- Email: [email protected]
Minimally Invasive Intragastric Approach to Gastroesophageal Junction Disease
Minimally Invasive Intragastric Approach to Gastroesophageal Junction Disease Anthony M. Villano, MD, Alexander Lofthus, MD, Thomas J. Watson, MD, Nadim G. Haddad, MD, and M. Blair Marshall, MD Departments of Surgery, Thoracic Surgery, and Gastroenterology, MedStar-Georgetown University Hospital, Washington, DC; and Regional Department of Surgery and Thoracic Surgery, MedStar Health, Washington, DC
Background. A minimally invasive intragastric approach to the gastroesophageal junction (GEJ) allows resection of intramural disease while avoiding disruption of the lower esophageal sphincter and vagus nerves. Few surgeons use this approach; thus little is known regarding its indications, feasibility, technical aspects, complication profile, and long-term outcomes. This study reviewed the experience with this technique. Methods. A retrospective review was performed of a prospectively maintained, Institutional Review Board– approved database covering the period from January 1, 2005 to August 1, 2017. Indications, operative details, postoperative complications, and outcomes were assessed. Results. There were 12 patients identified. The mean age of these patients was 51.9 years. The indications for resection included 10 symptomatic leiomyomas, one gastrointestinal stromal tumor, and three cancers of the GEJ. Mean and median length of stay were 4.9 and 2.5
days, respectively. There were two postoperative esophageal leaks managed with laparoscopic repair. Of the 3 patients with cancer, 2 underwent an R0 resection, whereas 1 patient underwent an R1 resection. There were no other complications or recurrences. Mean follow-up was 6.0 years (range, 0.5 to 12.6 years); no patients had stricture or symptomatic gastroesophageal reflux on long term follow-up. Conclusions. Resection of selected intramural GEJ disorders through a minimally invasive transgastric approach can be performed safely with acceptable morbidity and good long-term results. The approach allows preservation of the lower esophageal sphincter and vagus nerves, a potential advantage compared with other surgical alternatives to resection in this region.
I
specialized endoscopists, and they are especially challenging when the tumor sits near the GEJ [5–7]. An alternative to these approaches for the resection of GEJ tumors is laparoscopic intragastric resection (LIR). This procedure uses gastric insufflation and multiple port sites placed through the anterior gastric wall and provides a useful alternative to the wedge resection when size, location, or endophytic growth renders a standard laparoscopic approach challenging or impossible. Here we describe our experience with the LIR approach to a variety of intramural GEJ disorders, with an emphasis on long-term follow-up and patient outcomes.
ntramural disease of the gastroesophageal junction (GEJ) presents a unique therapeutic challenge. Submucosal gastric lesions (namely leiomyomas and gastrointestinal stromal tumors [GISTs]) and stage 1 GEJ carcinomas rarely involve lymph nodes and are thus candidates for local resection [1, 2]. Transgastric resection is the most common technique for submucosal lesions located in the gastric body and antrum, but it has lower applicability to GEJ tumors [3] because of the high likelihood of disruption of the lower esophageal sphincter or vagal injury. Without intragastric visualization, it is common to resect excessive normal gastric tissue, thus further limiting the transgastric approach near the GEJ [4]. Endoscopic submucosal dissection (ESD) and endoscopic mucosal resection (EMR) have been widely studied in Asia for the resection of GEJ tumors. However, these techniques are limited by tumor size, they require
Accepted for publication Aug 20, 2018. Presented at the Poster Session of the General Thoracic Surgery Club, Dove Mountain, AZ, March 8–11, 2018. Address correspondence to Dr Villano, Department of General Surgery, MedStar Georgetown University Hospital, 3800 Reservoir Rd, NW, PHC Bldg, 4th Flr, Washington, DC 20007; email: anthony.m.villano@gunet. georgetown.edu.
Ó 2018 by The Society of Thoracic Surgeons Published by Elsevier Inc.
(Ann Thorac Surg 2018;-:-–-) Ó 2018 by The Society of Thoracic Surgeons
Patients and Methods Data Source Deidentified patient information was maintained in a prospective, single-institution, Institutional Review Board–approved database. Information gathered included patient demographics (date of birth, age, sex), preoperative comorbidity status (American Society of Anesthesia classification, [ASA]), operative details (operative time, estimated blood loss, number of intraoperative transfusions), tumor pathologic features (size, histologic characteristics), outcome variables (length of stay, morbidity, mortality), and duration of available follow-up. 0003-4975/$36.00 https://doi.org/10.1016/j.athoracsur.2018.08.050
2
VILLANO ET AL MINIMALLY INVASIVE INTRAGASTRIC RESECTION
Medical records were incomplete for 1 patient and are noted accordingly in the referenced tables.
Operative Approach Specific details of our operative approach have been outlined elsewhere [8]. Briefly, the peritoneal cavity was entered and insufflation achieved as in standard laparoscopy. Flexible endoscopy was used to enter the stomach. A guidewire was used to cannulate the proximal duodenum, and over this a Coda balloon (Cook Incorporated, Bloomington, IN) was positioned and expanded to prevent insufflation of the small bowel. After achieving gastric insufflation, the anterior stomach could be inspected laparoscopically to allow for identification of ideal trocar sites. These trocar site locations varied by case, but they were placed far enough apart to triangulate three instruments consistently to the GEJ. Gastrotomies were made using hook cautery, and three 5-mm Kii balloon-anchored ports (Applied Medical, Rancho Santa Margarita, CA) were placed transperitoneally into the gastric lumen (Fig 1). These ports were secured with stay sutures. For tumors amenable to intragastric resection, a snare or stitch has been used to provide traction into the gastric lumen. Ligasure bipolar cautery (Valleylab Inc, Boulder, CO) was used to dissect and enucleate the tumor circumferentially. If enucleation was not feasible secondary to size or an exophytic component, the tumor could be stapled with a linear articulating Endo GIA (Medtronic, Minneapolis, MN), or a full-thickness resection was performed with cautery (Fig 2). The specimen could then be placed in an Endo Catch (Medtronic) bag and pulled through the oropharynx with the endoscope. Tumors too large for this approach were removed through one of the gastrostomy sites and subsequently through an abdominal wall port site in standard laparoscopic fashion. In the event of a residual defect at the GEJ, closure was performed with interrupted sutures using an Endo Stitch (Medtronic) or standard sutures on a curved needle as our skills improved. The gastrotomy sites were closed in a
Ann Thorac Surg 2018;-:-–-
similar fashion. Before exiting the peritoneal cavity, the endoscope was advanced, and the stomach was insufflated one final time to assess for a leak. The abdomen was then closed using standard procedures.
