|
|
||||||||
J Thorac Cardiovasc Surg 1999;117:1077-1085
© 1999 Mosby, Inc.
GENERAL THORACIC SURGERY |
From the Center for Swallowing and Esophageal Disorders and the Departments of Thoracic and Cardiovascular Surgery,a Anatomic Pathology, b Diagnostic Radiology,c Gastroenterology,d and Biostatistics and Epidemiology,e The Cleveland Clinic Foundation, Cleveland, Ohio.
Read at the Twenty-fourth Annual Meeting of The Western Thoracic Surgical Association, Whistler, British Columbia, June 24-27, 1998.
Received for publication July 15, 1998. Revisions requested Sept 30, 1998. Revisions received Feb 4, 1999. Accepted for publication Feb 9, 1999. Address for reprints: Thomas W. Rice, MD, The Cleveland Clinic Foundation, 9500 Euclid Ave, Cleveland, OH 44195.
| Abstract |
|---|
|
|
|---|
| Introduction |
|---|
|
|
|---|
The purpose of this study was to evaluate a protocol of esophageal resection and gastric reconstruction for end-stage achalasia.
| Patients and methods |
|---|
|
|
|---|
Pathology review
The esophageal portion of the resection specimens was evaluated and the following features were recorded: (1) presence or absence of ganglion cells and the number of ganglion cells; (2) presence or absence of myenteric neuritis, defined as inflammation within or immediately around myenteric nerves, graded as mild, moderate, or severe; and (3) presence or absence of myenteric neural fibrosis evaluated by the Masson trichrome stain, graded as mild, moderate, or severe.
1,2 In addition, the proximal esophageal margin and the gastric cardia were evaluated, and the presence of ganglion cells was recorded.
Outcome
Weight measured at each postoperative visit was obtained from medical records. Patients were contacted by telephone and completed a questionnaire to define outcome and gastrointestinal function. Symptom scores of 0 (absent), 1 (mild), 2 (moderate), 3 (severe), and 4 (incapacitating) were used at the time of follow-up to assess abnormal gastrointestinal function as described by dysphagia, early satiety, regurgitation, and acid reflux. Medications and procedures required to treat these symptoms were recorded. Dietary restrictions and the inability to eat at social events (social dietary restrictions) were also scored from 0 to 4. Patients were asked if they felt better after esophagectomy and whether they would undergo esophagectomy again. Complete symptom follow-up was obtained in all but 3 patients: 1 died, 1 refused to be interviewed, and 1 agreed to a partial interview.
Statistical review
Symptom scores were obtained at only one point in time for each patient. Ideally, these scores would be ascertained at multiple points across time for each patient. Because patients were treated over a 10-year period, we have used longitudinal data analysis of these single assessments to estimate the time-relatedness of symptom scores. This longitudinal data analysis used logistic regression analysis, with time interval since surgery used to assess temporal trends.
For each symptom, a sequential data analysis was done of preoperative variables, pathologic findings, operative techniques, and in-hospital complications. Preoperative variables included gender, age at operation, symptoms, duration of symptoms, dietary restrictions, prior therapy, complications of achalasia or therapy, maximum diameter of the esophagus, and maximum sigmoid width of the esophagus. Operative variables included surgical approach (transhiatal or thoracotomy), conversion of transhiatal approach to thoracotomy, variations from the treatment protocol, and whether the operation was elective or urgent. Postoperative variables included complications, need for blood transfusion, and intensive care unit and hospital lengths of stay. Pathologic variables included presence of ganglion cells, myenteric neuritis, eosinophils and granulomatous inflammation in the resection specimen, and presence of ganglion cells in the proximal esophagus and gastric cardia.
Although the scores were expressed on an ordered scale, few patients had symptom scores above 1. Thus scores were dichotomized in two ways: 0 versus 1 to 4, and 0 and 1 versus 2, 3, and 4 for separate analysis. The two results were consistent and, therefore, only the latter set is presented. Because the data set is small, assessment of symptom status used univariable statistics rather that multivariable statistics. Odds ratios, 95% confidence intervals for the odds ratios, and the P values were obtained from the logistic regression model. McNemar's test was used to determine whether symptom frequency changed after surgery.
Unlike symptom scores that were available at only one time for each patient, weight was measured before the operation and at each postoperative visit. A regression of weight versus time after the operation was calculated for each patient; slopes were obtained from the regression models. The slopes were averaged, and a 1-sample t test was used to determine whether the slopes differed significantly from 0. This is equivalent to a longitudinal data analysis of repeated measures. In addition, change in weight from preoperative to follow-up was calculated. An analysis of variance model was used to determine whether weight change was associated with age at operation, gender, or length of follow-up.
| Results |
|---|
|
|
|---|
|
|
Median intensive care unit stay was 2 days (range 1-12 days); median hospitalization was 14 days (range 8-95 days). Seventeen patients required a median of 2 transfusions (range 1-12 units). Postoperative complications are listed in Table II. There were no postoperative deaths. One patient who underwent emergency esophagectomy died 6 months after a complicated postoperative course.
