- Email: [email protected]
High-frequency probe EUS-assisted endoscopic mucosal resection: A therapeutic strategy for submucosal tumors of the GI tract
High-frequency probe EUS-assisted endoscopic mucosal resection: a therapeutic strategy for submucosal tumors of the GI tract Irving Waxman, MD, Yusuke Saitoh, MD, Gottumukkala S. Raju, MD, MRCP, Jiro Watari, MD, Kinichi Yokota, MD, Angela L. Reeves BSN, CGRN, Yutaka Kohgo, MD Galveston, Texas, and Asahikawa, Japan
Background: The aim of this study was to evaluate the efficacy and safety of high-frequency probe EUS (HFPE)-assisted endoscopic mucosal resection in the management of submucosal tumors of the GI tract. Methods: HFPE-assisted endoscopic mucosal resection was attempted in 28 patients with submucosal tumors less than 2 cm in diameter. HFPE was performed with a 20-MHz “through-thescope” probe. Saline solution was injected into the submucosa. After confirming detachment of the lesion from the muscularis propria by repeat HFPE, endoscopic mucosal resection was performed with a lift-and-cut or endoscopic mucosal resection cap technique. Follow-up endoscopy was performed in all patients. Results: Submucosal tumors from the following areas were included: esophagus 3, stomach 4, duodenum 3, and colon 18. The submucosal tumors were located in the upper third (n = 3), middle third (n = 18), and lower third (n = 7) of the submucosa. Twenty-one submucosal tumors were removed by the lift-and-cut technique and 6 by the cap method. One patient required surgical resection after unsuccessful endoscopic mucosal resection. The origin and depth of penetration of all lesions was accurately depicted by HFPE. Median tumor diameter was 9 mm (range 3-20 mm). Resection was successful and complete in 93% of the cases. There were no immediate postprocedure complications (exact 95% CI [0%, 12.3%]). During a median follow-up of 21.5 months (range 2-74 months) no recurrence was found. Conclusions: HFPE-assisted endoscopic mucosal resection is safe and effective for the management of selected submucosal tumors of the GI tract. A management algorithm based on endoscopic and HFPE findings is proposed. (Gastrointest Endosc 2002;55:44-9.)
Submucosal tumors (SMTs) of the GI tract are uncommon and are usually found incidentally at endoscopy. Until recently, management consisted of 2 relatively unsatisfactory options: (1) observation without a definitive tissue diagnosis or (2) surgical resection. Two important technical advances in endoscopy during the last decade, EUS and endoscopic mucosal resection (EMR), have the potential to impact management of SMTs. Although EUS has revolutionized the staging of GI tumors, imaging of small SMTs can be technicalReceived December 21, 2000. For revision March 6, 2001. Accepted April 12, 2001. From the University of Texas Medical Branch at Galveston, Texas, Asahikawa Medical College, Asahikawa, Japan, and Kansas University Medical Center, Kansas City, Kansas. Presented at the annual Digestive Disease week, May 16-19, 1999, Orlando, Florida (Gastrointest Endosc 1999;49:AB133). Supported in part by a technological grant from Fujinon Photo Optical, Omiya, Japan. Reprint requests: Irving Waxman, MD, University of Chicago, DCAM Room 66011, 5758 S. Maryland Ave., MC 9028, Chicago, IL 60637. Copyright © 2002 by the American Society for Gastrointestinal Endoscopy 0016-5107/2002/$35.00 + 0 37/1/119871 doi:10.1067/mge.2002.119871 44
GASTROINTESTINAL ENDOSCOPY
ly demanding with conventional instruments. In addition, up to 3 intubations are required: the first for initial identification of the submucosal lesion with a forward-viewing endoscope, the second for US to delineate transmural extent, and the third for therapeutic intervention. With recent advances in US technology, all of these steps can potentially be accomplished with a single intubation with a “through-the-scope” catheter US probe. These devices provide high-resolution images of the GI wall and allow accurate assessment of lesions to a depth of 15 to 20 mm under direct endoscopic visualization.1-3 In addition, the catheter US probe may be a useful adjunct to EMR.4 EMR, a major therapeutic technical advancement in the management of superficial GI tumors, is widely used in Japan5-10 and increasingly in the United States.11 EMR involves lifting of a lesion from the muscular coat of the GI tract, either by injection of saline solution or suction of the lesion into a cap fitted to the endoscope, followed by electrosurgical snare resection. The saline solution cushion between the lesion and the muscle layer not only facilitates wide and deep excision but also VOLUME 55, NO. 1, 2002
GI submucosal tumors: high-frequency probe EUS-assisted EMR
Figure 1. Schematic representation of classification of SMTs: SM1, limited to upper third of submucosa; SM2, involves middle third of submucosa; SM3, involves lower third of submucosa. m, Mucosa; mm, muscularis mucosa; sm, submucosa; mp, muscularis propria.
