- Email: [email protected]
Total Fundoplication Gastroplasty (Nissen Gastroplasty): Five-Year Review
Total Fundoplication Gastroplasty (Nissen Gastroplasty): Five-Year Review Robert D. Henderson, M.B., F.R.C.S.(C), and Gary V. Marryatt ABSTRACT Gastroplasty was introduced by Collis in 1961 and has undergone several modifications. The combination of total fundoplication with gastroplasty was reported in 1977 and referred to as total fundoplication gastroplasty; however, the term Nissen gustroplusfy also is commonly used. This article is an extension of the original 1977 report and, to our knowledge, represents the first 5year review of total fundoplication gastroplasty. Three hundred fifty-one consecutive patients with intractable reflux were preoperatively evaluated by history, radiographic studies, manometric studies with determination of pH, and esophagogastroduodenoscopyprior to surgical management by total fundoplication gastroplasty. There were no operative deaths. Follow-up averaged 6.5 years with an effective clinical review available for 95.4% of the patients, radiographic studies for 92.3%, and manometric studies with pH evaluation for 70.7%. Among the 335 patients with 5 or more years of follow-up, 93.1% had excellent results with normal eating and no investigative evidence of recurrence of reflux, 4.0% had mild residual symptoms, and 2.9% had persistent or recurrent symptoms. With this technique, the problems of overcompetence and dysphagia are substantially reduced because the completion fundoplication is tailored to a length of 1 cm while anatomical stability is maintained with a long intraabdominal segment.
Gastroplasty was reported in 1961 by Collis [l] as a method of esophageal lengthening in patients with esophageal stricture. Although the operation produced a stable intraabdominal segment, the long-term followup was associated with a 59% incidence of moderate to severe radiographic reflux. The procedure was modified in 1971 [2] with the introduction of the Belsey gastroplasty. Long-term followup of patients undergoing this operation indicates satisfactory results, but a high incidence of continued or recurrent reflux has been reported [3], particularly in patients requiring surgical revision [4]. In general, though, a low incidence of anatomical recurrence is reported for gastroplasty [5]. Total fundoplication (Nissen) From the Women's College Hospital, University of Toronto, Toronto, Ont, Canada. Presented at the Twentieth Annual Meeting of The Society of Thoracic Surgeons, San Antonio, TX, Jan 23-25, 1984. Address reprint requests to Dr. Henderson, Chief of Surgery, Women's College Hospital, 76 Grenville St, Toronto, Ont, Canada M5S 182.
74
gastroplasty was designed to simultaneously incorporate anatomical stability and provide the benefits of the more effective reflux control inherent in total fundoplication [6].
Material and Methods Three hundred fifty-one consecutive patients were treated for symptoms of intractable reflux by total fundoplication gastroplasty. There were 123 male and 228 female patients. The average age was 45.5 years (range, 17 to 75 years). A preoperative history was obtained using a prepared history format, and all patients underwent preoperative evaluation by radiography, flexible fiberoptic esophagogastroduodenoscopy, and manometry with pH determination. Three hundred twenty-one patients also had acid perfusion studies. The manometric studies were conducted using three polyethylene tubes with side holes. Infusion was maintained at 6.75 ml per minute by means of a modified Harvard pump. Data were recorded with a Honeywell 1508 Visicorder. A Beckman pH probe and recorder were used to assess pH changes. All follow-up radiographic studies were conducted in the same department. Water siphon studies were routinely performed for evaluation of reflux.
Operative Technique A transthoracic approach was used for primary hernia repair in the majority of patients. A thoracoabdominal incision was used in patients with previous esophagogastric repair and in those requiring simultaneous abdominal and thoracic surgical procedures. The transthoracic approach was through the seventh or eighth interspace. Mobilization of the esophagogastric junction was completed medially to the aortic arch and along its aortic surface to the inferior pulmonary vein. The esophagogastric fat pad was excised. Posterior crural sutures were placed, and care was taken to preserve the peritoneal covering to add strength to the repair. The sutures were tied at the completion of the gastroplasty for loose approximation of the crurae. Gastroplasty tubes were cut using snugly fitting rightangle DeBakey clamps over a 60F Maloney bougie. The tubes were closed with an interlocking 2-0 chromic suture and continuous 3-0 silk. The fundus of the stomach was reapproximated to the gastroplasty tube and distal 1.5 cm of the esophagus with 2-0 silk mattress sutures through the gastroplasty suture lines. This gives very strong approximation and is important for the anatomical stability of the repair. Total fundoplication was achieved by the lateral and medial
75 Henderson and Marryatt: Total Fundoplication Gastroplasty
Fig 1 . Prepared gastroplasty tube with the new gastric fundus sutured to the 5-cm tube and distal 1.5 cm of esophagus. Fol[oaiing fundoplication, the stomach is reduced below the diaphragm and the crural sutures are tied.
spread of the new gastric fundus. The completion fundoplication was restricted in length to 1 cm and was sutured with three 2-0 silk mattress sutures. In patients with poor esophageal motor function or previous gastric repair, the length of the completion fundoplication was reduced to 0.5 cm or less. Before reduction of the gastroplasty tube into the abdomen, two sutures were placed from the fundus to the undersurface of the diaphragm; these were tied following reduction, as were the posterior crural sutures (Figs 1, 2). When a thoracoabdominal approach was employed, the technique of preparation of the gastroplasty tube was similar; however, four to five short gastric vessels were divided for fundic mobilization. The total fundoplication was passed clockwise behind the gastroplasty tube and sutured to the tube and distal 1.5 cm of esophagus before the performance of a 1-cm completion fundoplication. Several points should be emphasized concerning the technique. The intraabdominal segment is 6.5 cm long5 cm of gastroplasty tube and 1.5 cm of distal esophagus. Overreduction of the esophagus should be avoided. Although the completion fundoplication is very loose, the length of approximated stomach must be carefully tailored to 1 cm. An excessively long completion fundoplication is the cause of overcompetence, so the completion fundoplication must be tailored to reduce its length in patients with poor esophageal motor function (Fig 3).
Results Among the 351 patients treated for reflux by total fundoplication gastroplasty, the average follow-up was 6.5 years (range, 5 to 8.5 years).
Preoperative Evaluation Reported symptoms, obtained from the preoperative history, are listed in the Table.
Fig 2. Following resuturing of the fundus to the gastroplasty tube and distal esophagus, three gastrogastric sutures are placed, approximating the stomach. A total lax fundoplication is achieved and carefully tailored to prevent the fundus from accidentally approximating and increasing the length of the fundoplication.
Radiographic studies demonstrated a hiatal hernia in 222 patients and reflux alone in 48. Sixty-eight patients had normal studies. No studies were available for review of the remaining 13 patients. On manometric examination, the average tone in the high-pressure zone (HPZ) was 12.26 cm H 2 0 and disordered motor activity (DMA) in the lower esophagus averaged 46.2%. A manometric diagnosis of hiatal hernia was made in 68.9% of the patients. Acid perfusion studies were positive in 295 (91.9%) of 321 studies. The major risk factors in operation for reflux include previous gastric or esophageal repair, severe ulceration or stricture, and scleroderma. Ninety-five patients (27.1%)had had a previous esophageal repair, 25 (7.1%) had had an esophageal repair combined with a gastric procedure, and 10 (2.9%) had had a previous gastric operation. Thus, a total of 130 patients (37%) had had previous surgical intervention. Severe ulceration or stricture was documented endoscopically in 68 patients (19.4%), while scleroderma with an adynamic esophagus was present in 5 (1.4%). One or more of these factors was present in 179 patients (51%).
Postoperative Evaluation Follow-up data are available on 335 (95.4%) of the 351 patients. In this group, all were evaluated by history, 324 (96.7%) by radiography, and 248 (74%)by manometry and pH studies. There were no operative deaths. Nine patients died late of unrelated causes. The major related complication was esophageal fistula, which occurred in 3 patients (0.9%). All of these fistulas closed early and spontaneously; two of them were in patients with recurrent hernia and the other, in a patient who had primary repair, was due to breakdown of staples. Other major complications included three instances of bowel obstruction (two
76 The Annals of Thoracic Surgery Vol 39 No 1 January 1985
Esophagus 1 112 cm 1 cm
Gastroplasty 5 cm
Fig 3. lntraabdominal length and fundoplication length are prepared by separate suturing techniques. When the completion fundoplication is reduced to one stitch in length, gastroesophageal sutures are added to maintain the stability of the fundoplication.
resulting from old adhesions and one from a hematobezoar) and one instance of wound dehiscence. There was one late ventral hernia in an alcoholic patient with poor tissue strength. Two patients had moderate wound infections. Three categories of clinical results are defined. Patients with excellent results include those who are asymptomatic or have only minor residual symptoms. Patients with moderate results are those who have notable residual symptoms but are improved following operation. Patients with poor results are those with major residual symptoms or in whom symptoms have developed during 5 or more years of follow-up. Among the 335 patients for whom follow-up data were available, 312 (93.1%)had excellent results, 13 (4.0%)had moderate residual symptoms, and 10 (2.9%)had poor results. The 13 with moderate residual symptoms had a variety of problems. Four had nonspecific complaints, for which an exact cause could not be found. Two patients had gas bloat syndrome. Two had bile gastritis and 5 , moderate persistent dysphagia. Seven of the 13 with residual problems had had previous gastric surgical procedures. Ten patients requiring surgical revision have been analyzed. Three had radiographic and symptomatic recurrence of reflux, and 1 had an intussusception recurrence that, while not recognized radiographically, occurred with increasing dysphagia. Two patients had a traumatic esophageal diverticulum with dysphagia, and 3 had dys-
phagia due to an overcompetent repair. One experienced severe gas bloat. In this group, 7 patients had had previous surgical intervention, including 3 who had both gastric and esophageal procedures. Two had major psychiatric overlay affecting interpretation of symptoms. Results of radiographic studies, including water siphon testing for reflux, are available for 324 of 351 patients (92.3%).Radiographic studies were conducted routinely five and one-half months after operation and
Preoperative Symptoms in 351 Patients Undergoing Total Fundoplication Gastroplasty Symptom Heartburn Pain" Epigastric Retrosternal Back Arm Reflux Aspiration Eructation Hiccough Excess salivation Dysphagia Cricopharyngeal Gastroesophageal Stricture Total motor dysphagia "Average duration was 4.9 years.
No. of Patients
Percent
344
98.0
332 321 142 43 282 161 249 91 81
94.6 91.5 40.5 12.3 80.3 45.9 70.9 25.9 23.1
140 200 36 212
39.9 57.0 10.3 60.4
77 Henderson and Marryatt: Total Fundoplication Gastroplasty
TONE PREOPS
0
CM H20
30
POST-OPS
I
20
10
TONE RANGE
0-6
7-10
11-15
16-20
21+
CM H20
AV.TONE
5
9.1
12.5
17.8
26
CM H20
43
70
61
52
17
PATIENT NO.
Fig 4. Preoperative tone in the high-pressure zune (HPZ). Those with low HPZ tone had a greater average tone increase than those with higher preoperative tone. Following total fundoplication gastroplasty, HPZ tune in all patients was within the normal range of 15 to 20 cm HZO.
were repeated only if clinically indicated. Two patients had an asymptomatic trace of reflux, 2 had a traumatic diverticulum, and 3 had a radiographically demonstrated anatomical recurrence (one recurrence was not recognized). Follow-up manometric and pH studies were conducted in 243 of the 351 patients (69.2%).These studies were conducted five and one-half months after operation. The average HPZ tone increased postoperatively to 18.1 cm H 2 0 (+47.6%), and average lower esophageal DMA decreased to 29.1% (-37%). The actual increase in HPZ tone depended on the preoperative tone (Fig 4). In patients with a low preoperative HPZ tone of 0 to 6 cm H20, the tone increase averaged 240%. As the preoperative HPZ tone increased, the percentage of tone increase became smaller. In a small group of patients with high preoperative HPZ tone, this value fell after surgical correction of reflux. The average postoperative tone in all groups was within the normal range of 15 to 20 cm H20.
Comment The operative technique of gastroplasty combined with total fundoplication was developed in 1975, and early results were reported in 1977 [7]. To our knowledge, the present report is the first 5-year review of total fundoplication gastroplasty, although undoubtedly others will follow. Several variations on this operation have already been described, including an uncut total fundoplication gastroplasty [8] and alternative methods of fundoplication [9, 101. The Collis-Nissen procedure described in 1978 by Orringer and Sloan [lo] is similar in principle, but the method of fundoplication varies considerably in fixation and length. Comparison of results of these procedures depends on both the severity of preoperative disease and the methods of objective and subjective analysis after operation. Preoperatively, 37% of our patients had had previous esophageal or gastric procedures and 10% had had gastric operations. In addition, 19.4% had ulceration or
stricture and 1.4%, scleroderma. These complicated cases account for most of the patients with moderate results or operative failure. Excellent results were reported only for patients with minimal or no symptoms and no objective radiographic, manometric, or pH evidence of failure. Those labeled as having moderate results are much improved and have no evidence of recurrence, so the decision to classify them in the moderate category is inevitably subjective. In the group with poor results are all patients whose symptoms were not improved or who experienced late failure of the operation. The overall incidence of excellent results (93.1%) is quite acceptable, and the failure rate of 2.8% is low. The rate of actual anatomical recurrence of reflux (1.1%)is very low compared with rates reported for other methods of gastroplasty. Radiographic studies are of limited value because with total fundoplication, severe reflux is rare in the absence of anatomical recurrence. In the 3 patients with anatomical recurrence, there was no reflux; the patient with an intussusception recurrence was considered radiographically normal. Manometric studies with determination of pH are of value in assessing reflux and showing the resolution of reflux-induced secondary motor damage. The motor improvement detected is better with total fundoplication gastroplasty than with partial fundoplication procedures. Preventing reflux is easy; however, little attention has been paid to methods of calibrating the fundoplication to avoid overcompetence. Therefore, total fundoplication gastroplasty has been designed to avoid overcompetence and dysphagia and to allow the patient to burp and vomit. There are two major components of the total fundoplication gastroplasty: anatomical stability and the creation of a tailored degree of competence that avoids dysphagia. Stability is maintained by using a gastroplasty tube, which is held below the diaphragm together with 1.5 cm of distal esophagus. Standard hernia repairs require 4 cm of esophagus for fundoplication so that 2.5 cm is spared and there is no tension on the esophagus following reduction. When severe shortening is present, the HPZ may remain in the chest and competence can be achieved by fundoplication of the gastroplasty tube below the diaphragm. The major anchoring sutures are in the stomach, adding considerable strength to the repair. The left gastric artery fixes the gastroplasty tube below the diaphragm. A recurrence rate of 1%is consistently reported with gastroplasty and almost certainly reflects the anatomical stability of the repair. Avoidance of overcompetence is the most important feature of the design of total fundoplication gastroplasty. Recognition of these problems has evolved with growing experience. The application of this knowledge in the past few years has made overcompetence and dysphagia problems of the past. None of the 300 patients operated on from 1980 to 1983 has required revision for dysphagia. The length of the completion fundoplication is deter-
78 The Annals of Thoracic Surgery Vol 39 No 1 January 1985
mined by direct gastrogastric suturing. Completion length is 1 cm in the majority of patients, but is reduced to 0.5 cm in patients with severe DMA (61 to 80%)in the lower esophagus. In the presence of very severe DMA (81 to 100%) in patients with previous gastric resection or scleroderma, the wrap length is kept to less than 0.5 cm and the fundoplication is stabilized by adding sutures to the gastroplasty tube and esophagus (see Fig 3). This type of fundoplication, in addition to minimizing dysphagia, allows 85% of patients to eructate normally. Ability to vomit is difficult to assess as most individuals have no desire to vomit.
References 1. Collis JL: Gastroplasty. Thorax 16:197, 1961 2. Pearson FG, Langer B, Henderson RD: Gastroplasty and Belsey hiatus hernia repair: an operation for the management of peptic stricture with acquired short esophagus. J Thorac Cardiovasc Surg 61:50, 1971 3. Orringer MB, Sloan H: Complications and failings of the combined Collis-Belsey operation. J Thorac Cardiovasc Surg 74726, 1977 4. Henderson RD: The gastroplasty tube as a method of reflux control. Can J Surg 21:264, 1978 5. Pearson FG, Cooper JD, Nelms JM: Gastroplasty and fundoplication in the management of complex reflux problems. J Thorac Cardiovasc Surg 76:665, 1978 6. DeMeester TR, Johnson LF, Kent AH: Evaluation of current operations for the prevention of gastroesophageal reflux. Ann Surg 180:511, 1974 7. Henderson RD: Reflux control following gastroplasty. Ann Thorac Surg 24:206, 1977 8. Evangelist FA, Taylor FH, Alford JD: The modified CollisNissen operation for control of gastroesophageal reflux. Ann Thorac Surg 26:107, 1978 9. Bingham JA: Hiatus hernia repair combined with the construction of an anti-reflux valve in the stomach. Br J Surg 64:460, 1977 10. Orringer MB, Sloan H: Combined Collis-Nissen reconstruction of the esophagogastric junction. Ann Thorac Surg 25:16, 1978
Discussion (Ann Arbor, MI): I thank Dr. Henderson and Mr. Marryatt for the opportunity to review their manuscript. They are clearly to be credited with outstanding 5-year follow-up results in a large group of patients operated on for reflux. Having been charged with the task of critically reviewing this work, however, I wish to ask Dr. Henderson’s comments on several points. As many know, both Dr. Henderson and I have advocated a 360-degree fundoplication for reflux control after construction of the Collis gastroplasty. The gastroplasty reduces the volume of the remaining fundus available for the wrap by the amount used to fashion the tube. If the surgeon tries to use the residual elongated fundus for a Belsey-type repair, he or she is simply tacking the stomach back against the tube and can achieve only a limited fundoplication. This was the rationale for our abandonment of the Collis-Belsey combination and for our preference for passing the elongated gastric fundus behind the gastroplasty tube and performing a 360-degree Nissen-type wrap. In his operation, Dr. Henderson is actually reapproximating the gastric fundus to the gastroplasty tube, a technique reminiscent DR. MARK B. ORRINGER
of the combined Collis-Belsey procedure, and then performing the I-cm completion fundoplication. I am surprised that he has had no difficulty drawing the edges of the remaining fundus all the way around the tube. I wonder if this difficulty has been responsible for his progressively decreasing the length of the wrap, now down to only 1 cm. Second, it is clear that Dr. Henderson includes two quite different operations under the heading of total fundoplication gastroplasty. One involves no division of short gastric vessels, reapproximation of the fundus to the gastroplasty tube, a 1-cm fundoplication, and anchoring of the gastric fundus to the undersurface of the diaphragm a la the second row of sutures of the standard Belsey operation. The second operation requires division of four to five short gastric vessels, a true wraparound Nissen-type fundoplication like the one my colleagues and I perform, and apparently no anchoring of the stomach beneath the diaphragm. Dr. Henderson, how many patients underwent which operation? Is there a difference in the operative results between these procedures? Why is it necessary to divide short gastric vessels or anchor the wrap beneath the diaphragm on some occasions but not others? Finally, as I am sure Dr. Henderson will acknowledge, objective assessment of the efficacy of an antireflux operation with preoperative and postoperative manometry and acid reflux testing with the intraesophageal pH electrode has become a mandatory standard in our specialty. While Dr. Henderson indicates that all of his patients were evaluated preoperatively and 71% postoperatively with manometric and pH studies, the preoperative and postoperative pH reflux testing data have not been presented. Their absence is a serious omission when attempting to “sell” a relatively new antireflux procedure. It would also be helpful to have some idea of the relative frequency of adverse postoperative results following total fundoplication gastroplasty: dysphagia requiring dilation as well as intermittent untreated dysphagia; severe gas bloat syndrome as well as less troublesome, but still disturbing, early satiety; postvagotomy diarrhea; and postthoracotomy incisional pain. Again, I congratulate Dr. Henderson and Mr. Marryatt on their continued fine work. DR. NICHOLAS J. DEMOS (Jersey City, NJ): I wish to congratulate Dr. Henderson and Mr. Marryatt on their tremendous series and the evangelistic fervor with which they study this disease. Personally, I am committed to the uncut gastroplasty. I started studying the stapled, uncut gastroplasty plus total fundoplication more than 15 years ago in animal experiments. I performed successful canine experiments for 2 years, making all my mistakes in dogs. I have applied this technique for the last 14 years in human beings. My results are somewhat similar to those of Dr. Henderson, but I cannot understand why surgeons cut the stomach and reapproximate it. When one cuts the stomach and sutures it in two layers, more dissection and transection of short gastric vessels are necessary. This has resulted in splenectomies, ischemia, and leakage. In follow-up ranging from 1 to 13 years of almost 100 patients who had stapled, uncut gastroplasty, there has been no recurrence of reflux, no need for reoperation or splenectomy, and no fistula formation. One patient had a stricture that had to be dilated every six months for 6 years. Because of its simplicity and, the reduced morbidity it affords, the stapled, uncut gastroplasty with total fundoplication remains my operation of choice. DR. HENDERSON: I thank Dr. Orringer and Dr. Demos for their comments. There is a great deal of agreement among the three of us, and we have discussed these issues many times.
79 Henderson and Marryatt: Total Fundoplication Gastroplasty
Dr. Orringer's suggestion that two operations have been described is misleading. Basically, 32% of the patients had a thoracoabdominal approach for a recurrent hiatal hernia. I have described this method before. The thoracoabdominal approach permits dissection of scar tissue under direct vision and also allows a pyloromyotomy. The addition of pyloromyotomy is important because even if the nerves are well preserved, it is unlikely they will function normally in people with from one to five previous hernia repairs. There is no problem with fundus size, and creation of a total fundoplication is always possible even in patients with previous radical gastrectomies. We did start with a completion fundoplication of 2.5 cm and have deliberately reduced this to a 1cm wrap. This length is further reduced in patients with poor esophageal peristalsis. The key to a stable repair is to suture the
intraabdominal segment separately which in turn enables careful tailoring of the completion fundoplication. This approach almost totally eliminates dysphagia. As noted in our paper, pH testing was carried out both preoperatively and postoperatively. In terms of Dr. Demos's discussion of the cut versus the uncut gastroplasty, I do not think there is an absolute answer. Long-term studies are necessary. In my experience, a better fundoplication is achieved with a cut gastroplasty. gingham, who described the uncut gastroplasty in 1977, reported a 7% incidence of breakdown of staples, which one has to expect when stapling healthy stomach to healthy stomach. This has been a recognized problem with the transverse gastroplasty for obesity, and suturing is now added to stapling to increase stability in this procedure.
Notice from the Editor and the Editorial Board The Editor and members of the Editorial Board are grateful for the assistance given in manuscript review during the past year by the following individuals: Hendrick B. Barner, M.D. St. Louis, MO
George C. Kaiser, M.D. St. Louis, MO
W. Gerald Rainer, M.D. Denver, CO
Mr. Richard Beauchamp St. Louis, MO
A. Ercument Kopman, M.D. St. Louis, MO
Francis Robicsek, M.D. Charlotte, NC
David Bregman, M.D. Paterson, NJ
Ronald D. Leidenfrost, M.D. St. Louis, MO
Benson B. Roe, M.D. San Francisco, CA
John Codd, M.D. St. Louis, MO
Harold V. Liddle, M.D. Salt Lake City, UT
Charles L. Roper, M.D. St. Louis, MO
Lawrence H. Cohn, M.D. Boston, MA
Floyd D. Loop, M.D. Cleveland, OH
Stuart S. Sagel, M.D. St. Louis, MO
Jack G. Copeland, 111, M.D. Tucson, AZ
Gerald Medoff, M.D. St. Louis, MO
William F. Sasser, M.D. St. Louis, MO
James L. Cox, M.D. St. Louis, MO
Robert M. Mentzer, Jr., M.D. Charlottesville, VA
Mrs. Louise Sper Detroit, MI
Robert W. M. Frater, M.D. Bronx, NY
Mark B. Orringer, M.D. Ann Arbor, MI
Thomas L. Spray, M.D. St. Louis, MO
J. Alex Haller, Jr., M.D. Baltimore, MD
D. Glenn Pennington, M.D. St. Louis, MO
Clarence S. Weldon, M.D. St. Louis, MO
Alexis F. Hartmann, Jr., M.D. St. Louis, MO
Salvatore Pizzo, M.D. Durham, NC
Gastroplasty was reported in 1961 by Collis [l] as a method of esophageal lengthening in patients with esophageal stricture. Although the operation produced a stable intraabdominal segment, the long-term followup was associated with a 59% incidence of moderate to severe radiographic reflux. The procedure was modified in 1971 [2] with the introduction of the Belsey gastroplasty. Long-term followup of patients undergoing this operation indicates satisfactory results, but a high incidence of continued or recurrent reflux has been reported [3], particularly in patients requiring surgical revision [4]. In general, though, a low incidence of anatomical recurrence is reported for gastroplasty [5]. Total fundoplication (Nissen) From the Women's College Hospital, University of Toronto, Toronto, Ont, Canada. Presented at the Twentieth Annual Meeting of The Society of Thoracic Surgeons, San Antonio, TX, Jan 23-25, 1984. Address reprint requests to Dr. Henderson, Chief of Surgery, Women's College Hospital, 76 Grenville St, Toronto, Ont, Canada M5S 182.
74
gastroplasty was designed to simultaneously incorporate anatomical stability and provide the benefits of the more effective reflux control inherent in total fundoplication [6].
Material and Methods Three hundred fifty-one consecutive patients were treated for symptoms of intractable reflux by total fundoplication gastroplasty. There were 123 male and 228 female patients. The average age was 45.5 years (range, 17 to 75 years). A preoperative history was obtained using a prepared history format, and all patients underwent preoperative evaluation by radiography, flexible fiberoptic esophagogastroduodenoscopy, and manometry with pH determination. Three hundred twenty-one patients also had acid perfusion studies. The manometric studies were conducted using three polyethylene tubes with side holes. Infusion was maintained at 6.75 ml per minute by means of a modified Harvard pump. Data were recorded with a Honeywell 1508 Visicorder. A Beckman pH probe and recorder were used to assess pH changes. All follow-up radiographic studies were conducted in the same department. Water siphon studies were routinely performed for evaluation of reflux.
Operative Technique A transthoracic approach was used for primary hernia repair in the majority of patients. A thoracoabdominal incision was used in patients with previous esophagogastric repair and in those requiring simultaneous abdominal and thoracic surgical procedures. The transthoracic approach was through the seventh or eighth interspace. Mobilization of the esophagogastric junction was completed medially to the aortic arch and along its aortic surface to the inferior pulmonary vein. The esophagogastric fat pad was excised. Posterior crural sutures were placed, and care was taken to preserve the peritoneal covering to add strength to the repair. The sutures were tied at the completion of the gastroplasty for loose approximation of the crurae. Gastroplasty tubes were cut using snugly fitting rightangle DeBakey clamps over a 60F Maloney bougie. The tubes were closed with an interlocking 2-0 chromic suture and continuous 3-0 silk. The fundus of the stomach was reapproximated to the gastroplasty tube and distal 1.5 cm of the esophagus with 2-0 silk mattress sutures through the gastroplasty suture lines. This gives very strong approximation and is important for the anatomical stability of the repair. Total fundoplication was achieved by the lateral and medial
75 Henderson and Marryatt: Total Fundoplication Gastroplasty
Fig 1 . Prepared gastroplasty tube with the new gastric fundus sutured to the 5-cm tube and distal 1.5 cm of esophagus. Fol[oaiing fundoplication, the stomach is reduced below the diaphragm and the crural sutures are tied.
spread of the new gastric fundus. The completion fundoplication was restricted in length to 1 cm and was sutured with three 2-0 silk mattress sutures. In patients with poor esophageal motor function or previous gastric repair, the length of the completion fundoplication was reduced to 0.5 cm or less. Before reduction of the gastroplasty tube into the abdomen, two sutures were placed from the fundus to the undersurface of the diaphragm; these were tied following reduction, as were the posterior crural sutures (Figs 1, 2). When a thoracoabdominal approach was employed, the technique of preparation of the gastroplasty tube was similar; however, four to five short gastric vessels were divided for fundic mobilization. The total fundoplication was passed clockwise behind the gastroplasty tube and sutured to the tube and distal 1.5 cm of esophagus before the performance of a 1-cm completion fundoplication. Several points should be emphasized concerning the technique. The intraabdominal segment is 6.5 cm long5 cm of gastroplasty tube and 1.5 cm of distal esophagus. Overreduction of the esophagus should be avoided. Although the completion fundoplication is very loose, the length of approximated stomach must be carefully tailored to 1 cm. An excessively long completion fundoplication is the cause of overcompetence, so the completion fundoplication must be tailored to reduce its length in patients with poor esophageal motor function (Fig 3).
Results Among the 351 patients treated for reflux by total fundoplication gastroplasty, the average follow-up was 6.5 years (range, 5 to 8.5 years).
Preoperative Evaluation Reported symptoms, obtained from the preoperative history, are listed in the Table.
Fig 2. Following resuturing of the fundus to the gastroplasty tube and distal esophagus, three gastrogastric sutures are placed, approximating the stomach. A total lax fundoplication is achieved and carefully tailored to prevent the fundus from accidentally approximating and increasing the length of the fundoplication.
Radiographic studies demonstrated a hiatal hernia in 222 patients and reflux alone in 48. Sixty-eight patients had normal studies. No studies were available for review of the remaining 13 patients. On manometric examination, the average tone in the high-pressure zone (HPZ) was 12.26 cm H 2 0 and disordered motor activity (DMA) in the lower esophagus averaged 46.2%. A manometric diagnosis of hiatal hernia was made in 68.9% of the patients. Acid perfusion studies were positive in 295 (91.9%) of 321 studies. The major risk factors in operation for reflux include previous gastric or esophageal repair, severe ulceration or stricture, and scleroderma. Ninety-five patients (27.1%)had had a previous esophageal repair, 25 (7.1%) had had an esophageal repair combined with a gastric procedure, and 10 (2.9%) had had a previous gastric operation. Thus, a total of 130 patients (37%) had had previous surgical intervention. Severe ulceration or stricture was documented endoscopically in 68 patients (19.4%), while scleroderma with an adynamic esophagus was present in 5 (1.4%). One or more of these factors was present in 179 patients (51%).
Postoperative Evaluation Follow-up data are available on 335 (95.4%) of the 351 patients. In this group, all were evaluated by history, 324 (96.7%) by radiography, and 248 (74%)by manometry and pH studies. There were no operative deaths. Nine patients died late of unrelated causes. The major related complication was esophageal fistula, which occurred in 3 patients (0.9%). All of these fistulas closed early and spontaneously; two of them were in patients with recurrent hernia and the other, in a patient who had primary repair, was due to breakdown of staples. Other major complications included three instances of bowel obstruction (two
76 The Annals of Thoracic Surgery Vol 39 No 1 January 1985
Esophagus 1 112 cm 1 cm
Gastroplasty 5 cm
Fig 3. lntraabdominal length and fundoplication length are prepared by separate suturing techniques. When the completion fundoplication is reduced to one stitch in length, gastroesophageal sutures are added to maintain the stability of the fundoplication.
resulting from old adhesions and one from a hematobezoar) and one instance of wound dehiscence. There was one late ventral hernia in an alcoholic patient with poor tissue strength. Two patients had moderate wound infections. Three categories of clinical results are defined. Patients with excellent results include those who are asymptomatic or have only minor residual symptoms. Patients with moderate results are those who have notable residual symptoms but are improved following operation. Patients with poor results are those with major residual symptoms or in whom symptoms have developed during 5 or more years of follow-up. Among the 335 patients for whom follow-up data were available, 312 (93.1%)had excellent results, 13 (4.0%)had moderate residual symptoms, and 10 (2.9%)had poor results. The 13 with moderate residual symptoms had a variety of problems. Four had nonspecific complaints, for which an exact cause could not be found. Two patients had gas bloat syndrome. Two had bile gastritis and 5 , moderate persistent dysphagia. Seven of the 13 with residual problems had had previous gastric surgical procedures. Ten patients requiring surgical revision have been analyzed. Three had radiographic and symptomatic recurrence of reflux, and 1 had an intussusception recurrence that, while not recognized radiographically, occurred with increasing dysphagia. Two patients had a traumatic esophageal diverticulum with dysphagia, and 3 had dys-
phagia due to an overcompetent repair. One experienced severe gas bloat. In this group, 7 patients had had previous surgical intervention, including 3 who had both gastric and esophageal procedures. Two had major psychiatric overlay affecting interpretation of symptoms. Results of radiographic studies, including water siphon testing for reflux, are available for 324 of 351 patients (92.3%).Radiographic studies were conducted routinely five and one-half months after operation and
Preoperative Symptoms in 351 Patients Undergoing Total Fundoplication Gastroplasty Symptom Heartburn Pain" Epigastric Retrosternal Back Arm Reflux Aspiration Eructation Hiccough Excess salivation Dysphagia Cricopharyngeal Gastroesophageal Stricture Total motor dysphagia "Average duration was 4.9 years.
No. of Patients
Percent
344
98.0
332 321 142 43 282 161 249 91 81
94.6 91.5 40.5 12.3 80.3 45.9 70.9 25.9 23.1
140 200 36 212
39.9 57.0 10.3 60.4
77 Henderson and Marryatt: Total Fundoplication Gastroplasty
TONE PREOPS
0
CM H20
30
POST-OPS
I
20
10
TONE RANGE
0-6
7-10
11-15
16-20
21+
CM H20
AV.TONE
5
9.1
12.5
17.8
26
CM H20
43
70
61
52
17
PATIENT NO.
Fig 4. Preoperative tone in the high-pressure zune (HPZ). Those with low HPZ tone had a greater average tone increase than those with higher preoperative tone. Following total fundoplication gastroplasty, HPZ tune in all patients was within the normal range of 15 to 20 cm HZO.
were repeated only if clinically indicated. Two patients had an asymptomatic trace of reflux, 2 had a traumatic diverticulum, and 3 had a radiographically demonstrated anatomical recurrence (one recurrence was not recognized). Follow-up manometric and pH studies were conducted in 243 of the 351 patients (69.2%).These studies were conducted five and one-half months after operation. The average HPZ tone increased postoperatively to 18.1 cm H 2 0 (+47.6%), and average lower esophageal DMA decreased to 29.1% (-37%). The actual increase in HPZ tone depended on the preoperative tone (Fig 4). In patients with a low preoperative HPZ tone of 0 to 6 cm H20, the tone increase averaged 240%. As the preoperative HPZ tone increased, the percentage of tone increase became smaller. In a small group of patients with high preoperative HPZ tone, this value fell after surgical correction of reflux. The average postoperative tone in all groups was within the normal range of 15 to 20 cm H20.
Comment The operative technique of gastroplasty combined with total fundoplication was developed in 1975, and early results were reported in 1977 [7]. To our knowledge, the present report is the first 5-year review of total fundoplication gastroplasty, although undoubtedly others will follow. Several variations on this operation have already been described, including an uncut total fundoplication gastroplasty [8] and alternative methods of fundoplication [9, 101. The Collis-Nissen procedure described in 1978 by Orringer and Sloan [lo] is similar in principle, but the method of fundoplication varies considerably in fixation and length. Comparison of results of these procedures depends on both the severity of preoperative disease and the methods of objective and subjective analysis after operation. Preoperatively, 37% of our patients had had previous esophageal or gastric procedures and 10% had had gastric operations. In addition, 19.4% had ulceration or
stricture and 1.4%, scleroderma. These complicated cases account for most of the patients with moderate results or operative failure. Excellent results were reported only for patients with minimal or no symptoms and no objective radiographic, manometric, or pH evidence of failure. Those labeled as having moderate results are much improved and have no evidence of recurrence, so the decision to classify them in the moderate category is inevitably subjective. In the group with poor results are all patients whose symptoms were not improved or who experienced late failure of the operation. The overall incidence of excellent results (93.1%) is quite acceptable, and the failure rate of 2.8% is low. The rate of actual anatomical recurrence of reflux (1.1%)is very low compared with rates reported for other methods of gastroplasty. Radiographic studies are of limited value because with total fundoplication, severe reflux is rare in the absence of anatomical recurrence. In the 3 patients with anatomical recurrence, there was no reflux; the patient with an intussusception recurrence was considered radiographically normal. Manometric studies with determination of pH are of value in assessing reflux and showing the resolution of reflux-induced secondary motor damage. The motor improvement detected is better with total fundoplication gastroplasty than with partial fundoplication procedures. Preventing reflux is easy; however, little attention has been paid to methods of calibrating the fundoplication to avoid overcompetence. Therefore, total fundoplication gastroplasty has been designed to avoid overcompetence and dysphagia and to allow the patient to burp and vomit. There are two major components of the total fundoplication gastroplasty: anatomical stability and the creation of a tailored degree of competence that avoids dysphagia. Stability is maintained by using a gastroplasty tube, which is held below the diaphragm together with 1.5 cm of distal esophagus. Standard hernia repairs require 4 cm of esophagus for fundoplication so that 2.5 cm is spared and there is no tension on the esophagus following reduction. When severe shortening is present, the HPZ may remain in the chest and competence can be achieved by fundoplication of the gastroplasty tube below the diaphragm. The major anchoring sutures are in the stomach, adding considerable strength to the repair. The left gastric artery fixes the gastroplasty tube below the diaphragm. A recurrence rate of 1%is consistently reported with gastroplasty and almost certainly reflects the anatomical stability of the repair. Avoidance of overcompetence is the most important feature of the design of total fundoplication gastroplasty. Recognition of these problems has evolved with growing experience. The application of this knowledge in the past few years has made overcompetence and dysphagia problems of the past. None of the 300 patients operated on from 1980 to 1983 has required revision for dysphagia. The length of the completion fundoplication is deter-
78 The Annals of Thoracic Surgery Vol 39 No 1 January 1985
mined by direct gastrogastric suturing. Completion length is 1 cm in the majority of patients, but is reduced to 0.5 cm in patients with severe DMA (61 to 80%)in the lower esophagus. In the presence of very severe DMA (81 to 100%) in patients with previous gastric resection or scleroderma, the wrap length is kept to less than 0.5 cm and the fundoplication is stabilized by adding sutures to the gastroplasty tube and esophagus (see Fig 3). This type of fundoplication, in addition to minimizing dysphagia, allows 85% of patients to eructate normally. Ability to vomit is difficult to assess as most individuals have no desire to vomit.
References 1. Collis JL: Gastroplasty. Thorax 16:197, 1961 2. Pearson FG, Langer B, Henderson RD: Gastroplasty and Belsey hiatus hernia repair: an operation for the management of peptic stricture with acquired short esophagus. J Thorac Cardiovasc Surg 61:50, 1971 3. Orringer MB, Sloan H: Complications and failings of the combined Collis-Belsey operation. J Thorac Cardiovasc Surg 74726, 1977 4. Henderson RD: The gastroplasty tube as a method of reflux control. Can J Surg 21:264, 1978 5. Pearson FG, Cooper JD, Nelms JM: Gastroplasty and fundoplication in the management of complex reflux problems. J Thorac Cardiovasc Surg 76:665, 1978 6. DeMeester TR, Johnson LF, Kent AH: Evaluation of current operations for the prevention of gastroesophageal reflux. Ann Surg 180:511, 1974 7. Henderson RD: Reflux control following gastroplasty. Ann Thorac Surg 24:206, 1977 8. Evangelist FA, Taylor FH, Alford JD: The modified CollisNissen operation for control of gastroesophageal reflux. Ann Thorac Surg 26:107, 1978 9. Bingham JA: Hiatus hernia repair combined with the construction of an anti-reflux valve in the stomach. Br J Surg 64:460, 1977 10. Orringer MB, Sloan H: Combined Collis-Nissen reconstruction of the esophagogastric junction. Ann Thorac Surg 25:16, 1978
Discussion (Ann Arbor, MI): I thank Dr. Henderson and Mr. Marryatt for the opportunity to review their manuscript. They are clearly to be credited with outstanding 5-year follow-up results in a large group of patients operated on for reflux. Having been charged with the task of critically reviewing this work, however, I wish to ask Dr. Henderson’s comments on several points. As many know, both Dr. Henderson and I have advocated a 360-degree fundoplication for reflux control after construction of the Collis gastroplasty. The gastroplasty reduces the volume of the remaining fundus available for the wrap by the amount used to fashion the tube. If the surgeon tries to use the residual elongated fundus for a Belsey-type repair, he or she is simply tacking the stomach back against the tube and can achieve only a limited fundoplication. This was the rationale for our abandonment of the Collis-Belsey combination and for our preference for passing the elongated gastric fundus behind the gastroplasty tube and performing a 360-degree Nissen-type wrap. In his operation, Dr. Henderson is actually reapproximating the gastric fundus to the gastroplasty tube, a technique reminiscent DR. MARK B. ORRINGER
of the combined Collis-Belsey procedure, and then performing the I-cm completion fundoplication. I am surprised that he has had no difficulty drawing the edges of the remaining fundus all the way around the tube. I wonder if this difficulty has been responsible for his progressively decreasing the length of the wrap, now down to only 1 cm. Second, it is clear that Dr. Henderson includes two quite different operations under the heading of total fundoplication gastroplasty. One involves no division of short gastric vessels, reapproximation of the fundus to the gastroplasty tube, a 1-cm fundoplication, and anchoring of the gastric fundus to the undersurface of the diaphragm a la the second row of sutures of the standard Belsey operation. The second operation requires division of four to five short gastric vessels, a true wraparound Nissen-type fundoplication like the one my colleagues and I perform, and apparently no anchoring of the stomach beneath the diaphragm. Dr. Henderson, how many patients underwent which operation? Is there a difference in the operative results between these procedures? Why is it necessary to divide short gastric vessels or anchor the wrap beneath the diaphragm on some occasions but not others? Finally, as I am sure Dr. Henderson will acknowledge, objective assessment of the efficacy of an antireflux operation with preoperative and postoperative manometry and acid reflux testing with the intraesophageal pH electrode has become a mandatory standard in our specialty. While Dr. Henderson indicates that all of his patients were evaluated preoperatively and 71% postoperatively with manometric and pH studies, the preoperative and postoperative pH reflux testing data have not been presented. Their absence is a serious omission when attempting to “sell” a relatively new antireflux procedure. It would also be helpful to have some idea of the relative frequency of adverse postoperative results following total fundoplication gastroplasty: dysphagia requiring dilation as well as intermittent untreated dysphagia; severe gas bloat syndrome as well as less troublesome, but still disturbing, early satiety; postvagotomy diarrhea; and postthoracotomy incisional pain. Again, I congratulate Dr. Henderson and Mr. Marryatt on their continued fine work. DR. NICHOLAS J. DEMOS (Jersey City, NJ): I wish to congratulate Dr. Henderson and Mr. Marryatt on their tremendous series and the evangelistic fervor with which they study this disease. Personally, I am committed to the uncut gastroplasty. I started studying the stapled, uncut gastroplasty plus total fundoplication more than 15 years ago in animal experiments. I performed successful canine experiments for 2 years, making all my mistakes in dogs. I have applied this technique for the last 14 years in human beings. My results are somewhat similar to those of Dr. Henderson, but I cannot understand why surgeons cut the stomach and reapproximate it. When one cuts the stomach and sutures it in two layers, more dissection and transection of short gastric vessels are necessary. This has resulted in splenectomies, ischemia, and leakage. In follow-up ranging from 1 to 13 years of almost 100 patients who had stapled, uncut gastroplasty, there has been no recurrence of reflux, no need for reoperation or splenectomy, and no fistula formation. One patient had a stricture that had to be dilated every six months for 6 years. Because of its simplicity and, the reduced morbidity it affords, the stapled, uncut gastroplasty with total fundoplication remains my operation of choice. DR. HENDERSON: I thank Dr. Orringer and Dr. Demos for their comments. There is a great deal of agreement among the three of us, and we have discussed these issues many times.
79 Henderson and Marryatt: Total Fundoplication Gastroplasty
Dr. Orringer's suggestion that two operations have been described is misleading. Basically, 32% of the patients had a thoracoabdominal approach for a recurrent hiatal hernia. I have described this method before. The thoracoabdominal approach permits dissection of scar tissue under direct vision and also allows a pyloromyotomy. The addition of pyloromyotomy is important because even if the nerves are well preserved, it is unlikely they will function normally in people with from one to five previous hernia repairs. There is no problem with fundus size, and creation of a total fundoplication is always possible even in patients with previous radical gastrectomies. We did start with a completion fundoplication of 2.5 cm and have deliberately reduced this to a 1cm wrap. This length is further reduced in patients with poor esophageal peristalsis. The key to a stable repair is to suture the
intraabdominal segment separately which in turn enables careful tailoring of the completion fundoplication. This approach almost totally eliminates dysphagia. As noted in our paper, pH testing was carried out both preoperatively and postoperatively. In terms of Dr. Demos's discussion of the cut versus the uncut gastroplasty, I do not think there is an absolute answer. Long-term studies are necessary. In my experience, a better fundoplication is achieved with a cut gastroplasty. gingham, who described the uncut gastroplasty in 1977, reported a 7% incidence of breakdown of staples, which one has to expect when stapling healthy stomach to healthy stomach. This has been a recognized problem with the transverse gastroplasty for obesity, and suturing is now added to stapling to increase stability in this procedure.
Notice from the Editor and the Editorial Board The Editor and members of the Editorial Board are grateful for the assistance given in manuscript review during the past year by the following individuals: Hendrick B. Barner, M.D. St. Louis, MO
George C. Kaiser, M.D. St. Louis, MO
W. Gerald Rainer, M.D. Denver, CO
Mr. Richard Beauchamp St. Louis, MO
A. Ercument Kopman, M.D. St. Louis, MO
Francis Robicsek, M.D. Charlotte, NC
David Bregman, M.D. Paterson, NJ
Ronald D. Leidenfrost, M.D. St. Louis, MO
Benson B. Roe, M.D. San Francisco, CA
John Codd, M.D. St. Louis, MO
Harold V. Liddle, M.D. Salt Lake City, UT
Charles L. Roper, M.D. St. Louis, MO
Lawrence H. Cohn, M.D. Boston, MA
Floyd D. Loop, M.D. Cleveland, OH
Stuart S. Sagel, M.D. St. Louis, MO
Jack G. Copeland, 111, M.D. Tucson, AZ
Gerald Medoff, M.D. St. Louis, MO
William F. Sasser, M.D. St. Louis, MO
James L. Cox, M.D. St. Louis, MO
Robert M. Mentzer, Jr., M.D. Charlottesville, VA
Mrs. Louise Sper Detroit, MI
Robert W. M. Frater, M.D. Bronx, NY
Mark B. Orringer, M.D. Ann Arbor, MI
Thomas L. Spray, M.D. St. Louis, MO
J. Alex Haller, Jr., M.D. Baltimore, MD
D. Glenn Pennington, M.D. St. Louis, MO
Clarence S. Weldon, M.D. St. Louis, MO
Alexis F. Hartmann, Jr., M.D. St. Louis, MO
Salvatore Pizzo, M.D. Durham, NC