Transhiatal versus transthoracic esophagectomy for esophageal cancer

Transhiatal Versus Transthoracic Esophagectomy for Adenocarcinoma Distal Esophagus and Cardia

of the

Steven P. Stark, MD, Kansas City, Missouri, Michael S. Romberg, MD, George E. Pierce, MD, Ado S. Hermreck, MD, William FL Jewell, MD, Jon F. Moran, MD, Kansas City, Kansas, George Cherian, MD, Roman0 Delcore, MD, Kansas City, Missouri, James H. Thomas, MD. Kansas City, Kansas

BACKGROUND: Transhiatal esophagectomy is a popular method of resection because of its reported lower morbidity and mortality and similar survival rates compared to transthoracic esophagectomy. A review of recent experience with these two procedures for resection of distal esophageal and cardia adenocarcinoma is reported. METHODS: From 1988 to 1994,48 patients with adenocarcinoma of the distal esophagus and gastric cardia were resected with intent to cure, 32 by transhiatal esophagectomy (group I) and 16 by transthoracic esophagectomy (group II). The two groups were comparable in terms of patient demographics, preoperative risk factors, tumor stage, tumor differentiation, and involvement of resection margins (all not significant [NS]). RESULTS: There was no significant difference in median intensive care unit stay, median hospital stay, incidence of postoperative anastomotic leak, and stricture. Respiratory complications were higher in group I (41% versus 6%, P = 0.01). Hospital mortality was not significantly different for the two groups (group I 3.1% versus group II 0%, NS). Actuarial 5-year survival rates (KaplanMeier) were 12% for group I and 39% for group II U’W. CONCLUSIONS: These results suggest that when compared with transhiatal esophagectomy, the transthoracic approach is at least as safe, has comparable complication and survival rates, and remains an acceptable procedure for resection of adenocarcinomas of the distal esophagus and gastric cardia. 0 7996 by Excerpfa Medica, Inc. Am J Surg. 1996;172:478-482.

From the Department of Veterans Affairs Medical Center (SPS, GC), Kansas City, Missouri; and the Department of Surgery, the University of Kansas Medical Center (MSR, GEP, ASH, WRJ, JFM, RD, JHT), Kansas City, Kansas. Requests for reprints should be addressed to Steven P. Stark, MD, Department of Surgery, Department of Veterans Affairs Medical Center, 4801 E. Linwood Blvd., Kansas City, Missouri 64128. Presented at the 48th Annual Meeting of the Southwestern Surgical Congress, Scottsdale, Arizona, April 28-May 1, 1996.

T

he incidence of adenocarcinoma of the esophagus and gastric cardia has risen recently at an alarming rate to become one of the 15 most common malignancies among white men in this country.‘*Z This trend is in contrast to the incidence of adenocarcinoma of the gastric body and antrum, which has declined over the past several decades. Some authors propose that adenocarcinoma of the esophagus and of the gastric cardia represent the same disease, with most tumors at both locations arising in Barrett’s esophagus.3*4 Despite a low cure rate, surgical resection allows for effective palliation, immediate restoration of swallowing, and a reasonable quality of life, and remains standard treatment for adenocarcinoma of the esophagus and gastric canlia. Different technical approaches to esophagectomy for cancer are available. Prior to the 1980s the most popular approach was esophagogastrectomy via combined laparotomy and right thoracotomy as originally described by Lewis.’ Recently, more esophageal resections in this country are performed by the transhiatal technique as described by Turner6 and popularized by Orringer.‘*’ Contemporary preference for the transhiatal approach relates mainly to the lower morbidity reported with this procedure which avoids a thoracotomy. Several studies have shown a lower incidence of pulmonary complications with tran.shiatal esophagectomy, as well as lower mortality rates for cervical versus thoracic anastomotic leaks. s-l* Some authors argue, however, that the transhiatal technique has no benefit over transthoracic esophagectomy.‘3-‘7 The purpose of this study was to compare the results of transhiatal versus transthoracic esophagectomy specifically for the treatment of adenocarcinoma of the distal esophagus and gastric cardia.

