Extensive (8 to 12 cm 2 ) Noncircumferential Endoscopic Mucosal Resection for Early Esophageal Cancer

Transcription

1 SECTION V: ESOPHAGUS MALIGNANT Extensive (8 to 12 cm 2 ) Noncircumferential Endoscopic Mucosal Resection for Early Esophageal Cancer Philippe Monnier, MD, Yves Jaquet, MD, Alexandre Radu, MD, Raphaelle Pilloud, MD, Pierre Grosjean, MD, Anette Escher, MD, Elsa Piotet, MD, and Snezana Andrejevic Blant, MD Otolaryngology, Head and Neck Surgery Department and Institute of Pathology, University Hospital Centre Hospitalier Universitaire Vaudois, Lausanne, Switzerland Background. Endoscopic mucosal resection (EMR) is an appealing method for treating intramucosal esophageal cancer but must comply with the following stringent requirements: proper preoperative staging, complete resection of the lesion, obtaining a resected specimen for histologic analysis of safety margins, and squamous reepithelialization without stricture formation. Methods. A rigid esophagoscope was created to resect up to 12 cm 2 of esophageal mucosa in a single specimen and at a constant depth through the submucosa. Under visual control, the esophageal mucosa is sucked into a transparent window and resected with a thin diathermy wire loop in 10 seconds. After extensive preclinical studies in a sheep model, this article reports our early experience in humans. Results. Twenty-one hemi-circumferential EMRs were performed for 11 dysplastic Barrett s esophagi and 10 early squamous cell carcinomas with no perforation, one hemorrhage controlled by embolization of the left gastric artery, and one incomplete resection. Deep safety margins were clear in 19 of 21 resected specimens (2 patients, unfit for operations, had submucosal invasion of squamous cell carcinoma and adenocarcinoma, respectively). Lateral margins were not clear by definition in 7 circumferential Barrett s esophagi, but were clear in 4 incomplete Barrett s esophagi and 10 early squamous cell carcinomas. Conclusions. Large EMRs of 12 cm 2 can safely be performed at the submucosal level in the esophagus. Although feasible in one session, circumferential EMR in humans is not yet advisable because of the risk of stricture formation during the healing phase. The rate of complications of this series of 21 EMRs in humans is acceptable. (Ann Thorac Surg ) 2010 by The Society of Thoracic Surgeons Presented at the 2 nd International Bi-Annual Minimally Invasive Thoracic Surgery Summit, Boston, MA, October 9 10, Address correspondence to Dr Monnier, Otolaryngology, Head and Neck Surgery Department, University Hospital CHUV Lausanne, Bugnon 46, Lausanne, CH-1011 Switzerland; philippe.monnier@chuv.ch. Endoscopic mucosal resection (EMR) is an appealing method for treating intramucosal (T1a) squamous cell carcinomas and high-grade intraepithelial neoplasia or intramucosal adenocarcinomas arising in Barrett s esophagus. The risk of lymph node metastasis is reported to be about 6% for intramucosal squamous cell carcinomas [1] and less than 10% for early adenocarcinomas arising in Barrett s esophagus [2 5]. Although esophagectomy remains the gold standard of treatment, EMR is an option in elderly patients or in those with severe comorbidities. Precise preoperative staging with 20- and 7.5-MHz endoscopic ultrasonography is a prerequisite for selecting the appropriate candidates for a minimally invasive procedure [6]. During the last decades, several ablation techniques using laser photoablation (potassium titanyl phosphate [KTP], neodymium-doped yttrium aluminium garnet laser) [7, 8], argon plasma coagulation [9], multipolar electrocoagulation [10], photodynamic therapy [11, 12], and spray cryotherapy [13] have been investigated. Most of the reported studies have involved a limited number of patients with short periods of follow-up, and several drawbacks of these techniques soon limited their use. The variable depth of tissue destruction has exposed patients to recurrence from buried glands left beneath the regenerated squamous epithelium in Barrett s esophagus [7, 10, 12] or to the potential risk of perforation [11] or stenosis [12]. Another major disadvantage of ablative techniques has been that a resected specimen for histologic analysis could not be obtained. In an attempt to circumvent these drawbacks, EMR has become increasingly popular in the gastroenterologic community for treating high-grade intraepithelial neoplasia or early adenocarcinoma arising in Barrett s esophagus [14 16]. However, technical limitations with Dr Monnier discloses that he has a financial relationship with Storz Company. Drs Jaquet, Radu, Pilloud, Grosjean, Escher, Piotet, and Blant have no conflicts of interest to disclose by The Society of Thoracic Surgeons /$36.00 Published by Elsevier Inc doi: /j.athoracsur

