Clinical presentation and comorbidities

Patients with heartburn as the main symptom who respond satisfactorily to PPIs are good candidates for antireflux surgery.

Statement endorsed, overall agreement 94.1%: A+ 67.6%, A 26.5%, A− 5.9%, D− 0.0%, D 0.0%, D+ 0.0%; GRADE A.

The vast majority of presurgery and postsurgery studies in the literature enrol patients with ‘typical’ GORD symptoms which include both heartburn and regurgitation, as well as patients with GORD who have typical symptoms refractory to acid suppression therapy. Several peer-reviewed studies investigating patients who reported a complete or partial response to PPI therapy prior to antireflux surgery, showed a benefit of antireflux surgery.18–21 Moreover, the response to PPI therapy, good compliance and objective preoperative evidence of acid reflux all predict a favourable outcome.22 However, specific data on heartburn as the main symptom preoperatively and the response of heartburn symptoms to antireflux surgery were (often) not provided.

There is also the issue of the terminology ‘satisfactorily’, which is very subjective. It has indeed been shown that responses to satisfactorily relief could possibly be influenced by baseline severity.23 24 However, this terminology is easy to understand by patients and it fits within the practice of medicine in the office setting. Furthermore, in IBS therapy trials, the usefulness of ‘satisfactorily relief’ as an outcome parameter was linked to its ability to integrate various symptoms and the impact of therapy on various symptoms.25

2. Patients with regurgitation as the main symptom are good candidates for antireflux surgery, regardless of the response pattern to PPI therapy.

Statement not endorsed, overall agreement 79.4%: A+ 14.7%, A 64.7%, A−- 11.8%, D− 5.9%, D 2.9%, D+ 0.0%; GRADE B.

The ability for PPIs to adequately improve regurgitation appears to be much less than their ability to improve heartburn.26 27 In the literature, there is a lack of solid evidence to support the statement above. However, in a systematic analysis, surgery does appear to be superior to PPIs in alleviating symptomatic regurgitation, although dysphagia, rectal flatulence and the inability to belch or vomit were significantly more common in patients treated surgically.8 28 29 Important to notice is that symptoms of regurgitation due to primary oesophageal motility disorders (eg, achalasia, rumination syndrome) have to be ruled out by means of oesophageal motility testing (preferably using high-resolution manometry (HRM)) before referring a patient for antireflux surgery.

3. Patients with reflux-hypersensitive oesophagus (normal acid exposure but positive symptom association with reflux events) are good candidates for antireflux surgery.

Statement not endorsed, overall agreement 55.9%: A+ 5.9%, A 50.0%, A− 26.5%, D− 11.8%, D 2.9%, D+ 2.9%; GRADE C.

Reflux hypersensitivity is categorised as a functional disorder in the latest Rome IV criteria and defined as ‘patients with oesophageal symptoms who lack evidence of reflux on endoscopy or abnormal acid burden on reflux monitoring, but show triggering of symptoms by physiological reflux’.3 Symptoms in patients with reflux hypersensitivity are caused by reflux events (main difference with functional heartburn); therefore antireflux surgery can theoretically improve symptoms as it minimises oesophageal reflux. The majority of studies suggest that patients with a hypersensitive oesophagus are possibly good candidates for antireflux surgery,4 30–34 while a few reports suggest the opposite.35 More specifically, outcome in patients with hypersensitive oesophagus where reflux has been documented in the past (either by the presence of oesophagitis or a pathological acid exposure time) is similar as to patients with documented reflux without being hypersensitive.31 32 However, a recent study by Patel et al showed that pure acid sensitivity was a negative predictor for symptom improvement with antireflux therapy, including surgical management.35 Moreover, patients with reflux hypersensitivity often display a high level of anxiety. Blondeau et al demonstrated that psychosocial factors and somatisation might contribute to symptom perception in patients with reflux hypersensitivity.36 This confounding factor has to be taken into consideration before referring these patients for antireflux surgery.

4. Patients with functional heartburn (Rome III/IV criteria, who have no association of symptoms with documented episodes of reflux events) are poor candidates for surgery.

Statement endorsed, overall agreement 100%: A+ 91.2%, A 8.8%, A− 0.0%, D− 0.0%, D 0.0%, D+ 0.0%; GRADE B.

Functional heartburn is defined according to the Rome IV criteria as ‘a burning retrosternal discomfort or pain refractory to optimal antisecretory therapy in the absence of gastro-oesophageal acid reflux (GOR), histopathological mucosal abnormalities, major motor disorders or structural explanations’.3 In other words, in functional heartburn, symptoms manifest themselves without association to reflux events. From the mechanistic point of view, it is therefore unlikely that functional heartburn would be improved by antireflux surgery. The few available studies do not support the efficacy of surgery.34 37

5a. Patients with non-cardiac chest pain (NCCP) are good candidates for antireflux surgery only if symptoms can be attributed to reflux.

Statement not endorsed, overall agreement 79.4%: A+ 14.7%, A 64.7%, A− 14.7%, D− 2.9%, D 2.9%, D+ 0.0%; GRADE C.