reduces the risk of perforation by separating the lesion from the muscular layer, which thereby excludes muscle from the snare resection. Complete separation of the lesion from the muscle layer can be quickly confirmed by imaging the lesion with a catheter US probe. Despite these theoretical considerations, there is little information on the role of high-frequency probe EUS (HFPE)-assisted EMR in the management of SMTs. The primary aims of this study were to assess the role of HFPE-assisted EMR in tissue diagnosis and prediction of complete resection of SMTs with tumor-free margins. Secondary aims were to evaluate the efficacy of resection of SMTs depending on the extent of submucosal involvement and location in the GI tract, the complications of SMT resection, and tumor recurrence during follow-up. All images obtained by HFPE were reviewed to identify specific sonographic criteria that could be used to select lesions for EMR and those that can be observed without treatment. PATIENTS AND METHODS Twenty-eight patients (15 men, 13 women; age range: 33-75 years) with submucosal lesions of the GI tract underwent HFPE and HFPE-assisted EMR. Two investigators (I.W., Y.S.) performed all procedures and reviewed all images. Criteria for HFPE-assisted EMR were as follows: (1) diameter 2 cm or less, (2) normal overlying mucosa without ulceration, and (3) lack of extension into the muscularis propria. Exclusion criteria were as follows: (1) size larger than 2 cm (because of attenuation of the US signal leading to poor imaging of the deepest part of the tumor) and (2) bleeding diathesis. Informed consent was obtained from all patients for all procedures described. Endoscopy was performed with either a single or a double channel therapeutic endoscope with a 3.2-mm diameter accessory channel (GIF-1T-140, GIF-2T-140, Olympus America, Inc., Melville, N.Y.). The submucosal lesion was imaged at 20 MHZ with a 7F catheter probe (SP-501 or 701, Fujinon Corp., Tokyo, Japan; UM3-R, Olympus) by using the water-filling technique. The size, site of origin, extent of submucosal involvement, and ultrasonographic characteristics of the lesion were noted. VOLUME 55, NO. 1, 2002
I Waxman, Y Saitoh, G Raju, et al.
Figure 2. Diagrammatic representation of HFPE-assisted EMR: 1, after HFPE, injection needle is placed at base of lesion; 2, saline solution injected beneath SMT to lift it from muscularis propria; 3, HFPE to confirm detachment of SMT from muscularis propria; 4, SMT lifted with grasping forceps; 5, electrosurgical snare resection; 6, completion of EMR. m, Mucosa; sm, submucosa; mp, muscularis propria. The submucosal lesions were categorized according to the Japanese Cancer Society Classification, which recognizes 3 types depending on location with the submucosa: (1) SM1, tumor limited to the upper third; (2) SM2, limited to middle third; and (3) SM3, tumor involving the deep portion of submucosa (Fig. 1).12,13 Once HFPE was completed, saline solution was injected with a 23-G needle (NM-201-L-0423, Olympus) just beneath the tumor to detach it from the muscularis propria. After confirming complete detachment by repeat HFPE (infiltration of saline solution beneath SMT), EMR was performed by using the lift-and-cut technique or with the use of an EMR cap9,11 (Fig. 2). Histologic diagnosis was confirmed in all cases. All patients were closely observed for any immediate procedure-related complications and followed endoscopically for local recurrence of the tumor.
RESULTS HFPE-assisted EMR of an SMT was successful in 27 of 28 patients. Of these lesions, 21 were resected with the lift-and-cut method and 6 with the cap technique. The 1 patient in whom HFPE-assisted EMR was unsuccessful required surgical resection. The volume of saline solution required to lift the SMTs from the muscularis propria varied between 5 and 20 mL. All resections were performed during a single endoscopic session. SMT characteristics including location (histologic layer), site (in relation to sonographic layer structure of gut wall), histology, and status of the resection margin are shown in Table 1. Twenty-seven of the 28 lesions were completely resected. There was no difference in resection rate with respect to location within the submucosa (SM1, SM2, and SM3 lesions). The resection margin was tumor free in 26 of 28 cases. In the 1 patient in whom HFPE-assisted EMR was unsuccessful, a rectal GASTROINTESTINAL ENDOSCOPY
45
I Waxman, Y Saitoh, G Raju, et al.
GI submucosal tumors: high-frequency probe EUS-assisted EMR
Table 1. Characteristics and resectability of SMTs by HFPE-assisted EMR
Patient No. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28
Location Esophagus Esophagus Esophagus Stomach Stomach Stomach Stomach Duodenum Duodenum Duodenum Descending colon Sigmoid colon Sigmoid colon Sigmoid colon Sigmoid colon Rectum Rectum Rectum Rectum Rectum Rectum Rectum Rectum Rectum Rectum Rectum Rectum Rectum
SM status by HFPE
Size (sq mm)
Histology
Resection margin involvement
SM2 SM2 SM2 SM1 SM2 SM3 SM3 SM2 SM3 SM3 SM2 SM2 SM2 SM2 SM3 SM1 SM2 SM2 SM2 SM2 SM2 SM2 SM3 SM2 SM2 SM3 SM2 M3
7 7 11 6 20 10 20 10 5 7 9 20 5 5 11 3 6 3 5 7 5 11 9 5 20 10 10 3
Leiomyoma Leiomyoma Lymphangioma Carcinoid Carcinoid Fibrovascular polyp Heterotopic pancreas Lipoma Carcinoid Carcinoid* Fibrovascular polyp Lipoma Leiomyoma Leiomyoma Leiomyoma Carcinoid Carcinoid Carcinoid Carcinoid Carcinoid Carcinoid Carcinoid Carcinoid† Carcinoid Hemangioma Leiomyoma Carcinoid Carcinoid
Negative Negative Negative Negative Negative Negative Negative Negative Negative Negative Negative Negative Negative Negative Negative Negative Positive Negative Negative Negative Negative Negative Surgery Negative Negative Negative Negative Negative
*EMR specimen could not be retrieved. †Transanal resection.
