Curet MJ, Josloff RK, Schoeb O, Zucker KA. Laparoscopic Reoperation for Failed Antireflux Procedures. Arch Surg. 1999;134(5):559-563. doi:10.1001/archsurg.134.5.559
Laparoscopic fundoplication has become the criterion standard for the surgical treatment of gastroesophageal reflux disease. Recently, several patients were referred with recurrent symptoms of gastroesophageal reflux disease or severe dysphagia following previous antireflux surgery for possible laparoscopic reoperation.
To determine the safety and efficacy of this procedure.
Case series, consecutive sample.
University-affiliated and community tertiary care hospitals.
Prospective study of 27 consecutive patients undergoing attempted laparoscopic reoperation for symptoms of recurrent gastroesophageal reflux disease or intractable dysphagia following antireflux surgery. Patients were available for follow-up for 1 to 60 months postoperatively.
All patients underwent preoperative workup and attempted laparoscopic reoperation for treatment of symptoms.
Main Outcome Measures
Data were collected on preoperative symptoms and evaluation, operative time, blood loss, time to regular diet, length of hospitalization, morbidity, mortality, and long-term results.
Twenty-six patients underwent successful laparoscopic operations, with no mortality and minimal morbidity. One patient underwent conversion to open laparotomy and then developed a proximal gastric leak, which was treated conservatively. Twenty-four patients began a liquid diet by postoperative day 1, and most were discharged from the hospital by postoperative day 3. One patient required dilation for postoperative dysphagia. The remaining patients are doing well and none have required treatment with acid-reducing medication.
Although technically challenging, laparoscopic reoperation for recurrent gastroesophageal reflux disease can be performed safely and with excellent results. In the hands of experienced endoscopic surgeons, patients who have undergone unsuccessful antireflux surgery should be offered laparoscopic reoperation.
NISSEN fundoplication has long been a popular and successful procedure for the treatment of severe gastroesophageal reflux disease (GERD), with control of symptoms in 85% to 90% or more of patients undergoing fundoplication.1- 3 Recently, several prospective studies have demonstrated that the operation can be performed laparoscopically in more than 95% of patients.4- 7 These studies have indicated that when the operation is performed by experienced laparoscopic surgeons, excellent symptomatic and physiological results can be achieved with minimal mortality and morbidity rates that are comparable to conventional open antireflux procedures.4- 7 In addition, many of the benefits of minimally invasive techniques, such as decreased pain, a shorter hospital stay, and quicker return to regular activity, have been recognized.4- 7 An interesting subset of patients has been identified in our practice: patients referred for laparoscopic surgery who have undergone previous antireflux surgery and who then developed recurrent symptoms of GERD or intractable dysphagia. Reoperation for failed antireflux surgery performed prior to the era of minimally invasive surgery has been discussed thoroughly in the literature. Data concerning reoperative laparoscopic antireflux surgery have been sparse. After performing more than 100 primary laparoscopic operations for GERD, we believed that patients undergoing reoperative correction could also benefit from a minimally invasive approach to their disease. The purposes of this study were to evaluate closely 27 patients who developed recurrent symptoms of GERD or intractable dysphagia after primary antireflux surgery to determine which patients were appropriate candidates for a laparoscopic reoperation, to determine if reoperative antireflux surgery was technically feasible, and to evaluate the safety and efficacy of the procedure.
