Antireflux surgery, highly selective vagotomy and duodenal switch procedure: post-operative evaluation in patients with complicated and non-complicated Barrett's esophagus

Antireflux surgery, highly selective vagotomy (HSV) and Roux-en-Y duodenojejunostomy have been suggested for control of pathophysiological factors involved in patients with Barrett's esophagus (BE). The aim of this study was to evaluate prospectively the results of this technique in patients with complicated (n = 21) and noncomplicated (n=45) BE. Complete evaluation of esophageal function, endoscopic histologic and clinical control was carried out before and 2 years after surgery. Post-operative results show recurrence of ulcer in patients with complicated BE, but no recurrence in patients with non-complicated BE. Preoperative esophageal ulcer and stricture were present in 85.3% and 14.3%, respectively, of patients with complicated BE. In this group, recurrence of these complications was 38.1% and 9.5% respectively. The technique offers excellent results in patients with non-complicated BE. However, in patients with complicated BE, the recurrence rate is higher, mainly because of the persistence of acid reflux into the esophagus.

Intraperitoneal bile collections after laparoscopic cholecystectomy: causes, clinical presentation, diagnosis, and treatment

BACKGROUND

Bile leakage is more common after laparoscopic cholecystectomy than after open surgery. In our department, the rate of postoperative bile collections after open surgery is 0.2% vs 0.6% after laparoscopic cholecystectomy.

METHODS

We studied 13 cases of intraperitoneal bile collection without common bile duct damage drawn from a total of 5,200 laparoscopic cholecystectomies (0.23%). Clinical presentation, symptoms, method of diagnosis, causes, time of diagnosis, correlation of time of diagnosis with definitive treatment, and postoperative results were analyzed.

RESULTS

The symptoms appeared between the 5th and 8th postoperative days. They were observed in patients with either chronic or acute cholecystitis. The main causes were misapplication of clips at the cystic duct and open Luschka's duct. Ultrasound failed for early recognition of bile collections. The definitive diagnosis was made by repeat ultrasonography, CAT scan, and ERCP.

CONCLUSION

The ideal treatment in these cases is a minimally invasive procedure, but since the diagnosis is frequently delayed, open surgery is performed in the majority of patients. However, there were no mortalities in this group of patients.

Anatomic dilatation of the cardia and competence of the lower esophageal sphincter: a clinical and experimental study

Anatomic and clinical data suggest that the gastroesophageal junction or cardia in patients with gastroesophageal reflux disease GERD) may be dilated. We hypothesized that anatomic dilatation of the cardia induces a lower esophageal sphincter dysfunction that may be corrected by narrowing the gastroesophageal junction (i.e., calibration of the cardia). We measured the perimeter of the cardia during surgery in control subjects and patients with GERD and Barrett's esophagus. We then tested our hypothesis in a mechanical model. The model was based on a pig gastroesophageal specimen with perpendicularly placed elastic bands around the cardia simulating the action of the "sling" and "clasp" fibers. "Dilatation" of the cardia was induced by displacing the sling band laterally and decreasing its tension. "Calibration" of the cardia was performed by reapproximation of the sling band toward the esophagus but maintaining the same tension as the dilated model. In the "basal," "dilated," and "calibrated" states, the perimeter of the cardia was noted and rapid mechanized pullback manometry with a water-perfused catheter was performed. The opening pressure was determined, and three-dimensional sphincter pressure images were analyzed. The average cardia perimeter was 6.3 cm in control subjects, 8.9 cm in GERD patients, and 13.8 cm in patients with Barrett's esophagus. The arrangement of the bands in the experimental model generated a manometric high-pressure zone similar to that in the human lower esophageal sphincter. Dilatation of the cardia resulted in a decrease in the resting pressure, length, and vector volume of the high -pressure zone, and reduced the opening pressure. Calibration restored the resting and opening pressure, and normalized the three-dimensional pressure image. In patients with GERD and Barrett's esophagus, the cardia is dilated. Our model supports the hypothesis that lower esophageal sphincter function is compromised by anatomic dilatation of the cardia and can be restored by approximation of the "sling" fibers toward the lesser curvature "clasp" fibers). This provides evidence for a correlation between gastroesophageal sphincter dysfunction in reflux disease and its correction by antireflux surgery.

Cardiomyotomy in achalasia: which fibers do we cut?

