43: Management of the Failed Reflux Operation

Management of the failed reflux operation is emerging as an important challenge in modern surgical foregut practice. Over the last decade and a half, the number of patients referred for antireflux surgery has increased eightfold. Approximately 70,000 operations are performed annually in the United States.¹ The increased use of minimally invasive techniques to treat gastroesophageal reflux disease (GERD) has resulted from the lower perceived morbidity associated with laparoscopy in comparison with the open approach.

Most patients who undergo laparoscopic antireflux surgery experience good long-term outcome. Specialty centers report 90% to 95% “sustained benefit” after initial surgery, although not all centers see their own complications.^2,3 The results published by the broader surgical community are less favorable.^3,4 This finding is similar for laparoscopic and open surgery. However, the results are subjective and depend on the definition of failure and the experience of the surgeon.^5-9

The failure of antireflux surgery may occur early or late. The etiology of the failure is associated with and is sometimes revealed by the timing of symptoms. Early failures can be attributed to poor patient selection and technical error.⁹ For example, the misdiagnosis of an unrecognized primary esophageal motility disorder (PEMD) may lead to improper choice of surgical procedure, which dooms the procedure to fail.^10,11 Late failures may be secondary to the progression of underlying disease or attributed to the length of the procedure.¹²

After several decades of experience, multiple reports of transient increases in failed antireflux procedures have ascribed these failures to the initial learning-curve effect,^13-15 modifications of surgical technique during the initial transition to laparoscopic approach,¹⁶ and relaxation of patient selection criteria. With growing experience in the thoracic community, however, these sorts of failures are expected to diminish.

Despite good surgical results after initial operation, some patients present with recurrent symptoms or mechanical failure. Most of these patients can be managed medically with good results. However, 4% to 10% of patients become or remain symptomatic with a poor quality of life and seek additional surgical therapy.^2,10,17-21 Success rates for reoperations range between 50% and 89%.⁶ Second and third reoperations traditionally are associated with lower success rates, decreasing as much as 20% with each subsequent operation.²² The technical difficulty of reoperation has led some surgeons to advocate an open approach.²³ Evidence supporting the safety and efficacy of laparoscopic reoperation, however, is increasing.^1,2,6,24 In our experience, the laparoscopic approach to reoperation is feasible in over 95% of patients regardless of the approach used for the primary or previous surgeries (e.g., open or laparoscopic, thoracic, or abdominal).

Determining the cause is the difficult aspect of reevaluating patients with recurrent reflux. First, one must establish whether the patient's reflux or procedure-related symptoms are from surgical failure or attributable to some other etiology. In this regard, the diagnosis always should be reexamined to rule out previously undiagnosed or misdiagnosed conditions such as PEMD.

A thorough history is essential to distinguishing the patient's current symptoms from symptoms experienced preoperatively. The relevant symptoms include heartburn, dysphagia (e.g., generalized liquids or solids), postprandial pain, and respiratory symptoms, in particular, recurrent pneumonia and aspiration. If the patient has chest pain, it should be evaluated to discern whether it is secondary to reflux or to some other cause of typical or atypical chest pain (Table 43-1). In our practice, several patients referred for antireflux surgery were found on preoperative workup to have postprandial angina. Another common cause of chest pain is dysphagia that often can be fixed with endoscopic dilatation.

It is essential to understand the mechanics of the patient's previous antireflux procedure by obtaining the operative report and previous diagnostic studies. Understanding the neoanatomy and expected physiologic change is vital to interpreting new diagnostic examinations and procedures. Possible reasons for postoperative reflux include failure of surgical technique or incorrect choice of surgical procedure for the patient's initial problem. The failure in surgical technique may take many forms: obstruction (e.g., hiatus, wrap too tight, wrap too long, or wrap poorly placed), herniation (e.g., hiatus not closed adequately or short esophagus), dehiscence (e.g., poor stitch placement, knot tie, and depth), and poor gastric emptying secondary to vagal injury. Consequently, the surgeon performing the evaluation should be familiar not only with the technical details of the patient's original operation but also with all of the procedure-related pitfalls and late complications.

Hinder et al.⁵ described four failure patterns after open fundoplication: the slipped or misplaced fundoplication, the disrupted fundoplication, the herniated fundoplication, and the excessively tight or long fundoplication. Two additional failure patterns have emerged in the laparoscopic era: the twisted fundoplication and the two-compartment stomach⁹ (Fig. 43-1). The incidence of wrap failure linked to cause is variable. In our opinion, it is usually related to surgical technique and largely preventable. For example, routine hiatal repair at initial operation has been shown to reduce the incidence of recurrent herniation by 80%, and this practice has become standard of care.²

The evaluation of patients with recurrent reflux at our institution includes the four standard studies used for GERD: 24-hour esophageal pH probe study, manometry, esophagogastroduodenoscopy (EGD), and esophagram or cine. In addition, for patients with recurrent reflux, we routinely obtain esophageal and gastric emptying studies (i.e., nuclear scintigraphy). If nonacid reflux is suspected, an impedance study is also obtained.

The aim of the evaluation is to differentiate the anatomic versus physiologic cause of the patient's symptoms. Tailoring an effective surgical solution depends on this evaluation. Problems to look for include the presence of a previously undiagnosed PEMD, other esophageal anomaly, postoperative obstruction at the gastroesophageal junction (GEJ), or a primary delayed gastric emptying disorder. A large series of 104 patients demonstrated that a thorough preoperative evaluation can predict the mechanism of postfundoplication failure found at reoperation in 78% of patients.¹ Mechanisms of postfundoplication failure in this cohort are summarized in Table 43-2.

The preoperative risk-benefit assessment should take into account the patient's underlying health condition, symptoms, quality of life, and disability. The evaluation also must weigh the benefits of a second operation in terms of the surgical risk and higher failure rate. Overall symptom resolution after reoperation is significantly lower (~70%) than after an initial antireflux procedure. This finding may be secondary to the presence of more complex or advanced disease. Reoperation, regardless of the surgeon's experience, is technically more demanding. Adhesions from previous surgery can obscure the anatomy and increase the chance of operative perforation of the esophagus or stomach, pneumothorax, diaphragmatic injury, vagal nerve injury, splenic trauma, and bleeding. Reoperation is associated with an elevated morbidity (15%–40%) and mortality (0%–2%).

All patients are offered a tailored surgical approach based on the preoperative evaluation. Consequently, surgeons who perform redo antireflux procedures should have experience with the complete range of antireflux operations (see Chapters 38-40). We routinely perform intraoperative EGD before making an incision to confirm the preoperative evaluation and assess esophageal length, structural changes, hiatal hernia, active peptic disease, and other concurrent pathologies that would delay, contraindicate, or alter the planned reoperative approach.

We prefer to use the abdominal approach whenever possible. Laparoscopic Nissen fundoplication remains the “gold standard” for reoperation on recurrent GERD; however, approximately 8% to 20% of patients will require a Collis lengthening procedure, with or without hiatal repair, as described in Chapter 42. Rarely, we use other operations, such as Toupet or Dor, with a partial wrap if the patient has dysmotility (see Chapter 40). Regardless of the number of prior surgeries or the technique used (open versus laparoscopic), we approach all reoperations for failed antireflux surgery using minimally invasive technique. Rarely, a frozen abdomen is encountered, rendering the abdominal approach technically prohibitive, and we offer a Belsey–Collis procedure (see Chapter 38). Since procedure length can vary from 1.5 to 8 hours, cases should be scheduled early in the morning. In our experience, over 95% of these surgeries can be completed minimally invasively without resort to open technique.

Abdominal Approach

The patient is induced using combined anesthesia (i.e., general anesthesia with endotracheal intubation and epidural analgesia). Endoscopy is performed. The patient is placed supine in the lithotomy position. Both arms are tucked in at the sides, and the bed is placed in a shallow anti-Trendelenburg position. A nasogastric tube is inserted for drainage and decompression of the stomach. Our reoperative approach is a synthesis of antireflux procedures mentioned in previous chapters.

Approach Through The Left Chest

A left thoracotomy approach can be used for patients with recurrent failure of fundoplication. The esophagus usually is less adhesed and easier to control and dissect from this approach. As in the abdomen, the surgeon must be able to tolerate some bleeding from the liver during the dissection. The same parameters are addressed as in the abdominal approach. The old wrap is taken down, and the cause of the recurrence is identified and fixed. Collis extension is used liberally. The repair can be performed using the Belsey–Mark IV operation or the Nissen operation.

Practices That Differ From Initial Antireflux Repair

Scope Insertion, Trocar Placement, and Adhesiolysis

We place an 11-mm camera port using a modified Hassan technique to minimize inadvertent injury to the abdominal viscera. Placing a periumbilical trocar in a previously operated patient can be technically challenging. If this is not possible because of adhesions, an alternate initial camera port can be placed. After abdominal insufflation with CO₂ (the insufflation pressure is 15 mm Hg; when needed, a maximal pressure of 20 mm Hg can be used transiently), a 10-mm laparoscope with a deflectable tip capable of 360-degree angulation is inserted.