NCCP is a common condition, with a prevalence of up to 25% in the US adult population.38 After excluding a cardiac cause, reflux is the most common underlying mechanism for this disorder. Patients with NCCP might be referred to antireflux surgery after ruling out oesophageal motility disorders such as hypercontractile oesophagus and functional chest pain.39 Although literature on GORD-related NCCP as the sole indication for surgical treatment is non-existent, fundoplication has been performed in this patient group. Improvement after surgical treatment is better in patients with a clear correlation between reflux events and symptoms, in patients who also display typical reflux symptoms such as heartburn and third when there is a satisfactory response to PPIs prior to the surgery.40–46

5b. Patients with extra-oesophageal syndromes (asthma, chronic cough or laryngitis) are good candidates for antireflux surgery only if symptoms can be attributed to reflux.

Statement not endorsed, overall agreement 44.1%: A+ 8.8%, A 35.3%, A− 44.1%, D− 8.8%, D 2.9%, D+ 0.0%; GRADE C.

The vast majority of data is reported in patients with typical GORD symptoms and coexisting extra-oesophageal symptoms that seem to respond to surgery. Varying degrees of symptom improvement has been shown mainly in case series in respiratory symptoms, asthma, cough and laryngopharyngeal symptoms.43 47–69 Few data are available on the outcome of antireflux surgery for isolated atypical symptoms. Moreover, patient selection remains uncertain as there is no well-established method for demonstrating that these symptoms can be attributed to reflux. The use of symptom markers is valid for typical reflux symptoms; however, in case of extra-oesophageal symptoms, it has been subject of debate for a long time. An objective monitoring for chronic cough can be added through ambulatory manometry or acoustic monitoring.70–73 Outcomes of Nissen fundoplication in patients with chronic cough attributable to reflux were good although in uncontrolled and often retrospective studies, but these were selected patients who also displayed a positive pH monitoring.42 50–59 61–68 It has to be stressed out that none of the studies were placebo/sham controlled, which is pivotal in studying the exact effect of antireflux surgery in patients with chronic cough.

6. Patients with eosinophilic oesophagitis (EOO) on oesophageal biopsies are poor candidates for antireflux surgery.

Statement endorsed, overall agreement 88.2%: A+ 61.8%, A 26.5%, A− 8.8%, D− 2.9%, D 0.0%, D+ 0.0%. GRADE C

There is evidence in the literature that eosinophilic oesophagitis in children and adults does not respond to antireflux surgery.74–77 Obtaining oesophageal biopsies in all patients evaluated for Nissen fundoplication is debatable, as reports suggest a low prevalence of eosinophilic oesophagitis in adults with refractory heartburn.78

7. Patients with scleroderma (and/or other severe smooth muscle disease) are poor candidates for antireflux surgery.

Statement not endorsed, overall agreement 64.7%: A+ 11.8%, A 52.9%, A− 26.5%, D− 5.9%, D 2.9%, D+ 0.0%; GRADE C.

Data on the outcome of antireflux surgery in patients with scleroderma (systemic sclerosis) is contradictory: there are a few non-randomised studies reporting (partial or full) resolution of reflux symptoms, while other studies suggest that surgery is of limited success in these patients.79–84 Although the severity of reflux symptoms improved after Nissen fundoplication, postoperative dysphagia was present in 38%–71% of patients with scleroderma.79 81 85 A retrospective study suggested laparoscopic Roux-en-Y gastric bypass might be a better option in patients with systemic sclerosis-associated reflux: less dysphagia and improved reflux control was seen after Roux-en-Y gastric bypass compared with fundoplication.81 86 Although the oesophagus is not always affected in patients with systemic sclerosis, the majority of patients with oesophageal involvement is found to have aperistalsis, which is a risk factor for postoperative dysphagia also in patients without scleroderma.87–89

8. Patients with concomitant functional disorders such as dyspepsia and IBS are good candidates for antireflux surgery, only if symptoms can be attributed to reflux.

Statement not endorsed, overall agreement 64.7%: A+ 23.5%, A 41.2%, A− 26.5%, D− 8.8%, D 0.0%, D+ 0.0%; GRADE B.

According to Rome IV criteria, functional dyspepsia (FD) is defined as ‘a medical condition that significantly impacts on the usual activities of a patient and is characterised by one of the following symptoms: postprandial fullness, early satiation, epigastric pain or epigastric burning that are unexplained after a routine clinical evaluation’.90 Within patients with FD, a distinction between postprandial distress syndrome (PDS), with predominant postprandial fullness and early satiation and epigastric pain syndrome (EPS) characterised by epigastric pain or epigastric burning is made. There are currently no data about the influence of Rome III/IV FD comorbidity on the outcome of antireflux surgery, but it may parallel the inferior response to PPIs.91 Studies suggest that antireflux surgery is not contraindicated in patients with dyspepsia comorbidity since these symptoms also tend to improve, but they do point out that the expected outcome is worse in patients with dyspepsia comorbidity.92