Until now, it has not been quite clear which muscular fibers are cut when a cardiomyotomy for achalasia is carried out. In the present report, in a human achalasic gastroesophageal specimen, the mucosa of the stenotic segment was stripped off, allowing the fibers of the inner muscular coat to be seen. In addition, three cardiomyotomies at different sites were simulated. In achalasic specimens, the stenotic area is formed by the semicircular ('clasp') and oblique ('sling') muscular fibers. Different myotomies section these two muscular bands in distinct proportions. The stenotic segment in achalasia coincides topographically with the anatomic lower esophageal sphincter area. The site of cardiomyotomy is not irrelevant because this sphincter is not an annular muscle and the two muscular components of the sphincter can be sectioned in different ways. This may be important in post-operative results with regard to the relief of dysphagia and the appearance of gastroesophageal reflux.

Late results of a randomized clinical trial comparing total fundoplication versus calibration of the cardia with posterior gastropexy

BACKGROUND

The aim was to perform a prospective randomized study in patients with chronic gastro-oesophageal reflux treated either by total fundoplication or calibration of the cardia with posterior gastropexy. Late follow-up considered subjective and objective parameters, and related outcome to the presence of Barrett's oesophagus.

METHODS

A total of 164 patients were randomized to fundoplication (n = 76) or calibration of the cardia (n = 88). They were evaluated by clinical questionnaire, upper gastrointestinal endoscopy with biopsies, oesophageal manometry and gastro-oesophageal reflux studies, including scintigraphy and 24-h oesophageal pH monitoring.

RESULTS

There were no operative deaths. There was 95 per cent follow-up at a mean of 85 months. The mean recurrence rate for both operations was near 40 per cent at 10 years, but patients without Barrett's oesophagus had a recurrence rate after both operations of around 23 per cent compared with 83 per cent after 10 years for those with Barrett's oesophagus (P < 0.0001). Low-grade dysplasia developed in 13 per cent of the patients with Barrett's oesophagus. There were significant differences in all objective parameters in a comparison of patients with Visick I or II and those with Visick III or IV disease at the late assessment.

CONCLUSION

Both total fundoplication and calibration of the cardia with posterior gastropexy had similar subjective and objective late results. However, results were significantly worse in patients with Barrett's oesophagus.

[Survival of patients with esophageal cancer subjected to total thoracic esophagectomy]

BACKGROUND

Esophageal carcinoma has a dismal prognosis. Several authors have reported a very low survival in Chile.

AIM

To report the survival of patients with esophageal carcinoma, subjected to esophageal resection.

MATERIAL AND METHODS

Analysis of 108 patients subjected to thoracic esophageal resection between 1985 and 1996. Patients were classified according to the location of the tumor and its staging.

RESULTS

Eleven patients died in the immediate postoperative period and 90 patients were followed. In 53 the exact cause of death was determined. Global five years survival was 29% and median survival was 18 months. Survival was 100% in stage I tumors. Adjuvant therapy resulted in a better survival of stage III tumors. Survival of stage IV tumors was worst than stage I to III tumors. There was no survival difference between squamous carcinoma or adenocarcinoma. Tumors located in the superior third of the esophagus had a worst prognosis. Causes of death were mediastinic metastases, local recidivism, pleural or pulmonary metastases and less frequently, brain, bronchial or bone metastases.

CONCLUSIONS

The survival of these, patients with esophageal carcinoma did not differ from the figures reported abroad.

Gallbladder carcinoma during laparoscopic cholecystectomy: is it associated with bad prognosis?

Laparoscopic cholecystectomy is the treatment of choice for gallstone disease. The ultrasonogram has failed for the early detection of gallbladder cancer, especially if inflammation (chronic or acute) is present. Incidental gallbladder could be an important cancer finding during laparoscopic cholecystectomy, due to the potential cancer cell dissemination during the procedure. In our Department, 6500 laparoscopic cholecystectomies have been performed in the last 5 years and in 15 cases (0.23%) gallbladder cancer was found during surgery or after histological examination of the resected gallbladder. In none of these 15 patients was pre-operative diagnosis of gallbladder carcinoma postulated. When re-evaluation of the pre-operative ultrasonograms was done, it was possible to observe signs suggesting the presence of neoplastic infiltration in 4 of them (28.6%). During videoscopic exploration, also in 4 patients, the suspicion of gallbladder cancer was noted. Laparoscopic cholecystectomy was completed in 9 patients. In 2 of them, in situ or mucosal invasion was demonstrated with a long survival. One patient presented recurrence at the biliary hilum 2,5 years after surgery. Six patients were re-operated and in 4 of them peritoneal or port site metastasis was found; all died early (4.5 month median survival). The other 2 patients were submitted to liver bed resection and lymph node dissection. These patients are free of cancer recurrence after 15 months of follow-up. Six patients were converted to open surgery, performing palliative procedures and died before the 12 month follow-up. The suspicion of pre-operative gallbladder cancer is generally unlikely to be confirmed based on ultrasonographic signs; but, in some cases with high suspicion, further investigation (TAC, tumor markers, etc.) must be indicated in order to avoid poor results. Laparoscopic cholecystectomy could be associated with bad prognosis, and then, when gallbladder cancer is suspected during the laparoscopic procedure, conversion to open surgery could be the best choice.