lesion could not be lifted completely from the muscularis propria by injection of saline solution. The lesion, a carcinoid tumor, was excised transanally. In another patient with a rectal submucosal lesion, histopathologic evaluation of the EMR specimen revealed a carcinoid tumor with positive margins, probably because of inadequate separation of the lesion from the muscularis propria. Histologic diagnoses, confirmed in all cases, were as follows: carcinoid tumor 15, leiomyoma 6, lipoma 2, hemangioma 1, lymphangioma 1, fibrovascular polyp 1, inflammatory fibroid polyp 1, and heterotopic pancreas 1. All leiomyomas arose from the muscularis mucosa. There were no immediate postprocedure complications of EMR (exact 95% CI [0%, 12.3%]). Median follow-up was 21.5 months (range 2-74 months). No local recurrence was noted at follow-up endoscopy. Review of the HFPE images revealed that all carcinoid tumors, whether arising in the upper or lower GI tract, had similar sonographic characteristics (well demarcated, round, hypoechoic “salt and pepper” appearance; Fig. 3). In contrast, all stromal tumors were homogenous and hypoechoic (less 46
GASTROINTESTINAL ENDOSCOPY
echogenic than carcinoid tumors) and could be traced to the muscularis mucosae in 4 out of 6 cases (all colonic) (Fig. 4). Lipomas appeared as hyperechoic and homogeneous. DISCUSSION Recent advances in endoscopic technology and techniques, namely HFPE and EMR, have had a major impact on the diagnosis and management of superficial and early stage cancers of the esophagus, stomach, and colon. Because of their ease of application at initial diagnostic endoscopy, these 2 modalities are being used increasingly in Japan and the United States.11,14-18 However, there is scant data on certain aspects of the endoscopic diagnosis and, in particular, management of SMTs of the GI tract. Matsui et al.19 performed EUS-guided fine needle aspiration (FNA) with a convex array echoendoscope and a 22-G needle in 24 patients with SMTs. Mean tumor diameter was 40 mm (range 18-95 mm). Specimens adequate for histopathologic evaluation were obtained in 20 (83%) patients. Seventeen patients underwent surgical resection; in 16, EUSFNA had provided a correct diagnosis. Kojima et VOLUME 55, NO. 1, 2002
GI submucosal tumors: high-frequency probe EUS-assisted EMR
A
C
I Waxman, Y Saitoh, G Raju, et al.
B
D
Figure 3. A, Endoscopic image of rectosigmoid carcinoid tumor. B, HFPE image showing well circumscribed, “salt and pepperlike” hypoechoic SMT (SM2). C, Photomicrography showing carcinoid tumor in submucosa with clear deep margin (H&E, orig. mag. ×5). D, Endoscopic image of mucosal filling defect after EMR sealed with endoclips.
al.20 performed conventional EUS-guided EMR in 51 patients with SMTs to delineate the lesion and rule out malignancy. EMR of lesions originating in the muscularis mucosa or submucosa was attempted. EUS and histopathologic findings coincided in 73% of cases. Kawamoto et al.10 endoscopically resected submucosal tumors after saline solution injection in 45 patients. EUS was performed with a conventional echoendoscope. Complete resection was achieved in all cases. In 5 patients (11%), mild to moderate bleeding was encountered. HFPE accurately delineated the nature and depth of penetration of all submucosal lesions in the present study. Based on this information, EMR was successful and complete for 93% of the SMTs (26/28) and without complication. Inability to separate the lesion from the muscularis propria by saline solution injection was associated with incomplete resection in 2 cases. The latter phenomenon might be used to predict resectability and thereby obviate the need for further imaging. VOLUME 55, NO. 1, 2002
The results in the present study suggest that deeply seated submucosal lesions (SM3), as delineated by HFPE, are equally amenable to EMR as compared with more superficial lesions. Tumor resection was complete with resection margins free of tumor in all but 2 patients. Early in our experience, the resection of a 6-mm diameter rectal carcinoid tumor was incomplete (positive margin). This could have been due to technical problems, including leakage of saline solution with loss of the submucosal cushion or insufficient separation of the lesion from the muscularis propria. In another patient with a 9-mm diameter SM3 carcinoid tumor in the distal rectum near to the dentate line, there was no detachment of the lesion from the muscularis propria after injection of saline solution into the submucosa. The procedure was aborted and the patient underwent transanal resection without immediate complication. On follow-up, there was no tumor recurrence. From a technical standpoint, HFPE-assisted EMR has several distinct advantages over EUSGASTROINTESTINAL ENDOSCOPY
47
I Waxman, Y Saitoh, G Raju, et al.
A
C
GI submucosal tumors: high-frequency probe EUS-assisted EMR
B
D
Figure 4. A, Endoscopic view of colonic submucosal lesion. B, HFPE image showing hypoechoic lesion in middle third of submucosa (SM2). C, HFPE image demonstrating separation of lesion from muscularis propria after submucosal injection of saline solution. D, Photomicrograph of leiomyoma arising from muscularis propria (H&E, orig. mag. ×5).
assisted EMR. The entire procedure can be completed with a single intubation. The lesion can be imaged with a “through-the-scope” US probe and then resected by using the same endoscope, thereby reducing procedure time and patient discomfort. As has been demonstrated for lesions under 2 cm in diameter, imaging with a catheter US probe under simultaneous endoscopic guidance is technically less demanding compared with imaging with a conventional echoendoscope because it is easier to locate the lesion.16 For lesions over 2 cm in diameter that cannot be fully evaluated with a 20-MHZ probe because of attenuation of the US signal, a “throughthe-scope” probe with a lower imaging frequency can be used. HFPE-assisted EMR provides a large specimen for histopathologic diagnosis as well as assessment of the margin of resection. As demonstrated in the present study, this procedure offers definitive therapy for SMTs up to 20 mm in diameter. For larger lesions for which HFPE-assisted EMR is not possi48
GASTROINTESTINAL ENDOSCOPY
ble, a tissue diagnosis can often be obtained by EUSguided FNA or by obtaining biopsy specimens after removal of the overlying mucosa. Based on the results of the present study, the following therapeutic strategy is proposed for SMTs based on a combination of endoscopic and HFPE characteristics. Well-demarcated hypoechoic lesions up to 2 cm in diameter without umbilication (suggestive of pancreatic rest) on endoscopic examination should be removed by EMR given the high likelihood that the lesion is a carcinoid tumor. In the colon, in addition to carcinoid tumors, stromal tumors, which have the potential for malignant degeneration, should also be resected. Hyperechoic homogeneous lesions or SMTs with or without a clearly evident cystic component can be followed because these are highly likely to be benign histopathologically (e.g., pancreatic rest, lipomas). Kameyama et al.,21 based on experience with conventional EUS and catheter US probe EUS in patients with colonic SMTs, concluded that informaVOLUME 55, NO. 1, 2002
GI submucosal tumors: high-frequency probe EUS-assisted EMR
I Waxman, Y Saitoh, G Raju, et al.