2 S2152 MINIMALLY INVASIVE THORACIC SURGERY SUMMIT 2009 MONNIER ET AL Ann Thorac Surg Fig. 1. Rigid esophagoscope for endoscopic mucosal resection: (A) the diagnostic esophagoscope has an inner tube that seals the distal aperture of the outer tube (see close-up view) and is used as a conventional rigid esophagoscope, with a remote control for inflation, suction and irrigation. (B) Resectoscope mounted on the outer shaft of the diagnostic esophagoscope. The transparent, perforated window and the diathermy loop are seen in the close-up view. Mucosal resection is carried out from distal to proximal. flexible videoendoscopes only permit piecemeal removal of about 2-cm 2 mucosal patches with the cap-fitted flexible endoscope [16]. Although removal of the whole metaplastic mucosa of circumferential Barrett s esophagus with the flexible endoscope has been attempted [17, 18], these trials demonstrated that piecemeal resections are fastidious and take a long time to complete. They often required multiple treatment sessions. Furthermore, the histologic study of the numerous specimens renders the analysis of the surgical margins difficult, and the risk of perforation increases with possible overlaps of the resected areas. Further attempts at large, en bloc mucosal resection with the flexible gastroscope have fallen far short of the aimed goals [19] and have stimulated further research, combining EMR of the early neoplastic lesions with thermal ablation of the residual Barrett s mucosa with radiofrequency energy such as the HALO system (BARRX Medical Inc, Sunnyvale, CA) [20, 21]. Whatever the technique, the only rational treatment for early squamous cell carcinoma, multicentric foci of highgrade intraepithelial neoplasia, or early adenocarcinomas arising in Barrett s esophagus is the entire resection of the lesion and of the columnar-lined segment. In Barrett s esophagus, complete removal of the metaplastic mucosa will simultaneously eliminate all foci of dysplasia or early cancer potentially missed by random biopsies [22]. Because reepithelialization arises from the margins of the mucosal defect, any persisting area of the specialized intestinal metaplasia will lead to regrowth of metaplastic mucosa. If an incomplete Barrett s esophagus presenting as finger-like projections from the Z-line is fully resected, then reepithelialization occurs with squamous cell mucosa. Therefore, Barrett s esophagus should be fully eradicated in a single session. The development of a safe technique for performing large mucosal resections in a single piece, allowing extensive (12 cm 2 specimens) and even circumferential removal of Barrett s esophagus in 2 or 3 resected specimens at most, was the aim of our studies. Material and Methods This study received institutional approval from the Ethics Committee of our hospital. Patients provided written informed consent before being enrolled in the study. EMR Rigid Esophagoscope The EMR esophagoscope (Fig 1) and technique of mucosal resection have already been published in detail [23]. Briefly, the principle consists of sucking the esophageal mucosa against a large, transparent window created at the distal extremity of a rigid esophagoscope. With a thin diathermy loop, the mucosa is resected under visual control at a predetermined depth through the submucosa [24] (Fig 2). A 12-cm 2 resected specimen is obtained for histologic study. To perform a circumferential mucosal resection, a second resection is done on the opposite esophageal side during the same endoscopic session. Fig. 2. Principle of endoscopic mucosal resection with the rigid esophagoscope: Negative pressure is applied to the inside portion of the resectoscope, and under visual control, the esophageal mucosa is sucked against the transparent window at a predetermined depth. Resection with a diathermy wire loop (dashed line) is performed at a regular depth through the submucosa in less than 10 seconds. (LMM lamina muscularis mucosae; MPE muscularis propria externa; MPI muscularis propria interna).

