An Early Experience Using the Technique of Transoral OrVil EEA Stapler for Minimally Invasive Transthoracic Esophagectomy

Transcription

1 An Early Experience Using the Technique of Transoral OrVil EEA Stapler for Minimally Invasive Transthoracic Esophagectomy Dawn E. Jaroszewski, MD, Dustin G. Williams, MBBCh BAO, David E. Fleischer, MD, Helen J. Ross, MD, Yvonne Romero, MD, and Kristi L. Harold, MD Divisions of Cardiothoracic Surgery and Minimally Invasive Surgery, Department of Surgery and the Divisions of Gastroenterology and Hematology/Oncology, Department of Medicine, Mayo Clinic Arizona, Phoenix, Arizona; and the Division of Gastroenterology and Hepatology, Mayo Clinic Rochester, Rochester, New York Background. Minimally invasive esophagectomy (MIE) has been performed using a variety of techniques evolving during the past decade. We present our initial experience and outcomes of patients undergoing transthoracic MIE using a circular end-to-side anastomosis created with the transorally directed EEA circular stapler OrVil (Covidien, New Haven, CT). Complications, including anastomotic leak and stricture occurrence, are reviewed. Methods. A retrospective review evaluated consecutive patients undergoing MIE for esophageal cancer or related disease with intrathoracic end-to-side anastomoses using the transorally directed EEA circular stapler from December 2007 to May Medical records were reviewed for demographics, staging, neoadjuvant chemoradiotherapy, comorbidities, adjuvant therapy, complications, and survival. Results. During this period, 51 consecutive patients (84% male; mean age, 65 years) underwent MIE. Neoadjuvant chemoradiotherapy was performed in 32 patients, and 4 had intraoperative radiotherapy. Mean operative time was 338 minutes (range, 211 to 565 minutes), including the 4 patients with intraoperative radiotherapy. Operative time improved with experience (excluding intraoperative radiotherapy) from a mean of 378 minutes (patients 1 to 14) to 300 minutes (patients 37 to 51). Median hospital stay was 11 days (range, 6 to 48 days). Anastomotic leaks occurred in 5 patients (9.8%). Postoperative deaths included 1 in-hospital (2.0%) and 2 (3.9%) after discharge. Stricture was diagnosed and treated in 7 patients (13.7%). Follow-up was a median of 12 months (range, 1 to 31 months). Conclusions. Transthoracic MIE using an end-to-side anastomosis with the transorally directed EEA circular stapler resulted in acceptable stricture and leak rates with good outcomes comparable to published outcomes for open surgical resections. (Ann Thorac Surg 2011;92:1862 9) 2011 by The Society of Thoracic Surgeons The feasibility and safety of minimally invasive esophagectomy (MIE) for patients with esophageal cancer has been demonstrated in multiple centers worldwide [1 3]. Different techniques of MIE have been proposed and have evolved during the past decade [4 6]. Unless contraindicated by tumor location or a previous operation, we currently favor transthoracic MIE (Ivor Lewis) over cervical anastomosis, which may have increased risks, including leak, stricture, recurrent laryngeal nerve injury, and pharyngoesophageal swallowing dysfunction [3, 7, 8]. The intrathoracic anastomosis can be technically challenging and time consuming, and a variety of techniques have been described to meet and address these challenges [6, 9 11]. Optimal methods for anastomotic formation are debated, and several institutions have presented different techniques to minimize leak risk and postoperative stricture formation [9, 10, 12]. We present the early experience and outcomes of a Accepted for publication July 12, Address correspondence to Dr Jaroszewski, Division of Cardiothoracic Surgery, Mayo Clinic Arizona, 5777 E Mayo Blvd, Phoenix, AZ 85054; jaroszewski.dawn@mayo.edu. consecutive series of transthoracic MIE using an circular end-to-side anastomosis created with a standardized XL EEA circular stapler with an OrVil transorally directed anvil (Covidien, New Haven, CT) placed in the proximal esophageal stump. Primary outcome points included anastomotic leak and stricture occurrence. Material and Methods This retrospective review from a prospectively maintained database evaluated consecutive patients of a single surgeon undergoing MIE for esophageal cancer, or related disease, with am intrathoracic circular-stapled end-to-side anastomosis created with the transoral OrVil anvil from December 2007 to May This was the exclusive technique used for all thoracic anastomosis during this time. Approval was obtained from the Institutional Review Board. Medical records were reviewed for demographics, staging, neoadjuvant chemoradiotherapy (ncrt), comorbidities, adjuvant therapy, complications, and survival. Biopsy reports and pathologic assessments were reviewed to determine the stage of disease using the 2011 by The Society of Thoracic Surgeons /$36.00 Published by Elsevier Inc doi: /j.athoracsur