tion provided by EUS was useful in choosing therapy for SMTs of the large intestine. These investigators recommended observation for SMTs with benign ultrasonographic characteristics (lipoma, lymphangioma), and EMR for other lesions depending on size and depth of penetration. In summary, SMTs of the GI tract can be managed safely, quickly, and easily with HFPE-assisted EMR. A treatment algorithm for SMTs has been suggested, based on HFPE and endoscopic findings. However, this proposed approach must be validated in prospective studies.
tumorectomy for gastrointestinal submucosal tumors restricted to the submucosa: a new form of endoscopic minimal surgery. Gastrointest Endosc 1997;46:311-7. Waxman I, Saitoh Y. Clinical outcome of endoscopic mucosal resection for superficial gastrointestinal lesions and the role of high frequency ultrasound probe sonography in an American population. Gastrointest Endosc 2000;52:322-7. Japanese Research Society for Cancer of the Colon and Rectum. General rules for clinical and pathological studies on cancer of the colon, rectum and anus: histopathological classification. 5th ed. Tokyo: Kauehara Syuppan; 1994. p. 50-73. Kudo S. Endoscopic mucosal resection of flat and depressed types of early colorectal cancer. Endoscopy 1993;25:455-61. Hasegawa N, Niwa Y, Arisawa T, Hase S, Goto H, Hayakawa T. Preoperative staging of superficial esophageal carcinoma: comparison of an ultrasound probe and standard endoscopic ultrasonography. Gastrointest Endosc 1996;44:388-93. Yanai H, Yoshida T, Harada T, Matsumoto Y, Nishiaki M, Shigemitsu T, et al. Endoscopic ultrasonography of superficial esophageal cancers using a thin ultrasound probe system equipped with switchable radial and linear scanning modes. Gastrointest Endosc 1996;44:578-82. Chak A, Canto M, Stevens PD, Lightdale CJ, de Mierop FV, Cooper G, et al. Clinical applications of a new through-the scope ultrasound probe: prospective comparison with an ultrasound endoscope. Gastrointest Endosc 1997;45:291-5. Yoshida M, Tsukamoto Y, Niwa Y, Goto H, Hase S, Hayakawa T, et al. Endoscopic assessment of invasion of colorectal tumors with a new high-frequency ultrasound probe. Gastrointest Endosc 1995;41:587-92. Saitoh Y, Obara T, Einami K, Nomura M, Taruishi M, Ayabe T, et al. Efficacy of high-frequency ultrasound probes for preoperative staging of invasion depth in flat and depressed colorectal tumors. Gastrointest Endosc 1996;44:34-9. Matsui M, Niwa Y, Arisawa T, Hirooka Y, Kameyama H, Hayakawa M, et al. Endoscopic ultrasound-guided fine needle aspiration biopsy (EUS-FNAB) in the diagnosis of upper gastrointestinal submucosal tumors [abstract]. Gastroenterology 1998;114:A220. Kojima T, Takahashi H, Parra-Blanco A, Ohguri S, Fujita R. Diagnosis of submucosal tumors of the upper gastrointestinal tract by endosonography assessed endoscopic resection [abstract]. Gastroenterology 1998;114:A23. Kameyama H, Niwa Y, Arisawa T, Goto H, Hayakawa T. Endoscopic ultrasonography in the diagnosis of submucosal lesions of the large intestine. Gastrointest Endosc 1997;46: 406-11.
11.
12.
13. 14.
REFERENCES 1. Lightdale CJ. Diagnosis, staging, and cure of early gastrointestinal cancer. Gastrointest Endosc 1996;44:95-6. 2. Silverstein FE, Martin RW, Kimmey MB, Jiranek GC, Franklin DW, Proctor A. Experimental evaluation of an endoscopic ultrasound probe. Gastroenterology 1989;96:1058-62. 3. Rösch T, Classen M. A new ultrasonic probe for endosonographic imaging of the upper GI tract. Endoscopy 1990;20:41-6. 4. Waxman I. Clinical impact of high-frequency ultrasound probe sonography during diagnostic endoscopy—a prospective study. Endoscopy 1998;30:A166-8. 5. Karita M, Tada M, Okita K, Kodam T. Endoscopic therapy for early colon cancer: the strip biopsy resection technique. Gastrointest Endosc 1991;37:128-32. 6. Tada M, Inoue H, Yabata E, Okabe S, Endo M. Colonic mucosal resection using a transparent cap-fitted endoscope. Gastrointest Endosc 1996;44:63-5. 7. Sakai P, Filho FM, Iryia K, Moura EGH, Tomishigue T, Scabba A, et al. An endoscopic technique for resection of small gastrointestinal carcinoma. Gastrointest Endosc 1996; 44:65-8. 8. Fleischer DE, Wang GQ, Dawsey S, Tio TL, Newsome J, Kidwell J, et al. Tissue band ligation followed by snare resection (band and snare): a new technique for tissue acquisition in the esophagus. Gastrointest Endosc 1996;44:68-72. 9. Inoue H, Takeshita K, Hori H, Muraoka Y, Yoneshima H, Endo M, et al. Endoscopic mucosal resection with a cap-fitted panendoscope for esophagus, stomach, and colon mucosal lesions. Gastrointest Endosc 1993;39:58-62. 10. Kawamoto K, Yamada Y, Furukawa N, Utsunomiya T, Hataguchi Y, Mizuguchi M, et al. Endoscopic submucosal
VOLUME 55, NO. 1, 2002
15.