|
|
Symptom scores were available in 29 of 31 patients. The assessment of symptom scores and postoperative diet are listed in Table IIIa. With esophagectomy and gastric reconstruction, dysphagia was reduced significantly from 93% to 17% (P < .001) and regurgitation from 90% to 28% (P < .001), with minimal change in reflux from 24% to 31% ( P = .5). Postoperatively, there was a marked reduction of dietary restrictions from 59% to 17% (P < .001). Postoperative treatment necessary to obtain these scores is shown in Table IIIb.
|
|
|
| Discussion |
|---|
|
|
|---|
End-stage achalasia may be characterized clinically, radiographically, or pathologically. The majority of patients have disabling dysphagia (including aphagia) and regurgitation with associated weight loss and profound dietary restrictions. Aggressive treatment and its complications may further aggravate dysphagia caused by the primary disorder. The radiographic findings in end-stage achalasia are massive esophageal dilatation and tortuosity (sigmoid esophagus) in the majority of patients. Pathologically, there is a marked reduction or absence of ganglion cells with fibrous replacement of the myenteric plexus.
Treatment options in end-stage achalasia are limited, and esophagectomy offers the potential for improved swallowing. The principal function of the esophagus is the rapid unidirectional transit of food from the hypopharynx to the stomach. Unlike other segments of the gastrointestinal tract, the esophagus has no digestive, absorptive, metabolic, or endocrine activities. Despite this seemingly rudimentary task and the deceptively simple arrangement of a muscle pump between two sphincters, it is difficult to replicate esophageal function.
Esophagectomy and colon interposition.
Colon interposition is a true esophageal replacement; it spans the thoracic cavity to connect the cervical esophagus to the stomach that remains below the diaphragm. In expert hands, colon replacement offers restoration of gastrointestinal function in patients with end-stage achalasia.
5 However, most surgeons have abandoned colon interposition when the stomach is adequate for reconstruction.
6-10 There are many disadvantages to the use of colon as an esophageal replacement,
11 including a propensity to late complications and a tendency for the interposition to dilate and form redundant loops (Fig. 3).
12-15 This late complication results in the same problem for patients with end-stage achalasia that initially prompted esophageal resection.Reoperation for a failed colon interposition was reported in 6 of 85 (7%) patients undergoing esophagectomy with colon interposition for benign esophageal diseases.
16
|
Dysphagia, the hallmark of end-stage achalasia and the primary indication for most operations, was relieved or minimized in the majority of patients. However, patients who were younger reported more persistent dysphagia than older ones. Patients with achalasia have a disordered perception of swallowing. Some of this reflects the disease process and some the social aspects of eating. The fact that younger patients were more likely to report dysphagia than older ones is probably composed of multiple factors: shorter duration of disease, a memory of normal swallowing, social activities that may foster dysphagia (such as active lifestyles, eating fast food, eating rapidly), high expectations of normal swallowing after gastric reconstruction, and more social situations in which eating and drinking are focal points. A patient with an anastomotic leak was more likely to have dysphagia.
A lesser sigmoid width was associated with an increased chance of postoperative dysphagia, regurgitation, and reflux. These unexplained findings were not related to any variable or the pathologic findings in the proximal esophagus or gastric cardia. However, this observation suggests that esophagectomy for end-stage achalasia should be considered only after alternative treatment options have been exhausted and the esophagus is dilated and has obtained a large sigmoid width. With the advent of minimally invasive esophageal surgery, laparoscopic Heller myotomy has been performed generally without fundoplication in patients without prior therapy or inadequate therapy regardless of the sigmoid quality of the esophagus. Esophagectomy with gastric reconstruction is reserved for patients with documented failure of adequate myotomy who have a sigmoid esophagus. This delay may avoid esophagectomy in the younger patient and may decrease the prevalence of dysphagia.
| Conclusions |
|---|
|
|
|---|
| Appendix: Discussion |
|---|
|
|
|---|
The focus of Dr Banbury's presentation is that the stomach makes a good esophageal substitute. This is a subject about which I have given much thought, and I have the following comments: First, swallowing after esophageal replacement with either colon or stomach is never like swallowing with a normal esophagus. Consequently, 17% of Dr Banbury's patients have persistent dysphagia, 25% regurgitation, 31% reflux, and 44% early satiety. Yet 83% would elect to have the operation again if that were possible. This tells you how appreciative these patients are for any improvement in their ability to swallow. My question is why 17% have persistent dysphagia and why the younger patients seem to have more difficulty. Can you explain this?
Dr Banbury. To answer your first question, the advisability of positioning the stomach in the chest is a matter of debate. Many of these patients with dysphagia are younger but they do have the ability to swallow. Perhaps the anastomosis, perhaps the denervated intrathoracic organ is the cause. I cannot explain specifically why there is persistent dysphasia in some, but certainly it is improved from the preoperative status. The use of prokinetic agents helps with clearance of food and stasis. The specific reasons for dysphagia are difficult to pinpoint.
Dr DeMeester. Did any of the patients with dysphagia require dilation?
Dr Banbury. Yes. We are aggressive in dilation. All of these patients are monitored in the postoperative period with endoscopy, and often they are treated by dilation during endoscopy. Sixty percent of patients had postoperative dilation with a median of 5 dilations per patient.
Dr DeMeester. Was the dilation mainly required because of narrowing of their gastric esophageal anastomosis?
Dr Banbury. Any patient who had complaints with regard to the anastomosis was subjected to endoscopy. We routinely perform endoscopy in patients at a point in the postoperative period. If there were problems with the anastomosis that we identified on endoscopy, and in patients who had no complaints, then we dilated the esophagus during endoscopy.
Dr DeMeester. In our experience the need for dilation after a colon interposition is rare, less than 5% (J Thorac Cardiovasc Surg 1998;115:1241). Thus there is an issue here that I think needs to be evaluated in greater detail.
My second question deals with your emphasis on attention to detail in regard to the operation. I appreciate that emphasis and I think it is a very valid comment, particularly your emphasizing the need to prevent redundancy of the stomach in the thoracic cavity. Of course, the same admonition can be applied to the colon; you must make every effort to avoid redundancy of either substitute, colon or stomach. Could you explain why preventing redundancy is so important in your experience with the stomach?
Dr Banbury. There is concern about postoperative motor function of the stomach, and redundant stomach could complicate that situation. A straighter gastric tube allows gravity to help the food bolus to pass straight through into the abdominal cavity rather than to be hung up in folds in the chest. That is also true with colon.
Dr DeMeester. Do you make any effort to tube the stomach or reduce it in size, that is, to form a gastric tube as opposed to the whole stomach?
Dr Banbury. The stomach is not specifically tubed, but the proximal aspect is fashioned for the anastomosis.
Dr DeMeester. Do you anchor the stomach to the diaphragm?
Dr Banbury. Yes, we do.
Dr DeMeester. My last question is in regard to an observation we have made, and I would like your thoughts about it. In our experience it is very difficult to place gastric mucosa next to squamous mucosa without having a problem over the long term. These problems start about 5 years after the operation. Studies have shown that a denervated stomach regains its ability to produce acid similar to preoperative levels. Because of this, 19% of patients will eventually have esophagitis in the cervical esophagus on long-term follow-up, and 12% will have Barrett's esophagus in the cervical esophagus, with the mean time for its development being 7 years. This has not been associated with the use of the colon. My question to you, Dr Banbury, is this: Have you had the opportunity to examine some of your long-term patients with an endoscope? If so, what where the findings? I took note that you have made some adjustments in your operation because of concern over reflux, which suggests that the same thing is happening in your patients.
Dr Banbury. We do routinely examine these patients with an endoscope. We have not found esophagitis or endothelial changes so far, but reflux is a problem in some of our patients. We have changed our operative technique to try to further reduce the complication of reflux.
Dr DeMeester. Do the patients who have symptoms of reflux have irritation of the throat or are they having aspiration?
Dr Banbury. Irritation. We do use acid suppression therapy in any patient who has this problem.
Dr DeMeester. That ends my questions. I do believe that the best organ for long-term replacement of the esophagus is still an open issue.
Dr John R. Benfield (Sacramento, Calif). Dr Banbury, this is an important paper. My question is to you and perhaps to Dr DeMeester as well. I would like to hear about the conventional wisdom that you expressed, which is that the stomach remains peristaltic whereas the colon is an aperistaltic conduit. Do you actually have evidence that the stomach remains peristaltic?
Dr Banbury. No, we have not done studies of the stomach to determine specific motility.
Dr Benfield. I doubt that the stomach remains peristaltic, and I was wondering whether you had some evidence with which to dispel my doubt.
Dr DeMeester. Our studies on the stomach show that it does not regain peristaltic function; swallowing becomes pretty much a gravity feed system. Our studies of the colon show that it does provide some peristaltic function, but this occurs several years after the operation. I commonly say to the patients that the longer the colon is in the better it performs, in contrast to the stomach. From my own experience with these operations, about 5 years after a gastric pull-up the patients start having trouble. I will be interested to see Dr Banbury's data as follow-up becomes longer.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
R. F. Heitmiller Invited commentary Ann. Thorac. Surg., June 1, 2006; 81(6): 2049 - 2049. [Full Text] [PDF] |
||||
![]() |
S. V. Kostic, T. W. Rice, M. E. Baker, M. M. DeCamp, S. C. Murthy, L. A. Rybicki, E. H. Blackstone, and J. E. Richter Timed barium esophagogram: A simple physiologic assessment for achalasia J. Thorac. Cardiovasc. Surg., November 1, 2000; 120(5): 935 - 946. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |