- Email: [email protected]
Minimally Invasive Antireflux Surgery
RAPID COMMUNICATION
Minimally Invasive Antireflux Surgery J. Barry McKernan, MD, PhD, Marietta, Georgia, J. Kenneth Champion, MD, Macon, Georgia
BACKGROUND: Previous reports of minimally invasive antireflux surgery for gastroesophageal reflux disease (GERD) have been small, short-term series utilizing only a laparoscopic approach. We conducted a retrospective review and report our 66-month experience with more than 1,000 laparoscopic and thoracoscopic antireflux procedures. METHODS: Between September 1991 and October 1997, 968 adults underwent 1,003 minimally invasive antireflux procedures on a tailored basis, based on their preoperative evaluation. Procedures performed were laparoscopic Nissen (626), Toupet (348), paraesophageal (33), and thoracoscopic Belsey (22). A total of 23% (233) of patients underwent an ancillary procedure (esophageal myotomy 85, vagotomy 67, pyloromyotomy 13, and cholecystectomy 66). RESULTS: Follow-up averaged 33 months (range 1 to 66), operative mortality was 0.1%. Complications occurred in 2.7% with a 1% long-term dysphagia rate. Demonstrated recurrence rate was 3.8% to date, with an associated 3.4% reporting symptoms of GERD. CONCLUSION: Minimally invasive antireflux procedures provide sustained relief of GERD symptoms with low morbidity and rapid recovery. Am J Surg. 1998;175:271–276. © 1998 by Excerpta Medica, Inc.
G
astroesophageal reflux disease (GERD) is a common clinical problem and results in high rates of health care use. Surveys indicate that almost 45% of American adults experience heartburn, the cardinal symptom of GERD, at least once a month and 7% report having the symptom on a daily basis.1 The symptoms associated with GERD are caused by the reflux of gastric contents into the esophagus and its resultant complications. Approximately 50% of individuals with increased esophageal acid exposure experience complications such as erosive esophagitis, esophageal stricture, or Barrett’s esophagus.2 The ultimate therapeutic goal in managing GERD is to reduce reflux, minimize any damage to the esophageal
From the Department of Surgery (JBM), Medical College of Georgia, Augusta, Georgia, and the Department of Surgery (JKC), Mercer University School of Medicine, Macon, Georgia. Requests for reprints should be addressed to J. Barry McKernan, MD, PhD, 130 Vann Street, Suite 220, Marietta, Georgia 30060. Manuscript submitted December 16, 1997 and accepted in revised form February 3, 1998.
© 1998 by Excerpta Medica, Inc. All rights reserved.
mucosal lining, and provide complete symptomatic resolution without the introduction of new complaints. In many patients the symptoms of GERD can be effectively managed through diet and lifestyle modifications. Individuals whose symptoms are inadequately relieved by these measures or with severe, chronic esophageal reflux may require the use of medications that suppress gastric acid secretion such as histamine H2 receptor antagonists or proton pump inhibitors, or the use of pro-motility agents such as cisapride. A subset of patients are candidates for surgery because of severe reflux and a poor response to medical therapy. The indications for the surgical management of GERD are listed in Table I. Since the first published report of a fundoplication procedure performed laparoscopically appeared in 1991,3 several surgeons—including ourselves— have used minimally invasive techniques to treat GERD. Building upon our early experience with Nissen and Toupet fundoplications, we expanded our repertoire of minimally invasive techniques to include thoracoscopic Belsey partial fundoplication and videoscopic Collis esophagoplasty. While there have been numerous reports in the literature detailing experience with laparoscopic antireflux procedures, most of these have been relatively small series, involving fewer than 75 patients who were followed up, on average, for less than 1 year.4 –7 We have now analyzed our experience in performing 1,003 minimally invasive antireflux procedures on 968 patients, 695 of whom have been followed up for more than 1 year. In this paper we report our surgical techniques and results.
MATERIALS AND METHODS Patient Population Between September 1991 and October 1997, 1,003 minimally invasive antireflux procedures were performed on a total of 968 adults with symptomatic GERD who were refractory to or intolerant of medical therapy. Fifty-three percent of the cohort were male. The average age of patients was 52 years (range 19 to 85), and 20% were 65 years of age or older. Body weight ranged from 102 to 344 pounds, with the average weight being 178 pounds. Manometry performed prior to surgery revealed that the lower esophageal sphincter (LES) pressure in our patient group averaged 14 mm Hg (range 0 to 70). Sixty percent of our patients had a motility disorder, including nonspecific motility disorders (20%), low amplitude esophageal contraction (20%), hypertensive LES pressure (8%), nutcracker esophagus (8%), and diffuse esophageal spasm (2%). Achalasia was seen in 2% of our patients. Preoperative Evaluations Mandatory and selective tests. Esophagogastroduodenoscopy (EGD) and esophageal manometry are now routinely performed on every surgical candidate, with the exception 0002-9610/98/$19.00 PII S0002-9610(98)00034-8
271
MINIMALLY INVASIVE ANTIREFLUX SURGERY/MCKERNAN AND CHAMPION
TABLE I
TABLE II Indications for Antireflux Surgery
Refractory to medical therapy ● Persistant symptoms after 12 weeks of maximal medical therapy ● Recurrent symptoms upon cessation of medication ● Noncompliance with medical therapy Development of complications ● Stricture ● Barrett’s mucosa ● Esophagitis ● Aspiration ● Motility disorder
of patients having an associated stricture or paraesophageal hernia that will not permit passage of a manometry catheter. In patients seen before 1993 esophageal manometry was considered optional.8 During EGD, the presence of esophagitis is noted. Biopsies may be performed and complications such as ulceration, stricture, or Barrett’s mucosa are documented. This is particularly important in instances of Barrett’s mucosa as these patients require life-long monitoring since they are at increased risk for adenocarcinoma of the esophagus. Additionally, the length of the esophagus and the presence of a hiatal defect are noted. During esophageal manometry, we routinely measure LES pressure and perform an esophageal body motility study to characterize any motility disorder. Twenty-four-hour pH monitoring, barium swallow, and gastric emptying studies were used selectively in our series. A 24-hour pH monitoring study was indicated only in the absence of esophagitis upon EGD and normal esophageal manometry results. Only 11% of our patients underwent pH monitoring. A barium swallow test was used to confirm results of an EGD performed at another institution. This test is also useful in patients with stricture or with a history of esophageal diverticulum in order to accurately characterize the underlying pathology. A gastric emptying study was performed only if there was a history of gastric disease, frequent vomiting, or findings on EGD suggestive of gastric stasis. Operative Approaches We currently employ a selective approach to treating GERD, tailoring the antireflux procedure to each patient’s underlying anatomic and functional defect. The types of videoscopic antireflux procedures we have performed include Nissen fundoplication, Toupet fundoplication, paraesophageal hernia repair, and thoracoscopic Belsey repair. The general guidelines we use for determining which procedure to offer patients are detailed in Table II, while the number of procedures we have performed through October 20, 1997, are displayed by type in Table III. In addition to the primary antireflux procedure, 23% of our patients underwent an ancillary procedure (Table IV). A small number of patients (n 5 22) required a prosthetic crural repair when the crural size exceeded 5 cm in diameter after a prior failed surgery. In addition, 6% of our patients underwent cholecystectomy for coexisting cholelithiasis at the time of their GERD surgery. 272
Criteria for Selecting Minimally Invasive Antireflux Procedure Nissen fundoplication ● Indicated in patients without a primary motility disorder ● Indicated in patients with a nonspecific motility disorder Toupet fundoplication ● Indicated for cases where low amplitude peristalsis would interfere with esophageal emptying ● Indicated as the abdominal laparoscopic approach to be used in conjunction with esophageal myotomy for a primary motility disorder Thoracoscopic Belsey partial fundoplication ● Indicated in patients who had undergone extensive upper abdominal or previous antireflux surgery, particularly in the presence of a primary esophageal disorder requiring simultaneous repair ● Selected in cases where concomitant intrathoracic pathology could be accessed at the same time through a unilateral thoracoscopic approach
TABLE III Minimally Invasive Antireflux Procedures (9/30/91 to 10/20/97) Total Number of Procedures
1,003
Types of procedures Nissen fundoplication Toupet fundoplication Paraesophageal hernia repair Thoracoscopic Belsey repair
626* 348† 33‡ 22
* Eighteen were Collis-Nissen procedures. † Two were Collis-Toupet procedures. ‡ Seven of these repairs did not include a fundal wrap and thus are counted as unique cases in the overall total of 1,003 procedures. Twenty-three of the paraesophageal repairs were performed in conjunction with a Nissen procedure, two in conjunction with a Toupet procedure, and one in conjunction with a Belsey repair.
Over the 6 years of this series, 9 patients (0.9%) with previous gastric surgery (vertical banded gastroplasty or subtotal gastrectomy) were not deemed candidates for a fundoplication procedure and underwent a Roux-en-Y gastric diversion for intractable reflux. Operative Techniques Initial dissection. Patients are admitted on the morning of surgery, and are given prophylactic antibiotics and sequential compression hose to prevent deep venous thrombosis. Patients are placed in the supine position under general anesthesia. Six trocars are routinely used, with one 10-mm and five 5-mm ports. A 10-mm incision is made just to the left midline, 15 cm below the xiphoid process. A 0-degree scope with a 10-mm Optiview optical surgical obturator (Ethicon Endo-Surgery, Inc., Cincinnati, Ohio) is passed through the subcutaneous tissue, anterior rectus sheath, rectus muscle, and posterior rectus sheath, and the abdomen is entered. The Optiview trocar allows each fascial layer to be identified and any vessel or adherent bowel to be visualized, thereby preventing inadvertent injury. Once the Optiview trocar has been inserted, insufflation
THE AMERICAN JOURNAL OF SURGERY® VOLUME 175 APRIL 1998
MINIMALLY INVASIVE ANTIREFLUX SURGERY/MCKERNAN AND CHAMPION
TABLE IV Ancillary Surgical Procedures Procedure Esophageal Myotomy Total procedures in conjunction with: Toupet fundoplication (54) Belsey repair (14) Nissen fundoplication (14) Collis-Nissen fundoplication (1) No-wrap paraesophageal repair (2) Cricopharyngeal myotomy Vagotomy Highly selective vagotomy (HSV) HSV 1 Nissen fundoplication (30) HSV 1 Toupet fundoplication (12) HSV 1 pyloromyotomy (4) HSV 1 pyloroplasty (2) HSV 1 Jabouley gastroduodenostomy (1) Truncal vagotomy Truncal vagotomy 1 Nissen fundoplication (4) Truncal vagotomy 1 Belsey repair (2) Truncal vagotomy 1 esophagogastrectomy (2) Truncal vagotomy 1 pyloromyotomy (1) Pyloromyotomy Total procedures in conjunction with: Nissen fundoplication (9) Toupet fundoplication (3) Collis-Nissen fundoplication (1) Cholecystectomy Total
Number Performed 85
2 56
11
13
66 233
is begun to 15 mm Hg. The remaining 5-mm ports are inserted under direct visualization. An Allis clamp is placed through the diaphragm just above the apex of the esophageal hiatus to allow for liver retraction. This technique is less traumatic, requires no assistant to hold, and even works well in cases of hepatomegaly. The primary surgeon and assistant each utilize a twohanded technique, which enhances exposure and speeds the operation. The patient is placed in a reverse Trendelenberg position at this point, with the back elevated at approximately 30 degrees. The 10-mm, 0-degree scope is exchanged for a 45-degree angled scope, which is essential for performing the dissection. Dissection is begun along the left crus by dividing the peritoneal attachments along the fundus, angle of His, and lateral left crus to aid in the ultimate mobilization of the esophagus. Attention is then turned to the lesser omentum; this is opened directly over the caudate lobe, thereby exposing the caudate lobe and right crus. The hepatic branch of the vagus nerve is routinely sacrificed, but any vessel in the lesser omentum that is larger than 5 mm is typically preserved. This is important since in approximately 10% of patients, a large branch of the left hepatic artery can transverse this area and its sacrifice may result in necrosis of the liver. Once the right crus has been identified, the peritoneal attachment along its anterior border is divided. The dissection begins in the mid-body of the crus and is then carried inferiorly until the median arcuate ligament is identified. Blunt dissection is used to proceed up the anterior border of the left crus until a window has been created
behind the esophagus and the esophagus is completely mobilized. Turning the 45-degree scope laterally provides excellent visualization of this portion of the dissection and prevents inadvertent perforation of either the esophagus or posterior stomach. A blue vessel loop is passed around the esophagus for traction and is secured with a chromic Endoloop. The assistant can then use the Endo-loop for retraction to aid in the dissection. This dissection is performed with only an nasogastric tube in place since the presence of a bougie increases the likelihood of perforation. With the assistant retracting the esophagus bilaterally, dissection is begun along the anterior border of the right crus, and the phrenoeophageal ligament is circumferentially divided along the border of the crura until the esophagus is completely mobilized. Both the anterior and posterior vagus nerves are identified and preserved, and are usually left adherent to the esophagus. Using the 45-degree angled scope and the laparoscopic technique, we now know that it is possible to mobilize the esophagus all the way to the inferior pulmonary vein, which crosses the esophagus anteriorly at 12 o’clock. Care must be exercised in this area to avoid unintentional bleeding. Once the esophagus has been mobilized, the short gastric vessels are divided along the upper 10 cm of the fundus from the angle of His, extending laterally. Routine division of these short gastric vessels ensures a loose, floppy fundus and decreases the incidence of postoperative dysphagia. We have utilized metal clips, vascular staples, harmonic scalpel, and bipolar forceps to achieve hemostasis. The harmonic scalpel or bipolar forceps were used in our last 600 cases with no bleeding complications. Once the greater curve has been mobilized and the esophagus is mobilized, either a Nissen fundoplication or a Toupet fundoplication can be performed. We routinely measure the crural opening size with a endoscopic ruler and close the crura with interrupted sutures. With the #50 bougie in place, the crura are reapproximated with interrupted size 0, Ethibond Endoknot sutures, and tied extracorporally until they are closed loosely about the #50 bougie. Nissen fundoplication. If a Nissen fundoplication is being performed, the bougie is carefully withdrawn back into the middle chest. Placing a grasper behind the esophagus, the fundus is pulled posteriorly behind the esophagus, so that the two edges of the stomach will circumferentially encircle the esophagus just above the gastroesophageal junction. The #50 bougie is passed into the stomach under direct visualization. The Nissen repair is then undertaken by suturing the stomach laterally to the esophagus 2 cm above the gastroesophageal junction, taking care to avoid the anterior vagus nerve. The suture is passed through the stomach medially and tied down extracorporeally without tension. The next suture is placed through the stomach laterally, through the esophagus at the gastroesophageal junction anteriorly, and through the stomach medially. A size 0, Ethibond Endoknot is used to complete the repair, which is tied extracorporeally. Toupet fundoplication. In the event that a Toupet fundoplication is indicated, a grasper is placed behind the esophagus and the fundus is grasped and retracted medially
THE AMERICAN JOURNAL OF SURGERY® VOLUME 175 APRIL 1998
273
MINIMALLY INVASIVE ANTIREFLUX SURGERY/MCKERNAN AND CHAMPION
behind the esophagus so that a 270-degree wrap can be performed. The first suture in the Toupet fundoplication is placed through the fundus posteriorly and through the left crura using a size 0 Ethibond Endoknot, tied extracorporeally. A second suture is placed through the wrap posteriorly and through the right crus. At 6 o’clock the medial portion of the wrap is anchored using a size 0 Ethibond Endoknot so that the two sides will be sutured at the esophagus at approximately 10 o’clock and 2 o’clock to create the 270degree wrap. The first suture is placed through the esophagus, 2 cm above the gastroesophageal junction medially through the upper portion of the wrap, and then through the crura at 10 o’clock in order to anchor the wrap. To secure the medial portion of the wrap, the second suture is placed through the esophagus medially at the gastroesophageal junction and stomach. The lateral portion of the wrap is completed by placing a suture through the stomach laterally and then through the crus at 2 o’clock and through the lateral esophagus 2 cm above the gastroesophageal junction. The last suture is placed laterally through the stomach and esophagus at the gastroesophageal junction to complete the 270-degree wrap. Esophagomyotomy. Once dissection has been completed and prior to closure of the esophageal hiatus, the anterior esophagus is exposed between 11 and 12 o’clock. This area avoids the anterior vagus nerve. Beginning approximately 2 cm above the gastroesophageal junction, the longitudinal fibers in the first muscular layer of the esophagus are bluntly dissected and separated with a scissors tip. We use dissecting endoscopic scissors with no cautery since we find that cautery is rarely needed on the small vessels to the esophagus. Once the longitudinal fibers have been bluntly separated, the circular fibers become exposed. These are divided under direct visualization. The assistant uses a suction irrigator to keep the field clear for dissection. Once the circular fibers are divided down to the mucosa, the mucosa can be bluntly pushed inferiorly, and the dissection can proceed in a cephalad direction for approximately 2 cm more. The total length of the myotomy will depend on the indication for the procedure. For a primary motility disorder such as achalasia, nutcracker esophagus, or diffuse esophageal spasm, a length of 6 cm is usually sufficient. For a hypertensive LES, typically only a 4 cm myotomy is needed to relieve the obstruction. Once the proximal portion of the myotomy is completed, the more distal segment, which involves the gastroesophageal junction, is approached. Dissection is carried inferiorly until it impinges upon the dessicating fibers of the stomach wall and the presumed location of the gastroesophageal junction. Every patient who has an esophageal myotomy requires intraoperative esophagogastroscopy to accurately determine the location of the gastroesophageal junction since we have found that it is impossible to accurately determine this laparoscopically. The intraoperative esophagogastroscopy serves two functions: (1) to ensure that the myotomy extends beyond the gastroesophageal junction to totally relieve any distal obstruction, and (2) to ensure that there is no iatrogenic perforation of the mucosa prior to closure. Once the myotomy is complete, the muscular layer is swept laterally to expose approxi274
mately 1.5 cm of the mucosa. The site is inspected for bleeding and the partial fundoplication can then proceed. Collis esophagoplasty. Once the initial dissection has been completed, an endoscopic ruler is used to measure 6 cm from the angle of His interiorly along the lesser curve to mark the site for the distal most portion of the esophagoplasty. The anterior stomach is scored in this location to indicate where the 21-stapler will be passed through both walls of the stomach. A #50 bougie is passed along the lesser curve into the stomach to use as a stent for formation of the neoesophagus. A 2-cm incision is made in the midclavicular line, just below the costal margin, for insertion of the stapler and anvil and, ultimately, the circular stapler. The anvil is inserted initially with a blue sharp spike in place in order to position it through both walls of the stomach. The anvil is positioned posteriorly behind the stomach in the lesser sac, and the sharp point is placed against the #50 bougie. It is rolled off the bougie superiorly toward the operating room ceiling and pushed anteriorly through the stomach so that the anvil passes through both the posterior and anterior walls of the stomach. The blue spike is removed, and a #21 stapler is inserted through the 2-cm incision, mated with the circular anvil head, and fired. This creates a through-and-through opening in the stomach 6 cm below the angle of His directly adjacent to the #50 bougie. An 18-mm trocar is then inserted through the 2-cm incision. A 60-mm endoscopic stapler is inserted through the circular opening and pressed against the side of the bougie superiorly with traction by the assistant and then fired, usually for a total of two firings, to create the neoesophagus. At this point, the staple line is carefully inspected for bleeding and any weak areas that need to be oversewn. The new neofundus can then be used to perform the Nissen fundoplication at its inferior aspect. Thoracoscopic Belsey fundoplication. Our technique of thoracoscopic Belsey fundoplication has been described in detail previously.6 Paraesophageal hernia repair. Paraesophageal, or type 2, hiatal hernias are treated differently than the routine-type hernia. First, the abdominal contents are carefully withdrawn from the chest, pulling posteriorly. Furthermore, by working in either a clockwise or counter-clockwise method, the intrathoracic contents should be carefully removed in an almost stepwise manner. Once reduced, the crura are then carefully identified. Starting on the anterior portion of the crura, the sac is grasped and pulled into the abdominal cavity. It is then incised anteriorly from approximately 1 to 3 o’clock and 12 to 9 o’clock. The intrathoracic fibrous tissue is carefully dissected using electrocautery, taking care not to injure the pleura. Once the sac has been completely reduced from 9 to 3 o’clock anteriorly, a bougie and/or gastroscope is advanced down the esophagus to correctly identify the esophagus. Once the sac has been completely reduced, the right crus is dissected down to its coalescence posteriorly with the left crus (median arcuate ligament), thus freeing the window posteriorly to the esophagus. In some cases, one must first dissect the left crus, then the right crus in an alternate manner, until the crura are completely cleaned. Bipolar current is used to coagulate some of the larger vessels encountered in the chest.
THE AMERICAN JOURNAL OF SURGERY® VOLUME 175 APRIL 1998
MINIMALLY INVASIVE ANTIREFLUX SURGERY/MCKERNAN AND CHAMPION
TABLE V Short-Term Complications with Minimally Invasive Antireflux Surgery Nature of Complication Esophageal leak (3 Nissen, 3 Toupet, 2 Belsey) Splenic injury Conversion to open procedure Postoperative bleeding requiring reexploration Pneumonia requiring readmission Acute gastric perforation Acute gastric distention Pericardial syndrome Subphrenic abscess Overall complication rate
Number of Patients 8 3 4 2 2 1 1 3 3 27 (2.7%)
After the esophagus has been mobilized, the crura are completely closed posteriorly with permanent sutures. If the hiatus is extremely large, mesh can be used to either buttress a suture or close the hiatus. Care should be taken not to allow the esophagus to touch the mesh.
RESULTS Operative Results Operative time averaged 64 minutes, with a range of 23 to 376 minutes; this includes all patients, including those who had more than one procedure. Some variation was procedure-dependent, with the Nissen fundoplication typically requiring the least amount of time to perform and the Belsey repair requiring the longest time. Postoperative oral feedings were resumed within 4 hours in the majority of patients. In the relatively few patients for whom there was a concern concerning an esophageal injury or extensive dissection, the patient was held NPO overnight and a gastrographin swallow was performed the following morning before resuming liquids. Almost all patients (99%) were discharged within 24 hours of surgery, with instructions to observe a liquid diet on day 1, and a mechanical soft diet for the first 3 weeks. Most patients resumed a regular diet 3 weeks postoperatively. Complications Operative mortality was 0.1%. One patient died after suffering a myocardial infarction the day following surgery. The mean duration of follow-up was 33 months, with a range of 1 to 66 months. Postoperative morbidity can be separated into short-term and long-term morbidity. Overall, the complication rate related to the surgery itself was 2.7%. Short-term complications and the number of patients with each are listed in Table V. Long-term complication rates were calculated on the basis of postoperative dysphagia. Six percent of our patients required postoperative dilatation either for esophageal stenosis or as an ancillary procedure for an esophageal motility disorder. This complaint was generally short-lived, with only 2% of the patients reporting dysphagia more than 6 weeks after surgery and 1% experiencing persistent dysphagia for more than 6 months postoperatively after a Nissen fundoplication. Five patients (0.5%) required conversion
to a partial wrap due to persistant dysphagia that failed to respond to multiple dilatations. Recurrence Rate Thirty-eight of our patients (3.8%) demonstrated recurrent GERD on barium swallow, EGD, or 24-hour pH studies. Of these 38 recurrences, 19 were in patients who had undergone a Toupet fundoplication (5.5% of the cohort having this procedure), 14 (2.2%) were in patients who had had a Nissen fundoplication, 3 (13.6%) were the result of a failed Belsey repair, and 2 (6.3%) were the result of a failed paraesophageal repair. Twenty-seven of the 38 recurrences required reoperation, and 2 additional patients are awaiting a second procedure. The remaining 9 patients elected not to have a second repair. Evaluation for recurrent GERD was only performed in symptomatic patients; therefore, a follow-up questionnaire was sent to all patients in our cohort who had undergone a videoscopic antireflux procedure between February 1, 1994, and January 31, 1997. A total of 754 surveys were mailed, and the number of responses was 324 (43%). Of those patients who responded to the survey, 34 indicated that they had significant reflux symptoms. These patients have been contacted for testing for documentation of reflux; at this time, results are not available.
DISCUSSION Our experience in performing 1,003 videoscopic antireflux procedures confirms that these procedures are as safe and effective as their open counterparts. Only 3.8% of our patients had a recurrence of reflux or a hiatal hernia that was evaluated for reoperation despite a follow-up period that averaged almost 3 years. This compares favorably to the 3% recurrence rate reported for a large cohort of 298 patients undergoing a conventional operation (eg, Nissen fundoplication, Belsey repair) for reflux disease other than esophageal stricture.10 Only 6% of our patients reported dysphagia following laparoscopic antireflux surgery. This symptom was shortlived in most patients, with fewer than 1 in 100 patients reporting dysphagia more than 6 months after surgery. We experienced a 0.1% mortality rate with our minimally invasive antireflux techniques and our operative complication rate was low (2.7%). This is comparable to or lower than the morbidity rates reported in other larger series of laparoscopic Nissen procedures.4 – 6 The types of operative complications seen (ie, pneumonia, splenic injury, acute gastric distention, postoperative bleeding) were similar in type and reported incidence rate to those reported with open antireflux procedures11 or by those performing these repairs laparoscopically.4 –7 The only operative complication that occurred at a rate of 1% was esophageal leaks. Of the 8 patients with esophageal leaks, 3 were managed conservatively with antibiotics and 5 underwent operative exploration, with two repairs, one resection, and two drainage procedures. Both of the operative repairs subsequently broke down; however, both healed with conservative therapy consisting of intravenous hyperalimentation, antibiotics, and drainage. None of our conservatively managed esophageal leaks progressed to sepsis. Esophageal leaks have been reported in other large series involving antireflux surgery.12–14 In a 10-year retrospective review of 1,005 open antireflux procedures, Ur-
THE AMERICAN JOURNAL OF SURGERY® VOLUME 175 APRIL 1998
275
MINIMALLY INVASIVE ANTIREFLUX SURGERY/MCKERNAN AND CHAMPION
schel12 reported postoperative esophageal leaks in 12 (1.2%) patients. An incomplete fundal wrap (either a Belsey or partial fundoplication) was indicated as a significant predisposing factor for this complication, since 10 of Urschel’s 12 esophageal leaks occurred in patients who had undergone a partial fundoplication. Our experience was similar, with 5 of our 8 leaks associated with partial fundoplication techniques (3 with a Toupet fundoplication and 2 with a Belsey repair). The apparent preponderance of esophageal leaks associated with partial fundoplication procedures may be partially a function of the patient’s underlying disease, since we utilized the partial wrap primarily in those individuals with underlying motility disorders, strictures, or complex GERD. Our experience, as well as that of others, indicates that laparoscopic antireflux surgery provides many of the same benefits associated with other minimally invasive procedures. Virtually all of our patients could be discharged within 24 hours of surgery. Patients typically were able to resume oral feedings within 4 hours of surgery and a normal diet within 3 weeks. In directly comparing laparoscopic and open Nissen repairs, Peters et al6 reported that laparoscopic surgery was associated with a significantly shorter hospitalization stay, and notably, a significantly greater augmentation of LES pressure. In summary, minimally invasive antireflux procedures appear to provide sustained symptomatic relief for patients with refractory gastroesophageal reflux disease, with a rapid recovery and a low incidence of complications.
REFERENCES 1. Gallup Organization National Survey. Heartburn Across America. Princeton, NJ: Gallup Organization Inc.; 1988.
276
2. Stein HJ, Barlow AP, DeMeester TR, Hinder RA. Complications of gastroesophageal reflux disease. Role of the lower esophageal sphincter, esophageal acid and acid/alkaline exposure, and duodenogastric reflux. Ann Surg. 1992;216:35– 43. 3. Dallemagne B, Weerts JM, Jahaes C, et al. Laparoscopic Nissen fundoplication: preliminary results. Surg Laparosc Endosc. 1991;3: 138 –143. 4. Hinder RA, Filipi CJ, Wetscher G, et al. Laparoscopic Nissen fundoplication is an effective treatment for gastroesophageal reflux disease. Ann Surg. 1994;220:472– 481. 5. Pitcher DE, Curet MJ, Martin DT, et al. Successful management of severe gastroesophageal reflux disease with laparoscopic Nissen fundoplication. Am J Surg. 1994;168:547–553. 6. Peters JH, Heimbucher J, Kauer WKH, et al. Clinical and physiologic comparison of laparoscopic and open Nissen fundoplication. J Am College Surg. 1995;180:385–393. 7. Cuschieri A, Hunder J, Wolfe B, et al. Multicenter prospective evaluation of laparoscopic antireflux surgery. Preliminary report. Surg Endosc. 1993;7:505–510. 8. McKernan JB, Champion JK. Laparoscopic antireflux surgery. Am Surg. 1995;61:530 –536. 9. Champion JK, McKernan JB. Thoracoscopy and gastroesophageal reflux disease. Probl Gen Surg. 1996;13:97–103. 10. Skinner DB. Surgical management after failed antireflux operation. World J Surg. 1992;16:359 –363. 11. Urschel JD. Complications of antireflux surgery. Am J Surg. 1993;165:68 –70. 12. Urschel JD. Gastroesophageal leaks after antireflux operations. Am Thorac Surg. 1994;57:1229 –1232. 13. Polk HC. Fundoplication for reflux esophagitis: misadventures with the operation of choice. Ann Surg. 1976;183:645– 652. 14. Skinner DB, Belsey RHR. Surgical management of esophageal reflux and hiatus hernia. Long-term results in 1,030 patients. J Thorac Cardiovasc Surg. 1967;53:33–54.
THE AMERICAN JOURNAL OF SURGERY® VOLUME 175 APRIL 1998
Minimally Invasive Antireflux Surgery J. Barry McKernan, MD, PhD, Marietta, Georgia, J. Kenneth Champion, MD, Macon, Georgia
BACKGROUND: Previous reports of minimally invasive antireflux surgery for gastroesophageal reflux disease (GERD) have been small, short-term series utilizing only a laparoscopic approach. We conducted a retrospective review and report our 66-month experience with more than 1,000 laparoscopic and thoracoscopic antireflux procedures. METHODS: Between September 1991 and October 1997, 968 adults underwent 1,003 minimally invasive antireflux procedures on a tailored basis, based on their preoperative evaluation. Procedures performed were laparoscopic Nissen (626), Toupet (348), paraesophageal (33), and thoracoscopic Belsey (22). A total of 23% (233) of patients underwent an ancillary procedure (esophageal myotomy 85, vagotomy 67, pyloromyotomy 13, and cholecystectomy 66). RESULTS: Follow-up averaged 33 months (range 1 to 66), operative mortality was 0.1%. Complications occurred in 2.7% with a 1% long-term dysphagia rate. Demonstrated recurrence rate was 3.8% to date, with an associated 3.4% reporting symptoms of GERD. CONCLUSION: Minimally invasive antireflux procedures provide sustained relief of GERD symptoms with low morbidity and rapid recovery. Am J Surg. 1998;175:271–276. © 1998 by Excerpta Medica, Inc.
G
astroesophageal reflux disease (GERD) is a common clinical problem and results in high rates of health care use. Surveys indicate that almost 45% of American adults experience heartburn, the cardinal symptom of GERD, at least once a month and 7% report having the symptom on a daily basis.1 The symptoms associated with GERD are caused by the reflux of gastric contents into the esophagus and its resultant complications. Approximately 50% of individuals with increased esophageal acid exposure experience complications such as erosive esophagitis, esophageal stricture, or Barrett’s esophagus.2 The ultimate therapeutic goal in managing GERD is to reduce reflux, minimize any damage to the esophageal
From the Department of Surgery (JBM), Medical College of Georgia, Augusta, Georgia, and the Department of Surgery (JKC), Mercer University School of Medicine, Macon, Georgia. Requests for reprints should be addressed to J. Barry McKernan, MD, PhD, 130 Vann Street, Suite 220, Marietta, Georgia 30060. Manuscript submitted December 16, 1997 and accepted in revised form February 3, 1998.
© 1998 by Excerpta Medica, Inc. All rights reserved.
mucosal lining, and provide complete symptomatic resolution without the introduction of new complaints. In many patients the symptoms of GERD can be effectively managed through diet and lifestyle modifications. Individuals whose symptoms are inadequately relieved by these measures or with severe, chronic esophageal reflux may require the use of medications that suppress gastric acid secretion such as histamine H2 receptor antagonists or proton pump inhibitors, or the use of pro-motility agents such as cisapride. A subset of patients are candidates for surgery because of severe reflux and a poor response to medical therapy. The indications for the surgical management of GERD are listed in Table I. Since the first published report of a fundoplication procedure performed laparoscopically appeared in 1991,3 several surgeons—including ourselves— have used minimally invasive techniques to treat GERD. Building upon our early experience with Nissen and Toupet fundoplications, we expanded our repertoire of minimally invasive techniques to include thoracoscopic Belsey partial fundoplication and videoscopic Collis esophagoplasty. While there have been numerous reports in the literature detailing experience with laparoscopic antireflux procedures, most of these have been relatively small series, involving fewer than 75 patients who were followed up, on average, for less than 1 year.4 –7 We have now analyzed our experience in performing 1,003 minimally invasive antireflux procedures on 968 patients, 695 of whom have been followed up for more than 1 year. In this paper we report our surgical techniques and results.
MATERIALS AND METHODS Patient Population Between September 1991 and October 1997, 1,003 minimally invasive antireflux procedures were performed on a total of 968 adults with symptomatic GERD who were refractory to or intolerant of medical therapy. Fifty-three percent of the cohort were male. The average age of patients was 52 years (range 19 to 85), and 20% were 65 years of age or older. Body weight ranged from 102 to 344 pounds, with the average weight being 178 pounds. Manometry performed prior to surgery revealed that the lower esophageal sphincter (LES) pressure in our patient group averaged 14 mm Hg (range 0 to 70). Sixty percent of our patients had a motility disorder, including nonspecific motility disorders (20%), low amplitude esophageal contraction (20%), hypertensive LES pressure (8%), nutcracker esophagus (8%), and diffuse esophageal spasm (2%). Achalasia was seen in 2% of our patients. Preoperative Evaluations Mandatory and selective tests. Esophagogastroduodenoscopy (EGD) and esophageal manometry are now routinely performed on every surgical candidate, with the exception 0002-9610/98/$19.00 PII S0002-9610(98)00034-8
271
MINIMALLY INVASIVE ANTIREFLUX SURGERY/MCKERNAN AND CHAMPION
TABLE I
TABLE II Indications for Antireflux Surgery
Refractory to medical therapy ● Persistant symptoms after 12 weeks of maximal medical therapy ● Recurrent symptoms upon cessation of medication ● Noncompliance with medical therapy Development of complications ● Stricture ● Barrett’s mucosa ● Esophagitis ● Aspiration ● Motility disorder
of patients having an associated stricture or paraesophageal hernia that will not permit passage of a manometry catheter. In patients seen before 1993 esophageal manometry was considered optional.8 During EGD, the presence of esophagitis is noted. Biopsies may be performed and complications such as ulceration, stricture, or Barrett’s mucosa are documented. This is particularly important in instances of Barrett’s mucosa as these patients require life-long monitoring since they are at increased risk for adenocarcinoma of the esophagus. Additionally, the length of the esophagus and the presence of a hiatal defect are noted. During esophageal manometry, we routinely measure LES pressure and perform an esophageal body motility study to characterize any motility disorder. Twenty-four-hour pH monitoring, barium swallow, and gastric emptying studies were used selectively in our series. A 24-hour pH monitoring study was indicated only in the absence of esophagitis upon EGD and normal esophageal manometry results. Only 11% of our patients underwent pH monitoring. A barium swallow test was used to confirm results of an EGD performed at another institution. This test is also useful in patients with stricture or with a history of esophageal diverticulum in order to accurately characterize the underlying pathology. A gastric emptying study was performed only if there was a history of gastric disease, frequent vomiting, or findings on EGD suggestive of gastric stasis. Operative Approaches We currently employ a selective approach to treating GERD, tailoring the antireflux procedure to each patient’s underlying anatomic and functional defect. The types of videoscopic antireflux procedures we have performed include Nissen fundoplication, Toupet fundoplication, paraesophageal hernia repair, and thoracoscopic Belsey repair. The general guidelines we use for determining which procedure to offer patients are detailed in Table II, while the number of procedures we have performed through October 20, 1997, are displayed by type in Table III. In addition to the primary antireflux procedure, 23% of our patients underwent an ancillary procedure (Table IV). A small number of patients (n 5 22) required a prosthetic crural repair when the crural size exceeded 5 cm in diameter after a prior failed surgery. In addition, 6% of our patients underwent cholecystectomy for coexisting cholelithiasis at the time of their GERD surgery. 272
Criteria for Selecting Minimally Invasive Antireflux Procedure Nissen fundoplication ● Indicated in patients without a primary motility disorder ● Indicated in patients with a nonspecific motility disorder Toupet fundoplication ● Indicated for cases where low amplitude peristalsis would interfere with esophageal emptying ● Indicated as the abdominal laparoscopic approach to be used in conjunction with esophageal myotomy for a primary motility disorder Thoracoscopic Belsey partial fundoplication ● Indicated in patients who had undergone extensive upper abdominal or previous antireflux surgery, particularly in the presence of a primary esophageal disorder requiring simultaneous repair ● Selected in cases where concomitant intrathoracic pathology could be accessed at the same time through a unilateral thoracoscopic approach
TABLE III Minimally Invasive Antireflux Procedures (9/30/91 to 10/20/97) Total Number of Procedures
1,003
Types of procedures Nissen fundoplication Toupet fundoplication Paraesophageal hernia repair Thoracoscopic Belsey repair
626* 348† 33‡ 22
* Eighteen were Collis-Nissen procedures. † Two were Collis-Toupet procedures. ‡ Seven of these repairs did not include a fundal wrap and thus are counted as unique cases in the overall total of 1,003 procedures. Twenty-three of the paraesophageal repairs were performed in conjunction with a Nissen procedure, two in conjunction with a Toupet procedure, and one in conjunction with a Belsey repair.
Over the 6 years of this series, 9 patients (0.9%) with previous gastric surgery (vertical banded gastroplasty or subtotal gastrectomy) were not deemed candidates for a fundoplication procedure and underwent a Roux-en-Y gastric diversion for intractable reflux. Operative Techniques Initial dissection. Patients are admitted on the morning of surgery, and are given prophylactic antibiotics and sequential compression hose to prevent deep venous thrombosis. Patients are placed in the supine position under general anesthesia. Six trocars are routinely used, with one 10-mm and five 5-mm ports. A 10-mm incision is made just to the left midline, 15 cm below the xiphoid process. A 0-degree scope with a 10-mm Optiview optical surgical obturator (Ethicon Endo-Surgery, Inc., Cincinnati, Ohio) is passed through the subcutaneous tissue, anterior rectus sheath, rectus muscle, and posterior rectus sheath, and the abdomen is entered. The Optiview trocar allows each fascial layer to be identified and any vessel or adherent bowel to be visualized, thereby preventing inadvertent injury. Once the Optiview trocar has been inserted, insufflation
THE AMERICAN JOURNAL OF SURGERY® VOLUME 175 APRIL 1998
MINIMALLY INVASIVE ANTIREFLUX SURGERY/MCKERNAN AND CHAMPION
TABLE IV Ancillary Surgical Procedures Procedure Esophageal Myotomy Total procedures in conjunction with: Toupet fundoplication (54) Belsey repair (14) Nissen fundoplication (14) Collis-Nissen fundoplication (1) No-wrap paraesophageal repair (2) Cricopharyngeal myotomy Vagotomy Highly selective vagotomy (HSV) HSV 1 Nissen fundoplication (30) HSV 1 Toupet fundoplication (12) HSV 1 pyloromyotomy (4) HSV 1 pyloroplasty (2) HSV 1 Jabouley gastroduodenostomy (1) Truncal vagotomy Truncal vagotomy 1 Nissen fundoplication (4) Truncal vagotomy 1 Belsey repair (2) Truncal vagotomy 1 esophagogastrectomy (2) Truncal vagotomy 1 pyloromyotomy (1) Pyloromyotomy Total procedures in conjunction with: Nissen fundoplication (9) Toupet fundoplication (3) Collis-Nissen fundoplication (1) Cholecystectomy Total
Number Performed 85
2 56
11
13
66 233
is begun to 15 mm Hg. The remaining 5-mm ports are inserted under direct visualization. An Allis clamp is placed through the diaphragm just above the apex of the esophageal hiatus to allow for liver retraction. This technique is less traumatic, requires no assistant to hold, and even works well in cases of hepatomegaly. The primary surgeon and assistant each utilize a twohanded technique, which enhances exposure and speeds the operation. The patient is placed in a reverse Trendelenberg position at this point, with the back elevated at approximately 30 degrees. The 10-mm, 0-degree scope is exchanged for a 45-degree angled scope, which is essential for performing the dissection. Dissection is begun along the left crus by dividing the peritoneal attachments along the fundus, angle of His, and lateral left crus to aid in the ultimate mobilization of the esophagus. Attention is then turned to the lesser omentum; this is opened directly over the caudate lobe, thereby exposing the caudate lobe and right crus. The hepatic branch of the vagus nerve is routinely sacrificed, but any vessel in the lesser omentum that is larger than 5 mm is typically preserved. This is important since in approximately 10% of patients, a large branch of the left hepatic artery can transverse this area and its sacrifice may result in necrosis of the liver. Once the right crus has been identified, the peritoneal attachment along its anterior border is divided. The dissection begins in the mid-body of the crus and is then carried inferiorly until the median arcuate ligament is identified. Blunt dissection is used to proceed up the anterior border of the left crus until a window has been created
behind the esophagus and the esophagus is completely mobilized. Turning the 45-degree scope laterally provides excellent visualization of this portion of the dissection and prevents inadvertent perforation of either the esophagus or posterior stomach. A blue vessel loop is passed around the esophagus for traction and is secured with a chromic Endoloop. The assistant can then use the Endo-loop for retraction to aid in the dissection. This dissection is performed with only an nasogastric tube in place since the presence of a bougie increases the likelihood of perforation. With the assistant retracting the esophagus bilaterally, dissection is begun along the anterior border of the right crus, and the phrenoeophageal ligament is circumferentially divided along the border of the crura until the esophagus is completely mobilized. Both the anterior and posterior vagus nerves are identified and preserved, and are usually left adherent to the esophagus. Using the 45-degree angled scope and the laparoscopic technique, we now know that it is possible to mobilize the esophagus all the way to the inferior pulmonary vein, which crosses the esophagus anteriorly at 12 o’clock. Care must be exercised in this area to avoid unintentional bleeding. Once the esophagus has been mobilized, the short gastric vessels are divided along the upper 10 cm of the fundus from the angle of His, extending laterally. Routine division of these short gastric vessels ensures a loose, floppy fundus and decreases the incidence of postoperative dysphagia. We have utilized metal clips, vascular staples, harmonic scalpel, and bipolar forceps to achieve hemostasis. The harmonic scalpel or bipolar forceps were used in our last 600 cases with no bleeding complications. Once the greater curve has been mobilized and the esophagus is mobilized, either a Nissen fundoplication or a Toupet fundoplication can be performed. We routinely measure the crural opening size with a endoscopic ruler and close the crura with interrupted sutures. With the #50 bougie in place, the crura are reapproximated with interrupted size 0, Ethibond Endoknot sutures, and tied extracorporally until they are closed loosely about the #50 bougie. Nissen fundoplication. If a Nissen fundoplication is being performed, the bougie is carefully withdrawn back into the middle chest. Placing a grasper behind the esophagus, the fundus is pulled posteriorly behind the esophagus, so that the two edges of the stomach will circumferentially encircle the esophagus just above the gastroesophageal junction. The #50 bougie is passed into the stomach under direct visualization. The Nissen repair is then undertaken by suturing the stomach laterally to the esophagus 2 cm above the gastroesophageal junction, taking care to avoid the anterior vagus nerve. The suture is passed through the stomach medially and tied down extracorporeally without tension. The next suture is placed through the stomach laterally, through the esophagus at the gastroesophageal junction anteriorly, and through the stomach medially. A size 0, Ethibond Endoknot is used to complete the repair, which is tied extracorporeally. Toupet fundoplication. In the event that a Toupet fundoplication is indicated, a grasper is placed behind the esophagus and the fundus is grasped and retracted medially
THE AMERICAN JOURNAL OF SURGERY® VOLUME 175 APRIL 1998
273
MINIMALLY INVASIVE ANTIREFLUX SURGERY/MCKERNAN AND CHAMPION
behind the esophagus so that a 270-degree wrap can be performed. The first suture in the Toupet fundoplication is placed through the fundus posteriorly and through the left crura using a size 0 Ethibond Endoknot, tied extracorporeally. A second suture is placed through the wrap posteriorly and through the right crus. At 6 o’clock the medial portion of the wrap is anchored using a size 0 Ethibond Endoknot so that the two sides will be sutured at the esophagus at approximately 10 o’clock and 2 o’clock to create the 270degree wrap. The first suture is placed through the esophagus, 2 cm above the gastroesophageal junction medially through the upper portion of the wrap, and then through the crura at 10 o’clock in order to anchor the wrap. To secure the medial portion of the wrap, the second suture is placed through the esophagus medially at the gastroesophageal junction and stomach. The lateral portion of the wrap is completed by placing a suture through the stomach laterally and then through the crus at 2 o’clock and through the lateral esophagus 2 cm above the gastroesophageal junction. The last suture is placed laterally through the stomach and esophagus at the gastroesophageal junction to complete the 270-degree wrap. Esophagomyotomy. Once dissection has been completed and prior to closure of the esophageal hiatus, the anterior esophagus is exposed between 11 and 12 o’clock. This area avoids the anterior vagus nerve. Beginning approximately 2 cm above the gastroesophageal junction, the longitudinal fibers in the first muscular layer of the esophagus are bluntly dissected and separated with a scissors tip. We use dissecting endoscopic scissors with no cautery since we find that cautery is rarely needed on the small vessels to the esophagus. Once the longitudinal fibers have been bluntly separated, the circular fibers become exposed. These are divided under direct visualization. The assistant uses a suction irrigator to keep the field clear for dissection. Once the circular fibers are divided down to the mucosa, the mucosa can be bluntly pushed inferiorly, and the dissection can proceed in a cephalad direction for approximately 2 cm more. The total length of the myotomy will depend on the indication for the procedure. For a primary motility disorder such as achalasia, nutcracker esophagus, or diffuse esophageal spasm, a length of 6 cm is usually sufficient. For a hypertensive LES, typically only a 4 cm myotomy is needed to relieve the obstruction. Once the proximal portion of the myotomy is completed, the more distal segment, which involves the gastroesophageal junction, is approached. Dissection is carried inferiorly until it impinges upon the dessicating fibers of the stomach wall and the presumed location of the gastroesophageal junction. Every patient who has an esophageal myotomy requires intraoperative esophagogastroscopy to accurately determine the location of the gastroesophageal junction since we have found that it is impossible to accurately determine this laparoscopically. The intraoperative esophagogastroscopy serves two functions: (1) to ensure that the myotomy extends beyond the gastroesophageal junction to totally relieve any distal obstruction, and (2) to ensure that there is no iatrogenic perforation of the mucosa prior to closure. Once the myotomy is complete, the muscular layer is swept laterally to expose approxi274
mately 1.5 cm of the mucosa. The site is inspected for bleeding and the partial fundoplication can then proceed. Collis esophagoplasty. Once the initial dissection has been completed, an endoscopic ruler is used to measure 6 cm from the angle of His interiorly along the lesser curve to mark the site for the distal most portion of the esophagoplasty. The anterior stomach is scored in this location to indicate where the 21-stapler will be passed through both walls of the stomach. A #50 bougie is passed along the lesser curve into the stomach to use as a stent for formation of the neoesophagus. A 2-cm incision is made in the midclavicular line, just below the costal margin, for insertion of the stapler and anvil and, ultimately, the circular stapler. The anvil is inserted initially with a blue sharp spike in place in order to position it through both walls of the stomach. The anvil is positioned posteriorly behind the stomach in the lesser sac, and the sharp point is placed against the #50 bougie. It is rolled off the bougie superiorly toward the operating room ceiling and pushed anteriorly through the stomach so that the anvil passes through both the posterior and anterior walls of the stomach. The blue spike is removed, and a #21 stapler is inserted through the 2-cm incision, mated with the circular anvil head, and fired. This creates a through-and-through opening in the stomach 6 cm below the angle of His directly adjacent to the #50 bougie. An 18-mm trocar is then inserted through the 2-cm incision. A 60-mm endoscopic stapler is inserted through the circular opening and pressed against the side of the bougie superiorly with traction by the assistant and then fired, usually for a total of two firings, to create the neoesophagus. At this point, the staple line is carefully inspected for bleeding and any weak areas that need to be oversewn. The new neofundus can then be used to perform the Nissen fundoplication at its inferior aspect. Thoracoscopic Belsey fundoplication. Our technique of thoracoscopic Belsey fundoplication has been described in detail previously.6 Paraesophageal hernia repair. Paraesophageal, or type 2, hiatal hernias are treated differently than the routine-type hernia. First, the abdominal contents are carefully withdrawn from the chest, pulling posteriorly. Furthermore, by working in either a clockwise or counter-clockwise method, the intrathoracic contents should be carefully removed in an almost stepwise manner. Once reduced, the crura are then carefully identified. Starting on the anterior portion of the crura, the sac is grasped and pulled into the abdominal cavity. It is then incised anteriorly from approximately 1 to 3 o’clock and 12 to 9 o’clock. The intrathoracic fibrous tissue is carefully dissected using electrocautery, taking care not to injure the pleura. Once the sac has been completely reduced from 9 to 3 o’clock anteriorly, a bougie and/or gastroscope is advanced down the esophagus to correctly identify the esophagus. Once the sac has been completely reduced, the right crus is dissected down to its coalescence posteriorly with the left crus (median arcuate ligament), thus freeing the window posteriorly to the esophagus. In some cases, one must first dissect the left crus, then the right crus in an alternate manner, until the crura are completely cleaned. Bipolar current is used to coagulate some of the larger vessels encountered in the chest.
THE AMERICAN JOURNAL OF SURGERY® VOLUME 175 APRIL 1998
MINIMALLY INVASIVE ANTIREFLUX SURGERY/MCKERNAN AND CHAMPION
TABLE V Short-Term Complications with Minimally Invasive Antireflux Surgery Nature of Complication Esophageal leak (3 Nissen, 3 Toupet, 2 Belsey) Splenic injury Conversion to open procedure Postoperative bleeding requiring reexploration Pneumonia requiring readmission Acute gastric perforation Acute gastric distention Pericardial syndrome Subphrenic abscess Overall complication rate
Number of Patients 8 3 4 2 2 1 1 3 3 27 (2.7%)
After the esophagus has been mobilized, the crura are completely closed posteriorly with permanent sutures. If the hiatus is extremely large, mesh can be used to either buttress a suture or close the hiatus. Care should be taken not to allow the esophagus to touch the mesh.