16.
17.
18.
19.
20.
21.
GASTROINTESTINAL ENDOSCOPY
49
Background: The aim of this study was to evaluate the efficacy and safety of high-frequency probe EUS (HFPE)-assisted endoscopic mucosal resection in the management of submucosal tumors of the GI tract. Methods: HFPE-assisted endoscopic mucosal resection was attempted in 28 patients with submucosal tumors less than 2 cm in diameter. HFPE was performed with a 20-MHz “through-thescope” probe. Saline solution was injected into the submucosa. After confirming detachment of the lesion from the muscularis propria by repeat HFPE, endoscopic mucosal resection was performed with a lift-and-cut or endoscopic mucosal resection cap technique. Follow-up endoscopy was performed in all patients. Results: Submucosal tumors from the following areas were included: esophagus 3, stomach 4, duodenum 3, and colon 18. The submucosal tumors were located in the upper third (n = 3), middle third (n = 18), and lower third (n = 7) of the submucosa. Twenty-one submucosal tumors were removed by the lift-and-cut technique and 6 by the cap method. One patient required surgical resection after unsuccessful endoscopic mucosal resection. The origin and depth of penetration of all lesions was accurately depicted by HFPE. Median tumor diameter was 9 mm (range 3-20 mm). Resection was successful and complete in 93% of the cases. There were no immediate postprocedure complications (exact 95% CI [0%, 12.3%]). During a median follow-up of 21.5 months (range 2-74 months) no recurrence was found. Conclusions: HFPE-assisted endoscopic mucosal resection is safe and effective for the management of selected submucosal tumors of the GI tract. A management algorithm based on endoscopic and HFPE findings is proposed. (Gastrointest Endosc 2002;55:44-9.)
Submucosal tumors (SMTs) of the GI tract are uncommon and are usually found incidentally at endoscopy. Until recently, management consisted of 2 relatively unsatisfactory options: (1) observation without a definitive tissue diagnosis or (2) surgical resection. Two important technical advances in endoscopy during the last decade, EUS and endoscopic mucosal resection (EMR), have the potential to impact management of SMTs. Although EUS has revolutionized the staging of GI tumors, imaging of small SMTs can be technicalReceived December 21, 2000. For revision March 6, 2001. Accepted April 12, 2001. From the University of Texas Medical Branch at Galveston, Texas, Asahikawa Medical College, Asahikawa, Japan, and Kansas University Medical Center, Kansas City, Kansas. Presented at the annual Digestive Disease week, May 16-19, 1999, Orlando, Florida (Gastrointest Endosc 1999;49:AB133). Supported in part by a technological grant from Fujinon Photo Optical, Omiya, Japan. Reprint requests: Irving Waxman, MD, University of Chicago, DCAM Room 66011, 5758 S. Maryland Ave., MC 9028, Chicago, IL 60637. Copyright © 2002 by the American Society for Gastrointestinal Endoscopy 0016-5107/2002/$35.00 + 0 37/1/119871 doi:10.1067/mge.2002.119871 44
GASTROINTESTINAL ENDOSCOPY
ly demanding with conventional instruments. In addition, up to 3 intubations are required: the first for initial identification of the submucosal lesion with a forward-viewing endoscope, the second for US to delineate transmural extent, and the third for therapeutic intervention. With recent advances in US technology, all of these steps can potentially be accomplished with a single intubation with a “through-the-scope” catheter US probe. These devices provide high-resolution images of the GI wall and allow accurate assessment of lesions to a depth of 15 to 20 mm under direct endoscopic visualization.1-3 In addition, the catheter US probe may be a useful adjunct to EMR.4 EMR, a major therapeutic technical advancement in the management of superficial GI tumors, is widely used in Japan5-10 and increasingly in the United States.11 EMR involves lifting of a lesion from the muscular coat of the GI tract, either by injection of saline solution or suction of the lesion into a cap fitted to the endoscope, followed by electrosurgical snare resection. The saline solution cushion between the lesion and the muscle layer not only facilitates wide and deep excision but also VOLUME 55, NO. 1, 2002
GI submucosal tumors: high-frequency probe EUS-assisted EMR
Figure 1. Schematic representation of classification of SMTs: SM1, limited to upper third of submucosa; SM2, involves middle third of submucosa; SM3, involves lower third of submucosa. m, Mucosa; mm, muscularis mucosa; sm, submucosa; mp, muscularis propria.
reduces the risk of perforation by separating the lesion from the muscular layer, which thereby excludes muscle from the snare resection. Complete separation of the lesion from the muscle layer can be quickly confirmed by imaging the lesion with a catheter US probe. Despite these theoretical considerations, there is little information on the role of high-frequency probe EUS (HFPE)-assisted EMR in the management of SMTs. The primary aims of this study were to assess the role of HFPE-assisted EMR in tissue diagnosis and prediction of complete resection of SMTs with tumor-free margins. Secondary aims were to evaluate the efficacy of resection of SMTs depending on the extent of submucosal involvement and location in the GI tract, the complications of SMT resection, and tumor recurrence during follow-up. All images obtained by HFPE were reviewed to identify specific sonographic criteria that could be used to select lesions for EMR and those that can be observed without treatment. PATIENTS AND METHODS Twenty-eight patients (15 men, 13 women; age range: 33-75 years) with submucosal lesions of the GI tract underwent HFPE and HFPE-assisted EMR. Two investigators (I.W., Y.S.) performed all procedures and reviewed all images. Criteria for HFPE-assisted EMR were as follows: (1) diameter 2 cm or less, (2) normal overlying mucosa without ulceration, and (3) lack of extension into the muscularis propria. Exclusion criteria were as follows: (1) size larger than 2 cm (because of attenuation of the US signal leading to poor imaging of the deepest part of the tumor) and (2) bleeding diathesis. Informed consent was obtained from all patients for all procedures described. Endoscopy was performed with either a single or a double channel therapeutic endoscope with a 3.2-mm diameter accessory channel (GIF-1T-140, GIF-2T-140, Olympus America, Inc., Melville, N.Y.). The submucosal lesion was imaged at 20 MHZ with a 7F catheter probe (SP-501 or 701, Fujinon Corp., Tokyo, Japan; UM3-R, Olympus) by using the water-filling technique. The size, site of origin, extent of submucosal involvement, and ultrasonographic characteristics of the lesion were noted. VOLUME 55, NO. 1, 2002
I Waxman, Y Saitoh, G Raju, et al.
