Fundoplication in laparoscopic Heller's cardiomyotomy for achalasia

Abstract

Background

Laparoscopic Heller’s cardiomyotomy (LHC) is the preferred treatment of achalasia. It improves dysphagia by dividing muscles of the lower oesophageal sphincter, but this intervention can result in debilitating gastro‐oesophageal reflux symptoms in some patients. To prevent these reflux symptoms, most surgeons add a fundoplication to Heller’s cardiomyotomy, but there is no consensus regarding this or the type of fundoplication which is best suited for the purpose.

Objectives

To assess how the addition of a fundoplication affects postoperative reflux and dysphagia in people undergoing LHC and compare the different types of fundoplications used in combination with LHC to determine which is better at controlling reflux without worsening the dysphagia.

Search methods

We searched three databases (CENTRAL, MEDLINE and Embase) on 31 October 2021 and trial registers to identify all published and unpublished randomised controlled trials (RCTs) in any language, comparing different fundoplications used in combination with LHC to treat achalasia. We also included RCTs where LHC with a fundoplication is compared with LHC without any fundoplication.

Selection criteria

We only included RCTs which recruited adult participants with achalasia undergoing LHC with minimal hiatal dissection. We excluded non‐randomised studies or studies involving paediatric participants. We also excluded studies where the procedure was done by open surgery and where circumferential hiatal dissection of the oesophagus was carried out, unless it was necessary to reduce a hiatus hernia or to facilitate a Toupet or Nissen fundoplication.

Data collection and analysis

Two review authors independently identified studies to be included, assessed risk of bias using the Cochrane RoB 1 tool, and extracted the data. We calculated the risk ratio (RR) with 95% confidence interval (CI) using both fixed‐effect and random‐effect models with Review Manager (RevMan) software.

Main results

We included eight studies in this review, with a total of 571 participants with an average age of 45 years (range 33.5 to 50). LHC without any fundoplication was performed in 65 (11.3%) participants, 298 (52.1%) had Dor fundoplication, 81 (14.1%) had Toupet fundoplication, 72 (12.6%) had Nissen's fundoplication, and 55 (9.6%) participants had angle of His accentuation.

Three studies with a total of 143 participants compared LHC + Dor to LHC without fundoplication. We found that the evidence is very uncertain as to whether the addition of a Dor fundoplication made any difference to the outcome of postoperative pathological acid reflux (RR 0.37, 95% CI 0.07 to 1.89; I² = 56%; 2 studies, 97 participants; very low‐certainty evidence) and uncertain for severe postoperative dysphagia (RR 3.00, 95% CI 0.34 to 26.33; I² = 0%; 3 studies, 142 participants; low‐certainty evidence).

Three studies with 174 participants compared LHC + Dor to LHC + Toupet. The evidence suggests that there may be little to no difference in the outcomes of postoperative pathological acid reflux (RR 0.75, 95% CI 0.23 to 2.43; I² = 60%; 3 studies, 105 participants; low‐certainty evidence) and severe postoperative dysphagia (RR 0.78, 95% CI 0.19 to 3.15; I² = 0%; 3 studies, 123 participants; low‐certainty evidence) between the two interventions, but the certainty of the evidence is low.

One study with 138 participants compared LHC + Dor to LHC + Nissen. Nissen fundoplication caused increased severe postoperative dysphagia (RR 0.19, 95% CI 0.04 to 0.83; 1 study, 138 participants; high‐certainty evidence) when compared to Dor fundoplication. This study did not show a difference in postoperative pathological acid reflux (RR 4.72, 95% CI 0.23 to 96.59; 1 study, 138 participants; low‐certainty evidence), but the certainty of evidence is low.

One study with 110 participants compared LHC + Dor with LCH + angle of His accentuation, and reported that severe postoperative dysphagia was similar between the two interventions (RR 1.56, 95% CI 0.27 to 8.95; 1 study, 110 participants; moderate‐certainty evidence), with moderate certainty of evidence. This study did not report on postoperative pathological acid reflux.

Authors' conclusions

When LHC was performed with minimal hiatal dissection, we were very uncertain whether the addition of a Dor fundoplication made a difference in controlling postoperative reflux, and we were uncertain if it increased the risk of severe postoperative dysphagia. There may be little to no difference in the outcomes of postoperative pathological acid reflux or severe dysphagia between Dor and Toupet fundoplications when used in combination with LHC, but the certainty of the evidence is low. Nissen (total) fundoplication used in combination with LHC for achalasia increased the risk of severe postoperative dysphagia. The angle of His accentuation and Dor fundoplication had a similar effect on severe postoperative dysphagia when combined with LHC, but their effect on postoperative pathological acid reflux was not reported.

Plain language summary

Addition of a stomach wrap (fundoplication) around the food pipe (oesophagus) in keyhole (laparoscopic) surgery to treat achalasia

Review questions

1) Does the addition of a stomach wrap around the food pipe (oesophagus) reduce the backflow of food and acid from the stomach to the food pipe (regurgitation) when used with keyhole surgery to treat achalasia?

2) Which type of stomach wrap (partial or total) around the food pipe is better at controlling regurgitation without causing difficulty in swallowing (dysphagia) when used with keyhole surgery to treat achalasia?

