Transthoracic Esophagectom : A Safe Approach to Carcinoma of Je Esophagus

Transcription

1 Transthoracic Esophagectom : A Safe Approach to Carcinoma of Je Esophagus Douglas J. Mathisen, M.D., Hermes C. Grillo, M.D., Earle W. Wilkins, Jr., M.D., Ashby C. Moncure, M.D., and Alan D. Hilgenberg, M.D. ABSTRACT Transthoracic esophagogastrectomy is a safe operation. Mechanical staplers and a cervical anastomosis have been emphasized to avoid catastrophic consequences of anastomotic leaks in the chest. Transhiatal esophagectomy has been proposed to bring the anastomosis into the neck. It is meant to be a palliative procedure and consequently denies the patient the best chance for surgical cure. The emphasis should be on anastomotic technique and sound principles of surgical oncology. Since 198, we have performed 14 esophagectomies for carcinoma of the esophagus. We used a left thoracoabdominal incision for distal tumors (64) and the Ivor Lewis technique (4) for more proximal tumors. A two-layer inverting interrupted silk suture technique was used for all anastomoses. More than 9% of the procedures were performed by resident staff. The operative mortality was 2.9% (3 patients). There were no anastomotic leaks. Five patients required between one dilation and three dilations postoperatively. A positive smoking history was present in 83 patients and substantial alcohol use, in 33. Median estimated blood loss was 5 ml, and 6% of patients required no transfusions. Major complications included pneumonia (12 patients) and reexploration for bleeding (2). Minor complications included atelectasis (71 patients), atrial fibrillation (9), ventricular arrhythmias (9), urinary tract infection (31, and wound infection (2). Squamous cancer was present in 31 patients and adenocarcinoma, in 73. Positive lymph node metastases were present in 75%. Anastomotic recurrence was documented in 6 patients. Standard techniques of esophagogastrectomy and a two-layer anastomosis will give excellent results with low mortality and acceptable morbidity. In 1942, Churchill and Sweet [l] described a triple-layer technique of esophagogastrostomy and conventional en bloc resection of the cancer and adjacent lymph nodes. Five years later, Richard Sweet [2] published his initial experience with surgical management of carcinoma of the esophagus in 141 patients. Operating in an era without sophisticated postoperative monitoring devices, mechanical ventilation, or broad-spectrum antibiotics, his From the General Thoracic Surgical Unit, Massachusetts General Hospital, Boston, MA. Presented at the Twenty-third Annual Meeting of The Society of Thoracic Surgeons, Toronto, Ont, Canada, Sept 21-23, Address reprint requests to Dr. Mathisen, Warren 119, Massachusetts General Hospital, Boston, MA results were remarkable: an operative mortality of 15%, anastomotic leaks in 1.4% of patients, and overall 5-year survival of 11%. This served as a standard for many years. It was clear that surgical intervention alone was capable of only modest cure rates. Impressed by low rates of cure and even higher rates of morbidity and mortality, many surgeons began to view the role of surgery as purely palliative. Surgical techniques were developed to place the esophageal anastomosis in the neck to avoid the catastrophic consequences of intrathoracic anastomotic leaks [3]. These operations made no attempt at en bloc resection of the tumor and adjacent lymph nodes. Since much of the mortality and morbidity was attributed to the complications of anastomotic leaks, mechanically stapled anastomoses became increasingly popular in the hope of diminishing the leak rate [4-61. Esophageal cancer still presents a formidable problem for the thoracic surgeon. The fatalistic outlook of accepting palliation as the goal of operative intervention and selecting operations to lessen the consequences of frequent anastomotic leaks seems misguided. A multimethod approach that incorporates the best aspects of each discipline may improve the rate of cure for patients with esophageal cancer. Recent reports [7,8,8a] of treatment involving surgery, irradiation, and chemotherapy show promise. Surgery plays an integral part in these new approaches to esophageal cancer by removing the tumor and providing valuable staging information about lymph nodes and response of the tumor to prior treatment. The operations performed must be safe and must incorporate sound principles of surgical oncology to assess the impact of the combined treatment techniques. We believe that the standard left transthoracic or thoracoabdominal esophagectomy for distal esophageal tumors and Ivor Lewis s approach for more proximal tumors, both with a precise two-layer