Figure 2. Diagrammatic representation of HFPE-assisted EMR: 1, after HFPE, injection needle is placed at base of lesion; 2, saline solution injected beneath SMT to lift it from muscularis propria; 3, HFPE to confirm detachment of SMT from muscularis propria; 4, SMT lifted with grasping forceps; 5, electrosurgical snare resection; 6, completion of EMR. m, Mucosa; sm, submucosa; mp, muscularis propria. The submucosal lesions were categorized according to the Japanese Cancer Society Classification, which recognizes 3 types depending on location with the submucosa: (1) SM1, tumor limited to the upper third; (2) SM2, limited to middle third; and (3) SM3, tumor involving the deep portion of submucosa (Fig. 1).12,13 Once HFPE was completed, saline solution was injected with a 23-G needle (NM-201-L-0423, Olympus) just beneath the tumor to detach it from the muscularis propria. After confirming complete detachment by repeat HFPE (infiltration of saline solution beneath SMT), EMR was performed by using the lift-and-cut technique or with the use of an EMR cap9,11 (Fig. 2). Histologic diagnosis was confirmed in all cases. All patients were closely observed for any immediate procedure-related complications and followed endoscopically for local recurrence of the tumor.
RESULTS HFPE-assisted EMR of an SMT was successful in 27 of 28 patients. Of these lesions, 21 were resected with the lift-and-cut method and 6 with the cap technique. The 1 patient in whom HFPE-assisted EMR was unsuccessful required surgical resection. The volume of saline solution required to lift the SMTs from the muscularis propria varied between 5 and 20 mL. All resections were performed during a single endoscopic session. SMT characteristics including location (histologic layer), site (in relation to sonographic layer structure of gut wall), histology, and status of the resection margin are shown in Table 1. Twenty-seven of the 28 lesions were completely resected. There was no difference in resection rate with respect to location within the submucosa (SM1, SM2, and SM3 lesions). The resection margin was tumor free in 26 of 28 cases. In the 1 patient in whom HFPE-assisted EMR was unsuccessful, a rectal GASTROINTESTINAL ENDOSCOPY
45
I Waxman, Y Saitoh, G Raju, et al.
GI submucosal tumors: high-frequency probe EUS-assisted EMR
Table 1. Characteristics and resectability of SMTs by HFPE-assisted EMR
Patient No. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28
Location Esophagus Esophagus Esophagus Stomach Stomach Stomach Stomach Duodenum Duodenum Duodenum Descending colon Sigmoid colon Sigmoid colon Sigmoid colon Sigmoid colon Rectum Rectum Rectum Rectum Rectum Rectum Rectum Rectum Rectum Rectum Rectum Rectum Rectum
SM status by HFPE
Size (sq mm)
Histology
Resection margin involvement
SM2 SM2 SM2 SM1 SM2 SM3 SM3 SM2 SM3 SM3 SM2 SM2 SM2 SM2 SM3 SM1 SM2 SM2 SM2 SM2 SM2 SM2 SM3 SM2 SM2 SM3 SM2 M3
7 7 11 6 20 10 20 10 5 7 9 20 5 5 11 3 6 3 5 7 5 11 9 5 20 10 10 3
Leiomyoma Leiomyoma Lymphangioma Carcinoid Carcinoid Fibrovascular polyp Heterotopic pancreas Lipoma Carcinoid Carcinoid* Fibrovascular polyp Lipoma Leiomyoma Leiomyoma Leiomyoma Carcinoid Carcinoid Carcinoid Carcinoid Carcinoid Carcinoid Carcinoid Carcinoid† Carcinoid Hemangioma Leiomyoma Carcinoid Carcinoid
Negative Negative Negative Negative Negative Negative Negative Negative Negative Negative Negative Negative Negative Negative Negative Negative Positive Negative Negative Negative Negative Negative Surgery Negative Negative Negative Negative Negative
*EMR specimen could not be retrieved. †Transanal resection.