2 Ann Thorac Surg JAROSZEWSKI ET AL 2011;92: EEA FOR MINIMALLY INVASIVE ESOPHAGECTOMY 1863 American Joint Committee on Cancer system from 2002 based on the T N M classification. At Mayo Clinic, all potentially resectable patients are staged with computed tomography, 2-dexoy-2-18Ffluoro-D-glucose positron emission tomography, and endoscopic ultrasound imaging. If evidence of distant metastatic disease exists, confirmatory biopsies are performed. Endoscopic ultrasound-guided fine-needle aspiration is routinely performed on all suspicious lymph nodes, provided the primary tumor is not transgressed. Patients thought to have T1 N0 M0 or T2 N0 M0 esophageal cancer are offered esophagogastrectomy alone. Patients with T3 or deeper tumor, or fine needle aspiration confirmed local or celiac lymph node metastasis, are offered ncrt, followed by resection. Restaging after ncrt includes computed tomography and positron emission tomography to assess resectability. All operations were done with intent to cure. At our institution, ncrt consisted of radiotherapy given in 28 daily 160-cGy fractions, 5 days/week for a total of 5,040 cgy. Chemotherapy administered before mid-2008 consisted of cisplatin (75 mg/m 2 over 2 hours) given on the first and last 4 days of radiotherapy, and 5-flurouracil leucovorin given as a continuous 96-hour infusion (225 mg/m 2 /day). After mid-2008, chemotherapy consisted of paclitaxel (45 mg/m 2 over 1 hour) once a week, followed by carboplatin (area under the concentration 2 over 30 minutes) once a week, and 5-fluorouracil leucovorin given as a continuous 96-hour infusion (225 mg/m 2 /day). Both chemotherapy regimens were given over the 5- to 6-week course concurrent with radiotherapy. A 4- to 6-week recovery period followed. Patients were considered to have completed ncrt if they finished two chemotherapy cycles and radiotherapy. Dose reductions were not considered criteria to rule ncrt incomplete. Definition of Leak and Stricture Anastomotic leakage was defined as any identifiable contrast extravasation at the esophagogastric anastomosis on radiographic imaging by barium contrast medium esophagram or computed tomography, or clinical symptoms of leakage, or both. Stricture diagnosis was based on development of postoperative dysphagia combined with radiographic/ endoscopic evidence of significant narrowing at the anastomosis requiring dilation. This included any anastomotic opening of less than 15 mm not due to recurrent disease. The predilation diameter was measured to classify strictures as minimal ( 12 mm), mild (9 to 12 mm), moderate (5 to 8 mm), or severe ( 5 mm). Dilation was routinely performed using balloon dilation with increasing diameter at each balloon passage. Follow-up included clinical response to treatment and repeat endoscopy with staged dilations if strictures were moderate to severe or if symptoms did not improve. Mortality Operative mortality included all deaths within 90 days of operation, including patients still hospitalized and those who died later after hospital discharge. Statistical significance was determined using the Student t test, 2 test, or Fisher exact test. All tests were two-sided with p 0.05 for statistical significance. Surgical and Anastomotic Technique A thoracic epidural is placed in all patients before the operation for pain control that is used for 4 to 5 days for narcotic analgesia, which then can be administered through the feeding tube as needed. A double-lumen endotracheal tube is used to allow right-sided isolation of the lung. The operation involves laparoscopic formation of a gastric conduit and thoracoscopic resection of the esophagus with an esophagogastric anastomosis using the OrVil 25-mm anvil and the EEA XL circular stapler (25-mm with 4.8-mm staples). This procedure has been described previously [6, 10, 13]. Laparoscopic Gastric Conduit Formation and Jejunal Tube Placement The patient is first placed in the split-leg supine position and abdominal ports are placed (Fig 1A) [13]. The gastrohepatic ligament and short-gastric vessels are divided, and the crus of the diaphragm is exposed. Retrogastric attachments to the pancreas are divided, and the stomach is elevated to divide the left gastric artery at its base. Fig 1. (A) Port placement for the abdominal laparoscopy port sites and (B) thoracic port sites.

3 1864 JAROSZEWSKI ET AL Ann Thorac Surg EEA FOR MINIMALLY INVASIVE ESOPHAGECTOMY 2011;92: Celiac and associated lymph nodes are taken with the specimen or dissected out and sent separately. The stomach is fully mobilized to the proximal duodenum with care taken to avoid injury to the gastroepiploic vasculature. The esophageal dissection is extended through the hiatus into the lower mediastinum. Surgical or chemical pyloromyotomy is performed. The first 8 patients reviewed underwent surgical pyloromyotomy using the hook cautery, and subsequent patients had chemical pyloromyotomy using botulinum toxin A (Botox, Allergan, Inc. Irvine, CA) injections of 1 ml into 4 quadrants of the anterior pyloric muscle. A gastric conduit is formed with 4 to 6 cm of the proximal stomach excluded as a margin from the distal esophageal tumor. The lesser curvature of the stomach is divided using a 3.5-mm reticulating Endo-GIA universal stapler (United States Surgical Corp, Norwalk, CT) to create a conduit 4 to 5 cm wide. A portion of the proximal stomach is left connected, facilitating mobilization of the conduit into the thorax during the thoracoscopic portion of the case. Over-sewing of the gastric staple line is also performed using a running Vicryl suture (Ethicon, Somerville, NJ). All patients receive a jejunal feeding tube before or in conjunction with esophagectomy. Transthoracic Resection of Esophagus, Mediastinal Lymph Node Dissection, and Formation of Anastomosis The patient is positioned in left lateral decubitus with a break in the operating room table at the chest to allow maximal expansion between the ribs. Four port sites are placed from the axilla to the subscapular region from T4 to T7 (Fig 1B). The azygos vein is divided, and the esophagus, periesophageal fat, and lymph nodes are dissected circumferentially from the diaphragm to above the azygos vein. A reticulating 4.8-mm Endo-GIA stapler is used to divide the esophagus proximally. The specimen is removed in a 15-mm Endocatch retrieval bag (Covidien, New Haven, CT). After confirmation of adequate margins by pathology, the anastomosis is performed. The OrVil EEA anvil, attached to an orogastric tube, is passed transorally by the anesthesiologist until pressure from the orogastric tube is visualized at the staple line of the esophageal stump (Fig 2A, B). It is critical that the orogastric tube does not curl and that it creates a single focus of pressure on the esophageal stump against which a hook cautery is applied until the tip of the tube emerges through the newly made esophagotomy. It is critical to make the enterotomy only as large as necessary to allow passage of the orogastric tube. Whether this is made posterior to, exactly at, or in front of the staple line does not seem to be important, as long as the enterotomy is minimized. The orogastric tube is released from the anvil with cutting of the sutures and carefully pulled out of the thorax through a trocar to prevent contamination of the port site. The tube is stabilized and pulled with an instrument that prevents dilation and trauma to the esophagotomy. Fig 2. (A) An OrVile 25-mm anvil and orogastric tube is passed by the anesthesia personnel or operating room team into the patient s mouth and distal to the stapled esophageal stump. (B) A small enterotomy is made in the esophageal stump with a hook cautery along the stapled line through which the orogastric tube is passed and pulled into the thoracic space. A gastrotomy is made with the harmonic scalpel at the superior aspect of the stapled gastric conduit margin, avoiding the short gastric remnants (Fig 3). We have found that this can be made at a variety of sites, depending on the availability and required length of the conduit. Most of the distal aspect of the conduit will be transected off with closure of the gastrostomy, so the location should be planned to exclude any tissue with a questionable blood supply. The 25-mm EEA stapler is passed through the large trocar site into the thorax. This can be difficult in patients with small, narrow rib spaces. By maximizing the break in the operating room table, the stapler head can be wedged at an angle between the ribs and the EEA head will advance through the intercostal space. The rest of the stapler base unit will easily pass. The stapler head is placed into the conduit through the gastrotomy at the superior staple line of the gastric conduit. The pin is then advanced through the posterior gastric wall, centered between the staple line and greater