anastomosis, are optimal surgical approaches. Combining these surgical approaches with innovative strategies using radiation therapy, chemotherapy, and immunotherapy may offer the best chance for cure, can minimize morbidity and mortality, and allows evaluation of treatment methods. Material and Methods We examined the records of all patients undergoing esophagectomy in the General Thoracic Surgical Unit of the Massachusetts General Hospital from January 1, 198, to December 31, One hundred four esophagectomies were performed. This represents a consecutive series. Patients were considered for operation if their medical condition was deemed satisfactory and if 137 Ann Thorac Surg 45: , Feb Copyright 1988 by The Society of Thoracic Surgeons

2 138 The Annals of Thoracic Surgery Vol 45 No 2 February 1988 metastatic disease was not evident. All patients had a barium swallow and esophagoscopy with biopsy preoperatively. Since 1982, all patients have undergone computed axial tomography. Information is not available on the number of patients denied operation because of medical condition or metastatic disease. Surgical Technique Sweet [2] emphasized the details of technique and warned against factors predisposing to anastomotic leak. We think it is worth restating these principles, since little attention has been paid to them in recent reports. The focus has been on where the anastomosis is done rather than on how it is done. The lack of an esophageal serosal layer and the segmental blood supply of the esophagus make esophageal anastomosis more demanding than other intestinal anastomoses. Atraumatic handling of the tissues, preservation of the blood supply of both the esophagus and the stomach, avoidance of the use of crushing clamps, lack of tension on the anastomosis, use of fine interrupted sutures, cutting with a knife or other sharp instrument, and firm but gentle tying of sutures to avoid cutting tissues are all important details in the performance of an anastomosis. Few modifications have been made from the technique Churchill and Sweet [l] proposed 45 years ago. All operations are done with double-lumen endotracheal tubes to allow maximum exposure when performing the anastomosis. The patient is placed in true lateral thoracotomy position for a left thoracoabdominal esophagectomy. Exploration is accomplished through a small abdominal incision and if no or limited metastases are found, the incision is carried across the costal arch through the fifth or sixth interspace. The lateral thoracotomy position allows the anastomosis to be performed easily with the heart lying more medially than when the chest is elevated only 45 degrees. Pyloroplasty and the Kocher maneuver can be carried out easily with this exposure. The Kocher maneuver will add several centimeters of length and should be done whenever extra length is needed. The lateral position can also be helpful if the resection must be carried higher than anticipated. By extending the incision beyond the tip of the scapula, the fourth interspace can be entered if a lower interspace has been used or the fifth interspace incision can be lengthened to gain access to the apex of the chest if an anastomosis must be performed above the level of the aortic arch. Preservation of the blood supply is crucial when mobilizing the stomach and esophagus. The blood supply of the stomach will be from the right gastric and right gastroepiploic arteries. The esophagus should not be mobilized beyond a few centimeters above the proposed level of the anastomosis to avoid interference with the segmental blood supply. The stomach is divided with a TA-9 stapler. This suture line is then turned in with a layer of simple Lembert stitches. A circle approximately 2 cm in diameter is n / Fig 1. Surgical technique. (A) The serosa of the stomach has been scored, and the vessels have been ligated. The back row of sutures has been completed. (B) The button of stomach has been removed and the anterior wall of esophagus, opened. (C) The back row of the inner layer is completed, and the esophagus is transected. (D) The remainder of the inner layer is completed. A Connell stitch is used for closure of the final opening. A nasogastric tube is advanced across the anastomosis under direct vision before the inner layer is completed. (E) The outer layer is nearly finished. scored on a portion of the serosa of the stomach (Fig 1A). The circular defect in the stomach should be 2 cm away from the stapled edge of the stomach to avoid compromise of the blood supply. Individual vessels are identified and ligated with fine silk sutures (see Fig 1A). This minimizes bleeding while the anastomosis is performed, and allows for precise placement of sutures.