IBS is defined in the Rome IV criteria as ‘recurrent abdominal pain associated with defaecation or a change in bowel habits’. Disordered bowel habits are typically present (ie, constipation, diarrhoea or a mix of constipation and diarrhoea), as are symptoms of abdominal bloating/distention. Symptom onset should occur at least 6 months before diagnosis and symptoms should be present during the last 3 months.93 A study by Raftopoulos et al demonstrated that preoperative IBS is not a contraindication of antireflux surgery.94 Although Axelrod et al do not state that IBS is a contraindication of antireflux surgery, they showed that patients with a diagnosis of functional bowel disease or with preoperative symptoms of functional bowel disease were more likely to have a poor outcome compared with patients without the diagnosis or symptoms of functional bowel disease.95 The reported data—although scarce and variable in quality—indicate that neither FD nor IBS comorbidity is a contraindication for antireflux surgery because of a similar improvement of typical reflux symptoms. However, extensive counselling about the possibility of persistent functional GI symptoms and increased risk of gas-bloat syndrome postoperatively is warranted.

9. Patients with a body mass index >35 kg/m² are poor candidates for antireflux surgery.

Statement not endorsed, overall agreement 23.5%: A+ 2.9%, A 20.6%, A− 17.6%, D− 14.7%, D 38.2%, D+ 5.9%; GRADE B.

Obesity has been implicated as a major and independent risk factor for GORD by several mechanisms (increase of the intragastric pressure and of the abdominal–thoracic pressure gradient, increased gastric peptic secretion, abnormal gastric emptying).96 97 Losing weight should be the first pillar in GORD treatment for obese patients. Although Perez et al demonstrated that there was a 31% occurrence rate of GORD after antireflux surgery in 48 obese patients, other more recent studies have shown that preoperative obesity was not associated with a poorer outcome following laparoscopic Nissen fundoplication.98–102

10. Patients with psychiatric illness (major depression or anxiety disorder) are good candidates for antireflux surgery only if symptoms can be attributed to reflux.

Statement not endorsed, overall agreement 32.4%: A+ 2.9%, A 29.4%, A− 29.4%, D− 20.6%, D 14.7%, D+ 2.9%; GRADE C.

Only few studies investigated the influence of psychiatric comorbidity on the outcome of antireflux surgery in patients with GORD. It has been demonstrated that patients with GORD and concomitant psychiatric disorders (major depression or anxiety as defined by the DSM-IV) have more severe symptoms and lower quality of life at baseline. Even if a 24 hours pH-monitoring is normal after surgery, these patients report less symptom relief and less quality of life improvement compared with patients without psychiatric comorbidity.103–106

11. Patients known with substance abuse (such as alcohol abuse and drug abuse) are poor candidates for antireflux surgery.

Statement not endorsed, overall agreement 26.5%: A+ 8.8%, A 17.6%, A− 50.0%, D− 5.9%, D 17.6%, D+ 0.0%; GRADE D.

Alcohol and smoking may induce GOR by decreasing lower oesophageal sphincter (LOS) pressure and disturbance of oesophageal motility, although there is no evidence that lifestyle and dietary changes, for example, stopping smoking will improve symptoms.97 107–110 Impaired swallow-induced LOS relaxation and oesophageal body dysmotility were observed both in healthy volunteers and in symptomatic patients with dysphagia receiving opioids.111 112 Additionally, there is no evidence in the literature that all these factors may have an impact on the results of antireflux surgery.113 114

12. Patients with dental erosions related to documented reflux are good candidates for antireflux surgery.

Statement not endorsed, overall agreement 44.1%: A+ 0.0%, A 44.1%, A− 29.4%, D− 8.8%, D 5.9%, D+ 11.8%; GRADE D.

The literature on dental erosions related to GORD is limited. Most studies indicate an increased prevalence of dental erosions in patients with GORD.115 On average, 17%–68% with GORD have dental erosions.116–122 A study by Wilder-Smith et al showed that esomeprazole 20 mg twice a day significantly reduced the decrease in enamel thickness compared with placebo, suggesting that treatment of GORD may reduce the development of dental erosions.119 In a 1-year follow-up study, the same group described no further progression in erosive tooth wear in 74% of the patients.123 However, to date, there are no studies available evaluating the effect of antireflux surgery on dental erosions.

Oesophagogastroduodenoscopy

13. Endoscopy is mandatory and has to be carried out in the last year prior to antireflux surgery.

Statement endorsed, overall agreement 94.1%: A+ 82.4%, A 11.8%, A− 5.9%, D− 0.0%, D 0.0%, D+ 0.0%; GRADE B.

The literature on the use of and diagnostic output from endoscopy in the preoperative workup of patients with GORD before antireflux surgery is very extensive. Although there is no solid evidence that endoscopy is mandatory prior to antireflux surgery, there seems to be a general consensus that endoscopy shall be performed before antireflux surgery. The proper timing for endoscopy has not been studied so far and is therefore not well defined and the voting outcome reflects the opinion and clinical experience of the experts of the panel.

14. There is no need to wean the patient off PPI for an endoscopy in the preoperative workup for antireflux surgery.

Statement endorsed, overall agreement 88.2%: A+ 41.2%, A 47.1%, A− 2.9%, D− 5.9%, D 0.0%, D+ 2.9%; GRADE C.