To minimize cosmetic scarring, we attempt to use original port scars for trocar placement, but alternate sites can be used if the original ports are improperly placed or densely adhesed. If the abdomen has multiple adhesions, we attempt to place one of the four remaining ports in the least adhesed region of the abdomen in its intended location under direct visualization. Careful dissection of adhesions using scissors, ligature, or hook so that all working ports are inserted is paramount. The remaining ports are inserted sequentially. In our experience, the liberal use of an additional 5-mm port placed in an area free of adhesions can save time, aid in the dissection, and may decrease operative morbidity. Rarely, adhesions are so dense that port relocation is not possible, and the abdominal procedure must be terminated. Care should be taken to avoid inadvertent bowel injury. If this occurs, however, the bowel often can be repaired primarily laparoscopically.

Adhesiolysis is a crucial stage in reoperation; safety depends on retraction, visualization, and identification of the avascular planes. Unfortunately, the time required to perform adhesiolysis cannot be assessed preoperatively. In our experience, it can range from a few minutes to several hours. Care should be taken to avoid known pitfalls of antireflux surgery, such as bleeding, perforation, and vagal injury. Thus we perform surgery in a systematic manner (described below) to minimize these pitfalls. If they occur, most can be managed laparoscopically.

Dissection and Wrap Takedown

The conduct and design of the operation are similar to those described in Chapter 39 on laparoscopic Nissen fundoplication. However, greater care must be taken to avoid traction injuries that lead to bleeding or perforation of visceral organs. Especially relevant are the fibrous bands that attach to the spleen and dissection around the wrap and posterior distal esophagus. If the patient had a previous wrap, it typically will be adhesed to the left side of the liver or herniated into the chest. During this phase of the operation, an understanding of why the previous operation failed must be attained. Thus meticulous attention to detail during dissection is essential.

Dissection of the Gastrosplenic Ligament

Typically, dissection is begun by dividing the gastrosplenic ligament. This is carried out sequentially, anteriorly first and then posteriorly, using a 5-mm LigaSure (Valleylab, Boulder, CO) device. The highest two or three short gastric arteries are divided, if not already divided during the initial operation. After reaching the left GEJ, the left gastrophrenic ligament is opened, exposing the left crus. If a hiatal hernia is identified, its contents are reduced into the abdomen, and the peritoneal sac is amputated and removed.

Mobilization of the Distal Esophagus and Proximal Stomach on the Right

Prior fundoplication typically causes dense adhesion between the wrap and the left lobe of the liver. Dissection in this region can be technically challenging. Perforation of the stomach or esophagus and injury to the vagus nerve must be avoided at all costs. The gastrohepatic ligament is opened above the caudate lobe in an avascular region (near the liver), avoiding the left gastric artery, the possibility of a replaced hepatic artery, and vagal branches. The anatomy may be difficult to navigate because of the prior surgery and presence of adhesions. When separating the wrap from the liver, we find it is safer to err on the side of the liver as opposed to causing visceral injury to the stomach. Any bleeding is usually short-lived and easily controlled with retraction. The right crus is identified, and the overlying peritoneum is opened exposing its edge. We do not attempt to directly identify the esophagus before opening the crura because of the risk of perforation. In addition, the vagi are always found medial to the crura. The right crus is dissected from its confluence to the median arcuate ligament, where it joins the left crus. Dissection is done meticulously, taking only a thin layer of peritoneum.

The esophagus is revealed by its orientation, longitudinal muscle fibers, and the vagus. We routinely divide the esophagophrenic attachments and peritoneum over the crura in a semicircular fashion down to the median arcuate ligament, taking care to avoid injuring the anterior and posterior branches of the vagus. Finally, the wrap is opened along the plane where it was sewn in the previous operation. This is a difficult step because defining the plane may be technically challenging and remaining within that plane while not causing a perforation may be difficult. Occasionally, the general location of the suture line can be stapled to separate the wrap. Great care to prevent perforation of the esophagus or stomach should be taken during mobilization of the GEJ. The esophagus is retracted upward, and a window is opened and widened bluntly to reveal the left side of the abdomen. We avoid dividing any structure that could be the vagus until the two nerves are identified.

Assessing Esophageal Length

The fat pad is medialized (taking care not to injure the anterior vagus nerve) to accurately assess the level of the GEJ. This determination is made in the context of the esophageal length measurements that were made during intraoperative endoscopy, barium swallow, and the visualized location of the GEJ when no traction is placed on the stomach. If the patient is found to have a short esophagus, the Collis lengthening procedure is performed at this juncture (see Chapter 42).