Pathologic Analysis All tumors were processed and analyzed by two pathologists. Stromal tumors were routinely tested by immunohistochemistry to distinguish leiomyoma and GIST (CD117, CD34, smooth muscle actin, desmin, S-100, and MIB-1/Ki-67). Adenocarcinomas were evaluated for histologic features, size, grade, margins, lymphovascular invasion, and lymph node involvement (if included in the specimen).
Results Patient Characteristics Table 1 outlines the baseline patient characteristics. Mean age at resection was 51.9 years (range, 28 to 87 years). Most of the patients were in ASA class 2; however, all 3 of the patients with adenocarcinoma were in ASA class 3. One patient with leiomyoma underwent 5 weeks of neoadjuvant imatinib treatment for what was originally suspected to be a KIT/PDGFRA wild-type GIST. All 3 patients with adenocarcinoma had T1a tumors without evidence of nodal metastasis (N0) on the basis of preoperative endoscopic ultrasound. Additionally, 1 patient with adenocarcinoma underwent an earlier attempt at ablation with argon plasma coagulation (APC) twice with local recurrence before surgical referral.
Operative Metrics Operative variables are outlined in Table 2. Mean tumor size along the greatest dimension was 4.9 cm; however tumors with a wide range of sizes were resected (range, 0.8 to 8 cm). Tumor volume was of a similarly wide range (5.6 to 142.9 cm3). Mean operative time was 246.5 minutes (range, 106 to 425 minutes). No patients required a blood transfusion intraoperatively, and there were no conversions to an open procedure.
Histology and Outcomes
Fig 1. Laparoscopic view of two intragastric balloon-anchored ports (blue arrows) tenting up the stomach wall, with a third port in the process of being advanced into gastric lumen with silk suture tacking the gastric wall (white arrows).
Final histologic examination yielded 10 leiomyomas, one GIST, and three adenocarcinomas (Fig 3). Mean and medial length of stay was 4.8 and 2.5 days, respectively. In 2 patients, a postoperative leak developed on day 1 that required a return to the operating room for primary repair. There were no additional operative complications, and no patients had long-term sequelae such as dysphagia, odynophagia, esophageal stricture, or symptomatic reflux. Mean follow-up for the cohort was 5.4 years (Table 3). Detailed patient descriptions and outcomes are listed in Table 4. Patient 2 had a leak discovered on routine esophagram postoperative day 1. This patient was taken back to the operating room, and a small defect along the posterior aspect of the lesser curve of the stomach near the GEJ was noted. This defect was repaired primarily
Ann Thorac Surg 2018;-:-–-
VILLANO ET AL MINIMALLY INVASIVE INTRAGASTRIC RESECTION
3
Fig 2. (A) Tumor spanning the Z-line of the gastroesophageal junction and coursing across the distal esophagus and the stomach (dotted white line). (B) Completed full-thickness resection. The white arrow points to the muscular layer of the esophagus. Fat can be visualized at the base of the resection.
through a laparoscopic approach, and no leak was confirmed with repeat endoscopy. Repeat swallow study on postoperative day 7 demonstrated no leak. Patient 7 similarly had a leak identified by swallow study on postoperative day 1 and was taken to the operating room for laparoscopic repair. On exploration, the leak was identified at the proximal end of the staple line created when transecting the tumor from the gastric mucosa. This was oversewn with the Endo Stitch in an intermittent fashion. This patient was advanced to a regular diet before discharge on postoperative day 7. For patient 9, the intragastric portion of the procedure was performed with the use of the da Vinci surgical system (Intuitive Surgical, Sunnyvale, CA). Patient 12 underwent an R1 resection after three previous attempts at APC by gastroenterology with recurrence before surgical referral. Six months postoperatively, malignant ascites developed, and the patients subsequently died in the absence of locoregional recurrence.
Comment A transgastric approach to resection of GEJ disease was first described in 1995 by Ohashi [9]. Since this time, various surgical approaches have been described to Table 1. Patient Characteristics Characteristic Mean age (years) Sex Male Female Mean Preoperative ASA classa Neoadjuvant therapya Yes No Prior resection attempta Yes No a
Value (%) 53.2 6 (40) 6 (50) 2.4 1 (8) 11 (92) 1 (8) 11 (92)
Chart incomplete for patient 1. Values represent data for remaining 11 patients. ASA ¼ American Society of Anesthesiologists.
address this anatomically challenging subset of tumors. The most commonly used approaches have been exogastric and transgastric wedge resection, both of which have demonstrated good safety and efficacy in several small case series [10–12]. For large tumors with an endophytic growth pattern, this approach risks resection of unnecessarily large areas of gastric mucosa, with the possibility of injuring the vagus nerves, narrowing the GEJ, or converting to an open procedure [13]. Recognizing these limitations, modification of this strategy has led to the use of a minimally invasive intragastric approach through gastric insufflation and use of intragastric trocars. Advantages of this method include improved visualization (especially of endophytic tumors) and the ability to limit the extent of resection only to what is oncologically necessary. Traditional laparoscopic approaches often pose a challenge in accessing these high, posteriorly located tumors and can pose difficult angles for firing of the endoscopic stapler or safe application of electrocautery. LIR provides a unique visual perspective most surgeons are not used to, and it creates novel angles whereby the volume of tissue resected can be minimized and the risk of narrowing the GEJ or injuring the vagus nerves can be diminished. To date, this technique has now been used by various other groups in small case reports and has demonstrated safety and efficacy for benign tumors near the GEJ [4, 14–21]. Our present study represents one of the largest case series to date that used this method with a combination of benign and malignant disease. Although others have suggested that enucleation is likely the only way to
Table 2. Operative Metrics Characteristic
Value (Range)
Mean tumor size, greatest dimension (cm) Mean tumor volume (cm3) Mean operative time (minutes)a Mean estimated blood loss (mL)a Intraoperative transfusiona Conversion to open
4.8 33.1 265 45
a
(0.8–8.0) (5.6–142.9) (106–425) (5–100) 0 0
Chart incomplete for patient 1. Values represent data for remaining 11 patients.