From 1993 through 1998, our group evaluated 27 consecutive patients who had symptoms of recurrent GERD or intractable dysphagia after previous antireflux surgery. There were 14 women and 13 men ranging in age from 17 to 72 years (mean age, 46.2 years). Prior operations (Table 1) included 1 open Allison repair (anatomic), 1 open Allison repair with a concomitant truncal vagotomy and pyloroplasty, 21 Nissen fundoplications (13 open, 8 laparoscopic), 2 paraesophageal hernia repairs (1 laparoscopic, 1 open), 1 laparoscopic Toupet fundoplication, and 1 Heller myotomy and Belsey fundoplication done through a left thoracotomy. The initial antireflux operation had been performed 3 months to 30 years previously (mean, 7.2 years). Symptoms in all patients began 2 weeks to 25 years after the initial operation (mean, 2.3 years). Symptoms and preoperative evaluation are summarized in Table 1. Preoperative evaluation demonstrated a paraesophageal hernia in 4 patients, an excessively tight fundoplication in 2, a disrupted fundoplication in 9, a stricture in 2, a slipped wrap in 3 (1 of which had also herniated into the chest), an incomplete myotomy with a disrupted wrap in 1, and an intact fundoplication that had herniated into the chest in 2. We found no anatomic cause for the symptoms in 4 patients. It seemed that 3 patients simply had an inadequate fundoplication that now allowed for acid reflux, and 1 had severe gas bloating unresponsive to prokinetic agents or a laparoscopic pyloroplasty.
A laparoscopic Nissen fundoplication was performed in 14 patients, while 12 patients underwent a laparoscopic Toupet fundoplication (270° wrap). The remaining patient had a partial wrap performed via an open laparotomy after conversion secondary to gastric perforation and poor visualization. A partial wrap was performed in 3 patients with paraesophageal hernias, in 1 patient with recurrent achalasia, in 2 patients with previous tight Nissen fundoplication, in 1 patient with a stricture, in 2 patients with disrupted fundoplications where preoperative manometry demonstrated severely impaired esophageal peristalsis, in 2 patients who had undergone previous laparoscopic Nissen fundoplications that developed wrap migration (1 of whom also had a slipped fundoplication), and in 1 patient with severe gas bloating (Table 1).
A 5-puncture technique as described by Weerts et al4 and Pitcher et al5 was used in all cases. The surgical technique previously described for primary antireflux surgery was followed as closely as possible.5 In all cases, an open technique was used to enter the abdominal cavity with insertion of a Hasson trocar (Blunt Tip Trocar; Ethicon Endo-Surgery, Cincinnati, Ohio).8 Usually, adhesions needed to be lysed prior to insertion of the accessory trocars. A 301-angled laparoscope was routinely used for optimal visualization. Two high-flow carbon dioxide insufflators, each set at a maximum pressure of 15 mm Hg, were used to maintain optimal distension and visualization of the peritoneal cavity. A flexible upper endoscope was introduced and positioned at the distal esophagus to transilluminate important landmarks with the lighted tip, thus facilitating identification of the distal esophagus. In all patients, there were numerous adhesions to the left lobe of the liver, proximal stomach, distal esophagus, and diaphragmatic crura that needed to be dissected. Elevation of the liver and retraction of the stomach caudad and to the left were performed in the usual fashion. Blunt and sharp dissection were then used in the area of the distal esophagus and gastroesophageal junction to free up adhesions in this area. The distal esophagus was mobilized, and in all patients, the right and left crus of the diaphragm and both vagi were identified. An opening was then made in the retroesophageal space under direct visualization by alternately dissecting on the left and right side of the esophagus. A Penrose drain was placed around the esophagus to help retract the gastroesophageal junction inferiorly and to the patient's left. In all patients, the retroesophageal window was enlarged (6 × 6 cm minimum) to allow for a tension-free wrap. A minimum length of 5 cm of intra-abdominal esophagus was mobilized to prevent postoperative slippage or disruption of the wrap. In several instances, this required dissection well above the diaphragm. In all cases, the crura were closed with several interrupted sutures. Short gastric vessels and any adhesions found between the fundus of the stomach and the spleen were divided to achieve a tension-free wrap in all patients. The posterolateral edge of the greater curvature was then brought posteriorly through the retroesophageal window to create the fundoplication. A 56F to 60F bougie dilator was passed orally through the gastroesophageal junction to calibrate the lumen to prevent postoperative dysphagia. In the 14 patients undergoing a 360° wrap, a 2- to 3-cm fundoplication was then created using 3 interrupted sutures. Each suture incorporated the anterior wall of the esophagus while avoiding the left vagus nerve.
In the 12 patients undergoing a laparoscopic Toupet fundoplication, the fundus of the stomach was secured to each crus of the diaphragm and to the esophagus with interrupted sutures (a total of 9 stitches). At the termination of the procedure, the flexible endoscope was reintroduced in all patients and the esophagus and stomach were carefully examined. The upper abdomen was covered with a saline solution while the esophageal and gastric lumens were distended with air so that any full-thickness injuries would be apparent. All fascial openings were closed with absorbable sutures and skin edges were approximated with staples.
On termination of the procedure, the patients were taken to the floor and were not fed by mouth for the first night. The next day, an upper gastrointestinal gastrografin swallow procedure was performed, and if this demonstrated no leak or obstruction, patients were then fed clear liquids. Most patients were discharged from the hospital on day 2 or 3 and were allowed to return to their normal physical activities during the second postoperative week. The patients remained on full liquid diets for 2 to 3 weeks. All patients were seen in clinic within 2 weeks of discharge and were asked about symptoms of reflux and dysphagia. Patients were contacted by telephone at the time of this review and were questioned about use of acid-reducing medications or prokinetic agents. They were asked whether they were experiencing reflux symptoms, dysphagia, or other gastrointestinal symptoms. Finally, they were asked to rank their satisfaction with the results of the operation on a 3-point scale, where 1 indicated totally dissatisfied; 2, satisfied; and 3, very satisfied.
The fundoplication was completed laparoscopically in 26 patients. Mean operative time was 4.1 hours (range, 2.3-7.1 hours). Mean blood loss was 200 mL (range, 100-500 mL). No patient required perioperative transfusion. Complications are listed in Table 2. Five patients developed intraoperative pneumothoraces, which did not require a chest tube and did not interfere with completion of the laparoscopic fundoplication. One of these patients did require a Heimlich valve postoperatively. Another patient was noted to have an intraoperative lymph leak, which was controlled by clipping a branch of the thoracic duct. This patient experienced no postoperative problems. In 1 patient, the procedure was converted to open laparotomy because of poor visualization and 2 gastric perforations. The perforations were closed primarily after conversion. The patient later developed a proximal gastric leak that was treated successfully with closed suction catheter drainage. Four additional patients had either gastric or esophageal perforations, which were repaired intraoperatively. None of these patients developed a demonstrable postoperative leak, although 2 patients developed pleural effusions, 1 of whom eventually required a thoracotomy to treat an empyema.
The patient with the esophageal leak and 2 of the other 4 patients with an intestinal perforation were not fed by mouth postoperatively for 3 to 7 days. All other patients (including the other 2 patients with perforations) were given clear liquids on postoperative day 1 and were switched to a full liquid diet before discharge. Mean hospital stay was 3.7 days (range, 1-22 days). Three patients underwent prolonged hospitalization (>7 days). One patient, who had psychosis preoperatively, developed persistent nausea and vomiting following a Toupet fundoplication for a paraesophageal hernia. She eventually received biofeedback and was discharged from the hospital on postoperative day 9. The patient, whose operation had to be converted to open laparotomy and had a proximal gastric leak, was hospitalized for 22 days. A third patient with gastric perforation but no demonstrable leak postoperatively developed bilateral pleural effusions and respiratory distress. He was hospitalized for 9 days and was readmitted for surgical treatment of his left-sided thoracic empyema. Excluding these 3 patients, the mean length of hospital stay was 2.8 days (range, 1-6 days). The average time to return to normal activities after discharge was 9 days (range, 5-28 days). Follow-up data were available in all patients for a mean of 22 months (range, 1-60 months). All patients experienced complete relief of their reflux or obstructive symptoms. No patients are presently receiving acid-reducing medication. Four patients complained of transient gas bloating postoperatively, which resolved spontaneously 2 weeks to 4 months after surgery. One patient required dilation 5 weeks postoperatively for persistent dysphagia. One patient experienced delayed gastric emptying, which was controlled with prokinetic agents. All patients were satisfied or very satisfied with their results.