3 Ann Thorac Surg MINIMALLY INVASIVE THORACIC SURGERY SUMMIT 2009 MONNIER ET AL S2153 Table 1. Histologic Results of 11 Endoscopic Mucosal Resection Specimens for Barrett s Esophagus Variable Until now, circumferential mucosal resections have only been done in animal experiments, searching for ways to prevent cicatricial stenosis during the healing phase [25, 26]. Human Studies For Barrett s esophagus, the preoperative evaluation consisted of 4-quadrant biopsies performed each centimeter over the entire length of the columnar-lined mucosa. When high-grade dysplastic changes were confirmed by 2 independent pathologists, 7.5- and 20-MHz endoscopic ultrasonography was used to rule out submucosal infiltration of the esophageal wall and lymph node metastases. For squamous cell carcinoma, vital staining with Toluidine blue helped delineate the exact mucosal extent of the lesion, and 7.5 and 20 MHz endoscopic ultrasonography was also used to assess the in-depth invasion and to rule out possible lymph node metastases. Twenty-one hemi-circumferential EMRs were performed in 13 patients with a mean age of 61 years (range, 46 to 77 years). There were 10 early squamous cell carcinomas and 11 high-grade intraepithelial neoplasias in Barrett s esophagi. Because the animal study on local injection of cultured skin cells to avoid stenosis formation is still going on, no circumferential EMRs were performed in patients during this early feasibility study. The complication rate, histologic features of the resected specimens, reepithelialization process, and rate of stenosis formation were analyzed. Results No. Superficial adenocarcinoma 4 a High grade intraepithelial neoplasia 7 Free lateral margins Incomplete Barrett s esophagus 4/4 Complete Barrett s esophagus 0 b /7 Free deep margins 10 a /11 a One submucosal adenocarcinoma in a cirrhotic patient. b By definition, resection margins cannot be clear after hemi-circumferential resection for complete, circumferential Barrett s esophagus. Twenty of the 21 hemi-circumferential resections were performed adequately. In 1 patient, a technical problem related to a submucosal invasion of the tumor yielded a resected specimen in several pieces. The sizes of the resected specimens were 12 cm 2 in 9 cases and 8 cm 2 in 12 cases, depending on the initial extent of the pathology. There were no perforations, and one hemorrhage was controlled by embolization of the left gastric artery. As a late complication, there was no esophageal stenosis after full reepithelialization. In the analysis of the resected specimens, muscularis mucosae were present on all sections in 17 of 20 (85%) and partially present in 3 of 20 hemi-circumferential EMRs, respectively. In the last case, a technical problem due to a submucosal invasion of a squamous cell carcinoma did not allow precise assessment because the resected specimen was fragmented. Histologic results of hemi-circumferential EMRs for Barrett s esophagus and early squamous cell carcinoma are reported in Table 1 and Table 2, respectively. In the Barrett s esophagus group, there were four superficial adenocarcinomas with one submucosal extension in a cirrhotic, inoperable patient, and seven high-grade intraepithelial neoplasias. The resection margins were free in four cases of incomplete Barrett s esophagus. By definition, they could not be free in circumferential Barrett s esophagus because only hemi-circumferential resections were performed (Figs 3 and 4). The deep resection margins were clear in 10 of 11 patients in whom in-depth infiltration of the tumor did not reach beyond the muscularis mucosae. One patient with a submucosal extension of an adenocarcinoma was further treated endoscopically with KTP laser vaporization. The follow-up after 2 years has not shown any evidence of residual disease endoscopically and at 7.5- and 20-MHz endoscopic ultrasonography. In the group of early squamous cell carcinomas, there were 2 high-grade dysplasias, 5 carcinomas in situ, and 2 intramucosal carcinomas. One inoperable patient presented with a submucosal extension because of severe comorbidities. The lateral resection margins were free in 7 of 10 resected specimens, and foci of moderate dysplasia were present in the remaining 3. The deep resection margins were clear in 9 of 10 patients. The patient with submucosal extension of squamous cell carcinoma underwent further endoscopic treatment with KTP laser vaporization. At 1-year follow-up, he did not show any evidence of recurrent disease endoscopically and at 7.5- and 20-MHz endoscopic ultrasonography. Comment