4
VILLANO ET AL MINIMALLY INVASIVE INTRAGASTRIC RESECTION
Ann Thorac Surg 2018;-:-–-
Fig 3. Representative images of resected gastroesophageal junction disease. (A) Leiomyoma. (B) Gastrointestinal stromal tumor. (C) Adenocarcinoma.
remove tumors in close proximity to the GEJ (<3 cm) [14], we have very selectively performed full-thickness resection with cautery and primary repair with satisfactory results. Our results demonstrate that using such an approach is safe, affords a short length of stay, and confers minimal risk of long-term morbidity for the patient. Further, our series reaffirms that this approach confers acceptable operative times with minimal blood loss and can be performed in patients with multiple comorbidities. No patients in our series experienced stricture development or a vagal nerve injury. The 2 patients who had an esophageal leak postoperatively were able to undergo primary repaired without long-term complications. The tumors in both these patients were large (5.8 and 10 cm), and although the leaks were readily addressed by a second procedure, their complications highlight the point that large tumors at the GEJ carry greater risk for severe complications. Therefore, a challenging dissection may warrant modification of the surgical approach when difficulty is encountered by the operating surgeon. However, size should not be considered an absolute contraindication to this approach because three
Table 3. Postoperative Histologic Characteristics and Outcomes Characteristic Histopathology Leiomyoma Adenocarcinoma GIST Other Median length of stay (days) Complications No Postoperative leak Operative mortality Locoregional recurrences Mean follow-up (months) GIST ¼ gastrointestinal stromal tumor.
Value (%)
8 (67) 3 (25) 1 (8) 0 2.5 10 (83) 2 (17) 0 (0) 0 (0) 65
additional tumors 5 cm or larger were resected without such issues. Accumulation of additional patients with large tumors is warranted to clarify this point. One unique indication for LIR is early stage adenocarcinoma. Ours is the first study to describe technical and short-term oncologic outcomes of this population. From this standpoint, R0 resection was achieved in 2 of the 3 patients. The single cancer recurrence and subsequent mortality were in the setting of an R1 resection after two prior attempts at APC, which led to malignant ascites without gross locoregional recurrence. Patient selection is key to complete resection and long-term success of this operative strategy. In the setting of previous attempts at treatment, the ability to achieve an R0 resection is certainly more difficult and warrants further study before this can be recommended. When use of EMR is unavailable or not technically feasible, we have demonstrated that our approach represents a safe and effective alternative that achieves good oncologic results with a similar risk and outcome profile. LIR offers the ability to resect adenocarcinomas that fall outside currently accepted indications for endoscopic approaches, such as ESD and EMR in highly selected patients. Namely, tumors larger than 2 cm have traditionally been considered contraindications to ESD or EMR, as well as tumors that are poorly differentiated or have lymphovascular involvement [22]. This is largely attributable to the finding that any of these characteristics significantly heighten the risk of submucosal invasion and subsequent lymph node metastasis. By extension, submucosal invasion (T1b) is associated with at least one involved lymph node in 22% of cases as compared with 1.3% of those with only intramucosal extension (T1a) [23]. En-bloc resection and complete resection rates are typically low for ESD or EMR at 51.7% and 42.2%, respectively, when attempted on extended-criteria lesions larger than 2 cm [24]. All three adenocarcinomas fell out of standard ESD or EMR criteria and as such these patients would have otherwise been subjected to a major gastrectomy. The 3 patients in our series who underwent LIR for these tumors were highly comorbid and elderly, and they likely would have not tolerated a major oncologic resection. Therefore, in light of other favorable tumor characteristics
Ann Thorac Surg 2018;-:-–-
Table 4. Individual Patient Metrics
Patient
Sex
ASA
42 38 41 47 32 62 28 60 39 87 78 85
F M M F F F M F M F M M
N/A 2 3 2 2 2 2 2 2 3 3 3
No No No No No No Yesb No No No 0 0
Tumor Size, Greatest Dimension (cm)
Tumor Volume (cm3)
Histopathologic Feature
No No No No No No No No No Yesc No No
2.3 10 3 5 3.2 5.8 8 5.5 7 3.5 2.2 2.5
6.9 100.0 15.0 45.0 7.2 30.5 142.9 24.8 10.5 7.7 3.0 3.8
Leiomyoma Leiomyoma Leiomyoma Leiomyoma Leiomyoma GIST Leiomyoma Leiomyoma Leiomyoma Adenocarcinoma Adenocarcinoma Adenocarcinoma
LVI
Tumor Grade
Length of Stay (Days)
Complication
Follow-Up (Years)
. . . . . . . . . Yes Yes No
. . . . . . . . . Poor Poor Moderate
2 13 1 3 1 7 4 2 2 21 4 3
. Leak, return to ORa . . . Leak, return to ORa . . . Malignant ascites . .
12.6 10.3 9.6 9.3 4.9 7.7 2.4 1.8 0.6 0.5d 3.7 1.6
a
Both patients were taken back to the OR on postoperative day 1, with successful primary repair. Patient 2 was sent home on total parenteral nutrition and resumed a diet as an outpatient. Patient 6 was able to b be discharged with a full diet. The initial tumor was believed to be consistent with wild-type c-KIT/PDGFRA GIST given the age of patient and the rapid growth; thus, the patient was treated with 5 c months of imatinib. Postoperative pathologic examination confirmed a benign leiomyoma. Argon plasma coagulation was attempted three times by gastroenterology, given the significant patient’s d comorbidities and age. There were two recurrences of disease before referral for surgical resection. Patient 10 underwent an R1 resection, and malignant ascites developed at 6 months without gross locoregional recurrence, with subsequent mortality. ASA ¼ American Society of Anesthesiologists (class); able; OR ¼ operating room.
F ¼ female;
GIST ¼ gastrointestinal stromal tumor;
LVI ¼ postoperative lymphovascular invasion;
M ¼ male;
N/A ¼ not avail-
VILLANO ET AL MINIMALLY INVASIVE INTRAGASTRIC RESECTION
1 2 3 4 5 6 7 8 9 10 11 12
Age
Neoadjuvant Treatment
Previous Resection Attempt
5
6
VILLANO ET AL MINIMALLY INVASIVE INTRAGASTRIC RESECTION
(T1a, <3 cm, no preoperative lymphovascular involvement), our minimally invasive approach affords the potential to avoid major operation, the inherent complication profile of full-thickness gastric resection, and potential morbidity of an anastomosis while improving the ability to achieve an R0 resection. LIR is thus best applied to early stage adenocarcinoma very selectively. Namely, it is best viewed as an alternative approach for tumors that fall out of standard ESD or EMR criteria but meet extended criteria, for patients whose risk morbidity or mortality from esophagectomy outweighs the risk of potential lymph node involvement. Since the advent of LIR, there have been several modifications in the literature, such as a two-gastrostomy approach [15] and a single-incision approach [25]. In 1 patient, we performed the intragastric portion of the dissection with the da Vinci surgical system. To our knowledge this is the first documented use of robotic surgery in this manner. Robot-assisted surgery allows for binocular vision and the ability to angulate instruments, which is especially helpful when operating along the GEJ. Use of the robot did not add significant time to the procedure and had outcomes similar to those with our laparoscopic approach, thus making it a feasible option in selected patients. Further study is necessary to assess its safety and utility adequately in this venue. Although we herein have demonstrated efficacy of the LIR approach, our study has several limitations. First, this is a retrospective review and as such cannot control for study design biases that are inherent to the lack of prospective randomization. Notably, this technique represents a unique and novel approach to a rare subset of surgical disease, and as such it is likely not feasible ever to perform a randomized comparison that could be powered appropriately. Further, ascertainment of a large sample size from a single institution is not feasible. We continue to accrue patients, and future research may aim to expand to a multiinstitutional analysis or meta-analysis of all published series. Finally, we have only recently applied LIR to malignant disease, and as such long-term followup of our accrued patients is necessary to confirm the oncologic benefit we have seen in the short term. Aside from these limitations, the rich and granular level of the presented data affords promise for further study of this approach. In conclusion, we have demonstrated satisfactory long-term outcomes with the minimally invasive intragastric approach to GEJ disease. It offers a safe procedure with short hospital stay and minimal risk of long-term complications, with good oncologic outcomes. LIR is best applied to benign tumors of the GEJ in which nodal status is not of concern; however, we have demonstrated that it can be applied to patients with malignant lesions that meet expanded EMR or ESD criteria but who may have too many comorbidities to withstand esophagectomy. Large, multiinstitutional studies or meta-analysis should be the focus of future research to validate these findings in a heterogenous cohort.
Ann Thorac Surg 2018;-:-–-
References 1. Ishikawa K, Inomata M, Etoh T, et al. Long-term outcome of laparoscopic wedge resection for gastric submucosal tumor compared with open wedge resection. Surg Laparosc Endosc Percutan Tech 2006;16:82–5. 2. Pidhorecky I, Cheney RT, Kraybill WG, et al. Gastrointestinal stromal tumors: current diagnosis, biologic behavior, and management. Ann Surg Oncol 2000;7:705–12. 3. Hwang SH, Park DJ, Kim YH, et al. Laparoscopic surgery for submucosal tumors located at the esophagogastric junction and the prepylorus. Surg Endosc 2009;23:1980–7. 4. Tagaya N, Mikami H, Kogure H, et al. Laparoscopic intragastric stapled resection of gastric submucosal tumors located near the esophagogastric junction. Surg Endosc 2002;16:177–9. 5. Gotoda T. Endoscopic resection of early gastric cancer. Gastric Cancer 2007;10:1–11. 6. Lian J, Chen S, Zhang Y, et al. A meta-analysis of endoscopic submucosal dissection and EMR for early gastric cancer. Gastrointest Endosc 2012;76:763–70. 7. Oka S, Tanaka S, Kaneko I, et al. Advantage of endoscopic submucosal dissection compared with EMR for early gastric cancer. Gastrointest Endosc 2006;64:877–83. 8. Marshall MB, Haddad NG. Laparoscopic intragastric approach for gastroesophageal leiomyoma and cancer. J Thorac Cardiovasc Surg 2015;149:1210–2. 9. Ohashi S. Laparoscopic intraluminal (intragastric) surgery for early gastric cancer. A new concept in laparoscopic surgery. Surg Endosc 1995;9:169–71. 10. Liao YT, Yang CY, Lai IR, et al. Laparoscopic resection for submucosal tumors near the esophagogastric junction: feasibility and short-term outcome. Surg Innov 2013;20: 478–83. 11. Ryu KJ, Jung SR, Choi JS, et al. Laparoscopic resection of small gastric submucosal tumors. Surg Endosc 2011;25:271–7. 12. Lee CH, Hyun MH, Kwon YJ, et al. Deciding laparoscopic approaches for wedge resection in gastric submucosal tumors: a suggestive flow chart using three major determinants. J Am Coll Surg 2012;215:831–40. 13. Bedard ELR, Mamazza J, Schlachta CM, et al. Laparoscopic resection of gastrointestinal stromal tumors: not all tumors are created equal. Surg Endosc 2006;20:500–3. 14. Kanehira E, Kamei A, Umezawa A, et al. Long-term outcomes of percutaneous endoscopic intragastric surgery in the treatment of gastrointestinal stromal tumors at the esophagogastric junction. Surg Endosc 2016;30:2036–42. 15. Gabriel E, Rosso K, Shahab D, et al. Laparoscopic intragastric surgery with endoscopic assistance: a 2 gastrostomy approach with multiple applications. Surg Laparosc Endosc Percutan Tech 2017;27:e116–20. 16. Privette A, McCahill L, Borrazzo E, et al. Laparoscopic approaches to resection of suspected gastric gastrointestinal stromal tumors based on tumor location. Surg Endosc 2008;22:487–94. 17. Ghushe ND, Dulai PS, Trus TL. Laparoendoscopic transgastric resection of a submucosal mass at the gastroesophageal junction. J Gastrointest Surg 2012;16:2321. 18. Uchikoshi F, Ito T, Nishida T, et al. Laparoscopic intragastric resection of gastric stromal tumor located at the esophagocardiac junction. Surg Laparosc Endosc Percutan Tech 2004;14:1–4. 19. Hara J, Nakajima K, Takahashi T, et al. Laparoscopic intragastric surgery revisited: its role for submucosal tumors adjacent to the esophagogastric junction. Surg Laparosc Endosc Percutan Tech 2012;22:251–4. 20. Franklin MEJ, Portillo G, Trevino JM, et al. Laparoscopic intraluminal surgery for gastrointestinal malignancies. World J Surg 2008;32:1709–13. 21. Walsh RM, Ponsky J, Brody F, et al. Combined endoscopic/ laparoscopic intragastric resection of gastric stromal tumors. J Gastrointest Surg 2003;7:386–92.
Ann Thorac Surg 2018;-:-–-
22. Yamaguchi N, Isomoto H, Fukuda E, et al. Clinical outcomes of endoscopic submucosal dissection for early gastric cancer by indication criteria. Digestion 2009;80:173–81. 23. Leers JM, DeMeester SR, Oezcelik A, et al. The prevalence of lymph node metastases in patients with T1 esophageal adenocarcinoma a retrospective review of esophagectomy specimens. Ann Surg 2011;253:271–8.
VILLANO ET AL MINIMALLY INVASIVE INTRAGASTRIC RESECTION
7
24. Facciorusso A, Antonino M, Di Maso M, et al. Endoscopic submucosal dissection vs endoscopic mucosal resection for early gastric cancer: a meta-analysis. World J Gastrointest Endosc 2014;6:555–63. 25. Cazauran JB, Mercier F, Pasquer A, et al. Intragastric singleincision laparoscopic surgery for gastric leiomyoma: a stepwise approach. Ann Surg Oncol 2017;24:2281.
Background. A minimally invasive intragastric approach to the gastroesophageal junction (GEJ) allows resection of intramural disease while avoiding disruption of the lower esophageal sphincter and vagus nerves. Few surgeons use this approach; thus little is known regarding its indications, feasibility, technical aspects, complication profile, and long-term outcomes. This study reviewed the experience with this technique. Methods. A retrospective review was performed of a prospectively maintained, Institutional Review Board– approved database covering the period from January 1, 2005 to August 1, 2017. Indications, operative details, postoperative complications, and outcomes were assessed. Results. There were 12 patients identified. The mean age of these patients was 51.9 years. The indications for resection included 10 symptomatic leiomyomas, one gastrointestinal stromal tumor, and three cancers of the GEJ. Mean and median length of stay were 4.9 and 2.5
days, respectively. There were two postoperative esophageal leaks managed with laparoscopic repair. Of the 3 patients with cancer, 2 underwent an R0 resection, whereas 1 patient underwent an R1 resection. There were no other complications or recurrences. Mean follow-up was 6.0 years (range, 0.5 to 12.6 years); no patients had stricture or symptomatic gastroesophageal reflux on long term follow-up. Conclusions. Resection of selected intramural GEJ disorders through a minimally invasive transgastric approach can be performed safely with acceptable morbidity and good long-term results. The approach allows preservation of the lower esophageal sphincter and vagus nerves, a potential advantage compared with other surgical alternatives to resection in this region.
I
specialized endoscopists, and they are especially challenging when the tumor sits near the GEJ [5–7]. An alternative to these approaches for the resection of GEJ tumors is laparoscopic intragastric resection (LIR). This procedure uses gastric insufflation and multiple port sites placed through the anterior gastric wall and provides a useful alternative to the wedge resection when size, location, or endophytic growth renders a standard laparoscopic approach challenging or impossible. Here we describe our experience with the LIR approach to a variety of intramural GEJ disorders, with an emphasis on long-term follow-up and patient outcomes.
ntramural disease of the gastroesophageal junction (GEJ) presents a unique therapeutic challenge. Submucosal gastric lesions (namely leiomyomas and gastrointestinal stromal tumors [GISTs]) and stage 1 GEJ carcinomas rarely involve lymph nodes and are thus candidates for local resection [1, 2]. Transgastric resection is the most common technique for submucosal lesions located in the gastric body and antrum, but it has lower applicability to GEJ tumors [3] because of the high likelihood of disruption of the lower esophageal sphincter or vagal injury. Without intragastric visualization, it is common to resect excessive normal gastric tissue, thus further limiting the transgastric approach near the GEJ [4]. Endoscopic submucosal dissection (ESD) and endoscopic mucosal resection (EMR) have been widely studied in Asia for the resection of GEJ tumors. However, these techniques are limited by tumor size, they require
Accepted for publication Aug 20, 2018. Presented at the Poster Session of the General Thoracic Surgery Club, Dove Mountain, AZ, March 8–11, 2018. Address correspondence to Dr Villano, Department of General Surgery, MedStar Georgetown University Hospital, 3800 Reservoir Rd, NW, PHC Bldg, 4th Flr, Washington, DC 20007; email: anthony.m.villano@gunet. georgetown.edu.
Ó 2018 by The Society of Thoracic Surgeons Published by Elsevier Inc.
(Ann Thorac Surg 2018;-:-–-) Ó 2018 by The Society of Thoracic Surgeons
Patients and Methods Data Source Deidentified patient information was maintained in a prospective, single-institution, Institutional Review Board–approved database. Information gathered included patient demographics (date of birth, age, sex), preoperative comorbidity status (American Society of Anesthesia classification, [ASA]), operative details (operative time, estimated blood loss, number of intraoperative transfusions), tumor pathologic features (size, histologic characteristics), outcome variables (length of stay, morbidity, mortality), and duration of available follow-up. 0003-4975/$36.00 https://doi.org/10.1016/j.athoracsur.2018.08.050
2
VILLANO ET AL MINIMALLY INVASIVE INTRAGASTRIC RESECTION
Medical records were incomplete for 1 patient and are noted accordingly in the referenced tables.
Operative Approach Specific details of our operative approach have been outlined elsewhere [8]. Briefly, the peritoneal cavity was entered and insufflation achieved as in standard laparoscopy. Flexible endoscopy was used to enter the stomach. A guidewire was used to cannulate the proximal duodenum, and over this a Coda balloon (Cook Incorporated, Bloomington, IN) was positioned and expanded to prevent insufflation of the small bowel. After achieving gastric insufflation, the anterior stomach could be inspected laparoscopically to allow for identification of ideal trocar sites. These trocar site locations varied by case, but they were placed far enough apart to triangulate three instruments consistently to the GEJ. Gastrotomies were made using hook cautery, and three 5-mm Kii balloon-anchored ports (Applied Medical, Rancho Santa Margarita, CA) were placed transperitoneally into the gastric lumen (Fig 1). These ports were secured with stay sutures. For tumors amenable to intragastric resection, a snare or stitch has been used to provide traction into the gastric lumen. Ligasure bipolar cautery (Valleylab Inc, Boulder, CO) was used to dissect and enucleate the tumor circumferentially. If enucleation was not feasible secondary to size or an exophytic component, the tumor could be stapled with a linear articulating Endo GIA (Medtronic, Minneapolis, MN), or a full-thickness resection was performed with cautery (Fig 2). The specimen could then be placed in an Endo Catch (Medtronic) bag and pulled through the oropharynx with the endoscope. Tumors too large for this approach were removed through one of the gastrostomy sites and subsequently through an abdominal wall port site in standard laparoscopic fashion. In the event of a residual defect at the GEJ, closure was performed with interrupted sutures using an Endo Stitch (Medtronic) or standard sutures on a curved needle as our skills improved. The gastrotomy sites were closed in a
Ann Thorac Surg 2018;-:-–-
similar fashion. Before exiting the peritoneal cavity, the endoscope was advanced, and the stomach was insufflated one final time to assess for a leak. The abdomen was then closed using standard procedures.
Pathologic Analysis All tumors were processed and analyzed by two pathologists. Stromal tumors were routinely tested by immunohistochemistry to distinguish leiomyoma and GIST (CD117, CD34, smooth muscle actin, desmin, S-100, and MIB-1/Ki-67). Adenocarcinomas were evaluated for histologic features, size, grade, margins, lymphovascular invasion, and lymph node involvement (if included in the specimen).