RESULTS Operative Results Operative time averaged 64 minutes, with a range of 23 to 376 minutes; this includes all patients, including those who had more than one procedure. Some variation was procedure-dependent, with the Nissen fundoplication typically requiring the least amount of time to perform and the Belsey repair requiring the longest time. Postoperative oral feedings were resumed within 4 hours in the majority of patients. In the relatively few patients for whom there was a concern concerning an esophageal injury or extensive dissection, the patient was held NPO overnight and a gastrographin swallow was performed the following morning before resuming liquids. Almost all patients (99%) were discharged within 24 hours of surgery, with instructions to observe a liquid diet on day 1, and a mechanical soft diet for the first 3 weeks. Most patients resumed a regular diet 3 weeks postoperatively. Complications Operative mortality was 0.1%. One patient died after suffering a myocardial infarction the day following surgery. The mean duration of follow-up was 33 months, with a range of 1 to 66 months. Postoperative morbidity can be separated into short-term and long-term morbidity. Overall, the complication rate related to the surgery itself was 2.7%. Short-term complications and the number of patients with each are listed in Table V. Long-term complication rates were calculated on the basis of postoperative dysphagia. Six percent of our patients required postoperative dilatation either for esophageal stenosis or as an ancillary procedure for an esophageal motility disorder. This complaint was generally short-lived, with only 2% of the patients reporting dysphagia more than 6 weeks after surgery and 1% experiencing persistent dysphagia for more than 6 months postoperatively after a Nissen fundoplication. Five patients (0.5%) required conversion
to a partial wrap due to persistant dysphagia that failed to respond to multiple dilatations. Recurrence Rate Thirty-eight of our patients (3.8%) demonstrated recurrent GERD on barium swallow, EGD, or 24-hour pH studies. Of these 38 recurrences, 19 were in patients who had undergone a Toupet fundoplication (5.5% of the cohort having this procedure), 14 (2.2%) were in patients who had had a Nissen fundoplication, 3 (13.6%) were the result of a failed Belsey repair, and 2 (6.3%) were the result of a failed paraesophageal repair. Twenty-seven of the 38 recurrences required reoperation, and 2 additional patients are awaiting a second procedure. The remaining 9 patients elected not to have a second repair. Evaluation for recurrent GERD was only performed in symptomatic patients; therefore, a follow-up questionnaire was sent to all patients in our cohort who had undergone a videoscopic antireflux procedure between February 1, 1994, and January 31, 1997. A total of 754 surveys were mailed, and the number of responses was 324 (43%). Of those patients who responded to the survey, 34 indicated that they had significant reflux symptoms. These patients have been contacted for testing for documentation of reflux; at this time, results are not available.
DISCUSSION Our experience in performing 1,003 videoscopic antireflux procedures confirms that these procedures are as safe and effective as their open counterparts. Only 3.8% of our patients had a recurrence of reflux or a hiatal hernia that was evaluated for reoperation despite a follow-up period that averaged almost 3 years. This compares favorably to the 3% recurrence rate reported for a large cohort of 298 patients undergoing a conventional operation (eg, Nissen fundoplication, Belsey repair) for reflux disease other than esophageal stricture.10 Only 6% of our patients reported dysphagia following laparoscopic antireflux surgery. This symptom was shortlived in most patients, with fewer than 1 in 100 patients reporting dysphagia more than 6 months after surgery. We experienced a 0.1% mortality rate with our minimally invasive antireflux techniques and our operative complication rate was low (2.7%). This is comparable to or lower than the morbidity rates reported in other larger series of laparoscopic Nissen procedures.4 – 6 The types of operative complications seen (ie, pneumonia, splenic injury, acute gastric distention, postoperative bleeding) were similar in type and reported incidence rate to those reported with open antireflux procedures11 or by those performing these repairs laparoscopically.4 –7 The only operative complication that occurred at a rate of 1% was esophageal leaks. Of the 8 patients with esophageal leaks, 3 were managed conservatively with antibiotics and 5 underwent operative exploration, with two repairs, one resection, and two drainage procedures. Both of the operative repairs subsequently broke down; however, both healed with conservative therapy consisting of intravenous hyperalimentation, antibiotics, and drainage. None of our conservatively managed esophageal leaks progressed to sepsis. Esophageal leaks have been reported in other large series involving antireflux surgery.12–14 In a 10-year retrospective review of 1,005 open antireflux procedures, Ur-
THE AMERICAN JOURNAL OF SURGERY® VOLUME 175 APRIL 1998
275
MINIMALLY INVASIVE ANTIREFLUX SURGERY/MCKERNAN AND CHAMPION
schel12 reported postoperative esophageal leaks in 12 (1.2%) patients. An incomplete fundal wrap (either a Belsey or partial fundoplication) was indicated as a significant predisposing factor for this complication, since 10 of Urschel’s 12 esophageal leaks occurred in patients who had undergone a partial fundoplication. Our experience was similar, with 5 of our 8 leaks associated with partial fundoplication techniques (3 with a Toupet fundoplication and 2 with a Belsey repair). The apparent preponderance of esophageal leaks associated with partial fundoplication procedures may be partially a function of the patient’s underlying disease, since we utilized the partial wrap primarily in those individuals with underlying motility disorders, strictures, or complex GERD. Our experience, as well as that of others, indicates that laparoscopic antireflux surgery provides many of the same benefits associated with other minimally invasive procedures. Virtually all of our patients could be discharged within 24 hours of surgery. Patients typically were able to resume oral feedings within 4 hours of surgery and a normal diet within 3 weeks. In directly comparing laparoscopic and open Nissen repairs, Peters et al6 reported that laparoscopic surgery was associated with a significantly shorter hospitalization stay, and notably, a significantly greater augmentation of LES pressure. In summary, minimally invasive antireflux procedures appear to provide sustained symptomatic relief for patients with refractory gastroesophageal reflux disease, with a rapid recovery and a low incidence of complications.
REFERENCES 1. Gallup Organization National Survey. Heartburn Across America. Princeton, NJ: Gallup Organization Inc.; 1988.
276
2. Stein HJ, Barlow AP, DeMeester TR, Hinder RA. Complications of gastroesophageal reflux disease. Role of the lower esophageal sphincter, esophageal acid and acid/alkaline exposure, and duodenogastric reflux. Ann Surg. 1992;216:35– 43. 3. Dallemagne B, Weerts JM, Jahaes C, et al. Laparoscopic Nissen fundoplication: preliminary results. Surg Laparosc Endosc. 1991;3: 138 –143. 4. Hinder RA, Filipi CJ, Wetscher G, et al. Laparoscopic Nissen fundoplication is an effective treatment for gastroesophageal reflux disease. Ann Surg. 1994;220:472– 481. 5. Pitcher DE, Curet MJ, Martin DT, et al. Successful management of severe gastroesophageal reflux disease with laparoscopic Nissen fundoplication. Am J Surg. 1994;168:547–553. 6. Peters JH, Heimbucher J, Kauer WKH, et al. Clinical and physiologic comparison of laparoscopic and open Nissen fundoplication. J Am College Surg. 1995;180:385–393. 7. Cuschieri A, Hunder J, Wolfe B, et al. Multicenter prospective evaluation of laparoscopic antireflux surgery. Preliminary report. Surg Endosc. 1993;7:505–510. 8. McKernan JB, Champion JK. Laparoscopic antireflux surgery. Am Surg. 1995;61:530 –536. 9. Champion JK, McKernan JB. Thoracoscopy and gastroesophageal reflux disease. Probl Gen Surg. 1996;13:97–103. 10. Skinner DB. Surgical management after failed antireflux operation. World J Surg. 1992;16:359 –363. 11. Urschel JD. Complications of antireflux surgery. Am J Surg. 1993;165:68 –70. 12. Urschel JD. Gastroesophageal leaks after antireflux operations. Am Thorac Surg. 1994;57:1229 –1232. 13. Polk HC. Fundoplication for reflux esophagitis: misadventures with the operation of choice. Ann Surg. 1976;183:645– 652. 14. Skinner DB, Belsey RHR. Surgical management of esophageal reflux and hiatus hernia. Long-term results in 1,030 patients. J Thorac Cardiovasc Surg. 1967;53:33–54.
THE AMERICAN JOURNAL OF SURGERY® VOLUME 175 APRIL 1998