Our preclinical studies [23, 25, 26] indicate that the modified rigid esophagoscope offers the possibility of performing long and circumferential EMRs in the sheep esophagus. Each resection provides a single rectangular specimen of a maximum of 12 cm 2 in size. The results of our animal experiments are briefly summarized below: Table 2. Histologic Results of Endoscopic Mucosal Resection Specimens for 10 Early Squamous Cell Carcinomas Result No. High grade dysplasia 2 Carcinoma in situ 5 Intramucosal carcinoma 2 Submucosal carcinoma 1 Free lateral margins 7/10 With moderate dysplasia 3/10 Free deep margins 9 a /10 a One submucosal squamous cell carcinoma in an inoperable patient.

4 S2154 MINIMALLY INVASIVE THORACIC SURGERY SUMMIT 2009 MONNIER ET AL Ann Thorac Surg specimens (85%). Development of cicatricial stenosis during the reepithelialization process required a mean of 2.6 (range, 0 to 5) dilation sessions to prevent stenosis formation. Fig 3. Resected specimen for intramucosal adenocarcinoma arising in Barrett s esophagus, with superimposed histological mapping: the extent of Barrett s esophagus is outlined in white, low grade dysplasia in hatching, high grade dysplasia in light grey, and superficial adenocarcinoma in grey. Hemi-circumferential EMRs This experiment, published in 2004 [23], analyzed 55 hemi-circumferential EMRs performed in 21 sheep. No perforation was encountered. In 48 of 55 resected specimens, an accurate depth of resection at the submucosal level was obtained, and scarce superficial fibers from the internal layer of the muscularis propria were seen focally in the remaining 7 specimens. Complete reepithelialization occurred within 6 weeks. No stenosis formation occurred. Simple Circumferential Resections This study was published in 2005 [25]. Only 2 resections are needed for a 360 EMR in the esophagus. This surface corresponds to the usual extension of a 6-cm-long complete Barrett s esophagus. The specimens are easily oriented by the endoscopist and are thus adequate for histologic analysis. Circumferential resections of 2.2- to 5.5-cm in length were successfully performed in 23 of 24 animals. One perforation resulting from a technical error occurred directly after mucosectomy. Accurate depth of resection through the submucosa was obtained in 20 of 23 Circumferential Resections With Additional Multiple Injections of Cultured Skin Cells The preliminary results of this ongoing study were published in 2008 [26]. After 4-quadrant injection of cultured skin cells every 2 cm into the resected bed of a 6-cm-long circumferential EMR, none of the 10 animals needed any dilation. They all healed without stenosis formation. Extensive mucosal resection with the EMR rigid esophagoscope allows complete removal of dysplastic lesions and early squamous cell carcinoma or adenocarcinoma confined to the mucosa in humans. Moreover, the deep glands found in Barrett s esophagus are also treated because the cut is made at the submucosal level, and the mean depth of metaplastic mucosa is estimated from histologic studies to be 0.5 mm (range, 0.39 to 0.59 mm) [24]. The safety of the procedure has been demonstrated in the sheep model, showing a very low perforation rate of about 1.2% (1 of 86) [23, 25]. The main risk of circumferential EMR is the potential overlap of the two adjacent resections, which may lead to perforation. With saline injection into the submucosa and proper visual control before resection, this complication risk should be minimized. Avoidance of cicatricial stenosis warrants further investigation. Long-segment circumferential resections lead to cicatricial stenosis during the slow reepithelialization process. Multiple injections of cultured skin cells seem to prevent this complication in the sheep model [26], but further extensive studies are needed before this technique can be implemented in humans. The EMR rigid esophagoscope is a useful tool for the precise staging of early esophageal cancers and can identify more accurately than endoscopic ultrasonogra- Fig 4. Histology of the resected specimen displayed in Fig. 3 (hematoxylin-eosin staining): (A) Long-segment mucosal resection (2 original size): the thickness is fairly constant over the length of the resected specimen, except at both extremities. (B) (10 magnification) and (C) (20 magnification). Close-up views of high-grade intraepithelial neoplasia and early adenocarcinoma at different levels. The lamina muscularis mucosae is conspicuous in the close-up views and the deep resection margins at the submucosal level are clear.