3 139 Mathisen et al: Transthoracic Esophagectomy for Carcinoma of Esophagus Interrupted horizontal mattress sutures of fine suture material (we use 4- silk) are used to construct the back row of the anastomosis (see Fig 1A). Corner stitches are placed first, and the remaining sutures are evenly spaced between them. The sutures on the stomach involve the seromuscular layers and those on the esophagus, the longitudinal and circular muscle layers. The esophageal sutures should be deep enough to include both the longitudinal and circular muscles of the esophagus. The sutures should not be tied too tightly to avoid necrosis or cutting through the muscle. The esophagus is opened sharply from one comer stitch to the other. The circular button of stomach is removed (see Fig 1B). The inner layer is completed with simple sutures including just the mucosa of the esophagus and the full thickness of the stomach (see Fig 1C). The knots are on the inside, thereby allowing inversion or turning-in of the mucosa of both the esophagus and stomach. This is accomplished for the entire circumference of the anastomosis (see Fig 1D). A nasogastric tube is passed into the stomach under direct vision before a single Connell stitch is placed for closure of the final opening. Healing of the inverted mucosa is an important feature in preventing leakage, and the location of the knots on the luminal side minimizes foreign body reaction within the actual tissues of the anastomosis. The outer row is completed using horizontal mattress sutures as described for the back row of the outer layer (see Fig 1E). The omentum mobilized with the stomach is placed over the anastomosis anteriorly to provide an additional layer of coverage. The posterior part of the anastomosis lies between the esophagus and the more proximal stomach. A few sutures are placed between the stomach and the mediastinal pleura to avoid tension on the anastomosis when the patient is upright, particularly if the stomach is full. Sutures are also placed between the stomach and the diaphragmatic hiatus to prevent herniation of abdominal contents. Pyloromyotomy or pyloroplasty is usually performed. Viability of tissues on each edge of the anastomosis is best maintained if trauma is avoided. The edges are never crushed with clamps and, indeed, are handled with forceps as little as possible. Once the first stitch is placed and tied, traction on it permits placement of the next without the need for instrumental grasping of the mucosa. The sutures are tied by positioning the index finger cephalad to the anastomosis, lifting the stomach to the esophagus, and avoiding pulling down on the fixed and more fragile esophagus. This is especially important for the outer layer of the anastomosis because the esophagus lacks a peritoneal surface. A nasogastric tube passed through the anastomosis for a short time avoids distraction at the suture line by a distended stomach. Gentle, periodic irrigation of the tube ensures its patency. Temporary gastric decompression more than compensates for any potentially deleterious effect of an intraluminal foreign body lying against the suture line for a short period. Results There were 78 men and 26 women. The median age was 6.4 years and 64.7 years, respectively. Eighty percent of the patients smoked at least one pack of cigarettes per day, and 32% consumed at least one alcoholic beverage per day. In 7% of the patients, pulmonary function tests were performed preoperatively; the median forced expiratory volume in one second was 2.6 liters (range, 1.1 to 4.4 L/sec). Average preoperative weight loss was 6.6 kg. Adenocarcinoma was present in 73 patients and squamous carcinoma, in 31. Seventy-two carcinomas were considered to be in the distal esophagus (3 to 4 cm from the incisors) and 32, in the middle or upper esophagus. A left thoracoabdominal incision through the fifth or sixth interspace was used in 64 patients and a laparotomy and right thoracotomy (Ivor Lewis technique), in 4. Median estimated blood loss was 5 ml (range, 15 to 3, ml). Sixty percent of the patients required no transfusions. Median postoperative stay in the hospital was 13.7 days (range, 8 to 13 days). Nodal metastases were present in 78.6% of the patients with adenocarcinoma and 48.4% of those with squamous cancers. Transmural invasion was present in 77.1% of the patients with adenocarcinoma and 71% of those with squamous carcinomas. Postoperative pathological staging of the adenocarcinomas was as follows: Stage I, 9; Stage 11, 6; and Stage III,58. Of the squamous cancers, 6 were stage I, 1 was Stage 11, and 24 were