4 Ann Thorac Surg JAROSZEWSKI ET AL 2011;92: EEA FOR MINIMALLY INVASIVE ESOPHAGECTOMY 1865 curvature for docking with the anvil (Fig 4). Excellent visualization of the proper angle of insertion is critical without distortion or pressure on the prongs as they lock over the orange stripe (Fig 5). The stapler is then closed as the gastric tissues are held tightly and stretched across the EEA, preventing excessive tissue from rolling into the anastomosis. A distinct give in the tension of the conduit tissue can be visually noted, confirming that not too much tissue from the gastric conduit has been incorporated into the anastomosis. The EEA is fired and the tissue rings can be evaluated for completeness. The intraluminal surface of the newly formed gastroesophageal anastomosis can also be seen through the gastrotomy and evaluated before closure. A 3.5-mm linear stapler is used to close the gastrotomy, ultimately removing the tissue of the uppermost portion of the conduit that has the most potential for ischemia. Postoperative Care A chest tube is placed through one of the port sites, and a nasogastric tube is passed beyond the anastomosis and kept in place until a radiographic swallow evaluation confirms anastomotic integrity (typically between postoperative days 5 and 7). Port sites are thoroughly irrigated and closed with layered dissolvable suture, and dressings are placed. Patients are started on standard tube feeding within 48 to 72 hours. Oral intake is restricted to liquids until 10 to 14 days after the operation. Ambulation with physical Fig 4. The pin of the EEA stapler is advanced through the posterior gastric wall, centered between the staple line and greater curvature, for docking with the anvil. therapy on postoperative day 1 and aggressive pulmonary toilet are initiated. Results During the study period, 51 patients had MIE for distal esophageal disease and are further described in Table 1. Fig 3. A gastrotomy is made with the harmonic scalpel at the superior aspect of the stapled gastric conduit margin, avoiding the short gastric remnants. Fig 5. Careful parallel alignment and use of a grasper to stabilize and guide direction and angle is helpful for docking of the anvil with the pin advanced through posterior gastric wall.

5 1866 JAROSZEWSKI ET AL Ann Thorac Surg EEA FOR MINIMALLY INVASIVE ESOPHAGECTOMY 2011;92: Table 1. Demographic Data Variables Mean (range) or No. (%) Demographics Age 65 (32 88) Male 43 (84.3) Body mass index, kg/m ( ) Pre-op weight loss (lbs) 9 (0 36) Histology Adenocarcinoma 44 (86.3) Squamous cell carcinoma 2 (3.9) Dysplasia 3 (5.9) Metastatic breast cancer a 1 (2.0) Benign refractory stricture 1 (2.0) Preop clinical stage (T N M stage) 0 (Tis, N0, M0) 3 (5.9) I (T1, N0, M0) 1 (2.0) IIA (T2, N0, M0 or T3, N0, M0) 19 (37.3) IIB (T1, N1, M0 or T2, N1, M0) 2 (3.9) III (T3, N1, M0 or T4, any N, M0) 17 (33.3) IVA (any T, any N, M1a) b 7 (13.7) a Metastatic breast cancer was unknown until after diagnostic pathology postresection. b IVA, M1a metastasis in celiac lymph nodes. Of these 51 patients, ncrt was performed in 32 of the patients (63%) with esophageal cancer. Jejunostomy feeding tubes were placed in all patients, with 14 placed before or during ncrt. Mean operative time was 338 minutes (range, 211 to 565 minutes), which included 4 patients who underwent intraoperative radiotherapy (IORT). A significant learning curve and improvement of operating time was observed (cases without IORT), with the first 14 cases averaging 378 minutes vs the last 14 at 300 minutes (p 0.004). Mean estimated blood loss was 274 ml (range, 50 to 1,200 ml). There were no conversions to open surgery; however, 4 patients underwent IORT requiring a 4- to 5-cm laparotomy incision extending the midline-supraumbilical port for placement of the radiation cone to the tumor bed site. These patients received 12.5 Gy of radiation incorporating the celiac artery and diaphragmatic crus. IORT added, on average, more than 75 minutes to the operating room time. The mean of 18 lymph nodes (range, 7 to 49) were harvested. Of the 51 patients, 50 had an R0 resection, with one presumed R1 resection due to microscopic disease extending through the adventitia to the marked periesophageal fat on final pathology in a patient without prior ncrt. This patient was subsequently treated with adjuvant CRT. An additional 16 patients received adjuvant CRT due to positive nodal disease found on resected specimens or what was felt to be poor response to ncrt. Over all morbidity was 49.0%, and 90-day all-cause mortality was 5.9% (Table 2). There was only one inhospital death (2.0%), which occurred on postoperative day 36, from acute respiratory distress syndrome and sepsis subsequent to suspected aspiration pneumonia. Two patients (3.9%) who were discharged after uncomplicated hospital stays died within 90 days of the operation. One died of a suspected pulmonary embolism and the other, an 88-year-old man, was discharged to his care facility and died of unknown cause. No autopsy reports were authorized by remaining family to better assess the cause of death. Median hospital stay was 11 days (range, 6 to 48 days). Three patients had extended hospital stays due to respiratory distress syndrome, including one with an anastomotic leak and prolonged ventilator dependence. Excluding these outliers, median hospital stay was 9 days (range, 6 to 16 days). Anastomotic leaks were observed in 5 patients (9.8%). All but 2 were managed conservatively. Two patients with anastomotic leaks were treated surgically with evacuation of empyema and drainage tube placement (Table 3). Two patients diagnosed with leaks were clinically asymptomatic. One patient s leak was suspicious on the initial barium swallow, but a repeat swallow showed no leak. Stricture was diagnosed and treated in 7 patients (13.7%) using previously defined diagnostic criteria (Table 4). Follow-up was available in all 51 patients, with a median of 12 months (range, 6 to 31 months). Comment Performing the intrathoracic anastomosis can be the major limiting step of transthoracic MIE [1, 10]. Handsewn, linear, and the standard end-to-side anastomosis each have limiting factors [9]. Technical challenges, advanced thoracoscopic skills, time consumption, and the need to increase incision size for access into the chest cavity are all encountered with these techniques. The EEA stapler has been successfully used in colorectal and bariatric operations for years, and recent demonstrations of its utility in MIE have been shown [14 16]. Bizekis and colleagues [14] described minimally invasive Ivor Lewis esophagectomy involving transthoracic place- Table 2. In-Hospital Morbidity and Mortality Outcome No. (%) Morbidity Atrial fibrillation 16 (31.4) Pneumonia 8 (15.7) Acute respiratory distress syndrome 3 (5.9) Return to operating room 7 (13.7) Anastomotic leak 2 (3.9) Bleeding 2 (3.9) Chylothorax 1 (2.0) Tracheostomy 1 (2.0) Jejunal feeding tube revision 1 (2.0) 90-day mortality 3 (5.9) Acute respiratory distress syndrome 1 (2.0) Suspected pulmonary embolism 1 (2.0) Unknown cause 1 (2.0)