5 Ann Thorac Surg MINIMALLY INVASIVE THORACIC SURGERY SUMMIT 2009 MONNIER ET AL S2155 phy the true foci of submucosal tumor invasion that require an esophagectomy. It allows the curative resection of intramucosal squamous cell carcinoma and dysplastic incomplete Barrett s esophagus. Nodular and ulcerated lesions are usually associated with submucosal invasion lesions. They do not represent a good indication for EMR because of the high risk of lymph node metastasis. In the case of elderly, inoperable patients, the lesion is first vaporized with the KTP laser to make it flat, and a conventional EMR is then performed. Possible residual tumor at the level of the submucosa can further be vaporized until normal muscular fibers are identified. With the rigid esophagoscope, a cotton swab soaked in saline may be used to remove the char and better identify the normal muscular layers until the lesion has been fully vaporized. This was done in 2 of our inoperable patients, who showed no evidence of recurrence after short 2- and 1-year follow-up, respectively. Further animal studies are warranted before circumferential resection of complete Barrett s esophagus can be considered in humans without any risk of cicatricial stenosis formation. The main drawbacks of this technique include the need for general anesthesia, training in rigid esophagoscopy, and difficult access to the cardia in some elderly patients. General acceptance of this technique will require the design of a semi-rigid EMR esophagoscope with 2 chip cameras, 1 for front-viewing and another for sideviewing through the transparent window. The animal and human studies were conducted without any financial support from the industry but benefited from a grant of the Fonds de Service of the Otolaryngology, Head and Neck Surgery Department of the University Hospital CHUV Lausanne, Switzerland. The Karl Storz Company, Tuttlingen, Germany, assumed the financial development of the instrument. If the instrument is commercialized in the future, then the author and inventor of the EMR esophagoscope will hold a financial relationship with the Storz Company. References 1. Endo M, Kawano T. Detection and classification of early squamous cell esophageal cancer. Dis Esophagus 1997;10: DeMeester SR, DeMeester TR. The diagnosis and management of Barrett s esophagus. Adv Surg 1999;33: Thon KP. [Surgical treatment of early adenocarcinoma arising in Barrett s oesophagus]. Praxis (Bern 1994) 2004;93: Stein HJ, Feith M, Mueller J, Werner M, Siewert JR. Limited resection for early adenocarcinoma in Barrett s esophagus. Ann Surg 2000;232: Cense HA, van Eijck CH, Tilanus HW. New insights in the lymphatic spread of oesophageal cancer and its implications for the extent of surgical resection. Best Pract Res Clin Gastroenterol 2006;20: Manner H, May A, Pech O, et al. Early Barrett s carcinoma with low-risk submucosal invasion: long-term results of endoscopic resection with a curative intent. Am J Gastroenterol 2008;103: Bonavina L, Ceriani C, Carazzone A, Segalin A, Ferrero S, Peracchia A. Endoscopic laser ablation of nondysplastic Barrett s epithelium: is it worthwhile? J Gastrointest Surg 1999;3: Gossner L, May A, Stolte M, Seitz G, Hahn EG, Ell C. KTP laser destruction of dysplasia and early cancer in columnarlined Barrett s esophagus. Gastrointest