Background

Achalasia is a rare disease caused by a permanent state of contraction of a valve (lower oesophageal sphincter) at the lower end of the food pipe. It causes difficulty in swallowing which can be relieved by keyhole surgery to divide the muscles (Heller's cardiomyotomy) of this valve. The function of this valve is to stop the regurgitation of food and acid (reflux) from the stomach to the food pipe. Dividing the muscles of this valve, as a treatment for achalasia, can lead to increased acid regurgitation. To prevent this regurgitation, some surgeons, wrap the upper part of the stomach around the lower end of the food pipe. This stomach wrap acts as an artificial sphincter preventing acid regurgitation. The role of this wrap in controlling regurgitation is not universally accepted by all surgeons, and there is also disagreement as to the type of wrap (partially attached to the front (anterior) or back (posterior) of the stomach, or completely encircling the stomach (total)) that is best at controlling regurgitation, without worsening the difficulty in swallowing.

Study characteristics

We only included studies where adults with achalasia are treated with keyhole surgery to divide the muscles of the valve at the lower end of the food pipe; studies used different types of stomach wraps or no wraps at all. In all these studies, the participants were allocated to a treatment group at random (randomised controlled trial) to ensure uniform representation and reduce any error due to prejudice (bias). Comparing the results of these different techniques, we have tried to find out if the addition of a stomach wrap helps to prevent acid regurgitation, and also the type of wrap that is better at controlling regurgitation. We searched three databases (CENTRAL, MEDLINE, Embase) on 31 October 2021 and the evidence provided is current to October 2021. We included eight studies in this review (a total of 571 participants). All participants underwent keyhole surgery to divide the muscles of the valve at the lower end of the food pipe. In addition, 298 participants had partial anterior (Dor) stomach wrap, 81 had partial posterior (Toupet) wrap, 72 had total (Nissen) wrap, and 55 had the stomach hitched to the food pipe (angle of His augmentation) with stitches, and 65 had no additional procedures.

Key results

The evidence is very uncertain whether or not the addition of a partial anterior (Dor) wrap makes a change to the outcomes of acid regurgitation and is uncertain for difficulty in swallowing in these participants. There seems to be no difference in the outcomes between the partial anterior (Dor) and partial posterior (Toupet) wraps in terms of acid regurgitation and difficulty in swallowing, but there is uncertainty about the evidence. The total (Nissen) wrap causes increased swallowing difficulties when compared to the partial anterior (Dor) wrap. The study comparing partial anterior (Dor) wrap to a procedure where the stomach was simply hitched to the food pipe with a few stitches did not show any difference in difficulty of swallowing after the surgery, but did not report on acid regurgitation.

Certainty of evidence

All studies included in the review are well constructed. However, one study has only been published as an abstract, giving us limited information. The other limitations of this review are the small number of participants and short follow‐up periods in the studies. There are also differences in defining, measuring and assessing the outcomes in these studies. We have assessed the overall certainty of evidence in this review as low.

Summary of findings

Summary of findings 1. LHC + Dor compared to LHC without fundoplication for achalasia.

LHC + Dor compared to LHC without fundoplication for achalasia
Patient or population: adult patients (> 18 years) with achalasia Setting: tertiary hospitals Intervention: LHC + Dor Comparison: LHC without fundoplication
Outcomes	Anticipated absolute effects* (95% CI)		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)
	Risk with LHC without fundoplication	Risk difference with LHC + Dor
Postoperative pathological acid reflux assessed with: 24 hour pH study follow‐up: 6 months to 36 months	300 per 1000	189 fewer per 1000 (279 fewer to 267 more)	RR 0.37 (0.07 to 1.89)	97 (2 RCTs)	⊕⊝⊝⊝ Very lowa,b,c
Severe postoperative dysphagia assessed with: clinical scoring systems follow‐up: 6 months to 36 months	10 per 1000	20 more per 1000 (7 fewer to 253 more)	RR 3.00 (0.34 to 26.33)	143 (3 RCTs)	⊕⊕⊝⊝ Lowd,e,f
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; LHC: laparoscopic Heller's cardiomyotomy; RCT: randomised controlled trial; RR: risk ratio
GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: we are moderately confident in the effect estimate; the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: our confidence in the effect estimate is limited; the true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate; the true effect is likely to be substantially different from the estimate of effect.

^aStudies have different definitions for pathological acid reflux: certainty of evidence downgraded by 1 level for indirectness.
^bHeterogeneity (I² = 56%): certainty of evidence downgraded by 1 level for inconsistency.
^cStudies have different follow‐up periods: certainty of evidence downgraded by 1 level for indirectness.
^dStudies have different scoring system for dysphagia: certainty of evidence downgraded by 1 level for indirectness.
^eLarge 95% CI and small number of events: certainty of evidence downgraded by 1 level for imprecision.
^fAssumed baseline risk of 1%.

Summary of findings 2. LHC + Dor compared to LHC + Toupet for achalasia.

LHC + Dor compared to LHC + Toupet for achalasia
Patient or population: adult patients (> 18 years) with achalasia Setting: tertiary hospitals Intervention: LHC + Dor Comparison: LHC + Toupet
Outcomes	Anticipated absolute effects* (95% CI)		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)
	Risk with LHC + Toupet	Risk difference with LHC + Dor
Postoperative pathological acid reflux assessed with: 24 hour pH study follow‐up: 12 months to 36 months	294 per 1000	74 fewer per 1000 (226 fewer to 421 more)	RR 0.75 (0.23 to 2.43)	105 (3 RCTs)	⊕⊕⊝⊝ Lowa,b
Severe postoperative dysphagia assessed with: clinical scoring systems follow‐up: 12 months to 36 months	88 per 1000	19 fewer per 1000 (71 fewer to 189 more)	RR 0.78 (0.19 to 3.15)	123 (3 RCTs)	⊕⊕⊝⊝ Lowa,c
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; LHC: laparoscopic Heller's cardiomyotomy; RCT: randomised controlled trial; RR: risk ratio
GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: we are moderately confident in the effect estimate; the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: our confidence in the effect estimate is limited; the true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate; the true effect is likely to be substantially different from the estimate of effect.

^aLarge 95% CI: certainty of evidence downgraded by 1 level for imprecision.
^bHeterogeneity (I² = 60%): certainty of evidence downgraded by 1 level for inconsistency.
^cStudies have used different scoring systems for dysphagia: certainty of evidence downgraded by 1 level for indirectness.

Summary of findings 3. LHC + Dor compared to LHC + Nissen for achalasia.

LHC + Dor compared to LHC + Nissen for achalasia
Patient or population: adult patients (> 18 years) with achalasia Setting: tertiary hospitals Intervention: LHC + Dor Comparison: LHC + Nissen
Outcomes	Anticipated absolute effects* (95% CI)		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)
	Risk with LHC + Nissen	Risk difference with LHC + Dor
Postoperative pathological acid reflux assessed with: 24 hour pH study follow‐up: 60 months	1 per 1000	4 more per 1000 (1 fewer to 96 more)	RR 4.72 (0.23 to 96.59)	138 (1 RCT)	⊕⊕⊝⊝ Lowa,b
Severe postoperative dysphagia assessed with: clinical scoring system follow‐up: 60 months	149 per 1000	121 fewer per 1000 (143 fewer to 25 fewer)	RR 0.19 (0.04 to 0.83)	138 (1 RCT)	⊕⊕⊕⊕ High
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; LHC: laparoscopic Heller's cardiomyotomy; RCT: randomised controlled trial; RR: risk ratio
GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: we are moderately confident in the effect estimate; the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: our confidence in the effect estimate is limited; the true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate; the true effect is likely to be substantially different from the estimate of effect.

^aAssumed baseline risk of 0.1%.
^bVery large 95% CI and low number of events: certainty of evidence downgraded by 2 levels for imprecision.

Summary of findings 4. LHC + Dor compared to LHC + angle of His accentuation for achalasia.

LHC + Dor compared to LHC + angle of His accentuation for achalasia
Patient or population: adult patients (> 18 years) with achalasia Setting: tertiary hospitals Intervention: LHC + Dor Comparison: LHC + angle of His accentuation
Outcomes	Anticipated absolute effects* (95% CI)		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
	Risk with LHC + angle of His accentuation	Risk difference with LHC + Dor
Postoperative pathological acid reflux assessed with: 24 hour pH study	‐	‐	‐	‐	‐	Not measured
Severe postoperative dysphagia assessed with: clinical scoring system follow‐up: 36 months	36 per 1000	20 more per 1000 (26 fewer to 284 more)	RR 1.56 (0.27 to 8.95)	110 (1 RCT)	⊕⊕⊕⊝ Moderatea
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; LHC: laparoscopic Heller's cardiomyotomy; RCT: randomised controlled trial; RR: risk ratio
GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: we are moderately confident in the effect estimate; the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: our confidence in the effect estimate is limited; the true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate; the true effect is likely to be substantially different from the estimate of effect.

^aLarge 95% CI: certainty of evidence downgraded by 1 level for imprecision.

Background

Description of the condition

Achalasia is a rare primary motility disorder of the oesophagus, characterised by an absence of oesophageal peristalsis and incomplete relaxation of a frequently hypertensive lower oesophageal sphincter (LOS) (Cotran 1999). The incidence of achalasia is about 1 in 100,000 people per year, and the prevalence is about 10 per 100,000 people (Sadowski 2010).

The pathological changes found in achalasia consist of myenteric inflammation with injury to and subsequent loss of ganglion cells and fibres of myenteric nerves (Goldblum 1996). There is reduction in synthesis of nitric oxide and vasoactive intestinal polypeptide. The probable aetiology of the disease is thought to be an autoimmune mediated destruction of the inhibitory neurons in response to an unknown insult (Stefanidis 2012).

Affected people usually present with one or more of the following symptoms: dysphagia to both solid and liquid diet (> 90%), regurgitation of undigested food (76% to 91%), nocturnal cough (30%), aspiration (8%), chest pain (25% to 64%), heartburn (18% to 52%), and weight loss (35% to 91%) (Boeckxstaens 2014). Investigations performed to confirm diagnosis are:

1. barium swallow which shows a smooth tapering of the lower oesophagus (bird's beak appearance);

2. oesophageal manometry which shows reduced peristalsis of the oesophagus with hypertensive, non‐relaxing LOS;

3. oesophagogastroduodenoscopy to rule out any tumour at the gastro‐oesophageal junction (pseudoachalasia) (Pandolfino 2015; Stefanidis 2012; Vaezi 2013).

Available treatments are:

1. pharmacotherapy: calcium channel blockers and long‐acting nitrates;

2. botulinum toxin injections into the LOS;

3. pneumatic dilation of the LOS;

4. peroral endoscopic myotomy (POEM);

5. robotic Heller's cardiomyotomy;

6. laparoscopic Heller's cardiomyotomy (LHC).

Pharmacotherapeutic agents have incomplete symptom control, lose their potency over time, and have to be taken daily. Botulinum toxin injection relieves symptoms in 85% of people but its effects only last for 6 to 12 months (Pandolfino 2015; Stefanidis 2012; Vaezi 2013). Pneumatic dilation is currently considered the most effective non‐surgical treatment with a response rate ranging from 50% to 93% at 10 years with on‐demand repeat dilations (Pandolfino 2015; Stefanidis 2012; Vaezi 2013). Treatment with single dilation is only effective in 13% of people (Stefanidis 2012). POEM was first described in 2010 (Inoue 2010), and short‐term results show that it might be more effective than LHC in relieving dysphagia, but it is associated with very high incidence (29%) of pathological reflux (Familari 2016; Schlottmann 2018). The long‐term results of this procedure are still not available. Robotic Heller's cardiomyotomy was first reported in 2001 (Melvin 2001), and compared to LHC has reduced incidence of intraoperative oesophageal mucosal perforations, but there is no clear advantage in postoperative morbidity, symptom relief and long‐term outcomes (Maeso 2010). LHC provides symptomatic improvement in 89% (range 77% to 100%) of people on long‐term follow‐up (Campos 2009), and is superior to open abdominal, thoracic or thoracoscopic approaches of Heller's cardiomyotomy (Campos 2009; Stefanidis 2012). LHC with a fundoplication is considered the surgical treatment of choice for achalasia (Campos 2009; Stefanidis 2012).

See Appendix 1 for glossary of terms.

Description of the intervention

In 1913, Ernest Heller described extramucosal surgical myotomy of the lower oesophagus and proximal stomach in patients with chronic cardiospasm (Heller 1914). Laparoscopic Heller's procedure was first reported in Shimi 1991. The LHC combined with an anti‐reflux procedure is now consider the best treatment to control the symptoms of achalasia (Campos 2009; Patti 2010).

LHC is performed under general anaesthesia (Sharp 2002). Five ports are placed on the upper anterior abdomen to introduce a laparoscopic camera, a liver retractor and the instruments (Pellegrini 1992; Sharp 2002). The anterior phreno‐oesophageal ligament is divided at the hiatus to expose the anterior surface of the oesophagus. This limited hiatal dissection is preferred over circumferential hiatal dissection with disruption of posterior oesophageal attachments, as it produces less acid reflux postoperatively (Bonavina 2006; Braghetto 2007). The muscles of the oesophagus, including the LOS, are divided at the gastro‐oesophageal junction to expose the mucosa. The division of the muscles is carried proximally into the oesophagus for 4 cm to 6 cm and distally into the stomach for 1 cm to 2 cm (Sharp 2002; Stefanidis 2012). Any hiatal defect is then apposed using interrupted non‐absorbable sutures (Stefanidis 2012). Some surgeons perform simultaneous intraoperative endoscopy to assess the adequacy of the myotomy and rule out any oesophageal mucosal perforation (Sharp 2002).

A fundoplication is added to LHC to reduce the symptoms of gastro‐oesophageal reflux, produced by division of the LOS (Campos 2009; Stefanidis 2012). In fundoplication, the greater curvature of the stomach may be mobilised by dividing the short gastric vessels and wrapping around the lower oesophagus to produce an artificial sphincter (Hunter 1997). Depending on the amount of circumference of the oesophagus encircled by the stomach wrap, the fundoplications are classified as total (360° or Nissen) or partial. The most common partial fundoplications used with LHC are Dor (anterior 180° wrap) and Toupet (posterior 270° wrap). When a Dor fundoplication is created, the fundus is sutured to each edge of the myotomy and to the right crus with non‐absorbable sutures, overlying the anterior surface of the distal oesophagus, including the myotomy defect (Dempsey 2004). If a Toupet fundoplication is created, the mobilised fundus is pulled from left to right behind the oesophagus and is sutured to the edges of the myotomy wound on each side. It creates a wrap, covering the posterior and lateral aspects of the oesophagus, leaving the mucosa in the floor of the myotomy wound exposed (Katada 2006). In Nissen fundoplication, the fundus is pulled from left to right behind the oesophagus, and is sutured on to itself in front of the oesophagus, forming a complete wrap encircling the distal oesophagus (Rossetti 2005).

Angle of His accentuation is an alternative surgical procedure to fundoplication to reduce gastro‐oesophageal reflux. This procedure is carried out by suturing the gastric fundus to the left edge of the myotomy.

How the intervention might work

The main symptom of achalasia is dysphagia, which is caused by reduced peristalsis of the oesophageal muscles and a hypertensive, non‐relaxing LOS. The division of muscles (myotomy) of the hypertensive, non‐relaxing LOS relieves the symptoms of dysphagia but it can also cause considerable gastro‐oesophageal reflux (Burpee 2005; Kjellin 1999). A fundoplication is added to LHC (Campos 2009; Stefanidis 2012), to create an artificial sphincter and prevent gastro‐oesophageal reflux and its symptoms.

Why it is important to do this review

The role of fundoplication in combination with LHC is debatable. Many studies have reported that addition of a fundoplication does not improve reflux symptoms following LHC (Gupta 2005; Lyass 2003; Sharp 2002). However, the Society of American Gastrointestinal and Endoscopic Surgeons (SAGES) guidelines for the surgical treatment of achalasia strongly recommend that a person undergoing myotomy should have a fundoplication to prevent reflux (Stefanidis 2012). One systematic review and meta‐analysis of endoscopic and surgical treatment of achalasia concluded that laparoscopic myotomy with fundoplication was the most effective surgical treatment (Campos 2009).

There is no agreement on which type of fundoplication is better at controlling the reflux symptoms without worsening the dysphagia. An anterior 180° Dor fundoplication (Cuttitta 2011; Rice 2005), a posterior 270° Toupet fundoplication (Katada 2006; Kurian 2013), and complete 360° Nissen fundoplication (Di Martino 2011; Falkenback 2003; Rossetti 2005) have all been proposed to be effective in combination with Heller's cardiomyotomy. SAGES guidelines for the surgical treatment of achalasia have recommended a partial fundoplication over total fundoplication because of decreased dysphagia rates and similar reflux control, but has not recommended which partial fundoplication provides the best reflux control following LHC (Stefanidis 2012).

There are a few systematic reviews on this topic, but their conclusions are not clear and their evidence is limited due to lack of randomised controlled trials (RCTs) (Kurian 2013; Mayo 2012; Wei 2013). Since 2013, some new RCTs have been published comparing different types of fundoplications in association with LHC (Grubnik 2017; Gupta 2018; Torres‐Villalobos 2018). More recently, two systematic reviews compared Dor versus Toupet fundoplication in patients undergoing LHC (Aiolfi 2020; Siddaiah‐Subramanya 2019). While Siddaiah‐Subramanya 2019 found LHC + Toupet better in terms of length of hospital stay and quality of life, Aiolfi 2020 found there was no difference between the interventions. Moreover, there is no recent study comparing partial fundoplications to LHC without any fundoplication, hence there is an opportunity to revisit the topic and perform a systematic review on the role of fundoplication in people undergoing LHC and which fundoplication is best suited for this purpose.

Objectives

To assess how the addition of a fundoplication affects postoperative reflux and dysphagia in people undergoing LHC and compare the different types of fundoplications used in combination with LHC to determine which is better at controlling reflux without worsening the dysphagia.

Methods

Criteria for considering studies for this review

Types of studies

We only included RCTs in this review which have been published as full text, abstracts or any unpublished data. We included studies which have been published in any language.

Types of participants

We only included studies recruiting adults (aged 18 years or older) undergoing LHC for achalasia; we excluded studies with participants younger than 18 years.

Types of interventions

We only included studies where Heller's cardiomyotomy and fundoplication have been performed laparoscopically with minimal hiatal dissection. We excluded studies where circumferential oesophageal dissection was performed unless it was required to reduce a hiatal hernia or to facilitate Toupet or Nissen fundoplication.

We included studies where one type of fundoplication is compared with a different type of fundoplication in people undergoing LHC for achalasia. We included studies where one type of fundoplication combined with LHC is compared against LHC without any fundoplication. We excluded studies where LHC with or without fundoplication has been compared with non‐surgical interventions (e.g. endoscopic dilation, botulinum toxin injections or POEM).

Types of outcome measures

We included all studies irrespective of the outcomes they reported.

Primary outcomes

1. Postoperative pathological acid reflux: obtained from pH studies

2. Severe postoperative dysphagia (Eckardt score > 3) (Eckardt 1992) and those requiring medical or surgical intervention

Secondary outcomes

1. Health‐related quality of life scores (CDC 2000)

2. Major complications (Clavien‐Dindo grade III and above) (Clavien 2009; Dindo 2004): for example, perforation, pneumothorax and bleeding

3. LOS pressure: obtained from postoperative oesophageal manometry

4. DeMeester score (Johnson 1974): obtained from postoperative pH studies

5. Length of hospital stay

6. Operating time

Gastro‐oesophageal reflux and recurrent dysphagia are the two major postoperative problems in people undergoing LHC for achalasia. Pathological acid reflux obtained from pH studies is used to objectively diagnose reflux (Lacey 2010; Reynolds 2007). The DeMeester score is another composite outcome obtained from pH studies, and values above 14 usually indicate significant reflux. Recurrent dysphagia following LHC can be of varying severity, from mild intermittent symptoms which do not need treatment, to persistent severe dysphagia requiring medical or surgical treatment. A postoperative Eckardt score > 3, is considered as treatment failure. We only included severe dysphagia with an Eckardt score > 3 (treatment failure) and those requiring medical or surgical intervention as a primary outcome; milder dysphagia is difficult to measure and could be due to incomplete myotomy or aperistalsis of the oesophagus which is associated with achalasia. The LOS sphincter pressure obtained by postoperative oesophageal manometry is another indirect test to determine the constricting force of the fundoplication around the distal oesophagus, which could lead to dysphagia.

We also used other standard outcomes for comparing interventions, such as quality of life scores, major complications, length of hospital stay, and operating time, to compare the different types of fundoplications.

Reporting of the outcomes listed here was not an inclusion criterion for the review.

Search methods for identification of studies

We used the search strategies developed with the help of Ms Yuhong (Cathy) Yuan, Information Specialist for Cochrane Gut, before performing literature searches. There was no restriction on the language of publication when searching electronic databases; we also reviewed reference lists in identified studies.

Electronic searches

We conducted a literature search from inception of each database, on 31 October 2021, to identify all published and unpublished RCTs in all languages. We searched the following electronic databases:

1. Cochrane Central Register of Controlled Trials (CENTRAL, via Ovid Evidence‐Based Medicine Reviews (EBMR) Database (from inception; Appendix 2)

2. MEDLINE (via Ovid from 1966; Appendix 3)

3. Embase (via Ovid from 1974 to 31 October 2021; Appendix 4)

We also searched two clinical trial registers by combing terms related to achalasia and individual interventions: ClinicalTrials.gov (www.clinicaltrials.gov); World Health Organization International Clinical Trials Registry Platform (ICTRP; www.who.int/trialsearch).

Searching other resources

We cross‐checked the reference list of all primary studies and review articles for additional references which we may have missed from the electronic searches.

Data collection and analysis

Selection of studies

Two review authors (SM and DG) independently screened the titles and abstracts for inclusion. We coded all the potential studies we identified as a result of the search as 'retrieve' (eligible, potentially eligible or unclear) or 'do not retrieve'. We retrieved the full text of all the studies coded as 'retrieve' and the two review authors (SM and DG) independently screened the full texts. We identified studies for inclusion and also recorded reasons for exclusion of the ineligible studies. We resolved any disagreement through discussion or, if required, we consulted our third review author (MWM). We identified and excluded duplicates studies. We also collated multiple reports of the same study, so that each study rather than each report was the unit of interest in the review. We recorded the selection process in sufficient detail to complete a PRISMA (Page 2021) flow diagram and characteristics of excluded studies table.

Data extraction and management

We used a standardised data collection form for study characteristics and outcome data. Two review authors (SM and DG) independently extracted the following study characteristics from the included studies.

1. Methods: study design, total duration study and run in, number of study centres and location, study setting, withdrawal, date of study

2. Participants: number, mean age, age range, gender, type of achalasia, American Anaesthesiologist Association (ASA) status (ASA 2014), inclusion criteria, exclusion criteria

3. Interventions: intervention, comparison, concomitant intervention

4. Outcomes: primary and secondary outcomes specified and collected, time points reported

5. Notes: funding for trial, notable conflicts of interest of trial authors

One review author (SM) copied across the data from the data collection form into Review Manager 5 (Review Manager 2020). A second review author (DG) checked study characteristics for accuracy against trial reports.

Assessment of risk of bias in included studies

Two review authors (SM and DG) independently assessed the risk of bias for each included study, using the criteria outlined in Chapter 8 of the Cochrane Handbook forSystematic Reviews of Interventions (Higgins 2017). We assessed the risk of bias according to the following domains.

1. Random sequence generation

2. Allocation concealment

3. Blinding of participants and personnel

4. Blinding of outcome assessment

5. Incomplete outcome data

6. Selective outcome reporting

7. Other bias: funding source, conflict of interest of authors

We graded each potential source of bias as high, low or unclear. We summarised the risk of bias judgements across studies for each of the domains listed in Characteristics of included studies and provided a risk of bias graph and risk of bias summary.

When considering treatment effects, we took into account the risk of bias for the studies that contribute to that outcome, as a part of the GRADE methodology (Chapter 14 of the Cochrane Handbook of Systematic Reviews of Interventions; Higgins 2022).

Measures of treatment effect

We analysed dichotomous data (postoperative pathological acid reflux, severe postoperative dysphagia, and major complications) as risk ratios (RRs) and continuous data (LOS pressure, health‐related quality of life scores, length of hospital stay, and operating time) as mean difference (MDs) with 95% confidence intervals (CIs).

We performed meta‐analyses only where this is meaningful, that is, if the treatments, participants and the underlying clinical question are similar enough for pooling to make sense.

In instances where we used imputed values of mean and standard deviation (SD), we mentioned this in the text and downgraded the level of evidence in the summary of findings tables. Where we could not impute these values, we did not perform a meta‐analysis; we provided a narrative summary instead.

Where multiple arms are reported in a single trial, we included only the relevant arms.

Unit of analysis issues

The unit of analysis is the individual participant undergoing LHC (with or without a fundoplication) for achalasia. We did not come across any cluster‐RCTs or cross‐over RCTs which were eligible. If we had encountered cluster‐ or cross‐over RCTs, we would have used the appropriate variant of risk of bias assessment tool and would have analysed the data according to the guidance in Chapter 23 of the Cochrane Handbook for SystematicReviews of Interventions (Higgins 2022). We included one study with three treatment arms, where one of the arms did not fulfil our inclusion criteria. We excluded this arm and included the other two arms in our review and meta‐analysis.

Dealing with missing data

We contacted the study authors to verify key study characteristics and obtain missing numerical outcome data as indicated (e.g. when a study is identified as abstract only). Where we were unable to obtain the information from the authors, we imputed the mean from the median (i.e. consider median as mean) and SD from standard error, interquartile range or P values, according to the Cochrane Handbook forSystematic Reviews of Interventions (Higgins 2022). If we were unable to calculate the SD from standard error, interquartile range or P values, we did not perform a meta‐analysis and downgraded the evidence.

Assessment of heterogeneity

We used the I² statistic to measure heterogeneity among the trials in each analysis (Higgins 2003). Where we identified substantial or considerable heterogeneity (> 50%) as per the Cochrane Handbook forSystematic Reviews of Interventions (Higgins 2022), we mentioned this in the results and downgraded the evidence.

Assessment of reporting biases

As we included < 10 trials, we did not assess reporting bias in this review.

Data synthesis

We performed the meta‐analysis using Review Manager 5 (Review Manager 2020). We used the random‐effects model by default and presented only results from the random‐effects model. We grouped the studies according to the fundoplications they compare (e.g. Nissen versus Dor, Dor versus Toupet, etc.) and performed a separate meta‐analysis for each group. We also grouped all the RCTs comparing Dor fundoplication against no fundoplication and analysed the results.

Subgroup analysis and investigation of heterogeneity

We could not carry out subgroup analyses comparing age groups, gender, ASA groups and severity of achalasia as planned in the protocol, because the information required for these analyses was not available in the study reports. We used the Chi² test to examine heterogeneity between the studies.

Sensitivity analysis

We performed sensitivity analysis by excluding trials with imputed mean and SD values and this did not change our conclusions.

Summary of findings and assessment of the certainty of the evidence

We created summary of findings tables for the primary outcomes, postoperative pathological acid reflux and severe postoperative dysphagia in each group. We used the five GRADE assessment domains (risk of bias, inconsistency, imprecision, indirectness, and publication bias) to assess the certainty of the evidence for each outcome, classifying it as high, moderate, low or very low certainty. We used the methods and recommendations described in Chapter 14 of the Cochrane Handbook of Systematic Reviews of Interventions (Higgins 2022), and GRADEpro GDT software (GRADEpro GDT). We justified all decisions to downgrade or upgrade the certainty of the included studies in the footnotes, and where necessary, provided comments to aid the reader's understanding of the review. We considered whether there is additional outcome information that was not incorporated into the meta‐analyses, noted this in the comments, and stated if it supports or contradicts the information from the meta‐analyses.

Results

Description of studies

We only included RCTs where LHC was performed on adult participants with achalasia. The arms of the RCTs differed only by the presence or absence of fundoplication, or the type of fundoplication or anti‐reflux procedure used. Hence, any difference in outcomes could be attributed to the presence or absence of fundoplication, or the type of fundoplication (anti‐reflux procedure) used.

Results of the search

We identified 1122 studies by searching the electronic databases ‐ Cochrane Central Register of Controlled Trials (N = 263), MEDLINE (N = 481) and Embase (N = 378), on 31 October 2021. Of these, we excluded 1093 studies after reading the abstract as they did not comply with our inclusion criteria or were duplicate studies. We assessed 29 full‐text articles and found eight RCTs which we included in this review. The remaining studies were either non‐randomised or duplicates studies. The PRISMA flow diagram of the selection process is shown in Figure 1.

1.

PRISMA flow diagram

Included studies

We included eight studies in this review (Grubnik 2017; Gupta 2018; Kumagai 2014; Rawlings 2012; Rebecchi 2008; Richards 2004; Simic 2010; Torres‐Villalobos 2018). Two studies were multicentric RCTs (Kumagai 2014; Rawlings 2012); the remaining studies were conducted in a single centre.

All procedures were done laparoscopically except for one patient in Kumagai 2014, who underwent open Heller's procedure. It is not clear if this was a conversion of a laparoscopic procedure. The interventions compared in these studies are as follows.

• LHC + Dor versus LHC + Toupet (Kumagai 2014; Rawlings 2012; Torres‐Villalobos 2018)

• LHC + Dor versus LHC without fundoplication (Grubnik 2017; Richards 2004)

• LHC + Dor versus LHC + Nissen (Rebecchi 2008)

• LHC + Dor (with complete hiatal dissection) versus LHC + Dor (limited hiatal dissection) versus LHC alone (Simic 2010)

• LHC + Dor versus LHC + angle of His accentuation (Gupta 2018)

Simic 2010 involved three arms comparing LHC + Dor with complete hiatal dissection, LHC + Dor with limited hiatal dissection, and LHC alone. We excluded the arm where LHC + Dor is performed with complete hiatal dissection as it does not satisfy our inclusion criteria. We have explained in the section describing interventions that disrupting the posterior oesophageal attachments and mobilising the oesophagus circumferentially leads to increased acid reflux postoperatively (Bonavina 2006; Braghetto 2007). There is no advantage in disrupting the posterior oesophageal attachments while performing a LHC unless there is an associated hiatus hernia or where a Toupet/Nissen fundoplication is to be added.

This review includes a total of 571 participants from these eight studies with an average age of 45 years (range 33.5 to 50). All participants, except one in Kumagai 2014, had LHC with minimal hiatal dissection. LHC without any fundoplication was performed in 65 (11.3%) participants, 298 (52.1%) had Dor fundoplication, 81 (14.1%) had Toupet fundoplication, 72 (12.6%) had Nissen's fundoplication, and 55 (9.6%) participants had angle of His accentuation.

The details of these studies and their outcomes are summarised in the Characteristics of included studies section.