6 Ann Thorac Surg JAROSZEWSKI ET AL 2011;92: EEA FOR MINIMALLY INVASIVE ESOPHAGECTOMY 1867 Table 3. Anastomotic Leaks Age Sex Anastomosis Location Comorbidity Post-op Complications Stent in Situ ncrt Current Status (cm) 82 M 28 Hypertension, Hodgkin s lymphoma treated with chest RT 62 F 25 Breast cancer with esophageal metastasis treated with chest RT 62 M 25 Hypertension, vascular disease 59 F 25 Breast cancer and thymoma, treated with chest RT, valvular heart disease Return to OR for hemorrhage, A-Fib (wks) None RT No stricture No dysphagia Small contained leak 6 RT No stricture A-Fib, hypotension, chylothorax, prolonged intubation, sepsis, empyema Return to OR for leak repair with AlloDerm, a prolonged intubation, pleural effusion 58 M 25 None A-Fib; small contained leak that self-resolved 12 RT, CT No dysphagia Died of metastatic disease None RT No stricture None RT, CT No dysphagia Died of metastatic disease a LifeCell Corp, Branchburg, NJ. A-Fib atrial fibrillation; CT chemotherapy; F female; M male; ncrt neoadjuvant chemoradiotherapy; OR operating room; RT radiotherapy. ment of the anvil through a minithoracotomy with handsewing of the anvil into the proximal esophageal stump [14]. This method, although successful, requires time and technical expertise. Changes in the design of the OrVil anvil, in the form of a tilted head, allow passage through the oropharynx and upper esophageal sphincter permitting use of a 25-mm anvil. By attaching the anvil to an oropharyngeal tube, transoral guidance is easily performed. Once the anvil is seated and connected to the EEA stapler, the head unfolds and resumes its functional state. Staple height is crucial, and a 4.8-mm staple was used exclusively in our cases. The stapling device s technical challenges require practice. Having an experienced person available in the operating room for the initial cases to help with the Table 4. Strictures Age a Anastomosis Location Comorbidities Post-op Complications Time ncrt Dilations Current Status (cm) (mon) (No.) Obesity A-Fib, hypotension 3 Both 2 No further dysphagia Obesity, coronary artery disease, left ventricular outflow obstruction Obesity, coronary artery disease, hypertension Hypotension, reintubation for 3days, myocardial infarction Ileus, chronic pain, intermittent n/v 2 Neither 10 Intermittent self - dilations 6 Neither 1 No further dysphagia Hypertension, hypothyroidism Supraventricular tachycardia 2 Both 6 No further dysphagia Alcohol and tobacco A-Fib, hypotension 3 Both 5 No further dysphagia Obesity, hypertension, hyperthyroidism Post operative bleeding requiring blood transfusion 6 Both 3 No further dysphagia Hypertension None 2 Both 6 No further dysphagia Died of metastatic disease a All patients were male. A-Fib atrial fibrillation; ncrt neoadjuvant chemoradiotherapy; n/v nausea/vomiting.

7 1868 JAROSZEWSKI ET AL Ann Thorac Surg EEA FOR MINIMALLY INVASIVE ESOPHAGECTOMY 2011;92: Table 5. Risk Factors for Leak and Stricture Development Factor Leak p Value a Stricture Adjuvant chemoradiotherapy Atrial fibrillation Body mass index 30 kg/m Cardiovascular disease Intra-op radiotherapy Neoadjuvant chemoradiotherapy Post-op hypotension a None were found to be statistically significant. nuances is critical. Operative time of our first 14 cases compared with our last 14 decreased significantly (excluding IORT cases) from a mean of 378 vs 300 minutes (p 0.004). There were several major areas of difficulty: First, the anvil could not be passed through smaller patients esophagus. In these patients, the balloon of the endotracheal tube may be deflated or patient s head repositioned with neck extended, achieving passage into the stump. In 1 patient we were required to use the 21-mm anvil due to a small esophageal diameter and inability to pass the 25-mm anvil, despite these suggestions. Second, once the tip of the orogastric is against the proximal esophageal stump, it must be held steady, using a small amount of pressure, so that a single small opening in the esophagus stump can be accomplished for pull-through of the orogastric tube and positioning of the anvil tip. Third, docking the anvil and EEA stapler is technically challenging, especially higher up in the thoracic inlet where the angle creates difficult visualization and support of the anvil. We use the Prestige grasper (Aesculap Inc, Center Valley, PA), positioned from the most posterior 5-mm port, to grasp the anvil. The pin of the EEA staler must be guided at an exact angle because the metallic tips of the anvil can bend easily, causing it not to lock when docked (Fig 5). If this occurs, the