Results Patient Characteristics Table 1 outlines the baseline patient characteristics. Mean age at resection was 51.9 years (range, 28 to 87 years). Most of the patients were in ASA class 2; however, all 3 of the patients with adenocarcinoma were in ASA class 3. One patient with leiomyoma underwent 5 weeks of neoadjuvant imatinib treatment for what was originally suspected to be a KIT/PDGFRA wild-type GIST. All 3 patients with adenocarcinoma had T1a tumors without evidence of nodal metastasis (N0) on the basis of preoperative endoscopic ultrasound. Additionally, 1 patient with adenocarcinoma underwent an earlier attempt at ablation with argon plasma coagulation (APC) twice with local recurrence before surgical referral.
Operative Metrics Operative variables are outlined in Table 2. Mean tumor size along the greatest dimension was 4.9 cm; however tumors with a wide range of sizes were resected (range, 0.8 to 8 cm). Tumor volume was of a similarly wide range (5.6 to 142.9 cm3). Mean operative time was 246.5 minutes (range, 106 to 425 minutes). No patients required a blood transfusion intraoperatively, and there were no conversions to an open procedure.
Histology and Outcomes
Fig 1. Laparoscopic view of two intragastric balloon-anchored ports (blue arrows) tenting up the stomach wall, with a third port in the process of being advanced into gastric lumen with silk suture tacking the gastric wall (white arrows).
Final histologic examination yielded 10 leiomyomas, one GIST, and three adenocarcinomas (Fig 3). Mean and medial length of stay was 4.8 and 2.5 days, respectively. In 2 patients, a postoperative leak developed on day 1 that required a return to the operating room for primary repair. There were no additional operative complications, and no patients had long-term sequelae such as dysphagia, odynophagia, esophageal stricture, or symptomatic reflux. Mean follow-up for the cohort was 5.4 years (Table 3). Detailed patient descriptions and outcomes are listed in Table 4. Patient 2 had a leak discovered on routine esophagram postoperative day 1. This patient was taken back to the operating room, and a small defect along the posterior aspect of the lesser curve of the stomach near the GEJ was noted. This defect was repaired primarily
Ann Thorac Surg 2018;-:-–-
VILLANO ET AL MINIMALLY INVASIVE INTRAGASTRIC RESECTION
3
Fig 2. (A) Tumor spanning the Z-line of the gastroesophageal junction and coursing across the distal esophagus and the stomach (dotted white line). (B) Completed full-thickness resection. The white arrow points to the muscular layer of the esophagus. Fat can be visualized at the base of the resection.
through a laparoscopic approach, and no leak was confirmed with repeat endoscopy. Repeat swallow study on postoperative day 7 demonstrated no leak. Patient 7 similarly had a leak identified by swallow study on postoperative day 1 and was taken to the operating room for laparoscopic repair. On exploration, the leak was identified at the proximal end of the staple line created when transecting the tumor from the gastric mucosa. This was oversewn with the Endo Stitch in an intermittent fashion. This patient was advanced to a regular diet before discharge on postoperative day 7. For patient 9, the intragastric portion of the procedure was performed with the use of the da Vinci surgical system (Intuitive Surgical, Sunnyvale, CA). Patient 12 underwent an R1 resection after three previous attempts at APC by gastroenterology with recurrence before surgical referral. Six months postoperatively, malignant ascites developed, and the patients subsequently died in the absence of locoregional recurrence.
Comment A transgastric approach to resection of GEJ disease was first described in 1995 by Ohashi [9]. Since this time, various surgical approaches have been described to Table 1. Patient Characteristics Characteristic Mean age (years) Sex Male Female Mean Preoperative ASA classa Neoadjuvant therapya Yes No Prior resection attempta Yes No a
Value (%) 53.2 6 (40) 6 (50) 2.4 1 (8) 11 (92) 1 (8) 11 (92)
Chart incomplete for patient 1. Values represent data for remaining 11 patients. ASA ¼ American Society of Anesthesiologists.
address this anatomically challenging subset of tumors. The most commonly used approaches have been exogastric and transgastric wedge resection, both of which have demonstrated good safety and efficacy in several small case series [10–12]. For large tumors with an endophytic growth pattern, this approach risks resection of unnecessarily large areas of gastric mucosa, with the possibility of injuring the vagus nerves, narrowing the GEJ, or converting to an open procedure [13]. Recognizing these limitations, modification of this strategy has led to the use of a minimally invasive intragastric approach through gastric insufflation and use of intragastric trocars. Advantages of this method include improved visualization (especially of endophytic tumors) and the ability to limit the extent of resection only to what is oncologically necessary. Traditional laparoscopic approaches often pose a challenge in accessing these high, posteriorly located tumors and can pose difficult angles for firing of the endoscopic stapler or safe application of electrocautery. LIR provides a unique visual perspective most surgeons are not used to, and it creates novel angles whereby the volume of tissue resected can be minimized and the risk of narrowing the GEJ or injuring the vagus nerves can be diminished. To date, this technique has now been used by various other groups in small case reports and has demonstrated safety and efficacy for benign tumors near the GEJ [4, 14–21]. Our present study represents one of the largest case series to date that used this method with a combination of benign and malignant disease. Although others have suggested that enucleation is likely the only way to
Table 2. Operative Metrics Characteristic
Value (Range)
Mean tumor size, greatest dimension (cm) Mean tumor volume (cm3) Mean operative time (minutes)a Mean estimated blood loss (mL)a Intraoperative transfusiona Conversion to open
4.8 33.1 265 45
a
(0.8–8.0) (5.6–142.9) (106–425) (5–100) 0 0
Chart incomplete for patient 1. Values represent data for remaining 11 patients.
4
VILLANO ET AL MINIMALLY INVASIVE INTRAGASTRIC RESECTION
Ann Thorac Surg 2018;-:-–-
Fig 3. Representative images of resected gastroesophageal junction disease. (A) Leiomyoma. (B) Gastrointestinal stromal tumor. (C) Adenocarcinoma.
remove tumors in close proximity to the GEJ (<3 cm) [14], we have very selectively performed full-thickness resection with cautery and primary repair with satisfactory results. Our results demonstrate that using such an approach is safe, affords a short length of stay, and confers minimal risk of long-term morbidity for the patient. Further, our series reaffirms that this approach confers acceptable operative times with minimal blood loss and can be performed in patients with multiple comorbidities. No patients in our series experienced stricture development or a vagal nerve injury. The 2 patients who had an esophageal leak postoperatively were able to undergo primary repaired without long-term complications. The tumors in both these patients were large (5.8 and 10 cm), and although the leaks were readily addressed by a second procedure, their complications highlight the point that large tumors at the GEJ carry greater risk for severe complications. Therefore, a challenging dissection may warrant modification of the surgical approach when difficulty is encountered by the operating surgeon. However, size should not be considered an absolute contraindication to this approach because three
Table 3. Postoperative Histologic Characteristics and Outcomes Characteristic Histopathology Leiomyoma Adenocarcinoma GIST Other Median length of stay (days) Complications No Postoperative leak Operative mortality Locoregional recurrences Mean follow-up (months) GIST ¼ gastrointestinal stromal tumor.
Value (%)
8 (67) 3 (25) 1 (8) 0 2.5 10 (83) 2 (17) 0 (0) 0 (0) 65
additional tumors 5 cm or larger were resected without such issues. Accumulation of additional patients with large tumors is warranted to clarify this point. One unique indication for LIR is early stage adenocarcinoma. Ours is the first study to describe technical and short-term oncologic outcomes of this population. From this standpoint, R0 resection was achieved in 2 of the 3 patients. The single cancer recurrence and subsequent mortality were in the setting of an R1 resection after two prior attempts at APC, which led to malignant ascites without gross locoregional recurrence. Patient selection is key to complete resection and long-term success of this operative strategy. In the setting of previous attempts at treatment, the ability to achieve an R0 resection is certainly more difficult and warrants further study before this can be recommended. When use of EMR is unavailable or not technically feasible, we have demonstrated that our approach represents a safe and effective alternative that achieves good oncologic results with a similar risk and outcome profile. LIR offers the ability to resect adenocarcinomas that fall outside currently accepted indications for endoscopic approaches, such as ESD and EMR in highly selected patients. Namely, tumors larger than 2 cm have traditionally been considered contraindications to ESD or EMR, as well as tumors that are poorly differentiated or have lymphovascular involvement [22]. This is largely attributable to the finding that any of these characteristics significantly heighten the risk of submucosal invasion and subsequent lymph node metastasis. By extension, submucosal invasion (T1b) is associated with at least one involved lymph node in 22% of cases as compared with 1.3% of those with only intramucosal extension (T1a) [23]. En-bloc resection and complete resection rates are typically low for ESD or EMR at 51.7% and 42.2%, respectively, when attempted on extended-criteria lesions larger than 2 cm [24]. All three adenocarcinomas fell out of standard ESD or EMR criteria and as such these patients would have otherwise been subjected to a major gastrectomy. The 3 patients in our series who underwent LIR for these tumors were highly comorbid and elderly, and they likely would have not tolerated a major oncologic resection. Therefore, in light of other favorable tumor characteristics
Ann Thorac Surg 2018;-:-–-
Table 4. Individual Patient Metrics
Patient
Sex
ASA
42 38 41 47 32 62 28 60 39 87 78 85
F M M F F F M F M F M M
N/A 2 3 2 2 2 2 2 2 3 3 3
No No No No No No Yesb No No No 0 0
Tumor Size, Greatest Dimension (cm)
Tumor Volume (cm3)
Histopathologic Feature
No No No No No No No No No Yesc No No
2.3 10 3 5 3.2 5.8 8 5.5 7 3.5 2.2 2.5
6.9 100.0 15.0 45.0 7.2 30.5 142.9 24.8 10.5 7.7 3.0 3.8
Leiomyoma Leiomyoma Leiomyoma Leiomyoma Leiomyoma GIST Leiomyoma Leiomyoma Leiomyoma Adenocarcinoma Adenocarcinoma Adenocarcinoma
LVI
Tumor Grade
Length of Stay (Days)
Complication
Follow-Up (Years)
. . . . . . . . . Yes Yes No
. . . . . . . . . Poor Poor Moderate
2 13 1 3 1 7 4 2 2 21 4 3
. Leak, return to ORa . . . Leak, return to ORa . . . Malignant ascites . .
12.6 10.3 9.6 9.3 4.9 7.7 2.4 1.8 0.6 0.5d 3.7 1.6
a
Both patients were taken back to the OR on postoperative day 1, with successful primary repair. Patient 2 was sent home on total parenteral nutrition and resumed a diet as an outpatient. Patient 6 was able to b be discharged with a full diet. The initial tumor was believed to be consistent with wild-type c-KIT/PDGFRA GIST given the age of patient and the rapid growth; thus, the patient was treated with 5 c months of imatinib. Postoperative pathologic examination confirmed a benign leiomyoma. Argon plasma coagulation was attempted three times by gastroenterology, given the significant patient’s d comorbidities and age. There were two recurrences of disease before referral for surgical resection. Patient 10 underwent an R1 resection, and malignant ascites developed at 6 months without gross locoregional recurrence, with subsequent mortality. ASA ¼ American Society of Anesthesiologists (class); able; OR ¼ operating room.
F ¼ female;
GIST ¼ gastrointestinal stromal tumor;
LVI ¼ postoperative lymphovascular invasion;
M ¼ male;
N/A ¼ not avail-
VILLANO ET AL MINIMALLY INVASIVE INTRAGASTRIC RESECTION
1 2 3 4 5 6 7 8 9 10 11 12
Age
Neoadjuvant Treatment
Previous Resection Attempt
5
6
VILLANO ET AL MINIMALLY INVASIVE INTRAGASTRIC RESECTION
(T1a, <3 cm, no preoperative lymphovascular involvement), our minimally invasive approach affords the potential to avoid major operation, the inherent complication profile of full-thickness gastric resection, and potential morbidity of an anastomosis while improving the ability to achieve an R0 resection. LIR is thus best applied to early stage adenocarcinoma very selectively. Namely, it is best viewed as an alternative approach for tumors that fall out of standard ESD or EMR criteria but meet extended criteria, for patients whose risk morbidity or mortality from esophagectomy outweighs the risk of potential lymph node involvement. Since the advent of LIR, there have been several modifications in the literature, such as a two-gastrostomy approach [15] and a single-incision approach [25]. In 1 patient, we performed the intragastric portion of the dissection with the da Vinci surgical system. To our knowledge this is the first documented use of robotic surgery in this manner. Robot-assisted surgery allows for binocular vision and the ability to angulate instruments, which is especially helpful when operating along the GEJ. Use of the robot did not add significant time to the procedure and had outcomes similar to those with our laparoscopic approach, thus making it a feasible option in selected patients. Further study is necessary to assess its safety and utility adequately in this venue. Although we herein have demonstrated efficacy of the LIR approach, our study has several limitations. First, this is a retrospective review and as such cannot control for study design biases that are inherent to the lack of prospective randomization. Notably, this technique represents a unique and novel approach to a rare subset of surgical disease, and as such it is likely not feasible ever to perform a randomized comparison that could be powered appropriately. Further, ascertainment of a large sample size from a single institution is not feasible. We continue to accrue patients, and future research may aim to expand to a multiinstitutional analysis or meta-analysis of all published series. Finally, we have only recently applied LIR to malignant disease, and as such long-term followup of our accrued patients is necessary to confirm the oncologic benefit we have seen in the short term. Aside from these limitations, the rich and granular level of the presented data affords promise for further study of this approach. In conclusion, we have demonstrated satisfactory long-term outcomes with the minimally invasive intragastric approach to GEJ disease. It offers a safe procedure with short hospital stay and minimal risk of long-term complications, with good oncologic outcomes. LIR is best applied to benign tumors of the GEJ in which nodal status is not of concern; however, we have demonstrated that it can be applied to patients with malignant lesions that meet expanded EMR or ESD criteria but who may have too many comorbidities to withstand esophagectomy. Large, multiinstitutional studies or meta-analysis should be the focus of future research to validate these findings in a heterogenous cohort.