Reports of open reoperation for failed antireflux procedures (including Nissen, Hill, Belsey, and Alison fundoplications) have been published previously.9- 16 Reports of laparoscopic reoperation have been anecdotal.17- 21 The failure rate of open fundoplications is estimated at 6% to 20%, with the most common symptoms being recurrent reflux (30%-60%) or dysphagia (10%-20%).12,19,22- 24 Most patients can be managed medically, with only 4% to 6% requiring reoperation.12,18,23 Morbidity and mortality in these patients is higher than in primary antireflux surgery, with an average mortality rate of 3% (range, 2.5%-18%).9- 16,19,22,23 However, results are generally good, with success rates of 70% to 80% quoted for a first reoperation.11,13,22,23 Analysis of failures of laparoscopic fundoplications is limited by the short follow-up times available. However, it appears that failure rates, reoperation rates, and presenting symptomatology are similar in patients who undergo unsuccessful antireflux surgery whether the initial fundoplication was open or laparoscopic.18- 21
Reasons for failure have been well documented and generally are due to errors in technique, inexperience of surgeons, use of the wrong technique, or inappropriate patient selection.18,19,22,23 Symptoms generally begin within the first 2 years postoperatively.19,22 The most common cause for recurrent symptoms is breakdown of the wrap, which occurs in 25% to 35% of patients requiring reoperation. A wrap that is too tight or a stricture refractory to dilation leads to surgical revision 10% to 15% of the cases.10,12,13,19- 22 The wrap can herniate into the chest (5%-10%) or can slip down onto the stomach itself instead of staying around the gastroesophageal junction (10%-25%). Finally, patients can develop a paraesophageal hernia either because the crura were not closed or because they became disrupted (10%-15%).18- 21 Dallemagne et al20 and Watson et al21 found a higher incidence of early paraesophageal herniation in patients undergoing laparoscopic surgery (7%) than those undergoing open fundoplication (<2%), presumably due to poor or inadequate crural closure.
The reasons patients in our study required reoperation are similar to what has been reported previously: breakdown of the wrap, 37%; slipped wrap, 11%; paraesophageal hernia, 15%; too tight of a wrap or stricture that did not respond to dilation, 15%; and herniation of the intact wrap into the chest, 7%. In 3 patients (11%), an anatomic cause for recurrent reflux symptoms could not be found, but manometry studies demonstrated an incompetent lower esophageal sphincter.
Most patients with symptoms of recurrent GERD following antireflux surgery will not require reoperative surgery, as their symptoms can be effectively controlled with medications. Only patients with persistent, severe, refractory symptoms should be considered for reoperative exploration. We reoperated on 1 patient 3 months after his initial surgery because his workup demonstrated a clear technical problem that required surgical repair. Otherwise, we recommend aggressive long-term medical therapy prior to consideration of reoperation. In those patients selected for reoperative treatment, determination of the cause of failure of the original operation is critical to ensure that the patients receive the correct revision. Evaluation of the patient with symptoms that suggest failure of their antireflux surgery should include a combination of studies. Upper gastrointestinal tract series will rule out anatomic causes of failure such as paraesophageal hernia, wrap herniation into the chest, stricture, a wrap that is too tight, or a short esophagus. Endoscopy is helpful in determining if a wrap has been disrupted, or to document a recurrent hiatal hernia, poor gastric motility, or esophagitis. Manometry will confirm the presence of a wrap that is too tight, a disrupted wrap, an incompetent lower esophageal sphincter, missed achalasia, or poor esophageal motility. Patients with esophageal dysmotility will be candidates for a partial wrap because these patients experience an increased incidence of postoperative intractable dysphagia after a complete wrap. Studies of pH will confirm acid reflux and are especially important to obtain in patients with no demonstrable anatomic causes for their symptoms. Initially, if patients otherwise satisfied criteria for reoperation and demonstrated esophagitis on endoscopy, pH monitoring was not performed. However, performing full esophageal function tests including pH monitoring in all patients except those with paraesophageal hernias will lessen the chance of reoperation on patients for symptoms that are not due to their antireflux operation.
Most previous studies of reoperation after open antireflux surgery have used a Nissen fundoplication, although in large series, other gastric wraps and even esophagectomy with enteral interposition have been needed.11,13,15 Some authors have recommended operation through the chest if the patient previously underwent an operation through the abdomen, but most surgeons perform the procedure they are comfortable with. We offered a laparoscopic approach to these patients, who came to us for treatment after failure of their initial antireflux procedure, because we believed they would benefit from a minimally invasive approach and would experience less pain, a shorter hospitalization, and an earlier return to regular activity compared with patients who undergo an open reoperation. Patients who had undergone an earlier open fundoplication confirmed that the laparoscopic approach resulted in an earlier discharge and a quicker recovery period. We performed a Nissen fundoplication on most of our patients. We recommend a posterior partial wrap 270° in patients with stricture, a wrap that was too tight or esophageal dysmotility to decrease the incidence of postoperative dysphagia. We also perform the Toupet fundoplication in patients with a paraesophageal hernia or herniation of the wrap into the chest, as the partial wrap anchors the wrap in the abdomen better than the full wrap does.
Our results are consistent with other reports of reoperative antireflux surgery, whether open or laparoscopic. The morbidity rate was higher than for primary surgery.5 Operative times were significantly increased compared with patients undergoing a first time laparoscopic Nissen fundoplication.5 Lack of progress in dissection leading to prolonged operative time (>5) should be considered an indication for conversion to open laparotomy. Neither hospital stay nor postoperative recovery was prolonged in most of our patients.5 Symptomatic and clinical results were good to excellent in most patients.
Several technical points should be emphasized. Careful attention needs to be paid to dissection of the adhesions. Use of the endoscope will help identify the distal esophagus and prevent inadvertent gastric or esophageal perforation. If a perforation is identified, it should be repaired laparoscopically if possible. The fundoplication can cover the repaired perforation to help strengthen the repair. As with primary Nissen fundoplications, a tension-free wrap is important. This can be accomplished during the reoperation by dissecting out the retroesophageal space adequately. In addition, short gastrics must be taken down if they have not been ligated previously. Dissection of the stomach from the spleen may be necessary in patients who have had the short gastric vessels previously ligated. An adequate length of esophagus must be mobilized. This may necessitate dissection well up in the mediastinum. We have found that a 5- to 6-cm length of intra-abdominal esophagus should be mobilized. Closure of the crura is extremely important. In 3 of our patients, it appears that crural disruption with creation of a paraesophageal hernia was responsible for the recurrent reflux symptoms. It may be difficult to identify the crura or to close them without tension during the reoperation. In these cases, a patch can be sutured or stapled over the crural closure to reinforce it.
Alexander et al17 recently reported their experience with 2 patients who underwent successful reoperation laparoscopically. Average hospitalization was 3.5 days. Average operating time was 148 minutes. Both patients are free of symptoms. Croce et al18 described 5 patients who had recurrent symptoms of GERD following previous primary surgical therapy. Four of their patients underwent reoperation laparoscopically. The fifth required conversion to open laparotomy to manage a short esophagus and perform a total duodenal diversion. Clinical results were excellent in all patients undergoing laparoscopic reoperation. These studies, combined with our results, show that laparoscopic reoperation on patients who have had previous antireflux surgery for GERD can be done safely, with minimal morbidity and no mortality. Clearly, the advantages of laparoscopic surgery, including decreased pain, earlier discharge, and quicker return to regular activity, can be applied to this group of patients.
Laparoscopic reoperative surgery for previous failed antireflux procedures is technically possible and can be completed safely with good symptomatic results and improved recovery. Patients with this problem should be evaluated and treated only by an experienced laparoscopic surgeon. An appropriate preoperative workup to determine the cause of symptoms should be undertaken. The appropriate surgical technique should be applied. These patients need to be informed that there is a high risk of conversion and an increased risk of complications compared with a primary laparoscopic fundoplication.
Corresponding author: Myriam J. Curet, MD, FACS, Department of Surgery, University of New Mexico Health Sciences Center, ACC-Second Floor, 2211 Lomas Blvd NE, Albuquerque, NM 87131 (e-mail: firstname.lastname@example.org).