Integration of Knowledge

A crucial step in the operation is the moment when the operating surgeon integrates all the preoperative and operative findings and determines the cause of the failed antireflux procedure. In this regard, it is important to dissect the old wrap completely because this usually will verify the cause of the failure. An excellent summary of this decision-making process was published by Swanstrom et al.²⁴ A standard redo wrap then is fashioned, taking into account any lessons learned from the patient's preoperative workup or the intraoperative findings. For example, if the wrap is too loose, a small bougie can be used to calibrate the opening. If there is a short esophagus, a Collis extension is performed. If the wrap was “shoeshined” inadequately at the first operation, an attempt should be made to fashion a more accurate configuration. If the surgeon fails to determine the cause of the previous failure, chances are that the redo procedure also will fail to deliver good long-term palliation.

In addition, when performing a redo procedure, we tend to take a maximal approach when doubt exists. For example, if the esophagus is of equivocal length, we lengthen it. If the crura are of questionable strength, we reinforce them with bioabsorbable mesh, and if gastric emptying is delayed or vagal injury is suspected, we place a percutaneous gastrostomy tube (PEG).

For all redo procedures, we place a Jackson-Pratt no. 10 drain to the left of the wrap because in our experience the postoperative risk of leak is greater with redo surgery.

Closure of Hiatus

The hiatus is often fused and does not need to be closed. When it does require closure, we use the method described in previous chapters for simple closure or crural reinforcement (see Chapters 39 and 42). Choice of fundoplication depends on information gathered during the preoperative evaluation and intraoperative findings. The techniques for the various wraps also have been described in previous chapters.

Some patients who present with reflux after multiple surgeries are found to have a nearly frozen GEJ. In these patients we are often unable to completely dissect off the wrap but instead are able to tighten it anteriorly over a bougie by taking additional stitches to bunch it up. We find that the incidence of gastric leaks in these patients is higher. Rather than risk having to come back emergently, we leave Jackson-Pratt drains around the GEJ and insert a PEG to control the gastric output.

We routinely obtain a barium swallow to rule out leak after reoperation because of the higher incidence of esophageal perforation. If a Jackson-Pratt drain was placed in the OR, it is removed on the day of discharge, provided that the patient does not have a leak on barium swallow and is able to tolerate clear liquids. The Jackson-Pratt drain is removed on the day of discharge if the patient does not have a leak. In the event of a leak, the drain is left in place for 6 weeks. This enables a controlled fistula to develop, thus alleviating the need for reoperation in the event of a leak.

In all other respects, postoperative management is the same whether the patient is being operated on for the first time or the nth time for failed reflux. In our experience, postoperative nausea and delayed motility are more common in patients undergoing reoperation. Symptoms usually can be abated with Reglan or erythromycin or both. Patients are discharged after they are able to tolerate enough daily liquids to prevent dehydration. Postoperatively, we educate patients to eat slowly and small amounts, and patients are sent home on a full liquid diet for 2 weeks. If problems are not encountered, patients are advanced to a regular diet devoid of bread and carbonated beverages. The literature states that postoperative dysphagia is common and can be present for many months, but most patients will prefer some dysphagia to reflux.

The complications associated with redo laparoscopic Nissen fundoplication for recurrent or persistent reflux are, for the most part, identical to procedural complications for primary Nissen fundoplication (see Chapter 39). Briefly, these include splenic injury, esophageal or gastric perforation, and postoperative dysphagia. Less common complications include wrap failure, pancreatitis, gastric emptying problems, gastric necrosis, and bleeding. Generally speaking, the laparoscopic technique is less morbid, which underscores our preference for the laparoscopic approach even with complex reoperations.

Exponential growth of laparoscopic Nissen fundoplication over the last decade has caused a dramatic increase in the number of patients presenting with recurrent or persistent reflux symptoms owing to failure of their primary antireflux surgery. The causes of surgical failure are multifactorial. A thorough evaluation and diagnostic workup are necessary to discern the etiology and identify a tailored, effective surgical solution. All patients should be counseled that reoperation carries a greater risk for complications and poorer overall outcome than a primary antireflux procedure. However, approximately 70% of patients report symptom resolution, and overall patient satisfaction is 7 on a scale of 1 to 10. Patient selection is key to successful outcome. All patients must have an esophagus that is of normal length, or steps must be taken to create a normal-length esophagus, and the esophagus must exhibit normal motility (i.e., without evidence of achalasia, DES, or scleroderma). During reoperation, a clear understanding of the mechanics of the failed procedure is paramount, and the importance of fully unwrapping the prior fundoplication cannot be overstated. A tailored approach to this complex reoperation is necessary and should be based on the preoperative evaluation and intraoperative findings. Reoperation can be accomplished safely by means of minimally invasive technique in most patients regardless of the previous mode of surgery and is the preferred approach at our institution.

The patient must understand that the results of redo surgery may be less than perfect because of the potential for vagal injury or imperfect calibration of the wrap. On occasion, postoperative dysphagia can be managed with repeated endoscopic dilatations.