Long-term follow-up after laparoscopic cholecystectomy without routine intraoperative cholangiography

The indications for routine intraoperative cholangiography remain controversial. We present here our recent results concerning the frequency of unknown retained common bile duct stones in 253 consecutive patients who underwent laparoscopic cholecystectomy without intraoperative cholangiography in whom the presence of preoperative choledocholithiasis had been excluded by clinical, biochemical, and ultrasonographic evaluation. These patients were followed up for at least 4 years after surgery with evaluations similar to those made preoperatively. Freedom from symptoms and normal test results were found in 96.8% of patients. Jaundice and abnormal liver function test results were demonstrated in 3.2% of patients, but retained common bile duct stones were found in only 2.3% of patients. We conclude that laparoscopic cholecystectomy without routine intraoperative cholangiography can be performed safely without the discovery of a high percentage of retained common bile duct stones at later follow-up.

Long-term results of classic antireflux surgery in 152 patients with Barrett's esophagus: clinical, radiologic, endoscopic, manometric, and acid reflux test analysis before and late after operation

BACKGROUND

The classic surgical procedure for patients with Barrett's esophagus (BE) has been either Nissen fundoplication or posterior gastropexy with calibration of the cardia.

METHODS

The purpose of our study was to determine late subjective and objective results of these classic surgical techniques in a large number of patients with BE. A total of 152 patients were included in this prospective protocol.

RESULTS

There was 1 death (0.7%) after operation. The late follow-up of 100 months demonstrated a high percentage of failures among patients with noncomplicated BE (54%) and an even higher figure in patients with complicated BE (64%). In 15 patients low grade dysplasia appeared at 8 years of follow-up and an adenocarcinoma in 4 patients. Twenty-four-hour pH monitoring demonstrated a decrease in acid reflux into the esophagus, and Bilitec studies also demonstrated a decrease of duodenoesophageal reflux, but in all cases with a higher value than the normal limit.

CONCLUSIONS

Classic antireflux surgery in patients with BE results in a high percentage of failures at very late follow-up because it cannot completely avoid acid and duodenal reflux into the esophagus.

Esophagocardioplasty, vagotomy-antrectomy and Roux-en-Y gastrojejunostomy: indication in cases with severe esophageal motor disfunction

Almost 10% of patients with Crest syndrome associated with severe gastroesophageal reflux and 5-10% of patients with failed cardiomyotomy for achalasia present with cardial or distal esophageal organic stricture. Some of these cases are poor risk patients for surgery and therefore the surgeon must offer a safe procedure with low morbimortality, keeping in mind the pathophysiological motor pattern of these patients. In order to treat the stricture to improve the esophageal transit we treated patients with esophagocardioplasty associated with vagotomy-antrectomy and Roux-en-Y gastrojejunostomy, thereby avoiding the potential acid or biliary reflux in poor risk patients in whom esophagectomy would be a very deleterious procedure. All four patients had a good postoperative evolution and late control demonstrated good esophagogastric transit with no postoperative esophagitis.

A new physiologic approach for the surgical treatment of patients with Barrett's esophagus: technical considerations and results in 65 patients

OBJECTIVE

To determine the results of a new surgical procedure for patients with Barrett's esophagus.

SUMMARY BACKGROUND DATA

In addition to pathologic acid reflux into the esophagus in patients with severe gastroesophageal reflux and Barrett's esophagus, increased duodenoesophegeal reflux has been implicated. The purpose of this study was to establish the effect of a new bile diversion procedure in these patients.

METHODS

Sixty-five patients with Barrett's esophagus were included in this study. A complete clinical, radiologic, endoscopic, and bioptic evaluation was performed before and after surgery. Besides esophageal manometry, 24-hour pH studies and a Bilitec test were performed. After surgery, gastric emptying of solids, gastric acid secretion, and serum gastrin were determined. All patients underwent highly selective vagotomy, antireflux procedure (posterior gastropexy with cardial calibration or fundoplication), and duodenal switch procedure, with a Roux-en-Y anastomosis 60 cm in length.

RESULTS

No deaths occurred. Morbidity occurred in 14% of the patients. A significant improvement in symptoms, endoscopic findings, and radiologic evaluation was achieved. Lower esophageal sphincter pressure increased significantly (p < 0.0001), as did abdominal length and total length of the sphincter (p < 0.0001). The presence of an incompetent sphincter decreased from 87.3% to 20.9% (p < 0.0001). Three of seven patients with dysplasia showed disappearance of this dysplasia. Serum gastrin and gastric emptying of solids after surgery remained normal. Basal and peak acid output values were low. Twenty-four hour pH studies showed a mean value of 24.8% before surgery, which decreased to 4.8% after surgery (p < 0.0001). The determination of the percentage time with bilirubin in the esophagus was 23% before surgery; this decreased to 0.7% after surgery (p < 0.0001). Late results showed Visick I and II gradation in 90% of the patients and grade III and IV in 10% of the patients.

CONCLUSIONS

This physiologic approach to the surgical treatment of patients with Barrett's esophagus produces a permanent decrease of acid secretion (and avoids anastomotic ulcer), decreases significantly acid reflux into the esophagus, and abolishes duodenoesophageal reflux permanently. Significant clinical improvement occurs, and dysplastic changes at Barrett's epithelium disappear in almost 50% of the patients.

Correlation among ultrasonographic and videoscopic findings of the gallbladder: surgical difficulties and reasons for conversion during laparoscopic surgery

Laparoscopic cholecystectomy is currently the standard procedure for chronic and acute cholecystitis. The purpose of this prospective study was to evaluate the preoperative ultrasound findings of the gallbladder and correlate those aspects with surgical videoscopic and histopathologic findings and the results concerning intraoperative complications and the conversion index to open surgery. Gallbladder findings were classified into three categories according to the gallbladder wall characteristics and the presence of visible lumen and stones. Simple chronic cholecystitis (type I) and acute cholecystitis, with gallbladder wall thickness <5 mm (type IIa) presented significantly lower intraoperative complications without conversion to open surgery. Scleroatrophic (type III) and acute cholecystitis with gallbladder wall thickness >5 mm (type IIB) presented significantly more surgical difficulties and a higher conversion rate to open surgery (p < 0.01). We postulate that this classification will be useful for surgeons in predicting potential problems in individual patients, at least at the initial laparoscopic cholecystectomy experience, and in advising patients of the potential risks of and conversion to open surgery.

Laparoscopic highly selective vagotomy: technical considerations and preliminary results in 119 patients with duodenal ulcer or gastroesophageal reflux disease

The technical considerations and preliminary results of 119 patients submitted to laparoscopic highly selective vagotomy are presented. There were 33 with duodenal ulcers, 31 with duodenal ulcers plus gastroesophageal reflux, and 55 with gastroesophageal reflux. Operating time varied from 120 to 160 minutes. Six complications occurred: four perforations of the gastric fundus and two bleeding episodes. Conversion to open surgery was done in four cases and reoperation in one case. No deaths occurred, and the mean hospital stay was 3 days. The mean follow-up was 16 months, being 94% of the cases with Visick I or II and 6% with Visick III or IV. This technique is completely feasible by laparoscopic procedure and reproduces exactly what has been done with the laparotomy approach.

[Patients with pathologic gastroesophageal reflux without erosive esophagitis: correlation of the endoscopic and histological aspect of the esophagus]

Eighty four patients with chronic gastroesophageal reflux in whom endoscopy showed a normal or minimally altered esophageal mucosa (hyperemia, erythema or mucosal congestion) were prospectively studied. In each patient, two esophageal biopsies were obtained (1 and 3 cm above the mucosal change zone). Histological esophagitis was found in 28% of patients with endoscopically normal mucosa and in 26% of patients with minimal endoscopical mucosal alterations. It is concluded that the description of these minimal esophageal alterations during endoscopy is subjective, unreproducible and do not clearly indicate the presence of esophagitis. A classification of endoscopical findings is proposed, based on the presence of objective lesions such as erosions or Barrett esophagus.

[Magnitude of acid gastroesophageal reflux measured by 24-hour esophageal pH monitoring compared to the degree of endoscopic esophagitis]

Twenty four hour esophageal pH monitoring was performed in 110 patients, placing a pH electrode 5 cm above the proximal border of the lower esophageal sphincter. This test allows to measure the total lapse in which there is an acid pH in the esophagus, the ability of the esophagus to get rid of the acid reflux and documents the relationship between esophageal pH and symptoms. Upper GI endoscopy showed a normal esophagus in 38 patients, and esophagitis grade I in 25, grade II in 11, grade III in 25 and grade IV in 11. There was a high correlation between the severity of esophagitis and the total time in which the esophagus was exposed to an acid pH. It is concluded that this test can be used as a "gold standard" for gastroesophageal reflux detection.

Prospective randomized study comparing three surgical techniques for the treatment of gastric outlet obstruction secondary to duodenal ulcer

A prospective randomized clinical trial was performed in order to evaluate the results of three surgical techniques for the treatment of gastric outlet obstruction secondary to duodenal ulcer. Ninety patients with clinical and laboratory evidence of gastric retention were enrolled. After laparotomy, patients underwent either highly selective vagotomy (HSV) + gastrojejunostomy, HSV + Jaboulay gastroduodenostomy, or selective vagotomy (SV) + antrectomy. One patient died after HSV + Jaboulay gastroduodenostomy due to postoperative acute pancreatitis. There were no differences in the postoperative course of the three groups. Patients were followed for a mean of 98 months (range: 30 to 156 months). There was a significantly better result after HSV + gastrojejunostomy than after Jaboulay anastomosis (p < 0.01), but not after SV + antrectomy. Gastric acid reduction was similar in the small group of patients studied. We propose HSV + gastrojejunostomy as the treatment of choice in patients with duodenal ulcer and gastric outlet obstruction.

Location of the lower oesophageal sphincter and the squamous columnar mucosal junction in 109 healthy controls and 778 patients with different degrees of endoscopic oesophagitis

In this study the location of the lower oesophageal sphincter measured by manometry and the location of the squamous columnar junction measured by endoscopy were determined in 109 healthy controls and 778 patients with different degrees of endoscopic oesophagitis. No significant differences in the prevalence and severity of the heartburn and regurgitation were observed when different degrees of oesophagitis were compared but dysphagia was more common and severe in patients with complicated Barrett's oesophagus (p < 0.001). This group also showed a male predominance and older age compared with other groups. The total length of the oesophagus, measured by the location of the distal end of the lower oesophageal sphincter was similar in all patients; however, the location of the squamous columnar junction extended more proximally and was related to the increasing severity of endoscopic oesophagitis. The manometric defects at the cardia were more frequent in severe oesophagitis (p < 0.001). These results suggest that, during the course of oesophagitis, the squamous columnar junction is displaced proximally. This displacement is limited to the mucosa, however, and does not involve the muscular layer, because the lower oesophageal sphincter undergoes no dislocation.

Surgical management of esophageal strictures

Benign esophageal strictures may be caused by numerous disorders, but more than 90% of them are due to severe gastroesophageal reflux or ingestion of lye. A new classification of the severity of the stricture is proposed, based on the endoscopic and radiological evaluation of three parameters: internal diameter, length of the stricture and ease or difficulty of dilatation. In patients with strictures secondary to reflux, initial treatment includes periodic dilatation. Grade I and II strictures require esophageal resection. In grade III patients, bile diversion or esophageal resection should be performed. Caustic ingestion produces a wide spectrum of tissue damage in the upper digestive tract ranging from minimal chemical burn to an extensive and massive necrotic lesion. The basic and main treatment in patients with an established esophageal stricture is periodic dilatation avoiding, if at all possible, any kind of surgery. In patients with grade III stricture, colonic interposition between cervical esophagus and stomach or duodenum is preferred, treating the damaged esophagus by resection or leaving it "in situ". Psychiatric evaluation is mandatory in these cases.

Gallstone ileus

A retrospective analysis of 74 patients with gallstone ileus detected during the period between 1975 and 1987 was performed at the Surgical Department. The group comprised 55 females and 19 males, with a mean age of 64.8 years. Previous biliary symptoms had been observed in 76% of the cases and in 58% there had been concomitant disease. The main duration of symptoms previous to admission was 3.4 days. In 85% of the cases complementary diagnostic procedures were performed. The triad of air in the biliary tract, air-fluid levels and ectopic stone was found in only 9.5% of the cases. The preoperative diagnosis of gallstone ileus was made in 31% of the patients. The preoperative period was 2.2 days. The main surgical procedure was enterolithotomy in 92% of the cases, the site of impaction being the terminal ileum in 65%. Only in 1 case was simultaneous biliary tract surgery and enterolithotomy performed. Overall, the 30-day postoperative mortality rate was 13.5%, with intra-abdominal sepsis as the main cause of death. Sixteen patients were submitted to biliary surgery 2 to 6 months later, and no deaths occurred.

[Characteristics of severe esophagitis in patients with collagen diseases]

Connective tissue disease may alter esophageal function generating symptoms due to gastroesophageal reflux and motor disturbances. Fifteen patients with connective tissue diseases and severe esophagitis defined by the presence of esophageal stenosis or ulcerations were studied. Diagnosis was made with radiologic, endoscopic and manometric studies. Dysphagia was present in 11 and gastroesophageal reflux in all. All patients has an hypotensive and shorter lower esophageal sphincter. Better therapeutic results were obtained with surgical treatment.

[A cooperative study on early and intermediate gastric cancer: clinical, diagnostic and therapeutic aspects]

A cooperative study involving 13 hospitals in Chile allowed the analysis of 353 patients with gastric cancer in early stages. 82 cancers were located at the mucosal level, 151 at the submucosa and 120 reached the muscular layer. There were no differences in age and sex among these groups. Compared to early stage, patients with intermediate stage had a greater incidence of bleeding, anemia and undernutrition and exhibited lower body weight. Endoscopy had a higher diagnostic yield compared to radiological study. The diagnosis was confirmed by biopsy in 95% of patients. Total or subtotal gastrectomy was performed according to the location of the lesion, with a low operative mortality rate. Early gastric cancer accounts for 8 to 10% of all patients with gastric cancer undergoing surgical treatment.

Histological studies of Auerbach's plexuses of the oesophagus, stomach, jejunum, and colon in patients with achalasia of the oesophagus: correlation with gastric acid secretion, presence of parietal cells and gastric emptying of solids

Histological changes in the Auerbach's plexuses of the oesophagus, stomach, jejunum, and colon were analysed in a prospective study in 34 patients with achalasia of the oesophagus. At the distal end of the oesophagus ganglia cells were absent in 91% of cases as well as in the middle third of the stomach (20%). The Auerbach's plexuses were normal in the jejunum and colon. The results of gastric acid secretion showed that the peak acid output was significantly lower in achalasia patients compared with controls (p less than 0.001). There was no correlation between the mean ganglion neuronal count in the gastric plexuses and the rate of gastric acid output (r = 0.33). Gastric emptying of solids was also evaluated, but there was no correlation between gastric emptying and the mean ganglion neuronal count in the gastric Auerbach's plexuses. The rate of gastric emptying of solids was similar in controls and patients with achalasia. These studies suggest that denervation of the oesophagus in patients with achalasia, which is a constant finding in several previous reports may extend beyond the oesophagus to the stomach in nearly half the cases.

Comparison of forceful dilatation and esophagomyotomy in patients with achalasia of the esophagus

In the present paper we have reviewed the results of forceful dilatation as compared with surgical esophagomyotomy in patients with achalasia. The review of 4 retrospective and uncontrolled studies revealed that in all, surgery produced a more effective late result than dilatation, with minimal side effects and very low or no mortality. In the only prospective randomized study comparing both treatments by the same group, good late results were seen after surgery in 95% of the cases, as compared with 65% after dilatation (p less than 0.001). The resting gastroesophageal sphincter pressure was predictive of the quality of the late results. Reflux occurred in 8% of the dilated and in 19% of the operated group as measured by standard acid reflux test. The old, classical concept that dilatation is the first choice and preferable method of treatment for patients with achalasia should be reviewed, and the idea that surgery should be reserved only for patients in whom dilatation has failed should be abandoned. We propose that surgical treatment should be the initial choice in the majority of patients with achalasia of the esophagus.