Stage 111. Staging of the squamous carcinomas might have been influenced by preoperative chemotherapy. Eighteen patients with squamous carcinoma received preoperative chemotherapy, 8 received postoperative radiation therapy, and 5 received neither preoperative nor postoperative treatment of any kind. Forty-four patients with adenocarcinoma received postoperative chemotherapy or radiation therapy. Mortality There were 3 postoperative deaths (2.9%). Two deaths were attributable to pneumonia and respiratory failure (patients aged 59 and 73 years). Both patients smoked at least one pack of cigarettes per day preoperatively. The third death was also due to pneumonia and respiratory failure. This 76-year-old patient was operated on emergently for massive gastrointestinal bleeding. Morbidity Major complications occurred in 16 patients (15%) (Table 1). PULMONARY. Radiographic evidence of pneumonia was present in 19 patients, all of whom had continuation of perioperative antibiotics for at least 1 days. The presence of pneumonia did not prolong hospitalization for 7 of these 19 patients. Six patients required a tracheostomy for prolonged mechanical ventilation (5) or pulmonary toilet (1). Three of these patients died. Radiographic evidence of atelectasis was seen in 71 patients. Pleural effusions following chest tube removal were

4 14 The Annals of Thoracic Surgery Vol 45 No 2 February 1988 Table 1. Major and Minor Complications in 14 Patients Complication Major Minor Pulmonary Pneumonia Atelectasis Pleural effusion Cardiovascular Myocardial infarction Congestive heart failure Premature ventricular contractions Atrial fibrillation Pulmonary embolus Axillary vein thrombosis Urinary tract Infection Retention Renal failure Gastric Outlet obstruction Bleeding Postoperative hemorrhage Wound infection Vocal cord paralysis present in 11 patients, but necessitated thoracentesis in only 1 patient. CARDIOVASCULAR. Myocardial infarction was documented in 2 patients. Congestive heart failure necessitating diuresis and digitalis occurred in 3 patients. Ventricular arrhythmias (one ventricular tachycardia) requiring lidocaine hydrochloride were present in 9 patients. Digitalization was necessary in 9 patients for atrial fibrillation. There was one documented pulmonary embolism and one episode of axillary vein thrombosis, both necessitating anticoagulation. RENAL. Renal failure occurred in 1 patient who ultimately died of multisystem failure. Three patients had a urinary tract infection, which was treated with antibiotic therapy. A transurethral resection for benign prostatic hypertrophy was performed in 2 patients. GASTRIC. Gastric outlet obstruction necessitating prolonged nasogastric suction occurred in 3 patients, 1 of whom required reoperation for revision of the diaphragmatic hiatus. One patient had gastrointestinal hemorrhage secondary to gastritis. ANASTOMOTIC STRICTURE. Dilation was necessary in 5 patients for anastomotic stricture three to six weeks postoperatively. One dilation to three dilations were required for successful resolution of dysphagia. Delayed anastomotic stricture was not apparent in this group of patients. ANASTOMOTIC LEAKS. All patients had postoperative barium swallows. There were no anastomotic leaks, even of localized type. ANASTOMOTIC RECURRENCE. Six anastomotic recurrences were proven. SPLENECTOMY. Incidental splenectomy was performed 3 2 i '"1 I... 1: / l...:... S uamous 33.2 %... Ademcarcinoma 27% OA " 1 " 2 " 3 " 4 ' 5" 6 ' MONTHS Fig 2. Actuarial 5-year survival of patients with adenocarcinoma and squamous carcinoma of the esophagus. in 5 patients. One of them required splenectomy because of a previous resection of an aneurysm of the splenic artery. POSTOPERATIVE HEMORRHAGE. Two patients were reoperated on for postoperative hemorrhage. A short gastric artery bled in 1 patient and an esophageal artery in the other. WOUND INFECTION. Minor wound infections occurred in 2 patients. Both were successfully managed with antibiotics and drainage of a small area of the wound. VOCAL CORD PARALYSIS. Vocal cord paralysis was documented postoperatively in 2 patients. Both of them had an extensive midesophageal tumor requiring an extended resection. The left recurrent nerve was included in the resection at the time of operation. Survival Actuarial 5-year survival for the 73 patients with adenocarcinoma was 8% and for the 31 patients with squamous carcinoma, 33.2% (Fig 2). Few conclusions can be drawn from the survival figures because uniform treatment was not given to either the patients with adenocarcinoma or those with squamous carcinoma. Since 1982, however, all patients with squamous carcinoma of the esophagus have received chemotherapy preoperatively. The results of this trial are preliminary, but they have been encouraging. The actuarial survival at 42 months of all patients undergoing resection is 54% [8aI. Comment Intrathoracic anastomotic leaks have long been recognized as potentially fatal problems [ Conversely, an anastomotic leak in the cervical area is much easier to treat and will usually heal with drainage alone. This seems to be the primary motivation for the development of techniques such as transhiatal esophagectomy [3]. Most reports concerning this technique focus on how the procedure is performed; rarely do they mention how the anastomosis is fashioned. Sweet [2] demonstrated 4

5 ~~~~ ~~~~~ 141 Mathisen et al: Transthoracic Esophagectomy for Carcinoma of Esophagus Table 2. Comparison of Techniques of Anastomosis No. of Leaks Stricture Technique Reference Pahents (%) Stapled 151, [4], [6], One-layer [9], [12], [3], Two-layer [13], [14], Present report, years ago that an intrathoracic esophageal anastomosis can be performed safely with a leak rate of less than 2%. The principles he emphasized to avoid anastomotic leaks are fundamental to all areas of surgery. The two-layer interrupted anastomosis currently performed at this institution has not changed appreciably from the one proposed by Churchill and Sweet [l] 45 years ago. Churchill and Sweet were master surgeons. However, their technique can be taught to and performed successfully by other surgeons; in this series, more than 9% of the operations were done by the resident staff under the guidance of the attending staff. The technique can be passed from one generation of surgeons to another; our results are not different from those of Dr. Sweet. One should not minimize the importance of this technique for resident training. Situations will arise when mechanical staplers are unavailable, will malfunction, or cannot be applied, for example, in a high cervical anastomosis. We are aware of no randomized studies that compare stapled, single-layer, and two-layer anastomoses. Many series fail even to mention the technique used or mention it only in passing. Reports describing the technique of anastomosis and the incidence of complications [3-6, 9, 12-14] are listed in Table 2. Absolute conclusions cannot be made by this kind of comparison, but the twolayer anastomosis seems to be associated with fewer anastomotic leaks, a finding not unique to the experience at our institution. Stapled anastomoses may offer a reliable alternative, especially to those untrained in the two-layer technique or to the occasional esophageal surgeon. Some series [5, 61 have reported a high rate of leaks or anastomotic strictures, and one must presume a learning curve as with other techniques. In 198, Earlam and Cunha-Melo [15] concluded after their review of 83,783 patients with esophageal cancer treated by surgical intervention that "esophagectomy was associated with the highest operative mortality of any commonly performed operation." Many of these deaths are attributed to anastomotic leaks; mortality of 5% or greater is common [ A safe anastomotic technique could certainly reduce the number of deaths, and proper patient selection could lower it even further. Other reports [13, 141 concur with our operative mortal- ity of less than 5%. This is a goal that should be attainable. Standard thoracoabdominal incisions or combined abdominal and right thoracotomy incisions have been implicated in being associated with prohibitive perioperative morbidity. Undoubtedly, operations of this magnitude place a great deal of stress on patients, especially those who have been heavy smokers or who are malnourished as is often the case for patients with esophageal cancer. Every effort should be made to have patients in optimal condition for such operations. We know of no randomized studies comparing transthoracic esophagectomy with transhiatal esophagectomy. Attempts have been made to compare groups [16, 171, but a convincing argument favoring one approach over another has not been made. Pulmonary complications have been the most frequent source of morbidity in patients undergoing esophagectomy [ll]. This was true in our series as well. However, of the 19 patients with pneumonia, only 12 had a prolonged hospitalization; in the remaining 7 patients, the pneumonia was apparent only radiographically and did not alter the hospital course. A 12% incidence of pneumonia is higher than we would like, and aggressive measures should be taken to lower it. Our 