Evaluation of resection of the gastroesophageal junction and jejunal interposition (Merendino procedure) as a rescue procedure in patients with a failed redo antireflux procedure. A single-center experience

Fundoplication in patients with chronic gastroesophageal reflux disease (GERD) is generally followed by excellent short- and long-term results [1‐5]. Failures are, however, unavoidable and present with recurrent reflux symptoms, postprandial pain, dysphagia and delayed gastric emptying [6]. A variety of different mechanisms may cause these symptoms such as recurrent hiatal hernia, dislocation of the fundoplication, vagus nerve damage and/or other morphological abnormalities in the hiatus [7, 8]. The problem with many of these symptoms (with the exception of acid reflux), appearing after a defective surgical repair, is that the result of conservative treatment is usually poor which consequently leads to the need for another fundoplication, with the aim to achieve a durable repair [9, 10]. With very few exceptions, the reported postoperative morbidity is significantly higher after redo surgery, which is due to the complexity of the anatomical region, postoperative scarring and deformation [11‐15]. If severe symptoms recur after a second or third fundoplication remedial surgical interventions have been advocated including total gastrectomy or gastric bypass [16, 17]. The outcome of these procedures has often been reported as good to excellent but a substantial publication bias in favour of respective surgical intervention may be present. In addition, a number of potentially relevant factors have to be taken into consideration. Maintenance of the gastrointestinal continuity with preservation of the duodenal and proximal jejunum contact with ingested food particles and prevention of reflux into the esophagus are critically important factors as well as the role of the gastric reservoir, which may be relevant for preventing post-gastrectomy symptoms. An additional symptom component, which may be prevalent, is chronic epigastric pain radiating to the back with or without exacerbation after ingestion of food. The mechanisms behind these complaints are poorly understood, but the need for surgical clearance and resection has been advocated [18].

Resection of the gastroesophageal junction with interposition of a jejunal segment between the distal esophagus and the remnant stomach (Merendino procedure) has been used in various clinical situations, mainly in patients with peptic strictures during the pre-proton pump inhibitor era and more recently for early Barrett cancer, where encouraging results have been reported [19‐22]. Proximal gastrectomy with jejunal interposition for non-advanced proximal gastric cancer, which is similar to the Merendino procedure, is performed mainly in the Far East and has been shown to induce fewer post gastrectomy symptoms as compared to total gastrectomy with Roux-en-Y reconstruction, without adding more postoperative complications [23, 24]. Therefore, the Merendino procedure may have the potential to offer significant advantages in challenging post-fundoplication situations. Hereby, we report a single institution’s experience of this reconstruction method in patients with a history of failed redo antireflux surgery.

To our knowledge the series of cases presented herein is the first describing the results of a Merendino reconstruction, i.e. jejunal interposition between the esophagus and stomach, in patients with previously failed redo-fundoplication. We found that the rate of postoperative complications was high and at least in the same range as in previous case series of RNY reconstruction in similar patients [16, 17]. In addition, re-interventions for postoperative symptoms were frequent. Alleged anastomotic narrowing were managed by endoscopic dilatations and some patients also presented with gastric outlet obstruction symptoms suggesting pyloric dysfunction as a result of vagal damage, which may occur after repeated surgery in the hiatus area. Dilatation of the pylorus was performed in these cases, but symptoms typically persisted. As many as 4 of the 12 patients, developed postoperative symptoms of intolerable severity and therefore a redo operation and conversion to RNY esophagojejunostomy was performed. Among the patients where a questionnaire-based follow-up was available, QoL was generally poor and gastrointestinal symptoms were common. Even dumping symptoms, the risk of which, from a theoretical point, should be minimized, given the preservation of the duodenal passage and maintenance of the gastric reservoir, were also frequently reported. Taken together these data, even though the number of patients is limited, suggests that the Merendino procedure is unsuitable for the complex group of patients where redo surgery after repeat fundoplication is considered. This is important to bear in mind, especially since an additional aspect in favour of resection of the gastroesophageal junction area, was the severe pain hypothetically related to the extensive scarring of the hiatal region. Redo surgery after fundoplication using other approaches than yet another fundoplication are rarely performed and there is no evidence supporting the use of any specific reconstruction method over the other. RNY reconstruction has been proposed as an alternative and more effective method in patients with obesity or esophageal dysmotility [31, 32]. In addition to this, resection of the distal esophagus, cardia and proximal stomach may be indicated if symptoms of the patient include dysphagia and pain. From a theoretical point, preservation of the duodenal passage by performing a Merendino procedure could reduce post gastrectomy symptoms as compared to RNY, but the data presented here are not supportive of this.

Previous studies of the Merendino reconstruction after resection for early Barrett cancer have shown varying short- and long-term results. In one of the largest series, Stein et al. [20] reported less postoperative complications compared to Ivor-Lewis esophagectomy and excellent long-term outcome as measured by the GIQLI (gastrointestinal quality of life index) instrument. In a more recent publication, where the EORTC QoL questionnaires were used, both short-term and functional results were similar after a Merendino reconstruction as compared to Ivor-Lewis procedure [33]. On the other hand, proximal gastrectomy with jejunal interposition in patients with non-advanced gastric cancer, which is a procedure similar to the Merendino, seems to have a superior functional outcome compared to total gastrectomy with RNY reconstruction [23, 24]. Obviously the importance of objectively assessing long-term data also in corresponding patient cohorts is warranted to assess the clinical place of the Merendino procedure. It is therefore particularly important to try to understand our experience of the poor results of the Merendino reconstruction among patients with a failed fundoplication. The high frequency of postoperative complications was expected, since this is frequent after most reconstructive procedure in the gastrointestinal tract. Objective findings of anatomical abnormalities among the patients were not always obvious. This may suggest that a general underlying pathology of gastrointestinal motility may have been present. Manometry of the esophagus and the interposed jejunal segment was available in 4 patients and in one of these cases aperistalsis of the jejunal segment was recorded. Barium swallow was performed postoperatively in 6 patients and showed a prolonged emptying of the jejunal segment or reflux in 5 and 4 cases respectively. The underlying mechanism for the poor function of the jejunal segment regardless of objective findings is obscure, but it can be hypothesized that the partial fundoplication that was performed may have contributed to the distal obstruction. However, also one patient with normal manometry findings had symptoms so severe, that a redo to RNY was performed. Thus, routine diagnostic methods may be inefficient in detecting possible motor disturbances of the gastrointestinal tract in this group of patients.

Finally, limitation of this study is that investigations done on a small sample of the group. Only four patients were investigated with HRM postoperatively and six patients with esophageal barium swallow. QOL data preoperatively don’t exist for these 12 patients and postoperatively were obtained from 6 patients. As a result it might be difficult to interpret the severity of the symptoms at the follow up.

The conclusion to be drawn from this case series is that the Merendino procedure is, until proven otherwise, unsuitable for patients who undergo redo surgery after previously failed re-fundoplications.