The literature to support a decision on whether there is greater value of maintaining or for stopping PPI therapy before making a decision regarding selection of a patient for antireflux surgery is scarce. Standard current practice seems to either perform endoscopy on PPI or not to specify.28 124 125 The information gained by preoperative endoscopy relates to the need to assess and grade dysplasia in Barrett’s oesophagus, identify the possibility of a short oesophagus and assess the size and configuration of hiatal hernia (HH).126 127 As interrupting PPI therapy for these assessments is unnecessary and unhelpful, patients can therefore continue their PPI treatment regimen for endoscopic assessment of reflux prior to a decision regarding the potential value of antireflux surgery.

15. Patients with GORD symptoms and an endoscopic diagnosis of a HH are good candidates for antireflux surgery.

Statement endorsed, overall agreement 82.4%: A+ 20.6%, A 61.8%, A− 14.7%, D− 2.9%, D 0.0%, D+ 0.0%; GRADE B.

A HH disrupts the anatomy and physiology of the normal antireflux mechanism (reducing the LOS length and LOS-pressure, impairing augmentation of the LOS by the right crus, impairing oesophageal peristalsis, increasing cross-sectional area of the oesophago-gastric junction (OGJ)). The herniated stomach acts as a reservoir allowing reflux into the lower oesophagus during swallowing. The presence of a HH is associated with increased symptoms of reflux, increased prevalence and severity of reflux oesophagitis, Barrett’s oesophagus, oesophageal adenocarcinoma and reduced efficacy of PPI.128 129 The severity of oesophagitis is best predicted by size of HH, followed by LOS pressure, in that order.130 Although patients with a large HH are more prone to have pathological reflux and more symptoms, not all patients with a HH have GORD.131

Up to date, there are no prospective studies reporting the influence of HH in recommending antireflux surgery. The Canadian Consensus Conference on the management of GORD in adults did suggest that a significant HH, because of its likely contribution to reflux in an individual patient, may tip the balance towards surgery.132 Most series looking for independent predictors of success for antireflux surgery did not find presence of HH to be significant in multivariate analysis.10 133

16a. Patients with GORD symptoms and unequivocal presence of reflux oesophagitis Los Angeles (LA) grade A or higher off PPI are good candidates for antireflux surgery.

Statement not endorsed, overall agreement 50.0%: A+ 17.6%, A 32.4%, A− 47.1%, D− 2.9%, D 0.0%, D+ 0.0%; GRADE B.

16b. Patients with GORD symptoms and unequivocal presence of reflux oesophagitis LA grade B or higher off PPI are good candidates for antireflux surgery.

Statement endorsed, overall agreement 91.2%: A+ 47.1%, A 44.1%, A− 8.8%, D− 0.0%, D 0.0%, D+ 0.0%; GRADE B.

Patients with GORD can be subdivided into those with erosive reflux disease (ERD, the presence of mucosal breaks) and those with non-erosive reflux disease (NERD, the absence of mucosal breaks) based on upper GI endoscopy findings.134 It has been demonstrated that progression from NERD to erosive oesophagitis occurs while regression from ERD to NERD is rare.135 136 The diminished response to medical treatment in patients with NERD would support a greater role for surgery in NERD than in ERD. It has been demonstrated that subjective and objective long-term outcomes of Nissen fundoplication were similar in ERD and NERD and results were sustained for up to 5 years after surgery.137 Moreover, in terms of symptoms and signs of erosive oesophagitis, a long-term study reported that surgery was superior to conservative management with modified lifestyle and medication.136

Historically, reflux oesophagitis off PPI was considered a good selection criterion, usually in combination with abnormal pH metry, for patient selection for an antireflux surgery. Active oesophagitis is a definite sign of ongoing pathological reflux and may help to select patients for surgery. More than 85% of the patients with documented oesophagitis were satisfied with the results of surgery (laparoscopic or open).138 However, previous studies demonstrated that up to 15% of the general population have oesophagitis LA grade A or higher. Almost half of these patients, in particular those with LA grade A are asymptomatic.2 139

Patient selection though remains mostly based on symptoms, as indicated by the recent LOTUS trial.8 140

17. Patients with GORD symptoms without reflux oesophagitis during endoscopy performed off PPIs are poor candidates for antireflux surgery.

Statement not endorsed, overall agreement 2.9%: A+ 0.0%, A 2.9%, A− 2.9%, D− 23.5%, D 44.1%, D+ 26.5%; GRADE C.

The diagnosis of NERD is based on upper GI endoscopy findings and a positive pH or pH-impedance study.3 134 Impairment of quality of life, however, and severity of symptoms are similar as for ERD.141 142 In theory, the diminished response to medical treatment in patients with NERD could support a greater role for surgery in NERD than in ERD. When comparing long-term outcome results of antireflux surgery in patients with PPI-refractory NERD and ERD, it was demonstrated that both subjective and objective long-term outcomes were similar in ERD and NERD and results were sustained for up to 5 years after surgery.32 143

18. Patients with GORD symptoms and Barrett’s oesophagus (non-dysplastic specialised intestinal metaplasia) on biopsies of the distal oesophagus are good candidates for antireflux surgery.