lesion could not be lifted completely from the muscularis propria by injection of saline solution. The lesion, a carcinoid tumor, was excised transanally. In another patient with a rectal submucosal lesion, histopathologic evaluation of the EMR specimen revealed a carcinoid tumor with positive margins, probably because of inadequate separation of the lesion from the muscularis propria. Histologic diagnoses, confirmed in all cases, were as follows: carcinoid tumor 15, leiomyoma 6, lipoma 2, hemangioma 1, lymphangioma 1, fibrovascular polyp 1, inflammatory fibroid polyp 1, and heterotopic pancreas 1. All leiomyomas arose from the muscularis mucosa. There were no immediate postprocedure complications of EMR (exact 95% CI [0%, 12.3%]). Median follow-up was 21.5 months (range 2-74 months). No local recurrence was noted at follow-up endoscopy. Review of the HFPE images revealed that all carcinoid tumors, whether arising in the upper or lower GI tract, had similar sonographic characteristics (well demarcated, round, hypoechoic “salt and pepper” appearance; Fig. 3). In contrast, all stromal tumors were homogenous and hypoechoic (less 46
GASTROINTESTINAL ENDOSCOPY
echogenic than carcinoid tumors) and could be traced to the muscularis mucosae in 4 out of 6 cases (all colonic) (Fig. 4). Lipomas appeared as hyperechoic and homogeneous. DISCUSSION Recent advances in endoscopic technology and techniques, namely HFPE and EMR, have had a major impact on the diagnosis and management of superficial and early stage cancers of the esophagus, stomach, and colon. Because of their ease of application at initial diagnostic endoscopy, these 2 modalities are being used increasingly in Japan and the United States.11,14-18 However, there is scant data on certain aspects of the endoscopic diagnosis and, in particular, management of SMTs of the GI tract. Matsui et al.19 performed EUS-guided fine needle aspiration (FNA) with a convex array echoendoscope and a 22-G needle in 24 patients with SMTs. Mean tumor diameter was 40 mm (range 18-95 mm). Specimens adequate for histopathologic evaluation were obtained in 20 (83%) patients. Seventeen patients underwent surgical resection; in 16, EUSFNA had provided a correct diagnosis. Kojima et VOLUME 55, NO. 1, 2002
GI submucosal tumors: high-frequency probe EUS-assisted EMR
A
C
I Waxman, Y Saitoh, G Raju, et al.
B
D
Figure 3. A, Endoscopic image of rectosigmoid carcinoid tumor. B, HFPE image showing well circumscribed, “salt and pepperlike” hypoechoic SMT (SM2). C, Photomicrography showing carcinoid tumor in submucosa with clear deep margin (H&E, orig. mag. ×5). D, Endoscopic image of mucosal filling defect after EMR sealed with endoclips.
al.20 performed conventional EUS-guided EMR in 51 patients with SMTs to delineate the lesion and rule out malignancy. EMR of lesions originating in the muscularis mucosa or submucosa was attempted. EUS and histopathologic findings coincided in 73% of cases. Kawamoto et al.10 endoscopically resected submucosal tumors after saline solution injection in 45 patients. EUS was performed with a conventional echoendoscope. Complete resection was achieved in all cases. In 5 patients (11%), mild to moderate bleeding was encountered. HFPE accurately delineated the nature and depth of penetration of all submucosal lesions in the present study. Based on this information, EMR was successful and complete for 93% of the SMTs (26/28) and without complication. Inability to separate the lesion from the muscularis propria by saline solution injection was associated with incomplete resection in 2 cases. The latter phenomenon might be used to predict resectability and thereby obviate the need for further imaging. VOLUME 55, NO. 1, 2002
The results in the present study suggest that deeply seated submucosal lesions (SM3), as delineated by HFPE, are equally amenable to EMR as compared with more superficial lesions. Tumor resection was complete with resection margins free of tumor in all but 2 patients. Early in our experience, the resection of a 6-mm diameter rectal carcinoid tumor was incomplete (positive margin). This could have been due to technical problems, including leakage of saline solution with loss of the submucosal cushion or insufficient separation of the lesion from the muscularis propria. In another patient with a 9-mm diameter SM3 carcinoid tumor in the distal rectum near to the dentate line, there was no detachment of the lesion from the muscularis propria after injection of saline solution into the submucosa. The procedure was aborted and the patient underwent transanal resection without immediate complication. On follow-up, there was no tumor recurrence. From a technical standpoint, HFPE-assisted EMR has several distinct advantages over EUSGASTROINTESTINAL ENDOSCOPY
47
I Waxman, Y Saitoh, G Raju, et al.
A
C
GI submucosal tumors: high-frequency probe EUS-assisted EMR
B
D
Figure 4. A, Endoscopic view of colonic submucosal lesion. B, HFPE image showing hypoechoic lesion in middle third of submucosa (SM2). C, HFPE image demonstrating separation of lesion from muscularis propria after submucosal injection of saline solution. D, Photomicrograph of leiomyoma arising from muscularis propria (H&E, orig. mag. ×5).
assisted EMR. The entire procedure can be completed with a single intubation. The lesion can be imaged with a “through-the-scope” US probe and then resected by using the same endoscope, thereby reducing procedure time and patient discomfort. As has been demonstrated for lesions under 2 cm in diameter, imaging with a catheter US probe under simultaneous endoscopic guidance is technically less demanding compared with imaging with a conventional echoendoscope because it is easier to locate the lesion.16 For lesions over 2 cm in diameter that cannot be fully evaluated with a 20-MHZ probe because of attenuation of the US signal, a “throughthe-scope” probe with a lower imaging frequency can be used. HFPE-assisted EMR provides a large specimen for histopathologic diagnosis as well as assessment of the margin of resection. As demonstrated in the present study, this procedure offers definitive therapy for SMTs up to 20 mm in diameter. For larger lesions for which HFPE-assisted EMR is not possi48
GASTROINTESTINAL ENDOSCOPY
ble, a tissue diagnosis can often be obtained by EUSguided FNA or by obtaining biopsy specimens after removal of the overlying mucosa. Based on the results of the present study, the following therapeutic strategy is proposed for SMTs based on a combination of endoscopic and HFPE characteristics. Well-demarcated hypoechoic lesions up to 2 cm in diameter without umbilication (suggestive of pancreatic rest) on endoscopic examination should be removed by EMR given the high likelihood that the lesion is a carcinoid tumor. In the colon, in addition to carcinoid tumors, stromal tumors, which have the potential for malignant degeneration, should also be resected. Hyperechoic homogeneous lesions or SMTs with or without a clearly evident cystic component can be followed because these are highly likely to be benign histopathologically (e.g., pancreatic rest, lipomas). Kameyama et al.,21 based on experience with conventional EUS and catheter US probe EUS in patients with colonic SMTs, concluded that informaVOLUME 55, NO. 1, 2002
GI submucosal tumors: high-frequency probe EUS-assisted EMR
I Waxman, Y Saitoh, G Raju, et al.
