa) are disrupted (arrow heads) and hypoechoic tumor (T) has penetrated into the third layer. The tumor invasion depth is thought to be limited to the submucosa (EUS-SM). C, Linearly scanned image of the tumor. The 2c layer is disrupted and the third layer is irregularly narrowed. The tumor was thought to penetrate to the submucosa (EUS-SM). D, Histologic examination (H&E) of the surgically resected specimen corresponded with the EUS findings. This is a CE-understaged case.
ent. CE was more likely to result in understaging for invasion depth. In contrast, EUS showed a lower (7.4%) understaging rate. When both CE and EUS classified a lesion as CE-M and EUS-M, the accuracy was improved to 92.2%. This suggests t h a t EUS can correct the understaging of ES. The main causes for overstaging in EUS were benign ulcerous change, fibrosis, benign cystic glands in the submucosal layer, inflammatory change, or an anomaly of the muscularis mucosae. Sano et al. 15 reported t h a t CE had an accuracy of 71.9% in 206 cases in the differentiation of mucosal from submucesal cancer. Shirao et al. ~s also reported t h a t CE was accurate in 73% (182 of 250) of EGC cases. Our CE accuracy rate of 72.2% was almost the same as in these reports. There are very few data on the differentiation between M and SM VOLUME 46, NO. 3, 1997
cancers using the high-frequency thin ultrasound probes. Yasuda 19 reported t h a t the accuracy rate for depth of gastric cancer invasion as determined by ultrasound probes was 83.7% in 43 cases. Our accuracy rate of 64.8% using a 20 MHz endoscopic ultrasound probe was lower. The accuracy for staging of tumor invasion depth by CE and EUS varied according to the location within the stomach. In the antrum-pylorus, CE showed a higher accuracy t h a n EUS and had low rates for both overstaging and understaging. Therefore, EUS m a y not be required for staging of lesions in this region t h a t appear at endoscopy to be early stage tumors. In contrast, in the gastric body-cardia, the accuracy rates of CE and EUS were similar, and CE showed a 23.9% understaging rate. Therefore, EUS may be useful to compensate for understaging GASTROINTESTINAL ENDOSCOPY 215
H Yanai, Y Matsumoto, T Harada, et al.
EUS and endoscopy for staging depth of invasion in early gastric cancer
by CE in the gastric body-cardia. This difference in CE accuracy between the antrum-pylorus and gastric body-cardia may be derived from the fact that the endoscopes used in this study are forward viewing. In the gastric body, it is difficult to achieve an en face view of the lesion with a forward viewing instrument. In contrast, lesions in this location are easily examined by EUS because the ultrasound beam is directed laterally. Furthermore CE can observe only the surface findings. When EGC lesions have a smooth surface and no swelling of fold tips, cross-sectional visualization using EUS may be useful. Our results suggest that the combination of CE and EUS is useful in evaluating depth of invasion in EGC. EUS may compensate for the understaging of lesions by CE, especially in the gastric body-cardia. When the same endoscopist performs both CE and E U S consecutively, each modality may result in similar staging data concerning invasion depth due to unconscious integration of both modalities. A prospective blinded study of EUS in EGC appears to be warranted. REFERENCES 1. Lambert R. Endoscopic treatment of esophagogastric tumors. Endoscopy 1996;28:27-37. 2. Yamao T, Shirao K, Ono H, Kondo H, Saito D, Yamaguchi H, et al. Risk factors for lymph node metastasis from intramucosal gastric carcinoma. Cancer 1996;77:602-6. 3. Oguro Y. Recent advances in endoscopic treatment. In: Oguro Y, Takagi K, editors. Endoscopic approaches to cancer diagnosis and treatment. Gann monograph on cancer research No. 37. London: Taylor & Francis; 1990. p. 89-99. 4. Longo WE, Zucker KA, Zdon MJ, Ballantyne GH, Cambria RP, Modlin IM. Role of endoscopy in the diagnosis of early gastric cancer. Arch Surg 1987;122:292-5. 5. Tio TL, Schouwink MH, Cikot RJLM, Tytgat GNJ. Preoperative TNM classification of gastric carcinoma by endosonography in comparison with the pathological TNM system: a prospective study of 72 cases. Hepatogastroenterology 1989; 36:51-6.
216 GASTROINTESTINAL ENDOSCOPY
6. Akahoshi K, Misawa T, Fujishima H, Chijiiwa Y, Maruoka A, Ohkubo A, et al. Preoperative evaluation of gastric cancer by endoscopic ultrasound. Gut 1991;32:479-82. 7. Lightdale CJ. Endoscopic ultrasonography in the diagnosis, staging and follow-up of esophageal and gastric cancer. Endoscopy 1992;24 (Suppl 1):297-303. 8. Grimm H, Binmoeller KF, Hamper K, Koch J, Henne-Bruns D, Soehendra N. Endosonography for preoperative locoregional staging of esophageal and gastric cancer. Endoscopy 1993;25:224-30. 9. Colin-Jones DG, RSsch T, Dittler HJ. Staging of gastric cancer by endoscopy. Endoscopy 1993;25:34-8. 10. Takemoto T, Yanai H, Tada M, Aibe T, Fujimura H, Murata N, et al. Application of ultrasonic probes prior to endoscopic resection of early gastric cancer. Endoscopy 1992;24 (Suppl 1):329-33. 11. Yanai H, Fujimura H, Suzumi M, Matsuura S, Awaya N, Noguchi T, et al. Delineation of the gastric muscularis mucosae and assessment of depth of invasion of early gastric cancer using a 20-megahertz endoscopic ultrasound probe. Gastrointest Endosc 1993;39:505-12. 12. Yanai H, Tada M, Karita M, Okita K. Diagnostic utility of 2O-megahertz linear endoscopic ultrasonography in early gastric cancer. Gastrointest Endosc 1996;44:29-33. 13. Ma~atka Z. Gastroscopy. In: Ma~atka Z, editor. Terminology, definitions and diagnostic criteria in digestive endoscopy. 2nd edition. Bad Homburg, Germany: Normed Verlag; 1989. p. 47-59. 14. Japanese Research Society for Gastric Cancer. Japanese Classification of Gastric Carcinoma. 1st English edition. Tokyo: Kanehara & Co., Ltd.; 1995. p. 37-71. 15. Sano T, Okuyama Y, Kobori O, Shimizu T, Morioka Y. Early gastric cancer: endoscopic diagnosis of depth of invasion. Dig Dis Sci 1990;35:1340-4. 16. Aibe T, Fuji T, Okita K, Takemoto T. A fundamental study of normal layer structure of the gastrointestinal wall visualized by endoscopic ultrasonography. Scand J Gastroenterol 1986;21 (Suppl 123):6-15. 17. Kimmey MB, Martin RW,Haggitt RC, Wang KY, Franklin DW, Silverstein FE. Histologic correlates of gastrointestinal ultrasound images. Gastroenterology 1989;96:433-41. 18. Shirao K, Saito D, Yamaguchi H, Kido M, Fukuda H, Yoshida S. Accuracy of preoperative diagnosis of early gastric cancer on submucosal invasion: a study for indications of endoscopic treatment. Stomach and Intestine 1992;27:1175-84. 19. Yasuda K. Endoscopic ultrasonic probes and mucosectomy for early gastric carcinoma. Gastrointest Endosc 1996;43:$29-31.
VOLUME 46, NO. 3, 1997