Statement endorsed, overall agreement 82.4%: A+ 17.6%, A 64.7%, A− 14.7%, D− 2.9%, D 0.0%, D +0.0%; GRADE B.

The presence of Barrett’s oesophagus can be considered proof of the presence of GORD. Many studies confirm that antireflux procedures in patients with Barrett’s oesophagus effectively reduce reflux-related symptoms and that uncomplicated Barrett’s oesophagus does not influence outcome of antireflux surgery.22 144–153 A meta-analysis found no evidence that antireflux surgery prevents the progression to carcinoma of the oesophagus; therefore, postoperative endoscopic follow-up should be maintained.154 In patients with refractory GORD eligible for antireflux surgery, the presence of Barrett’s oesophagus should not be a contraindication. On the other hand, antireflux surgery should not be suggested to asymptomatic patients with Barrett’s oesophagus or to patients with short segment Barrett’s oesophagus to prevent evolution to dysplasia or adenocarcinoma.

19. In patients considered for antireflux surgery, biopsies of the oesophageal body should be obtained during endoscopy.

Statement not endorsed, overall agreement 73.5%: A+ 58.8%, A 14.7%, A− 11.8%, D− 2.9%, D 2.9%, D+ 8.8%; GRADE C.

EOO and GORD are distinct clinical entities, theoretically with different pathophysiology and treatment. However, their differentiation may sometimes be problematic and disease previously thought to be associated with GORD may really be manifestations of EOO.155 156 Additionally, there may be a benefit of treatment of GORD in EOO, particularly in paediatric patients.157 158 Further confounding this issue of distinction is that patients not suspected of having EOO (those not undergoing preoperative biopsy) who receive antireflux surgery, have been reported to have poor outcomes.75 77 The finding of eosinophils on biopsy does not necessarily confirm the diagnosis of EOO or exclude other oesophageal diseases, therefore rendering the need for mandatory biopsies questionable.159 160 It has been shown that cost–benefit is only present when the prevalence of abnormal findings is expected to be 8% or more.161

Barium swallow

20. In patients with suspicion of HH or short oesophagus, a barium swallow is mandatory in the preoperative workup for antireflux surgery.

Statement endorsed, overall agreement 88.2%: A+ 44.1%, A 44.1%, A− 5.9%, D− 2.9%, D 2.9%, D+ 0.0%; GRADE B.

It is commonly accepted that 2.5 cm of intra-abdominal oesophagus is necessary to perform an effective antireflux procedure. Today, most antireflux operations are performed laparoscopically. The pneumoperitoneum necessary to perform laparoscopy elevates the diaphragm into the mediastinum and appears to ‘lengthen’ the oesophagus. Failure to recognise that the oesophagus is shortened may result in an inadequate length of intra-abdominal oesophagus at surgery. If a HH repair is constructed under tension on a short oesophagus, the hernia is reduced below the diaphragm at surgery and then retracts into the chest over time. The fundoplication may or may not remain subdiaphragmatically or it may disrupt or ‘slip’ onto the stomach. Slipped Nissen fundoplications may therefore result from the failure to recognise a shortened oesophagus before surgery. The occurrence of the true short oesophagus is indeed thought to be responsible for 20%–33% of the surgical failures after open or laparoscopic fundoplication.162 A study by Mattioli et al demonstrated that short oesophagus is present in about 20% of patients undergoing routine antireflux surgery, highlighting the importance of performing adequate testing.163

If the hernia is identified in the upright position, it is assumed that there is oesophageal shortening. In addition, the oesophagus is probably shortened when the HH length is 5 cm or greater alone or in combination with a stricture or a long segment (>3 cm) Barrett’s oesophagus.162 164 Other radiological findings that suggest a short oesophagus include severe extensive ulcerative oesophagitis, straightening or loss of the angle of His, the presence of a stricture alone and type III mixed or complex para-oesophageal hernias.165

In summary, if endoscopy reveals the presence of a large hernia and/or the presence of severe oesophagitis or long segment Barrett’s oesophagus, a barium swallow performed by a dedicated upper GI radiologist is strongly recommended before surgical intervention. This will allow to better plan the technical details of the surgery in order to eventually reduce the risk of anatomical and/or symptomatic recurrence.166 167

21. Patients with GORD symptoms and a small or medium size sliding HH on barium swallow are good candidates for antireflux surgery.

Statement endorsed, overall agreement 82.4%: A+ 20.6%, A 61.8%, A− 8.8%, D− 8.8%, D 0.0%, D+ 0.0%; GRADE B.

Very few studies have examined the effect of a HH on the outcome of antireflux surgery. A study by Power et al defined a HH size >3 cm at the time of the surgery as a predictor of failure.133 However, the presence and the size of a HH had no relationship with outcome according to several other studies.168–170

22. Patients with GORD symptoms and a large sliding HH on barium swallow are good candidates for antireflux surgery in the absence of short oesophagus.

Statement endorsed, overall agreement 85.3%: A+ 50%, A 35.3%, A− 8.8%, D− 2.9%, D 0.0%, D+2.9%; GRADE B.

Upper endoscopy and barium swallow are commonly used to diagnose short sliding HH. It has been demonstrated that in morbidly obese patients, barium swallow is superior to endoscopy in diagnosing sliding HH.171 Preoperative barium swallow can reveal more details on the sliding HH and contribute to better tailoring the antireflux surgery.166 Although there is currently no consensus on the definition for small, medium and large HH, often the cut-off of >3 cm or hernias belonging to categories II–IV have been used to define a large HH.94 172 173 As 2.5 cm of intra-abdominal oesophagus is mandatory to offer effective antireflux surgery, in large sliding HHs (larger intrathoracic component), a more comprehensive dissection is needed.164

23. Symptomatic patients with a para-oesophageal hernia on barium swallow are good candidates for antireflux surgery in addition to para-esophageal hernia repair.

Statement endorsed, overall agreement 97.1%: A+ 44.1%, A 52.9%, A− 2.9%, D− 0.0%, D 0.0%, D+ 0.0%; GRADE C.

Para-oesophageal hernias (POH) are subtypes of HH, defined as a herniation of the peritoneal cavity into the chest through the diaphragmatic hiatus.

Given the difficulty of distinguishing if reflux symptoms are from POH alone or independent of the POH, most surgeons routinely add an antireflux procedure (fundoplication) after POH repair in elective situations.174 A recent pilot trial by Muller-Stich et al showed a lesser degree of reflux and a less oesophagitis in patients where a fundoplication was added to the POH repair compared with those with a POH repair only.172

Some authors advocate a selective approach to antireflux procedures, with preoperative testing (including manometry, pH-metry or endoscopy) and patient symptoms determining whether or not to add a fundoplication.173 Others suggest always performing an antireflux procedure, but tailoring the type of fundoplication (eg, full or partial) depending on the patient.175 A minority suggest that fundoplication should be avoided due to the increased risk of dysphagia with antireflux procedures after POH repair.176 None of these approaches however have been proven superior to others in a prospective trial.

24. Patients with GORD symptoms and a short oesophagus on barium swallow are poor candidates for antireflux surgery.

Statement not endorsed, overall agreement 17.6%: A+ 2.9%, A 14.7%, A− 23.5%, D− 41.2%, D 11.8%, D+ 5.9%; GRADE C.

In the absence of adequate comparative studies, the question of the short oesophagus remains controversial, and there is insufficient evidence to preclude patients with radiological suspicion of a short oesophagus from antireflux surgery. If patients progress to surgery, there is also insufficient evidence to define the best surgical procedure in this scenario. Well-designed case–control or randomised clinical trials are needed to provide an evidence base to address this question.

Oesophageal manometry

25. Oesophageal manometry is mandatory to select patients for antireflux surgery.

Statement endorsed, overall agreement 94.1%: A+ 82.4%, A 11.8%, A− 5.9%, D− 0.0%, D 0.0%, D+ 0.0%; GRADE D.

Oesophageal manometry should be performed prior to antireflux surgery to rule out a major motor disorder, such as achalasia, OGJ outflow obstruction or absent contractility.177 178 There is no data to support that the manometric finding of distal oesophageal spasm (DOS), Jackhammer oesophagus or minor disorders of peristalsis, such as fragmented peristalsis predicts postoperative dysphagia. Incorporating HRM and impedance into pressure flow parameters might be helpful in predicting outcome since the dysphagia risk index appeared to be helpful in identifying patients at risk for post-fundoplication dysphagia.179

26. Patients with GORD symptoms and a hypercontractile oesophagus (Jackhammer and the previously described Nutcracker) oesophagus on manometry are good candidates for antireflux surgery if symptoms can be attributed to reflux.

Statement not endorsed, overall agreement 64.7%: A+ 11.8%, A 52.9%, A− 29.4%, D− 2.9%, D 2.9%, D+ 0.0%; GRADE D.

Data on outcome of antireflux surgery of patients with a hypercontractile oesophagus is scarce: there are no randomised, controlled trials available in literature. However, retrospective data on outcome of patients with nutcracker oesophagus (although no longer defined in the Chicago classification V.3.0) undergoing antireflux surgery show no difference compared with patients with a normal oesophageal motility pattern.180 Manometric abnormalities after a Nissen fundoplication were even improved in two patients with a Jackhammer oesophagus.181 Hypertensive oesophageal contraction patterns are not a contraindication for antireflux surgery; however, patients and clinicians should be aware of the risk of developing chest pain after the surgery.182

27. Patients with GORD symptoms and distal oesophageal spasm on manometry are poor candidates for antireflux surgery.

Statement not endorsed, overall agreement 64.7%: A+ 26.5%, A 38.2%, A− 20.6%, D− 11.8%, D 0.0%, D+ 0.0%; GRADE D.

Patients with DOS are poor candidates for antireflux surgery, provided that the motor disorder has been well characterised, preferably using HRM. Therapeutic approaches indicated for patients with DOS include medicines such as sildenafil, as well as endoscopic injection of botulin toxin and surgical myotomy.183–185 Although some patients may benefit from acid-suppressive therapy, antireflux surgery as the unique treatment should be avoided in patients with DOS.

28. In patients with GORD symptoms and hypocontractility of the oesophageal body on manometry, antireflux surgery should be tailored.

Statement not endorsed, overall agreement 47.1%: A+ 5.9%, A 41.2%, A− 41.2%, D− 0.0%, D 5.9%, D+ 5.9%; GRADE D.

There are no good data to suggest tailoring of antireflux surgery to oesophageal body hypomotility or hypocontractility.186 187 Provocative manoeuvres during manometry could in the future identify patients where peristaltic performance following fundoplication can modify the risk for postoperative dysphagia. Multiple rapid swallows (MRS) are often added to the manometric protocol as a marker for esophageal body peristaltic reserve. It has been shown that MRS testing before laparoscopic antireflux surgery is able to help predict late postoperative dysphagia.188 189

29. Patients with GORD symptoms and severe hypocontractility or failed peristalsis on manometry are poor candidates for antireflux surgery.

Statement not endorsed, overall agreement 64.7%: A+ 8.8%, A 55.9%, A− 23.5%, D− 2.9%, D 8.8%, D+ 0.0%; GRADE D.

Hypocontractility is not a contraindication for antireflux surgery, since surgery more often than not improved these manometric abnormalities.186 Further research is warranted since very little data exists on outcome of patients with the most severe hypocontractility or aperistalsis. Similar as described above (statement 28) is the importance of adding MRS during a manometric protocol, which is a marker of contractile reserve of the oesophagus.188 189 In addition, antireflux surgery can be tailored to each individual patient.

It has to be repeated that the main indication for manometry in patients with GORD considered for antireflux surgery is to identify patients with aperistalsis due to achalasia, who are candidates for fundoplication only when combined with myotomy of the LOS.

Reflux monitoring

30. Oesophageal pH (±impedance) monitoring off therapy is mandatory to select patients with NERD for antireflux surgery

Statement endorsed, overall agreement 97.1%: A+ 91.2%, A 5.9%, A− 0.0%, D− 2.9%, D 0.0%, D+ 0.0%; GRADE B.

In the absence of oesophagitis (ie, presence of mucosal breaks), pathological GOR and/or positive reflux symptom association ‘off’ therapy should be documented before embarking to antireflux surgery.22 33 190–193 In the preoperative setting, the added value of impedance in patients ‘off’ therapy remains to be determined.

Data on preoperative assessment ‘on’ PPIs are scarce. Few uncontrolled and short studies suggest that good postoperative outcomes can be achieved in patients who are refractory to PPIs in whom pH-impedance monitoring demonstrated either an abnormal number of reflux episodes or positive symptom association analysis.33 194

31. Oesophageal pH (±impedance) monitoring off therapy should be performed for selection for antireflux surgery of patients who have short Barrett’s oesophagus in the absence of erosive oesophagitis.

Statement endorsed, overall agreement 88.2%: A+ 41.2%, A 47.1%, A− 5.9%, D− 0.0%, D 5.9%, D+ 0.0%; GRADE B.

Oesophageal (impedance-) pH-monitoring off therapy should be performed in patients with short segment Barrett’s oesophagus as it provides an objective quantification of patient’s GOR.190 This evaluation of PPI therapy would provide a baseline comparator in assessing the efficacy of acid-suppressive therapy and/or reflux-reducing therapy.

32. Patients with GORD symptoms and normal reflux exposure on pH (±impedance) monitoring off PPI therapy are poor candidates for antireflux surgery.

Statement endorsed, overall agreement 82.4%: A+ 17.6%, A 64.7%, A− 17.6%, D− 0.0%, D 0.0%, D+ 0.0%; GRADE B.

There is very limited data examining the outcomes of surgery in patients with normal reflux monitoring. This is in large part due to the fact that most of the studies evaluating outcomes of antireflux surgery require abnormal reflux monitoring as a criteria to be eligible for surgery.56 Based on the available evidence, it would appear that patients with normal reflux exposure on pH (±impedance) monitoring off therapy are indeed poor candidates for antireflux surgery.

33a. Patients with GORD symptoms, a normal reflux exposure on pH (±impedance) monitoring off therapy and a positive symptom association are good candidates for antireflux surgery.

Statement not endorsed, overall agreement 58.8%: A+ 14.7%, A 44.1%, A− 23.5%, D− 17.6%, D 0.0%, D+ 0.0%.

33b. Patients with GORD symptoms, a normal reflux exposure on pH (±impedance) monitoring off therapy and a positive reflux symptom association are good candidates for antireflux surgery, only if symptoms respond to PPI therapy.

Statement not endorsed, overall agreement 73.5%: A+ 11.8%, A 61.8%, A− 14.7%, D− 5.9%, D 5.9%, D+ 0.0%.

There is a very limited data examining the outcomes of surgery in patients with normal reflux monitoring. This is in large part due to the fact that most of the studies evaluating outcomes of antireflux surgery require abnormal reflux monitoring as a criteria to be eligible for surgery.56 Some studies do suggest that reflux-hypersensitive patients with typical symptoms and an unsatisfactory response to PPIs may benefit from antireflux surgery with an outcome similar to the one of patients with pathological reflux.31 195 However, as mentioned above (statement 3), a recent study by Patel et al showed that pure acid sensitivity was a negative predictor for symptom improvement with antireflux therapy, including surgical management.35 Results should therefore be interpreted with caution.