Endosc 1999;49: Schulz H, Miehlke S, Antos D, et al. Ablation of Barrett s epithelium by endoscopic argon plasma coagulation in combination with high-dose omeprazole. Gastrointest Endosc 2000;51: Sampliner RE, Faigel D, Fennerty MB, et al. Effective and safe endoscopic reversal of nondysplastic Barrett s esophagus with thermal electrocoagulation combined with highdose acid inhibition: a multicenter study. Gastrointest Endosc 2001;53: Savary JF, Grosjean P, Monnier P, et al. Photodynamic therapy of early squamous cell carcinomas of the esophagus: a review of 31 cases. Endoscopy 1998;30: Overholt BF, Panjehpour M, Halberg DL. Photodynamic therapy for Barrett s esophagus with dysplasia and/or early stage carcinoma: long-term results. Gastrointest Endosc 2003;58: Johnston CM, Schoenfeld LP, Mysore JV, Dubois A. Endoscopic spray cryotherapy: a new technique for mucosal ablation in the esophagus. Gastrointest Endosc 1999;50: Ell C, May A, Gossner L, et al. Endoscopic mucosal resection of early cancer and high-grade dysplasia in Barrett s esophagus. Gastroenterology 2000;118: Lambert R. Endoscopic mucosectomy: an alternative treatment for superficial esophageal cancer. Recent Results Cancer Res 2000;155: May A, Gossner L, Behrens A, et al. A prospective randomized trial of two different endoscopic resection techniques for early stage cancer of the esophagus. Gastrointest Endosc 2003;58: Seewald S, Akaraviputh T, Seitz U, et al. Circumferential EMR and complete removal of Barrett s epithelium: a new approach to management of Barrett s esophagus containing high-grade intraepithelial neoplasia and intramucosal carcinoma. Gastrointest Endosc 2003;57: Satodate H, Inoue H, Yoshida T, et al. Circumferential EMR of carcinoma arising in Barrett s esophagus: case report. Gastrointest Endosc 2003;58: Rösch T, Sarbia M, Schumacher B, et al. Attempted endoscopic en bloc resection of mucosal and submucosal tumors using insulated-tip knives: a pilot series. Endoscopy 2004;36: Gondrie JJ, Pouw RE, Sondermeijer CM, et al. Stepwise circumferential and focal ablation of Barrett s esophagus with high-grade dysplasia: results of the first prospective series of 11 patients. Endoscopy 2008;40: Gondrie JJ, Pouw RE, Sondermeijer CM, et al. Effective treatment of early Barrett s neoplasia with stepwise circumferential and focal ablation using the HALO system. Endoscopy 2008;40: Nigro JJ, Hagen JA, DeMeester TR, et al. Prevalence and location of nodal metastases in distal esophageal adenocarcinoma confined to the wall: implications for therapy. J Thorac Cardiovasc Surg 1999;117:16 23; discussion Radu A, Grosjean P, Fontolliet C, Monnier P. Endoscopic mucosal resection in the esophagus with a new rigid device: an animal study. Endoscopy 2004;36: Ackroyd R, Brown NJ, Stephenson TJ, Stoddard CJ, Reed MW. Ablation treatment for Barrett oesophagus: what depth of tissue destruction is needed? J Clin Pathol 1999;52: Pilloud R, Jaquet Y, Monnier P. Extensive circumferential endoscopic mucosal resection with a new rigid esophagoscope: an animal study. J Thorac Cardiovasc Surg 2005;130: Escher A, Piotet E, Ikonomidis C, Jaquet Y, Monnier P. Mucosectomie circonférentielle de l oesophage chez le mouton: prévention des sténoses par injection de cultures de kératinocytes. Forum Med Suisse 2009;9:133 5.