Ann Thorac Surg 2018;-:-–-
References 1. Ishikawa K, Inomata M, Etoh T, et al. Long-term outcome of laparoscopic wedge resection for gastric submucosal tumor compared with open wedge resection. Surg Laparosc Endosc Percutan Tech 2006;16:82–5. 2. Pidhorecky I, Cheney RT, Kraybill WG, et al. Gastrointestinal stromal tumors: current diagnosis, biologic behavior, and management. Ann Surg Oncol 2000;7:705–12. 3. Hwang SH, Park DJ, Kim YH, et al. Laparoscopic surgery for submucosal tumors located at the esophagogastric junction and the prepylorus. Surg Endosc 2009;23:1980–7. 4. Tagaya N, Mikami H, Kogure H, et al. Laparoscopic intragastric stapled resection of gastric submucosal tumors located near the esophagogastric junction. Surg Endosc 2002;16:177–9. 5. Gotoda T. Endoscopic resection of early gastric cancer. Gastric Cancer 2007;10:1–11. 6. Lian J, Chen S, Zhang Y, et al. A meta-analysis of endoscopic submucosal dissection and EMR for early gastric cancer. Gastrointest Endosc 2012;76:763–70. 7. Oka S, Tanaka S, Kaneko I, et al. Advantage of endoscopic submucosal dissection compared with EMR for early gastric cancer. Gastrointest Endosc 2006;64:877–83. 8. Marshall MB, Haddad NG. Laparoscopic intragastric approach for gastroesophageal leiomyoma and cancer. J Thorac Cardiovasc Surg 2015;149:1210–2. 9. Ohashi S. Laparoscopic intraluminal (intragastric) surgery for early gastric cancer. A new concept in laparoscopic surgery. Surg Endosc 1995;9:169–71. 10. Liao YT, Yang CY, Lai IR, et al. Laparoscopic resection for submucosal tumors near the esophagogastric junction: feasibility and short-term outcome. Surg Innov 2013;20: 478–83. 11. Ryu KJ, Jung SR, Choi JS, et al. Laparoscopic resection of small gastric submucosal tumors. Surg Endosc 2011;25:271–7. 12. Lee CH, Hyun MH, Kwon YJ, et al. Deciding laparoscopic approaches for wedge resection in gastric submucosal tumors: a suggestive flow chart using three major determinants. J Am Coll Surg 2012;215:831–40. 13. Bedard ELR, Mamazza J, Schlachta CM, et al. Laparoscopic resection of gastrointestinal stromal tumors: not all tumors are created equal. Surg Endosc 2006;20:500–3. 14. Kanehira E, Kamei A, Umezawa A, et al. Long-term outcomes of percutaneous endoscopic intragastric surgery in the treatment of gastrointestinal stromal tumors at the esophagogastric junction. Surg Endosc 2016;30:2036–42. 15. Gabriel E, Rosso K, Shahab D, et al. Laparoscopic intragastric surgery with endoscopic assistance: a 2 gastrostomy approach with multiple applications. Surg Laparosc Endosc Percutan Tech 2017;27:e116–20. 16. Privette A, McCahill L, Borrazzo E, et al. Laparoscopic approaches to resection of suspected gastric gastrointestinal stromal tumors based on tumor location. Surg Endosc 2008;22:487–94. 17. Ghushe ND, Dulai PS, Trus TL. Laparoendoscopic transgastric resection of a submucosal mass at the gastroesophageal junction. J Gastrointest Surg 2012;16:2321. 18. Uchikoshi F, Ito T, Nishida T, et al. Laparoscopic intragastric resection of gastric stromal tumor located at the esophagocardiac junction. Surg Laparosc Endosc Percutan Tech 2004;14:1–4. 19. Hara J, Nakajima K, Takahashi T, et al. Laparoscopic intragastric surgery revisited: its role for submucosal tumors adjacent to the esophagogastric junction. Surg Laparosc Endosc Percutan Tech 2012;22:251–4. 20. Franklin MEJ, Portillo G, Trevino JM, et al. Laparoscopic intraluminal surgery for gastrointestinal malignancies. World J Surg 2008;32:1709–13. 21. Walsh RM, Ponsky J, Brody F, et al. Combined endoscopic/ laparoscopic intragastric resection of gastric stromal tumors. J Gastrointest Surg 2003;7:386–92.
Ann Thorac Surg 2018;-:-–-
22. Yamaguchi N, Isomoto H, Fukuda E, et al. Clinical outcomes of endoscopic submucosal dissection for early gastric cancer by indication criteria. Digestion 2009;80:173–81. 23. Leers JM, DeMeester SR, Oezcelik A, et al. The prevalence of lymph node metastases in patients with T1 esophageal adenocarcinoma a retrospective review of esophagectomy specimens. Ann Surg 2011;253:271–8.
VILLANO ET AL MINIMALLY INVASIVE INTRAGASTRIC RESECTION
7
24. Facciorusso A, Antonino M, Di Maso M, et al. Endoscopic submucosal dissection vs endoscopic mucosal resection for early gastric cancer: a meta-analysis. World J Gastrointest Endosc 2014;6:555–63. 25. Cazauran JB, Mercier F, Pasquer A, et al. Intragastric singleincision laparoscopic surgery for gastric leiomyoma: a stepwise approach. Ann Surg Oncol 2017;24:2281.