anvil will have to be removed and replaced with another. This is not easily done with endoscopy; therefore, to pull it backward out of the esophagus into the oropharynx, we now tie a long suture circumferentially to include the anvil (Fig 2A). This suture is kept out of the mouth by anesthesia so if necessary can permit gentle retrieval of the anvil back up into the pharynx if the prongs of the anvil are inadvertently bent and the stapler will not dock. Once the stapler is fired, the sutures will be severed and can be pulled directly out through the mouth. Anastomotic leaks are a concern with all types of esophagogastric anastomoses; however, our intrathoracic anastomotic leak rate was low (9.8%) and did not differ substantially from other types of anastomoses [17, 18]. Factors such as large body habitus, cardiovascular disease, and postoperative hypotension are suspected risk factors. Although we did not find any statistical correlation between previously defined risk factors, the small number of patient in our series may have masked statistical significance. Anastomotic strictures can have a significant affect on quality of life [19]. Dysphagia can have multiple causes after esophagectomy, including functional, neurologic, inflammatory, or fibrotic anastomotic. Stricture rate using different anastomotic techniques can be difficult to determine because there is no standardized reporting system for strictures. The presence of dysphagia alone does not define stricture. Anastomotic narrowing detected on barium swallow is commonly used, but not all patients with narrowing have dysphagia. For our purposes, dysphagia was evaluated with passage of an adult flexible endoscope. We define stricture not only as inability to pass the 9-mm endoscope beyond the anastomosis but also as any degree of anastomotic narrowing associated with dysphagia. The absence of a consensus definition and diagnostic criteria for anastomotic stricture limits comparisons between trials, with reported stricture rates ranging from 8.7% to 55% [9, 20]. Bizekis and colleagues [14] reported higher stricture rates with the 25-mm anvil; however, two recent studies comparing anvil sizes from 25 to 33 mm found no correlation between anvil size and dysphagia or stricture [6, 9]. Detailed analysis shows that significant stricture, defined as inability to pass the esophagoscopy scope thru the stricture, occurred in only 2 of our patients (3.9%), without including recurrent tumor. According to these criteria, 5 of our strictures were minimal ( 12 mm) and would not have been reported by some authors [21]. Stricture formation likely depends on patient characteristics and operative factors such as blood supply to the conduit and proximal esophageal stump, tension at the anastomosis, and immediate postoperative conditions, including hypotension, anemia, and the use of vasopressors. Postoperative atrial fibrillation and hypotension occurred in 16 patients (31.4%), and strictures subsequently developed in 3 of these patients. No statistically significant factors were associated with stricture development in our small series, including ncrt (Table 5). Limiting dissection above/around the proximal esophageal stump, preserving the pleura proximally, and minimizing tension with a sufficient length of gastric conduit may aid in preserving the anastomotic blood supply. In conclusion, transoral placement of the OrVil anvil facilitated the esophagogastric anastomosis in our transthoracic MIE cases with acceptable stricture and leak rates [22 24]. Our patients in this early experience achieved an acceptable morbidity and mortality rate with this MIE technique that compares favorably with comparative experiences of open esophageal resections; however, in the absence of randomized trials, suggested advantages cannot be made [22 24]. A standardized definition of stricture will better document occurrence in patients for evaluation of the optimal surgical technique for esophagogastric anastomosis. We thank Brie N. Noble for her help with statistical analysis.