PATIENTS

AND

METHODS

The records of all patients with adenocarcinoma of the esophagus and gastric cardia diagnosed between 1988 and 1994 at the University of Kansas Medical Center and Kansas City Department of Veterans Affairs Medical Center were reviewed. In all, I47 patients were evaluated during the study period, of which 65 (41%) came to operation. Of these 65 patients, 3 who had a palliative bypass only, 113who had distal esophagectomy plus partial or total gastrectomy, and 4 who had total gastrectomy and colon interposition were excluded. of the remaining 48 patients, 32 underwent transhiatal resection (group I), and 16 underwent transthoracic resection (group II), all with curative intent. The method of resection was chosen for individual patients by each surgeon according to personal preference in a nonrandomized fashion.

1

478

@ 1996 by Excerpta All rights reserved.

Medica,

Inc.

0002-961 O/96/$1 PII SOOO2-9610(96)00224-3

J

5.00

~

TRANSHIATAL

TABLE Hospital

37 Data, Group

OR time Transfusion Pyloroplasty/myotomy Mean duration of intubation Median ICU stay Median hospital stay Anastomotic Leak Pulmonary complication Anastomotic stricture Postop reflux 30 day mortality

Morbidity,

VS TRANSTHORACIC

ESOPHAGECTOMY/STARK

ET AL

and Mortality

I

Group

5.91 + 1.8 h 14 (39%) 18 (63%) 4.75 i 9.5 days 6d 15d 5 (16%) 13 (41%) 9 (28%) 3 (9%) 1 (3%)

II

P

7.56 + 2.9 h 4 (25%) 12 (75%) 0.96 2 0.6 days 3d 14d 1 (6%) 1 (6%) 5 (31%) 2 (13%) 0 (0%)

P < 0.048 NS NS P = 0.048 NS NS NS P = 0.01 NS NS NS

RESULTS

Figure 1. Kaplan-Meier overall group I (transhiatal esophagectomy) esophagectomy).

survival curves and group

for patients II (transthoracic

in

Transhiatal esophagectomy consisted of celiotomy for gastric conduit preparation, blunt esophagectomy through the esophageal hiatus and thoracic inlet, orthotopic placement of the gastric conduit through the posterior mediastinum, and cervical anastomosis. Transthoracic (Ivor Lewis) esophagectomy consisted of celiotomy for gastric conduit preparation, right posterolateral thoracotomy for esophagectomy, and intrathoracic anastomosis in the apex of the right hemithorax. All cervical anastomoses and 12 of 16 intrathoracic anastomoses were handsewn; four intrathoracic anastomoses were stapled. All intended transhiatal esophagectomies were completed; there were no conversions to thoracotomy. Resected specimens were evaluated histologically for lymph node involvement and resection margin status. The current American Joint Committee on Cancer staging system was used.‘” Data collected included patient demographics and preoperative risk assessment, which consisted of coronary artery disease, tobacco use, alcohol use, gastroesophageal reflux disease, biopsy proven Barrett’s changes, and the results of pulmonary function tests (forced expiratory volume in 1 second). Operative data, operative and postoperative complications, length of intensive care unit stay, length of hospital stay, 30-day mortality, and current patient status were also evaluated. Follow-up was obtained by direct telephone contact with the patient or an immediate family member for all 48 patients. Statistical differences between the two approaches were determined by the chi-square test and Student’s t test. Actuarial survival curves were constructed using the Kaplan-Meier method. Survival statistics were compared using the log rank test. THE

AMERICAN

Group I consisted of 30 (94%) men and 2 (6%) women. Group II was composed of 16 men. All patients in both groups were white. The mean age of group I patients was 63 z 12 years (range 17 to 80). Group II patients had a mean age of 56 2 10 years (range 41 to 76; P = 0.067). The two groups were comparable with respect to preoperative risk factors including coronary artery disease, pulmonary function tests, pack-years of tobacco use, gastroesophageal reflux disease, and Barrett’s esophagus. Ethanol use approached a statistically significant difference between the groups with more users in group I (P = 0.08). The two groups were similar with respect to tumor stage, tumor differentiation, and incidence of regional lymph node metastasis. The mean number of lymph nodes removed was 10.7 t 10.1 for group I and 10.8 ? 8.1 for group II. An equal percentage of patients in both groups had microscopically positive resection margins. Operative time was significantly longer in group II (P 5 days), and respiratory distress requiring reintubation. Thirteen (41%) group I patients versus 1 (6%) group II patient had one or more of these respiratory complications (P = 0.01). There were two transient recurrent laryngeal nerve injuries in group I and one in group II. Cardiac complications occurred infrequently and consisted mainly of atria1 fibrillation (4 patients in group I and 2 patients in group II). There were no documented myocardial infarctions. One group I patient developed postoperative pancreatitis and 1 suffered a postoperative stroke. One operative death occurred in group I, for a 2% ( 1 of 48) overall 30-day mortality rate. This death was due to a pulmonary embolus (Table). JOURNAL

OF

SURGERY”

VOLUME

172

NOVEMBER

1996

479

RANSHIATAL

VS TRANSTHORACIC

ESOPHAGECTOMY/STARK

ET AL 1

Ten (3 1%) group I patients and 8 (50%) group II patients received postoperative adjuvant chemotherapy and/or radiation therapy (NS). No patient in either group received preoperative adjuvant therapy. Ten (31%) group I patients and 7 (44%) group II patients were alive at follow-up (NS). Seventeen (53%) group I patients and 12 (75%) group II patients had developed a recurrence at follow-up (NS). There was a trend for better overall 5-year actuarial survival in group II (39% versus 12%), but this did not reach statistical significance (Figure 1). The actuarial disease-free survival at 2 years was 32% for group I and 43% for group II; at 5 years this figure dropped to 10% and 19%, respectively, for group I and group II (Figure 2; NS). The overall 5-year survival was 67% for stage I patients, 24% for stage II patients, and 9% for stage III patients. The l-year survival for stage IV disease was 15%, and none of these patients survived more than 2 years.

COMMENTS Surgical resection and reconstruction in patients with esophageal carcinoma, although infrequently curative, achieves palliation by immediately restoring swallowing and gastrointestinal continuity. Several techniques are available for resecting cancers of the esophagus; these include the transhiatal or blunt, the combined abdominal and right thoracic, the total thoracic, the left thoracoabdominal, and the radical or en bloc resection. Recently, the trend has been to perform more esophagectomies by the transhiatal or blunt technique.‘9~20 This procedure theoretically minimizes respiratory complications, avoids devastating mediastinal infections associated with leaks of an intrathoracic anastomosis, reduces the incidence of significant postoperative gastroesophageal reflux, and avoids a painful thoracotomy incision. However, it could be argued that the transhiatal approach may not be an adequate cancer operation because it is difficult to resect as much periesophageal tissue and that it is a potentially dangerous approach if the tumor is adherent to mediastinal structures since part of the dissection is done blindly. Most available comparisons of the various techniques involve mixtures of histologic types of cancers. In this study, the two most widely used techniques of esophagectomy were compared specifically for resection of adenocarcinoma of the esophagus and gastric cardia. The two groups in this study were well-matched in terms of preoperative risk factors, tumor characteristics including stage, and use of adjuvant treatment. Recognizing the limitations of a retrospective, nonrandomized review, we found that outcome measures for the transhiatal and transthoracic techniques did not reach statistically significant differences except for respiratory complications. Interestingly, significantly more respiratory complications occurred in the transhiatal group-the method touted to reduce pulmonary problems and extend the indications for esophagectomy to patients at higher pulmonary risk.s-12 A report by Orringer et al9 who compared transhiatal to transthoracic esophagectomy in patients with adenocarcinoma of the distal esophagus and gastric cardia found a fourfold higher rate of respiratory complications for thoracotomy patients. Other authors”-i’ have found no significant difference in the incidence of pulmonary complications following transthoracic versus transhiatal esophagectomy. There is no apparent explanation for the higher frequency of respiratory 480

THE AMERICAN

JOURNAL

OF SURGERY@

VOLUME

172

Figure 2. Kaplan-Meier tumor-free group I (transhiatal esophagectomy) esophagectomy).

survival curves for patients and group II (transthoracic

in

problems in the transhiatal group in the present study. Although there was no significant difference between the two groups with regards to objective pulmonary function tests and smoking history, there nevertheless may have been important differences in the patients’ respiratory status or general medical condition that could not be measured or detected by a retrospective review. Surgeon bias in favor of one procedure over the other also may have played a role. Operative time was significantly longer for group II than for group I. This difference can be explained by time taken to reposition, re-prepare, and drape the pati.ent for the thoracotomy and the additional time to close a thoracotomy as opposed to a cervical incision. In this series, the longer operative time for group II, however, did not appear to have an adverse effect on outcome. Another proposed advantage of the transhiatal technique is that it results in a lower incidence of postoperative gastroesophageal reflux, although there was no significant difference in the incidence of late reflux between the two groups in this study. Several reports have directly compared these two techniques of esophagectomy.‘-” Most of these s,tudies, however have included a mixture of histologic tumor types. A prospective, randomized study included only tumors of the squamous cell type.” This study revealed no significant difference between the two resection techniques in terms of complication rates, operative mortality, or long-term survival rates. Two retrospective reviews compared resection techniques specifically for adenocarcinomas of the distal esophagus and gastric cardia. One found no difference in morbidity or survival between transhiatal and transthoracic esophagectomy.16 The other found that there was no survival difference between techniques, but overall perioperative complications were significantly more frequent in thoracotomy patients (84% versus 48% overall complication rate) with a fourfold higher rate of respiratory complications, as previously noted.’ In general, most of the transthoracic resections in the aforementioned comparisons were performed in an earlier era since the transhiatal approach has only become popular more recently. Comparing a contemporary group with a relatively older historical group induces a potential bias. Most series, however, report similar long-term s.urvival rates irrespective of resection technique, with cure rates mainly dependent on tumor stage at presentation. The radical or “en bloc” esophagectomy, which involves a.n extensive mediastinal lymphadenectomy, may provide a survival benefit NOVEMBER

1996

1TRANSHIATAL

for a select group of patients with early lesions,” but this potential advantage has not been established. In summary, the two most commonly used techniques of esophagectomy were compared (transhiatal versus transthoracic) in well-matched contemporary groups of patients undergoing resection of adenocarcinoma of the distal esophagus and gastric cardia. With the exception of respiratory complications, which occurred less frequently in the transthoracic patients, there was no significant difference between the two groups in terms of morbidity, mortality, or long-term survival. Although the transthoracic approach required significantly longer operating time, it appeared to be just as safe as the transhiatal esophagectomy. Transthoracic esophagectomy therefore remains an acceptable approach for resection of adenocarcinomas of the distal esophagus and gastric cardia. Currently the prognosis for esophageal carcinoma depends mainly upon tumor stage at presentation, not on resection technique. Selection of the operative approach should be based primarily on the surgeon’s personal preference, experience, and the presence or absence of tumor adherence to mediastinal structures until a prospective, randomized trial determines the optimal surgical approach.

REFERENCES 1. Blot WJ, Devesa SS, Kneller RW, Fraurneni JF. Rising incidence of adenocarcinoma of the esophagus and gastric cardia. JAMA. 1991;263:1287-1289. 2. Blot W], Devesa SS, Fraumeni JF. Continuing climb in rates of esophageal adenocarcinoma: an update. JAMA. 1993;270:1320. Letter. 3. Lomboy CT, Pera M, Cameron A], et al. Adenocarcinoma of the esophagus and esophagogastric junction are both associated with Barrett’s epithelium. Gastroenterology. 1992;(suppl) 102:A373. 4. Clark GWB, Smyrk TC, Burdiles P, et al. Is Barrett’s metaplasia the source of adenocarcinomas of the cardia? Arch Surg. 1994;129:609-614. 5. Lewis I. The surgical treatment of carcinoma of the esophagus, with special reference to a new operation for growths of the middle third. Br J Surg. 1946;34:18-31. 6. Turner GG. Excision of thoracic esophagus for carcinoma, with construction of extra-thoracic gullet. Lancet. 1933;2:1315-1316.

DISCUSSION Alex G. Little, MD (Las Vegas, Nevada) : I think you have heard a very important presentation. 1 think that these data will, in fact, influence surgical treatment of this disease. The authors have made the point that there is, arguably, but in my mind clearly a superiority for the transthoracic approach to patients with adenocarcinoma of the esophagus compared to the transhiatal approach. I do find that the arguments made in this paper are convincing and, in fact, support my own trends in the care of these patients over these last years. I do have a few questions and comments which I would appreciate the authors addressing. I am not particularly bothered by the absence of randomization. That weakens the conclusions to some extent but the patients are clearly comparable. I would ask the authors to comment on the impact that the nonrandomization has on postoperative care. That is, surgeons who are going to prefer one operative technique over the other are also going THE AMERICAN

VS TRANSTHORACIC

ESOPHAGECTOMY/ST-1

7. Orringer MB, Sloan H. Esophagectomy without thoracotomy. J Thorax

Cnrdiocfusc

Surg.

1978;76:643-654.

8. Orringer MB, Marshall B, Stirling MC. Transhiatal

esophagec-

tomy for benign and malignant disease. J Thorac Curdiovasc Surg. 1993;105:265-277. 9. Goldfaden D, Orringer MB, Appelman HD, Kalish R. Adenocarcinoma of the distal esophagus and gastric cardia. Comparison of results of transhiatal esophagectomy and thoracoabdominal esophagogastrectomy. J Thorac Cardiooarc Surg. 1986;91:242-247. 10. Pat M, Basoglu A, Kocak H, et al. Transhiatal versus transthoracic esophagectomy for esophageal cancer. J Thorax Cnrdiovasc Surg. 1993;106:205-209. 11. Millikan KW, Silverstein J, Hart V, et al. A 15syear review of esophagectomy for carcinoma of the esophagus and cardia. Arch Surg. 1995;130:617-624. 12. Putnam JB, Sue11 DM, McMurtrey MJ, et al. Comparison of three techniques of esophagectomy within a residency training program. Ann Thorax Surg. 1994;57:319-325. 13. Fok M, Siu KF, Wong J. A comparison of transhiatal and transthoracic resection for carcinoma of the thoracic esophagus. Am J Surg. 1989;158:414-419. 14. Hankins JR, Attar S, Coughlin TR, et al. Carcinoma of the esophagus: a comparison of the results of transhiatal versus transthoracic resection. Ann Thorax Surg. 1989;47:700-705. 15. Gotley DC, Beard J, Cooper MK, et al. Abdorninocervical (transhiatal) oesophagectomy in the management of oesophageal carcinoma. Br J Surg. 1990;77:815-819. 16. Moon MR, Schulte WJ, Haasler GB, Condon RE. Transhiatal and transthoracic esophagectomy for adenocarcinoma of the esophagus. Arch Surg. 1992;127:951-955. 17. Goldminc M, Maddem G, LePrise E, et al. Oesophagectomy by a transhiatal approach or thoracotomy: a prospective randomized trial. Br J Surg. 1993;80:367-370. 18. Beahrs OH, Henson DE, Hutter RVP, Kennedy BJ. Mnnunl fur Smging of Cancer. 4th ed. Philadelphia: JB Lippincott; 1992:57-61. 19. Swisher SG, Hunt KK, Holmes EC, et al. Changes in the surgical management of esophageal cancer from 1970 to 1993. Am J Surg. 1995;169:609-614. 20. Bolton JS, Ochsner JL, Abdoh AA. Surgical management of esophageal cancer. A decade of change. Ann Surg. 1994;219:475-

480. 21. Hagen JA, Peters JH, DeMeester TR. Superiority of extended en bloc esophagogastrectomy for carcinoma of the lower esophagus and cardia. J Thorax Cnrdiovusc Surg. 1993;106:850-859.

to have their own style for postoperative care. I wonder if what is being compared postoperatively is not a result of the operative technique, but simply a comparison of individual surgeon preferences. There has to be some explanation for the difference in survival postoperatively. These patients had similar stages, similar risk factors and essentially similar resections it would appear. But there must be some difference and I wonder if the difference, in fact, is in the number of nodes that were removed from the mediastinum at the time of transthoracic operation. This might impact, of course, survival but also might affect staging. 1 wonder if the authors would comment on whether they think it affects survival or perhaps upstages some tumors who were in the transthoracic group? Similarly, I wonder if the extent of gastrectomy was ever compromised phagectomy. microscopically

in the patients undergoing transhiatal I appreciate that the number of patients positive margins were similar in. each

esowith group.

But having done transhiatal esophagectomies for this disease, I do know that one is very aware that you have got to JOURNAL

OF

SURGERY@

VOLUME

172

NOVEMBER

1996

481

TRANSHIATAL

VS TFIANSTHORACIC~

have enough stomach left to reach to the neck and this can make a difference. I recognize it has become politically correct to find Barretts everywhere, but I do wonder if this is not overcalled on occasion. I wonder how this group defines Barrett’s esophagus. Is it simply any specialized columnar epithehal cells in the specimen or did you actually have to see true columnar epithelium in the esophagus? Finally would you comment on adjuvant therapy? It is not clear whether this was preoperatively or postoperatively and I would appreciate hearing some more about that. In conclusion, I would say I think there are really two important things about this paper. One, is that it carries out the appropriate analyses and presents them very clearly and convincingly. Secondly, it is important that there is a focus on a very particular subset of patients with esophageal carcinoma, patients with adenocarcinoma of the lower third. It is very important as we continue to analyze patients with cancer of the esophagus that we compare apples to apples, etc. and this paper effectively does this. Merril T. Dayton, MD (Salt Lake City, Utah): I guess we are all just a little bit surprised that the incidence of pulmonary complications was actually higher in the group that had the transhiatal procedure. One of the putative advantages of that procedure is that because you do not do a thoracotomy and you do not enter the pleural space, the patient does not have to deal with postoperative pain that those who have a thoracotomy incision do. Would you explain to us why the incidence of pulmonary complications was higher in the transhiatal group? Also, would you tell us whether or not there were any aborted attempts to do a blunt or a transhiatal esophagectomy which then put that patient into the transthoracic category? There is also the suggestion that the transhiatal group had a better survival and that was a little bit surprising. Carey P. Page, MD (San Antonio, Texas): I think that perhaps another take home message from this paper is that regardless of a great operation, whether it be done by the transhiatal route or transthoracic route, that the patient outcome is just atrocious in the long run in terms of diseasefree survival. I suggest that we ought to try some additional alternate therapy. Neoadjuvant therapy with chemosensitizing radiotherapy has been shown in these junctional lesions to result in a no-evidence-of-disease state at exploration as high as 25% whether one is dealing with a junctional squamous or adenocarcinoma. I would make a plea that we develop and enter patients into protocols emphasizing neoadjuvant chemosensitizing radiotherapy for these junctional cancers. Otherwise, no matter how good we are and what operation we do, we are unlikely to impact the survival of these patients. Marco G. Patti, MD (San Francisco, California): You had a 0% mortality after transthoracic and 3% after transhiatal esophagectomy. Are these consecutive patients or were they carefully selected based on their preoperative status? Which are your exclusion criteria for patients with esophageal cancer?

482

THE AMERICAN

JOURNAL

OF

SURGERY”

VOLUME

172

You had a 16% rate of anastomotic leak :after transhiatal esophagectomy; do you think that these leaks affected the rate of pulmonary complications after this type of resection? Melvin W. Twiest, MD (Albuquerque, New Mexico): 1 have a question about the increased incidence of pulmonary complications in transhiatal esophagectomy. I think we continue to favor this operation because it is easier on us even if it is not easier on the patients. But I have been troubled by a couple of patients that 1 have seen that seem to have a severe problem with swallowing and aspiration, weeks and even months after their transhiatal esophagectomy, in the absence of vocal cord palsy. I wonder if the authors have seen such cases? Nicholas Lang, MD (Little Rock, Arkansas): I noticed some of your data had very wide ranges that differed in the two groups. I wondered if you looked to tte sure that you had normally distributed data because the statistical and analytical methods you were using are appropriate for normally distributed, not for non-normally distributed data?

CLOSING Michael S. Romberg, MD: I would like to thank Dr. Little and others for their comments and questions. We found that in asking questions regarding the pulmonary complications, the differences were due to several factors, including a surgeon bias based on their selection criteria which could not be looked at in our study, as well as a small cohort of patients, since there were only 16 in the transthoracic group. There was no randomization as this was a retrospective study, so surgeon preference could not be looked at other than that was their decision at the time of surgery based on their experience and the patient’s general overall condition, in addition to the specific factors which I have listed. The number of nodes which were removed was not looked into, only whether they were positive or not. We did find that the number of patients with positive nodes in both groups were similar and this did become statistically significant whether a patient was node positive or node negative for survival, which is well shown in the literature. Regarding the question of extent of gastrectomy, a total gastrectomy was done in a couple of patients, but these were excluded. Usually, just a 5-10 cm rim of Istomach was removed in these 48 patients. We found that our definition of Barrett’s was based on specialized epithelium seen in preoperative studies. AI1 radiation and chemotherapy was done postoperatively as the patients needed it. There were several patients who went into trials. No patients received preoperative radiation or chemotherapy. We did not have any cross over patients from the transhiatal into the transthoracic group. AI1 32 patients who began as transhiatal operations were subsequently completed by the transhiatal technique. There are reports in which tumor rupture or bleeding caused a cross over and we did not find this evident in our institution. There is a need for better adjuvant treatment and trials specifically for Adenocarcinoma.

NOVEMBER

1996