34a. Patients with GORD symptoms and pathological reflux exposure on pH (±impedance) monitoring off therapy and a negative reflux symptom association are eligible for antireflux surgery.

Statement not endorsed, overall agreement 58.8%: A+ 5.9%, A 52.9%, A− 29.4%, D− 8.8%, D 2.9%, D+ 0.0%.

34b. Patients with GORD symptoms and pathological reflux exposure on pH (±impedance) monitoring off therapy and a negative reflux symptom association are eligible for antireflux surgery, only if symptoms respond to PPI therapy.

Statement not endorsed, overall agreement 66.7%: A+ 6.1%, A 60.6%, A− 9.1%, D− 6.1%, D 18.2%, D+ 0.0%.

The literature available suggests that patients with proven pathological acid exposure who do not experience symptoms during pH (±impedance) monitoring or presenting a negative symptom–reflux association may still obtain good results from anti-reflux surgery.33 192 Moreover, there is a subgroup of patients that is truly refractory to PPIs, with ongoing acid secretion.193

35. Patients with pathological reflux exposure on pH (±impedance) monitoring on PPI who respond to baclofen therapy are good candidates for antireflux surgery.

Statement not endorsed, overall agreement 20.6%: A+ 5.9%, A 14.7%, A− 61.8%, D− 2.9%, D 11.8%, D+ 2.9%.

Baclofen, a GABA B-agonist, is known to reduce the number of transient LOS relaxations and subsequently, it reduces all types of reflux, including weakly acidic reflux.196 To date, there are no studies comparing baclofen with antireflux surgery, therefore it would be too speculative to say that patients responding to baclofen are good candidates for antireflux surgery. In the very recently published paediatric GOR clinical guidelines, the use of baclofen prior to antireflux surgery can be considered in children in whom other pharmacological treatments have failed (weak recommendation).197

Gastric emptying

36. A gastric emptying test for solid food is necessary to select patients with GORD with concomitant dyspeptic symptoms for antireflux surgery.

Statement not endorsed, overall agreement 5.9%: A+ 2.9%, A 2.9%, A− 8.8%, D− 5.9%, D 67.6%, D+ 11.8%; GRADE C.

Studies performed to assess the role of a preoperative gastric emptying test in antireflux surgery have generated controversial results: some studies have shown that this evaluation is useful to select the best type of surgery and to avoid surgical failures, while others have denied the validity of such an approach.198–200 However, so far no study has been performed to establish whether the assessment of gastric emptying is relevant or not to favour success of surgery in patients with GORD with concomitant dyspepsia symptoms.

37. If the gastric emptying test is abnormal for solid food, patients should not undergo an antireflux surgery.

Statement not endorsed, overall agreement 2.9%: A+ 0.0%, A 2.9%, A− 0.0%, D− 20.6%, D 67.6%, D+ 8.8%; GRADE C.

Literature shows that there is no evidence to suggest that preoperative slow gastric emptying for solids is associated with a poor outcome after surgery with regard to reflux parameters.201 202 A study by Lundell et al suggests that a slow preoperative gastric emptying for solids is weakly associated with symptoms of bloating.202 However, two other studies investigating the relationship between gastric emptying rates before and outcome after antireflux surgery could not confirm this.198 199 There is insufficient evidence to support the statement.

Recommendations

Based on the statements that generated consensus, a number of recommendations can be made for selecting patients for antireflux surgery. These are summarised in table 4.

The Delphi process also identified several areas of uncertainty, requiring further research. It is unclear whether patients with regurgitation as a main symptom, patients with NCCP, patients with extra-oesophageal manifestations of reflux and patients with dental erosions are good candidates for anti-reflux surgery (statements 2, 5 and 12). There is a lack of prospective controlled trials to support these statements. Patients with reflux hypersensitivity, patients with concomitant FD and IBS and patients with major psychiatric comorbidity are not considered good candidates for antireflux surgery (statements 3, 8 and 10). There is a need for additional markers of beneficial outcome of antireflux surgery in these patients, given the frequent overlap of GORD with FD and IBS symptoms. There is no consensus that patients with scleroderma are poor candidates for antireflux surgery (statement 7). It is unclear to which extent patients with Jackhammer (or Nutcracker) oesophagus or spasm on manometry are eligible for antireflux surgery (statements 26 and 27). The impact of oesophageal hypocontractility on the eligibility or type of antireflux surgery is unclear (statements 28 and 29). It is unclear whether patients with reflux hypersensitivity are eligible for antireflux surgery (statement 33). Finally, it is unclear whether patients with pathological reflux monitoring but negative symptom association are good candidates for antireflux surgery (statement 34).

It is important to stress that the decision of referring a patient for antireflux surgery has to take into account all positive as well as all negative support findings. Selecting patients suitable for antireflux surgery cannot be captured by one single statement and remains subject to guided clinical judgement and patient preference.