8 Ann Thorac Surg JAROSZEWSKI ET AL 2011;92: EEA FOR MINIMALLY INVASIVE ESOPHAGECTOMY 1869 References 1. Luketich JD, Alvelo-Rivera M, Buenaventura PO, et al. Minimally invasive esophagectomy: outcomes in 222 patients. Ann Surg 2003;238: Nguyen NT, Follette DM, Lemoine PH, et al. Minimally invasive Ivor Lewis esophagectomy. Ann Thorac Surg 2001; 72: Braghetto I, Csendes A, Cardemil G, et al. Open transthoracic or transhiatal esophagectomy versus minimally invasive esophagectomy in terms of morbidity, mortality and survival. Surg Endosc 2006;20: Nguyen NT, Longoria M, Chang K, et al. Thoracolaparoscopic modification of the Ivor Lewis esophagogastrectomy. J Gastrointest Surg 2006;10: Nguyen NT, Hinojosa MW, Smith BR, et al. Minimally invasive esophagectomy: lessons learned from 104 operations. Ann Surg 2008;248: Campos GM, Jablons D, Brown LM, et al. A safe and reproducible anastomotic technique for minimally invasive Ivor Lewis oesophagectomy: the circular-stapled anastomosis with the trans-oral anvil. Eur J Cardiothorac Surg 2010; 37: Dapri G, Himpens J, Cadiere GB. Minimally invasive esophagectomy for cancer: laparoscopic transhiatal procedure or thoracoscopy in prone position followed by laparoscopy? Surg Endosc 2008;22: van Heijl M, Gooszen JA, Fockens P, et al. Risk factors for development of benign cervical strictures after esophagectomy. Ann Surg 2010;251: Blackmon SH, Correa AM, Wynn B, et al. Propensitymatched analysis of three techniques for intrathoracic esophagogastric anastomosis. Ann Thorac Surg 2007;83: Nguyen TN, Hinojosa MW, Smith BR, et al. Thoracoscopic construction of an intrathoracic esophagogastric anastomosis using a circular stapler: transoral placement of the anvil. Ann Thorac Surg 2008;86: Walther B, Johansson J, Johnsson F, et al. Cervical or thoracic anastomosis after esophageal resection and gastric tube reconstruction: a prospective randomized trial comparing sutured neck anastomosis with stapled intrathoracic anastomosis. Ann Surg 2003;238: Kim RH, Takabe K. Methods of esophagogastric anastomoses following esophagectomy for cancer: a systematic review. J Surg Oncol 2010;101: Laxa BU, Harold KL, Jaroszewski DE. Minimally invasive esophagectomy. Innovations 2009;4: Bizekis C, Kent MS, Luketich JD, et al. Initial experience with minimally invasive Ivor Lewis esophagectomy. Ann Thorac Surg 2006;82: Kyzer S, Gordon PH. Experience with the use of the circular stapler in rectal surgery. Dis Colon Rectum 1992;35: Wittgrove AC, Clark GW, Tremblay LJ. Laparoscopic gastric bypass, Roux-en-Y: preliminary report of five cases. Obes Surg 1994;4: Law S. Minimally invasive techniques for oesophageal cancer surgery. Best Pract Res Clin Gastroenterol 2006;20: Blewett CJ, Miller JD, Young JE, et al. Anastomotic leaks after esophagectomy for esophageal cancer: a comparison of thoracic and cervical anastomoses. Ann Thorac Cardiovasc Surg 2001;7: Rice TW. Anastomotic stricture complicating esophagectomy. Thorac Surg Clin 2006;16: Kondra J, Ong SR, Clifton J, et al. A change in clinical practice: a partially stapled cervical esophagogastric anastomosis reduces morbidity and improves functional outcome after esophagectomy for cancer. Dis Esophagus 2008;21: Williams VA, Watson TJ, Zhovtis S, et al. Endoscopic and symptomatic assessment of anastomotic strictures following esophagectomy and cervical esophagogastrostomy. Surg Endosc 2008;22: Nagpal K, Ahmed K, Vats A, et al. Is minimally invasive surgery beneficial in the management of esophageal cancer? A meta-analysis. Surg Endosc 2010;24: Biere SSAY, Cuesta MA, Ven Der Peet AL. Minimally invasive versus open esophagectomy for cancer: a systemic review and meta-analysis. Minerva Chir 2009;64: Pennathur A, Zhang J, Chen H, Luketick J. The best operation for esophageal cancer? Ann Thorac Surg 2010;89: