Reflux Control by Fundoplication: of the Nissen Operation. A Clinical and Manometric Assessment

Transcription

1 Reflux Control by Fundoplication: A Clinical and Manometric Assessment of the Nissen Operation F. Henry Ellis, Jr., M.D., Ph.D., and Robert E. Crozier, M.D. ABSTRACT Ninety-two Nissen fundoplications were performed for relief of symptoms of reflux in patients without stricture and in whom the wrap was left in the abdomen. Of these, 25 were reoperations. Reflux secondary to a hypotensive lower esophageal sphincter was associated with a sliding esophageal hiatus hernia in 61 patients, and surgical or other manipulative maneuvers were responsible for a hypotensive lower esophageal sphincter in 5 patients. No obvious cause could be determined in the remaining patients. The operation was transabdominal in 74 patients and by thoracotomy in 18 patients. Fifteen patients required ancillary operative procedures. Follow-up studies averaged 5?4 years in 82 patients and revealed permanent control of reflux in 74 of them (90.2%). Eighty percent of the patients were able to belch after operation. Overall patient satisfaction was achieved in 67 patients (81.7%). Among those who underwent operation for the first time, 84.5% achieved satisfactory results, whereas only 75% of those who underwent reoperation were benefited. Poorest results were caused by too tight a wrap, a complication that became rare after the size of the indwelling stent was increased to 42F. Esophageal manometry documented effective and significant augmentation of the amplitude and length of the lower esophageal sphincter from preoperative values of 7.2 f 0.5 mm Hg and 2.4 * 0.1 cm to postoperative values of 15.6 f 0.6 mm Hg and 4.1 f 0.1 cm (p < 0.001). We conclude that if patients are properly selected and the wrap loosely fashioned, permanent control of reflux can be achieved by the classic Nissen fundoplication in 90% of patients with relatively few complications and a high degree of patient satisfaction. Gastroesophageal reflux is primarily the result of hypotension of the lower esophageal sphincter (LES). Awareness of this physiological abnormality has resulted in the widespread adoption of operations designed to raise the pressure of the LES and the abandon- From the Section of Thoracic and Cardiovascular Surgery and the Department of Gastroenterology, Lahey Clinic Medical Center, Burlington, MA, and the New England Deaconess Hospital, Boston, MA. Presented at the Twentieth Annual Meeting of The Society of Thoracic Surgeons, San Antonio, TX, Jan 23-25, Address reprint requests to Dr. Ellis, Lahey Clinic Medical Center, 41 Mall Road, Box 541, Burlington, MA ment of purely anatomically directed procedures. While the superiority of antireflux operations over anatomical repairs in relieving symptoms of reflux is now widely recognized, no uniformity of opinion exists as to which of the various antireflux procedures is best. Clinical and experimental studies have provided convincing evidence that the Nissen total fundoplication is superior to other fundoplication procedures in affording permanent control of reflux. This operation, however, has been criticized as producing an excessively competent LES, which may lead to complications that might restrict its usefulness. Awareness of these complications of the classic Nissen fundoplication procedure has led to the development of many modifications, and results of the originally described procedure are now difficult to evaluate. This article assesses clinical and manometric results in patients with symptoms of gastroesophageal reflux in whom Nissen total fundoplication was performed by the technique to be described. Technique of Operation The technique of fundoplication we have employed is similar tothat reported by Nissen [l] in his classic article and that we [2, 31 have reported on previously. An abdominal approach is preferred except when esophageal shortening is suspected or a previous thoracotomy has been performed. The left lobe of the liver is mobilized, and the gastric fundus is freed by division of the short gastric and posterior gastric vessels and part of the gastrohepatic omentum. A segment of abdominal esophagus is mobilized to provide an adequate length for the wrap, and care is taken to preserve the vagus nerves. The mobilized esophagus is encircled with a Penrose drain, traction is exerted on it, and it is encompassed for a variable distance by the previously freed gastric fundus. The adjacent margins of anterior and posterior gastric fundus are loosely approximated around the freed esophagus using heavy nonabsorbable sutures; a small portion of the underlying esophageal wall is also incorporated to prevent slippage. The suture line comes to lie on the right anterolateral aspect of the encircled esophagus. A loose wrap is ensured by prior placement of a large indwelling stent. Currently, a 42F Maloney dilator is preferred for this purpose. After completion of the fundoplication, the esophageal hiatus, if patulous, is narrowed with heavy nonabsorbable sutures. This maneuver does not reinforce the antireflux aspect of the procedure but prevents cephalad migration of the wrap into the thorax. 387

2 388 The Annals of Thoracic Surgery Vol 38 No 4 October 1984 w Secondary Procedure for Reflux Miscellaneous Fig 1. Type and distribution of antireflux operations performed at the Lahey Clinic between 2970 and Material and Methods Between January, 1970, and January, 1984, 181 fundoplication procedures were carried out at the Lahey Clinic by one of us (F. H. E.). Of these, 117 were of the Nissen type just described (Fig 1). Since our initial experience with use of this technique to surround a firm panmural fibrous stricture was unsatisfactory, as it was when the fundoplication was left in an intrathoracic position in instances of esophageal shortening, operations of this type have been eliminated from our analysis. Similarly, a number of fundoplications were performed in patients who underwent reoperation for such symptoms as postoperative dysphagia and gas-bloat syndrome (most were originally operated on elsewhere), and these procedures also were eliminated from the analysis. The present study includes only those patients in whom a Nissen fundoplication was performed for medically recalcitrant symptoms of reflux without stricture and in whom the fundoplication was left in the abdomen. Ninety-two such operations were performed in 89 patients. Three patients originally operated on by us required reoperation because of failure of the original procedure to relieve reflux symptoms permanently. Among these 92 operations, 25 were performed in patients who had been operated on previously; 14 had undergone one previous procedure, 7 had two previous operations, and 4 had had three previous operations and required a fourth operative procedure. The operation was a first one for 67 of these patients. Sixty-one patients in whom operation was performed for reflux secondary to a hypotensive LES also had a sliding esophageal hiatus hernia. In 26 patients no obvious cause for the physiological abnormality was present, whereas in 5 patients the LES had been rendered hypotensive by surgical or other manipulative maneuvers. In addition to routine studies, preoperative evaluation included radiographic examination of the upper gastrointestinal tract, endoscopy, and, in most patients, esophageal manometry. An abdominal approach was employed in 74 of these patients, and 18 underwent thoracotomy. Additional procedures were required in fifteen instances. Splenectomy was necessary in 5 patients. Incision and partial excision of a Schatzki ring were performed in 3 patients. Three had biliary procedures, including two cholecystectomies and one common bile duct exploration. Two patients who had previously had subtotal gastrectomy required fashioning of a jejunal pouch to act as a reservoir. One patient with a duodenal ulcer underwent parietal cell vagotomy at the time of fundoplication, and another had concomitant repair of a paraesophageal hiatus hernia. Methods of Assessment Clinical evaluation of the results of surgical intervention was based on careful assessment of the patient s condition during the year after operation either by personal interview or by letter. Patients were asked to respond to specific questions about their health, including their own evaluation of the overall result, their ability to belch, and the presence or absence and frequency of dysphagia, heartburn, regurgitation, postprandial bloating, diarrhea, and flatulence. The final overall assessment was determined by the physician and in practically every instance was of a lower grade than the patient s estimate of the result of the operation. An excellent result required that the patient be asymptomatic; a good result, that the patient have only occasional gastrointestinal symptoms; and a fair result, that the patient be improved compared with his or her preoperative status but remain symptomatic because of reflux or other symp.oms. A poor result indicated that the patient had persistent or recurrent symptoms or that other symptoms had developed that mitigated against a successful result. Results were tabulated according to the success of the operation in correcting reflux symptoms and to the overall result of the operation, which included not only control of reflux but the presence or absence of other symptoms that had not been present previously and might logically be related to the operative procedure itself. The technique of esophageal manometric examination was similar to that previously reported [4]. A constantly infused method was employed that used a triple-lumen catheter* with 5-cm spacings between the catheter openings circumventionally oriented to one another at intervals of 120 degrees. A special system* provided the constant infusion of degassed water at a rate of 1 ml per minute, and changes in pressure were recorded on a Hewlett-Packard multichannel recording device (model 88708) by Hewlett-Packard external transducers (model 1290C) I Pressures were recorded graphically and calibrated to create a 1-cm deflection for each 10 mm Hg. Resting intragastric pressure was used as a baseline from which pressures of the LES were measured. Maximal end-inspiratory and end-expiratory pressures were av- Amdorfer Medical Specialties, Greendale, WI.

3 389 Ellis and Crozier: Reflux Control by Fundoplication eraged to obtain the mean maximal pressure of the LES. Of the several values obtained, the highest one was selected in an effort to minimize inconsistencies of recording. Early postoperative studies were performed between ten days and a month after operation. While numerous late postoperative studies, that is, those performed more than a month after operation, were available, only the results of the most recent ones were used. Preoperative and early and late postoperative data were analyzed statistically by paired two-tailed t tests. Results Clinical Observations Recovery from the operation was usually rapid and without complication, although it is not uncommon for patients to experience mild dysphagia for a few days during the second postoperative week. In only 8 of the 92 patients was this of sufficient severity to require temporary adjustment of the patient s diet. Postoperative complications numbered 10, an incidence of 10.9%. Four patients experienced ileus. Four had wound infections; in 1 patient, wound dehiscence required resuture and subsequent repair of a ventral hernia. Two other patients experienced urinary tract infections. Clinical results are based on 82 operations. Eight patients were operated on too recently for evaluation, and 2 have been lost to follow-up review. During the followup period 4 other patients died of unrelated causes, including coronary artery disease in 2 patients, carcinoma of the breast in 1 patient, and homicide in 1. Since the clinical status of these patients at the time of death was known, they are included in the results. The followup interval ranged from 1 to 13 years and averaged 5% years. The operation was successful in permanently controlling the symptoms of reflux in 74 of the 82 patients (90.2%). The success rate for reoperations was similar to that for primary procedures. Among the patients for whom information was available, 80% were able to belch, with the percentage slightly smaller for those who underwent reoperation than for those for whom fundoplication was a primary procedure (Table 1). In evaluation of the overall result of the operation, both its success in preventing reflux and the patient s overall condition, defined by presence or absence of other symptoms, were assessed as already described. The success rate of the operation based on these criteria was lower than if only its effect on reflux had been considered. Thus, 67 of 82 patients under review had satisfactory results (81.7%). Fifteen of them were considered to have had poor results. The overall success rate was higher among patients who underwent operation for the first time than among those who underwent reoperation; the operation was classified as a failure in only 15.5% of the former group of patients and in 25% of the latter. When the poor results were analyzed, persistent or Table 1. Clinical Assessment of Operative Results (N = 82) Primary Total Operation Reoperation Assessment No. % No. % No. % Reflux control 74/ / / Ability to belch 61/ / / Overall result Excellent Good zi} 81.7 ;:} 84.5 i} Fair 12 Poor Total recurrent reflux was the most common reason for failure (8 patients). Three of these patients were operated on again with good results. One patient with the gas-bloat syndrome and 1 patient with a postoperative paraesophageal hernia with similar symptoms required reoperation, and both are now free from symptoms. Thus, we reoperated on 5 of 15 patients who had had poor results, all with success. Of the remaining 5 patients with poor results, 1 had severe diarrhea despite treatment with the usual medications, another had dysphagia that required dilations, and the remaining 3 complained of multiple, seemingly unrelated symptoms that defied analysis as to their cause. Because recurrent reflux and symptoms of gas-bloat syndrome were thought to be caused by too tight a wrap, the caliber of the indwelling stent was changed in 1977 from 32F to 42F. There has been only 1 poor result of operation since then, and that was in the patient with intractable diarrhea. Manometric Measurements Esophageal manometry was carried out preoperatively in all but 15 of the patients under review (Table 2). While there was considerable variation in the measured values for amplitude and length of the LES before and after operation, comparison of preoperative pressures with early and late postoperative values showed a significant difference (p < 0.001) (Fig 2). The same was true when the preoperative and postoperative lengths of the LES were compared (Fig 3). Although ph reflux testing was not routinely performed, it was negative preoperatively in only one instance and only rarely positive in the period after operation. Late studies were performed too infrequently for results to be significant. These results include data for all of the patients studied and are not as meaningful as paired data. Paired data for analysis of the amplitude and length of the LES were available for 58 and 55 patients, respectively. Statistical analysis of these data revealed identical findings with a significant difference between preoperative values and early and late postoperative values (p <

4 390 The Annals of Thoracic Sur,gery Vol 38 No 4 October 1984 Table 2. Manometric Measurementsa Variable No. of Patients Result an 5 4 SD I I. PREOPERATIVE FINDINGS Amplitude of LES (mm Hg) Length of LES (cm) t 0.1 ph Reflux test 32 Positive 31 Negative 1 EARLY POSTOPERATIVE FINDINGS Amplitude of LES (mm Hg) ? 0.6b Length of LES (cm) f O.lb ph Reflux test 25 Positive 5 Negative Fig 3. Length of the lower esophageal sphincter before and after Nissen fundoplication. (SD = standard deviation;* = p < compared with preoperative length.) LATE POSTOPERATIVE FINDINGS Amplitude of LES (mm Hg) f 1.5b Length of LES (cm) b ph Reflux test 7 Positive 2 Negative 5 "Where appropriate, values shown are mean? standard error of the mean. bsi gnificance:.' p < compared with preoperative values. L LES = lower esophageal sphincter. mm Hg *O T T Preop Late Postop Fig 2. Amplitude of pressure of the lower esophageal sphincter before and after Nissen fundoplication. (SD = standard deviation; * = p < compared with preoperative amplitude.) 0.001). No difference was detected between the effectiveness of the operative procedure in restoring manometric measurements to normal levels when patients who underwent operation for the first time were compared with those who underwent reoperation. In an effort to identify manometric findings that might distinguish between patients having good and excellent results and those with fair and poor results, no significant difference in either the amplitude of pressure or the length of the high pressure zone could be identified. However, among the 8 patients who experienced persistent or recurrent reflux as a reason for their poor result, all but 1 of the 7 for whom data were available had suboptimal levels of pressure of the LES. Comment Of the several antireflux procedures currently available, the Nissen fundoplication has received the widest acceptance throughout the world. This is not surprising, since considerable clinical [5-71 and experimental [8, 91 evidence suggests that this procedure surpasses other antireflux techniques in providing effective control of reflux. As with any relatively new procedure that is widely adopted by many surgeons with different backgrounds and levels of technical skill, a number of postoperative complications have been recognized that are believed by some to restrict the usefulness of the operation [lo]. These complications include disruption of the fundoplication [ll], paraesophageal hernia [12], gastric ulceration [13-151, the gas-bloat syndrome [16], postoperative dysphagia [17], gastric obstruction caused by a "slipped" Nissen fundoplication [ 181, and perforation of the wrap with fistula formation [19]. That these complications are rare if the operation is properly applied is evident from the results we are reporting and from large clinical reports of others [20, 211 that document improvement rates of about 90%. Concern has been expressed regarding the durability of the procedure because late results in some series exhibit an increase in the number of poor results [22, 231. This has not been true in our experience, in which an initially satisfactory result was maintained over an average follow-up interval of nearly 6 years. An exception is a patient in whom recurrent reflux developed 3 years after operation and required reoperation. The major cause of poor results in our patients was too tight a wrap, which resulted in persistent or recurrent reflux in 8 patients and serious gas-bloat symptoms in 1

5 391 Ellis and Crozier: Reflux Control by Fundoplication patient. Since enlarging the size of the wrap from 32F to 42F, we have not noted these complications. Others [24] have emphasized the necessity of fashioning a loose wrap. Four-fifths of our patients are able to belch, but most of them, if questioned, admit to excessive flatulence. Disillusionment with the classic Nissen fundoplication is becoming widespread and has led to the introduction of many modifications, including varying the extent of the wrap [25, 261, anchoring the plication sutures to the preaortic fascia [27, 281, and concomitant performance of parietal cell vagotomy [29, 301 and pyloroplasty [23]. Some have actually abandoned the conventional procedure entirely and use it only in conjunction with a Collis gastroplasty [31]. Our results certainly do not justify such pessimism because control of reflux was provided in 90% of our patients, and overall patient satisfaction was achieved in approximately 85% of those who underwent operation for the first time. The main technical recommendation we would make is that the wrap be fashioned over a large-bore indwelling stent to ensure a loose fundoplication. The manometric results we are reporting disclose a doubling of the amplitude and length of the LES both early and late after operation but do not reflect the tightness of the wrap. When reflux persists or recurs, however, the pressure of the LES usually is in the hypotensive range. For good results to be achieved with the Nissen fundoplication, it is essential that an accurate diagnosis be made by objective documentation of the presence of reflux. Our data do not bear on this point, but our experience with 38 patients who underwent reoperation for failed Nissen operations, most of which were originally performed elsewhere, is pertinent [32]. The most common symptom requiring reoperation was dysphagia, and in most of these patients the cause was an underlying motility disorder characterized by lack of a forceful peristaltic deglutitive sequence in the body of the esophagus. Our experience with the Nissen fundoplication suggests that if patients are properly selected and the wrap is loosely fashioned, permanent control of reflux can be achieved in 90% of patients with relatively few complications and with a high degree of patient satisfaction. We wish to acknowledge the assistance of Elton Watkins, Jr., M.D., of the Lahey Clinic Division of Research in the statistical analysis of data. References 1. Nissen R: Eine einfache Operation Beeinflussung der Refluxoesophagitis. Schweiz Med Wochenschr 86:590, Ellis FH Jr, El-Kurd MF, Gibb Sl? The effect of fundoplication on the lower esophageal sphincter. Surg Gynecol Obstet 143:1, Ellis FH Jr, Gibb Sl? Fundoplication for hypotensive lower esophageal sphincter. Hosp Pract 9:80, Ellis FH Jr, Leonardi HK, Dabuzhsky L, Crozier RE: Surgery for short esophagus with stricture: an experimental and clinical manometric study. Ann Surg 188:341, DeMeester TR, Johnson LF, Kent AH: Evaluation of current operations for the prevention of gastroesophageal reflux. Ann Surg 180:511, Sillin LF, Condon RE, Wilson SD, Worman LW: Effective surgical therapy of esophagitis: experience with Belsey, Hill, and Nissen operations. Arch Surg 114:536, Ferraris VA, Sube J: Retrospective study of the surgical management of reflux esophagitis. Surg Gynecol Obstet 152:17, Leonardi HK, Lee ME, El-Kurd MF, Ellis FH Jr: An experimental study of the effectiveness of various antireflux operations. Ann Thorac Surg 24:215, Bombeck CT, Coelho RG, Castro VA, Nyhus LM: An experimental comparison of procedures for the operative correction of gastroesophageal reflux. Bull SOC Int Chir 30:435, Negre JB: Post-fundoplication symptoms: do they restrict the success of Nissen fundoplication? Ann Surg 198:698, Hill LD, Ilves R, Stevenson JK, Pearson JM: Reoperation for disruption and recurrence after Nissen fundoplication. Arch Surg 114:542, Balison JR, Macgregor AM, Woodward ER Postoperative diaphragmatic herniation following transthoracic fundoplication: a note of warning. Arch Surg 106:164, Bremner CG: Gastric ulceration after a fundoplication operation for gastroesophageal reflux. Surg Gynecol Obstet 148:62, Herrington JL Jr, Meacham PW, Hunter RM: Gastric ulceration after fundic wrapping: vagal nerve entrapment, a possible causative factor. Ann Surg 195:574, Maher JW, Cerda JJ: The role of gastric stasis in the genesis of gastric ulceration following fundoplication (case report). World J Surg 6:794, Woodward ER, Thomas HF, McAlhany JC: Comparison of crural repair and Nissen fundoplication in the treatment of esophageal hiatus hernia with peptic esophagitis. Ann Surg 173:782, Rossman F, Branntigan CO, Sawyer RB: Obstructive complications of the Nissen fundoplication. Am J Surg 138:860, Olsen RC, Lasser RB, Ansel H: The slipped Nissen (abstract). Gastroenterology 70:924, Burnett HF, Read RC, Morris WD, Campbell GS: Management of complications of fundoplication and Barrett s esophagus. Surgery 82:521, Rossetti M, Hell K: Fundoplication for the treatment of gastroesophageal reflux in hiatal hernia. World J Surg 1:439, Polk HC Jr: Indications for, technique of, and results of findoplication for complicated reflux esophagitis. Am Surg 44:620, Brand DL, Eastwood IR, Martin D, et al: Esophageal symptoms, manometry, and histology before and after antireflux surgery: a long-term follow-up study. Gastroenterology 76:1393, Negre JB, Markkula HT, Keyrilainen 0, Matikainen M: Nissen fundoplication: results at 10-year follow-up. Am J Surg 146:635, Donahue PE, Bombeck CT: The modified Nissen fundoplication: reflux prevention without gas bloat. Chir Gastroenterol 11:15, 1977

6 392 The Annals of Thoracic Surgery Vol 38 No 4 October Menguy R: A modified fundoplication which preserves the ability to belch. Surgery 84:301, Guarner V, Martinez N, Gaviiio JF: Ten year evaluation of posterior fundoplasty in the treatment of gastroesophageal reflux: long-term and comparative study of 135 patients. Am J Surg 139:200, Cordiano C, Querci Della Rovere G, Agugiaro S, Mazzilli G: Technical modification of the Nissen fundoplication procedure. Surg Gynecol Obstet 143:977, Kaminski DL, Codd JE, Sigmund CJ: Evaluation of the use of the median arcuate ligament in fundoplication for reflux esophagitis. Am J Surg 134:724, Jordan PH Jr: Parietal cell vagotomy facilitates fundoplication in the treatment of reflux esophagitis. Surg Gynecol Obstet 147:593, Jones NA, Anders CJ: A new approach to the surgical treatment of reflux oesophagitis. Ann R Coll Surg Engl 61:48, Piehler JM, Payne WS, Cameron AJ, Pairolero PC: The uncut Collis-Nissen procedure for esophageal hiatal hernia and its complications. Probl Gen Surg 1:1, Leonardi HK, Ellis FH Jr: Complications of the Nissen fundoplication. Surg Clin North Am 63:1155, 1983 Discussion DR. TOM R. DEMEESTER (Omaha, NE): Drs. Ellis and Crozier have shown that symptoms of heartburn and regurgitation can be controlled by a classic Nissen fundoplication in 90% of patients. Based on personal experience with 108 similarly selected patients, I concur with their findings. The lesson learned from this experience, however, indicates that these results can be improved to more than 95%-a result that is necessary if surgical intervention is to compete effectively in the treatment of this disease. To do so requires attention to two factors. First is the selection of candidates for operation. Not all heartburn or esophagitis is caused by incompetence of the cardia. It can be caused by a decrease in esophageal clearance of a physiological reflux episode or the slow passage of medication having a low ph such as aspirin, vitamin C, or antibiotics, of which tetracycline is the most notorious example. Similarly a decrease in gastric emptying from duodenal, pyloric, or gastric abnormalities can cause pathological reflux across a mechanically competent cardia. Neither of these conditions are benefited by an antireflux procedure. I ask Dr. Ellis if he believes that a more detailed preoperative evaluation with 24hour ph monitoring to objectively prove the presence of reflux and radioisotope studies to assess esophageal clearance and gastric emptymg would improve his results and help sort out the cause of the preoperative symptoms that persisted postoperatively in 19% of his patients. Second, surgical technique has a major effect on results when performing operations to improve the function of an organ, as opposed to its removal. I concur with Dr. Ellis that one of the major causes of a poor result is too tight a gastric wrap. Either too tight or too long a gastric wrap can cause enough postoperative dysphagia or decreased esophageal clearance to change the patient s eating habits. It is important to realize that not all patients will complain of dysphagia; some will adjust their eating habits to avoid the symptoms. Simply asking about dysphagia can give a false impression as to the incidence of postoperative swallowing difficulties. This point is exemplified from my experience with the classic 4-cm Nissen fundoplication. Initially, there is a marked decrease of acid exposure to the esophagus following operation but with time, the long wrap begins to interfere with the clearance of physiological reflux episodes. This was reflected clinically by patient complaints about the necessity to eat slowly. In an attempt to prevent the delayed unphysiological condition, over the years I have progressively shortened the gastric wrap so that now it is only 1.5 cm long. Dr. Ellis, based on your experience, do you also get the impression that it might be wise to shorten the length of the gastric wrap? DR. MILTON v. DAVIS (Dallas, TX): I wish to mention that I have had a similar experience with this procedure, with mostly favorable results. I have done the same type of work that Drs. DeMeester and Ellis have described in about the same number of patients as reported by Dr. Ellis, but with a different set of indications-due, perhaps, to a different patient referral source. Approximately half of the patients on whom I have operated have had bronchopulmonary symptoms. However, these must be both severe and recurrent for the patient to be considered for operation. DR. ELLIS: I thank Drs. DeMeester and Davis for their comments. I agree that it is important to obtain objective evidence that what patients complain of really is reflux. Nearly all of our patients were evaluated preoperatively by esophageal manometry and the standard acid reflux test. We have not done 24- hour monitoring nor have we studied gastric emptying and esophageal clearance. Another aspect of the preoperative evaluation that I think is equally important is to avoid performing an antireflux operation in patients who suffer from a motility disorder other than a hypotensive LES. Dr. Leonardi and I recently reported on 38 patients who had undergone a Nissen wrap (fortunately, most of these were done at other institutions) who required reoperation. The major cause of failure was dysphagia, and the major cause of dysphagia was an underlying motility disorder characterized by lack of peristalsis-either achalasia, scleroderma, or diffuse spasm. So, an accurate preoperative diagnosis not only of the presence or absence of reflux but of some other possible disorder is essential if one is to achieve good results. The length of the wrap may be important, although our studies have not been able to identify any difference in amplitude or length of the high pressure zone postoperatively between those patients who have dysphagia or gas bloat and those who do not. There is some experimental evidence suggesting that a shorter wrap may be important, and the length of our wrap is now in the neighborhood of 2 to 3 cm. Dr. Davis reminds us that bronchopulmonary complaints are often the presenting symptoms of patients with reflux. I quite agree, since some of our patients were originally diagnosed as having asthma although they actually had nocturnal regurgitation and would wake up coughing and wheezing. Such symptoms are clearly manifestations of gastroesophageal reflux.