3 postoperative deaths were all direct results of respiratory failure and pulmonary sepsis. Acquired immunodeficiency syndrome and hepatitis bring new relevance to the need for hemostasis. Transthoracic esophagectomy allows careful hemostasis. Our median estimated blood loss was 5 ml, which is lower than that in recent reports [3, 171 for transhiatal esophagectomy. Forty-two patients required blood transfusions while in the hospital, and 14 of them had received preoperative chemotherapy. Use of a transthoracic esophagectomy is no guarantee that major bleeding will not occur; 2 of our patients required reoperation for postoperative hemorrhage. Incidental splenectomy can be a source of intraoperative blood loss, and should be avoided when possible. Vocal cord paresis or paralysis has been a common complication of the transhiatal technique [3]. Avoidance of retraction injuries to the recurrent nerve in the tracheoesophageal groove has lowered the rate of this

6 142 The Annals of Thoracic Surgery Vol 45 No 2 February 1988 complication in recent reports. We had only 2 patients with postoperative vocal cord paralysis. In both, the left recurrent nerve was removed because of an extensive middle-third tumor. The extent of the tumor necessitated the sacrifice of this nerve. Chylothorax and tracheal tears have been reported in most series of transhiatal esophagectomies [3]. In a population of malnourished patients undergoing extensive operations, chylothorax can be a serious complication. We did not encounter a single instance of chylothorax or tracheal injury in our series. Transthoracic esophagectomy allows precise dissection and direct ligation of tissues in areas where the thoracic duct is at risk. Anastomotic stricture was encountered in 5 patients. All were symptomatic within three to six weeks after operation. One dilation to three dilations were necessary to achieve resolution of dysphagia. An interrupted technique should minimize this complication. Absorbable suture material might also diminish the incidence. The excised gastric button must, of course, be of adequate size. The other complications reported in our series are not uncommon to elderly, malnourished patients undergoing a major operation and are not unique to transthoracic esophagectomy (see Table 1). Twenty-five major complications occurred in 16 patients, a major complication rate of 15.4%. There were six documented anastomotic recurrences in our series. This problem should be avoidable. A supraaortic anastomosis can be performed through a left thoracoabdominal incision. The esophagus above the tumor can be mobilized above and beneath the aortic arch and an anastomosis performed in the apex of the chest. Some [3] have argued that the transhiatal technique is superior on the grounds that it provides a greater margin. This might be true for middle-third tumors, but even for them, a cervical anastomosis is possible with transthoracic techniques by adding an incision in the neck to perform the anastomosis at that level. Tumors of the gastroesophageal junction, however, pose a more difficult problem for surgeons wanting to use transhiatal techniques, since distal margins must be compromised to allow the stomach to reach the neck. Wider resection for tumors at this level is possible by the transthoracic route. Until recently, surgical cure of esophageal cancer was uncommon. This led to the view that surgical intervention was principally palliative. Recent reports [7, 8, 8a] of multimethod treatments of esophageal cancer have given encouraging results, especially for squamous cancer. Our 33% 5-year actuarial survival of patients with squamous carcinoma represents a major improvement over past results (see Fig 2). Much of this is attributable to a regimen of preoperative chemotherapy, and current survival at 42 months is Progress in the treatment of adenocarcinoma has been much less impressive, but a few early reports may be encouraging [MI. As new treatments are being evaluated, we must return to sound principles of surgical oncology in managing esophageal cancer. Wide excision and en bloc resection of draining lymph nodes should offer the best surgical option. This approach also allows assessment of the response of the tumor to treatment and provides accurate staging information so important when evaluating new treatment protocols. Precise two-layer esophageal anastomosis and attention to those factors that predispose to anastomotic leak allow transthoracic esophagectomy to be performed safely. It can be done with low operative mortality and acceptable morbidity, and can allow precise evaluation as new treatment techniques are developed. Transhiatal esophagectomy is an alternative approach for patients with cervical cancers. It can also be used when the stated goal of the surgeon and the patient is to achieve palliation. References 1. Churchill ED, Sweet RH: Transthoracic resection of tumors of the stomach and esophagus. Ann Surg , Sweet RH: Carcinoma of the esophagus and cardiac end of the stomach: immediate and late results of treatment by resection of primary esophagogastric anastomosis. JAMA 135:485, Omnger MB, Orringer JS: Esophagectomy without thoracotomy: a dangerous operation? J Thorac Cardiovasc Surg 85:72, Donnelly RJ, Sastry MR, Wright CD: Oesophagogastrectomy using the end-to-end anastomotic stapler: results of the first 1 patients. Thorax 4:958, Hopkins RA, Alexander JC, Postlethwait RW: Stapled esophagogastric anastomosis. Am J Surg 147:283, Blum M, Kessler B, Bunte H: The influence of stapled anastomosis compared to handmade anastomosis on complication rate and mortality after resection of esophagus carcinoma. In Siewart IR, Holscher AH (eds): Disease of the Esophagus. Munich: Springer-Verlag, Popp MB, Hawley D, Reising J, et al: Improved survival in squamous esophagus cancer: preoperative chemotherapy and irradiation. Arch Surg 121:133, Leichman L, Steiger ZM, Seydel HG, et al: Preoperative chemotherapy and radiation therapy in patients with cancer of the esophagus: a potentially curative approach. J Clin Oncol 2:75, a. Hilgenberg AD, Carey RW, Wilkins EW, et al: Preoperative chemotherapy, surgical resection, and selective postoperative therapy for squamous cell carcinoma of the esophagus. Ann Thorac Surg (in press) 9. Belsey R, Hiebert CA: An exclusive right thoracic approach for cancer of the middle third of the esophagus. Ann Thorac Surg 18:1, Giuli R, Gignoux M: Treatment of carcinoma of the esophagus: retrospective study of 2,4 patients. Ann Surg 192:44, Postlethwait RW: Complications and death after operations for esophageal carcinoma. J Thorac Cardiovasc Surg 85:827, Skinner DB: En bloc resection for neoplasms of the esophagus and cardia. J Thorac Cardiovasc Surg 85:59, Akiyama H: Esophageal anastomosis. Arch Surg 17:512, Ellis FH, Gibb SP, Watkins E Jr: Esophagogastrectomy: a safe, widely applicable, and expeditious form of palliation

7 143 Mathisen et al: Transthoracic Esophagectomy for Carcinoma of Esophagus for patients with carcinoma of the esophagus and cardia. Ann Surg 198:531, Earlam R, Cunha-Melo JR: Oesophageal squamous cell carcinoma: I. A critical review of surgery. Br J Surg 67:381, Shahian DM, Neptune WB, Ellis FH Jr, Watkins E Jr: Transthoracic versus extrathoracic esophagectomy: mortality, morbidity, and long-term survival. Ann Thorac Surg 41: 237, Goldfaden D, Orringer MB, Appleman HD, Kalish R: Adenocarcinoma of the distal esophagus and gastric cardia: comparison of results of transhiatal esophagectomy and thoracoabdominal esophagogastrectomy. In Siewart IR, Holscher AH (eds): Disease of the Esophagus. Munich: Springer-Verlag, Forastiere AA, Gennis M, Orringer MB, Agha FP: Cisplatin, vinblastine, and mitoguazone chemotherapy for epidermoid and adenocarcinoma of the esophagus. J Clin Oncol 5:1143, 1987 Discussion DR. PETER c. PAIROLERO (Rochester, MN): I would like to congratulate the authors for their concise presentation and excellent clinical results. Clearly these authors have substantiated that transthoracic esophagogastrectomy is a safe operation. Unfortunately, despite modern surgical techniques and contemporary irradiation and chemotherapy, treatment of this cancer remains a constant challenge. Undoubtedly these poor long-term results reflect a delay in operation and diagnosis. Dissatisfaction with survival has produced variation in operative technique that has ranged from extended operation, as proposed by Skinner, to transhiatal esophagectomy, as championed by Orringer. As the Massachusetts General Hospital and other groups have clearly demonstrated, it is now clear that, in the hands of experienced surgeons, esophageal operations by any route can be performed with low operative mortality and low morbidity. Moreover, all of these operations equally relieve dysphagia, all have minimal associated gastroesophageal reflux, and all allow hospital discharge within several weeks. Thus, no longer should it be an issue whether the operation is best performed through the left chest, the right chest, or transhiatally. It is also time to stop evaluating the merits of these operations in terms of how extensive a resection is performed. How many lymph nodes or how many centimeters above the cancer are included in the resected specimen is not important, as the timehonored Halstedian principles of cancer treatment have not stood the scrutiny of time. Lymph node metastases are synonymous with systemic disease, and operation for cancer of the esophagus is local therapy. The debate about which operation is better should be laid to rest once and for all as we try to achieve better control of this malignancy. Clearly, early diagnosis is the key to improved survival. What then can we as surgeons do? Technically probably not much more, but as we have heard in this presentation and will hear later in this meeting, multimodality therapy, especially preoperative chemotherapy, appears promising. Randomized studies using chemotherapy, surgery, and radiation therapy need to be done. To make a meaningful comparison, we also need to have a common terminology that we can use in communication. Cancer staging is available for cancer of the esophagus but is infrequently reported in most recent reports. Unquestionably, esophageal cancer is stage dependent. Postsurgical stage I cancer patients have a five-year survival of 85%, stage 11, 34%, and stage 111, 15%. Although the Massachusetts General Hospital group did look at staging for these tumors, they did not report survival in terms of staging. Why not? Until staging is uniformly reported, meaningful evaluation of therapy can never be made. Again, I would like to thank the authors for a timely presentation and also the Society for the privilege of allowing me to discuss this paper. DR. EARLE w. WILKINS (Boston, MA): It is unusual for a secondary author to discuss a paper, but I do rise to reinforce an historical comment to which Dr. Mathisen referred. Somewhere along the line the concept or philosophy has arisen that a statistic less than 1% for leak rates is acceptable if you are doing esophageal surgery. We believe that is wrong. Dr. Edward Churchill, who was chief of surgery when I was a resident, wrote in his paper in 1942 [Churchill ED, Sweet RH: Transthoracic resection of tumors of the esophagus and stomach. Ann Surg , as follows: "Those mucous membrane sutures are placed with the exactitude and with the degree of tension that we would use in the fine plastic procedure on the lip. There must be no tension and there must be accuracy." Dr. Richard Sweet, who was my personal mentor for some 1 years, commented: "We have had no strictures. Just why I do not know but it remains a fact. Perhaps we have paid unusual attention to the detail of the anastomosis." And it is for that reason that I comment, because in an era when perhaps adjuvant therapy makes us look forward to better results, all the more attention must be paid to the anastomosis. As one of our surgeons at the Massachusetts General Hospital used to tell us, "you must do it right." DR. MATHISEN: I certainly want to thank Dr. Pairolero for his comments and also Dr. Wilkins. I think we all recognize the contributions the Mayo Clinic has made to general thoracic surgery and esophageal surgery in particular. We certainly would agree with everything he has said and that the key to even better survival is early diagnosis. The reason we did not report our results in terms of staging is that so many different treatments were given to these patients. The adenocarcinomas had a high number of patients who received postoperative chemotherapy and radiation therapy. As I mentioned, 18 of the patients with squamous cell carcinoma had received preoperative chemotherapy; some patients received postoperative radiation therapy, and therefore we did not stress the survival in this particular paper. The emphasis really was on the technique, the safety of transthoracic esophagogastrectomy, rather than an ultimate survivorship in this group of patients. We also would not argue with the premise that, when palliation is the goal, any safe technique for esophagogastrectomy is applicable. We think the transthoracic esophagectomy is as safe if not safer than any other reported technique. However, as we are now in an era when at least it seems as if there is promise on the horizon for new therapies incorporating chemotherapy, irradiation, or immunotherapy, the transthoracic techniques may offer important benefits in terms of staging, and only with this information will we be able to make sense of the new protocols that are developed. At the same time, we must balance more radical operations against the risks of these operations as we evaluate these treatment protocols. I would like to thank the Society for the privilege of presenting this paper.