tion provided by EUS was useful in choosing therapy for SMTs of the large intestine. These investigators recommended observation for SMTs with benign ultrasonographic characteristics (lipoma, lymphangioma), and EMR for other lesions depending on size and depth of penetration. In summary, SMTs of the GI tract can be managed safely, quickly, and easily with HFPE-assisted EMR. A treatment algorithm for SMTs has been suggested, based on HFPE and endoscopic findings. However, this proposed approach must be validated in prospective studies.
tumorectomy for gastrointestinal submucosal tumors restricted to the submucosa: a new form of endoscopic minimal surgery. Gastrointest Endosc 1997;46:311-7. Waxman I, Saitoh Y. Clinical outcome of endoscopic mucosal resection for superficial gastrointestinal lesions and the role of high frequency ultrasound probe sonography in an American population. Gastrointest Endosc 2000;52:322-7. Japanese Research Society for Cancer of the Colon and Rectum. General rules for clinical and pathological studies on cancer of the colon, rectum and anus: histopathological classification. 5th ed. Tokyo: Kauehara Syuppan; 1994. p. 50-73. Kudo S. Endoscopic mucosal resection of flat and depressed types of early colorectal cancer. Endoscopy 1993;25:455-61. Hasegawa N, Niwa Y, Arisawa T, Hase S, Goto H, Hayakawa T. Preoperative staging of superficial esophageal carcinoma: comparison of an ultrasound probe and standard endoscopic ultrasonography. Gastrointest Endosc 1996;44:388-93. Yanai H, Yoshida T, Harada T, Matsumoto Y, Nishiaki M, Shigemitsu T, et al. Endoscopic ultrasonography of superficial esophageal cancers using a thin ultrasound probe system equipped with switchable radial and linear scanning modes. Gastrointest Endosc 1996;44:578-82. Chak A, Canto M, Stevens PD, Lightdale CJ, de Mierop FV, Cooper G, et al. Clinical applications of a new through-the scope ultrasound probe: prospective comparison with an ultrasound endoscope. Gastrointest Endosc 1997;45:291-5. Yoshida M, Tsukamoto Y, Niwa Y, Goto H, Hase S, Hayakawa T, et al. Endoscopic assessment of invasion of colorectal tumors with a new high-frequency ultrasound probe. Gastrointest Endosc 1995;41:587-92. Saitoh Y, Obara T, Einami K, Nomura M, Taruishi M, Ayabe T, et al. Efficacy of high-frequency ultrasound probes for preoperative staging of invasion depth in flat and depressed colorectal tumors. Gastrointest Endosc 1996;44:34-9. Matsui M, Niwa Y, Arisawa T, Hirooka Y, Kameyama H, Hayakawa M, et al. Endoscopic ultrasound-guided fine needle aspiration biopsy (EUS-FNAB) in the diagnosis of upper gastrointestinal submucosal tumors [abstract]. Gastroenterology 1998;114:A220. Kojima T, Takahashi H, Parra-Blanco A, Ohguri S, Fujita R. Diagnosis of submucosal tumors of the upper gastrointestinal tract by endosonography assessed endoscopic resection [abstract]. Gastroenterology 1998;114:A23. Kameyama H, Niwa Y, Arisawa T, Goto H, Hayakawa T. Endoscopic ultrasonography in the diagnosis of submucosal lesions of the large intestine. Gastrointest Endosc 1997;46: 406-11.
11.
12.
13. 14.
REFERENCES 1. Lightdale CJ. Diagnosis, staging, and cure of early gastrointestinal cancer. Gastrointest Endosc 1996;44:95-6. 2. Silverstein FE, Martin RW, Kimmey MB, Jiranek GC, Franklin DW, Proctor A. Experimental evaluation of an endoscopic ultrasound probe. Gastroenterology 1989;96:1058-62. 3. Rösch T, Classen M. A new ultrasonic probe for endosonographic imaging of the upper GI tract. Endoscopy 1990;20:41-6. 4. Waxman I. Clinical impact of high-frequency ultrasound probe sonography during diagnostic endoscopy—a prospective study. Endoscopy 1998;30:A166-8. 5. Karita M, Tada M, Okita K, Kodam T. Endoscopic therapy for early colon cancer: the strip biopsy resection technique. Gastrointest Endosc 1991;37:128-32. 6. Tada M, Inoue H, Yabata E, Okabe S, Endo M. Colonic mucosal resection using a transparent cap-fitted endoscope. Gastrointest Endosc 1996;44:63-5. 7. Sakai P, Filho FM, Iryia K, Moura EGH, Tomishigue T, Scabba A, et al. An endoscopic technique for resection of small gastrointestinal carcinoma. Gastrointest Endosc 1996; 44:65-8. 8. Fleischer DE, Wang GQ, Dawsey S, Tio TL, Newsome J, Kidwell J, et al. Tissue band ligation followed by snare resection (band and snare): a new technique for tissue acquisition in the esophagus. Gastrointest Endosc 1996;44:68-72. 9. Inoue H, Takeshita K, Hori H, Muraoka Y, Yoneshima H, Endo M, et al. Endoscopic mucosal resection with a cap-fitted panendoscope for esophagus, stomach, and colon mucosal lesions. Gastrointest Endosc 1993;39:58-62. 10. Kawamoto K, Yamada Y, Furukawa N, Utsunomiya T, Hataguchi Y, Mizuguchi M, et al. Endoscopic submucosal
VOLUME 55, NO. 1, 2002
15.
16.
17.
18.
19.
20.
21.
GASTROINTESTINAL ENDOSCOPY
49