Excluded studies

Falkenback 2003 is a RCT conducted in Sweden to compare the outcomes of participants with achalasia who underwent Heller's cardiomyotomy with or without Nissen fundoplication. All procedures in this study were done by open surgery, hence it did not fulfil our inclusion criteria.

Torquati 2006 used the same group of patients as Richards 2004, to conduct a cost‐utility analysis using a mathematical simulation model and estimate the expected cost of each intervention over a 10‐year period. Hence, it is not a RCT.

Broman 2018 is a study where the authors tried to contact the participants included in the Richards 2004 RCT by mail and telephone to collect patient reported long‐term outcomes of the intervention. They only managed to contact 27 of the original 43 (62.7%) participants. This study was not a part of the original protocol of the RCT, and the methodology of collecting the data is different. Hence, we excluded the study as it is not a true RCT.

Risk of bias in included studies

See Figure 2 and Figure 3 for risk of bias summaries.

2.

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

3.

Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

All included studies, except Grubnik 2017, have been published as full‐text articles. We contacted all corresponding authors for any missing data, but unfortunately we only received a response from the authors of Gupta 2018. Grubnik 2017 was presented at SAGES 2017 conference and was published as an abstract, hence the available information is limited with this study.

Allocation

The participants in all eight studies were selected randomly. Grubnik 2017 has only been published as an abstract, and we do not have any information to comment on randomisation and allocation concealment. The remaining seven studies have described robust methods of randomisation, sequence generation and allocation concealment.

Blinding

In Richards 2004, all recruited participants and researchers involved in the evaluation of the end points were blinded to the treatment assigned throughout the whole period of study. In Kumagai 2014, each participant was allocated a numeric code on recruitment. The participants, care providers and assessors were blinded to the type of surgery performed. Only the principal investigator could access participants though the codes. Hence, in these two studies, the risk of bias due to blinding should be low. In Gupta 2018, the participants were blinded, but the healthcare providers were not. This information was obtained by contacting the authors directly. We could not obtain any information on blinding in the rest of the studies.

Incomplete outcome data

Four studies included all their randomised participants in their analysis (Gupta 2018; Kumagai 2014; Rebecchi 2008; Richards 2004), hence the risk of attrition bias is low in these studies. Eighteen participants in Simic 2010 and 25 participants in Rawlings 2012 were not included in the analysis as they did not attend follow‐up. There were also many participants in Torres‐Villalobos 2018 who did not attend follow‐up at two years, and it is not clear how this missing data were handled. In these three studies, the attrition bias is expected to be high.

Selective reporting

All studies reported the two most important outcomes of the intervention: postoperative reflux and recurrent dysphagia. They also reported mortality and other complications. We assessed the risk of reporting bias in the included studies as low.

Other potential sources of bias

The Rawlings 2012 study received financial support from SAGES and Torres‐Villalobos 2018 received some grants from Endostim, but it is unlikely to have led to any bias. The other studies have either not mentioned or denied having received any financial support. We did not find any other potential sources of bias in the eight studies included.

Effects of interventions

See: Table 1; Table 2; Table 3; Table 4

We grouped the studies according to the interventions they compared and discussed the outcomes in each of these groups.

LHC + Dor versus LHC without fundoplication

Three studies compared LHC + Dor to LHC without fundoplication (Grubnik 2017; Richards 2004; Simic 2010). As explained, we excluded one arm of Simic 2010 as it did not meet our inclusion criteria. Grubnik 2017 has only been published as an abstract and the available data are limited. In total, we included 143 participants from these three studies, of which 78 underwent LHC + Dor and 65 LHC only.

Postoperative pathological acid reflux

Pathological acid reflux is an objective measure of acid reflux and is calculated from the results of a 24‐hour pH study. It was reported in Richards 2004 and Simic 2010 but not in Grubnik 2017. Richards 2004 defined pathological acid reflux as when the total time pH < 4 was > 4.2% in the 24‐hour study, while Simic 2010 defined it as DeMeester score of > 14.7. The 24‐hour pH study was done three months to five months postoperatively in Richards 2004. Pathological acid reflux was found to be higher in participants who underwent LHC without fundoplication (47.6%) than in those who underwent LHC + Dor (9.1%). In Simic 2010, the 24‐hour pH test was done at one, two and three years postsurgery. After three years, this study did not find any statistical difference in pathological acid reflux between LHC without fundoplication (9.1%) and LHC + Dor (8.5%). In Grubnik 2017, participants underwent 24‐hour pH study 2 years postsurgery. Though the study did not report the number of participants who had pathological acid reflux, they mentioned that there is no difference in the prevalence of gastro‐oesophageal reflux in the two groups.

We performed a meta‐analysis using data from Richards 2004 and Simic 2010 and found that the evidence was very uncertain (RR 0.37, 95% CI 0.07 to 1.89; I²= 56%, Z= 1.20, P= 0.23; 2 studies, 97 participants; very low‐certainty evidence; Analysis 1.1); it is not possible for us to comment as to whether or not there is any difference in postoperative pathological acid reflux between the two interventions.

1.1. Analysis.

Comparison 1: LHC + Dor versus LHC without fundoplication, Outcome 1: Postoperative pathological acid reflux

Severe postoperative dysphagia

Postoperative dysphagia was reported in all three studies which compared these two interventions. Richards 2004 used a scoring system (scale of 0 to 10) by combining the frequency (score of 0 to 5) and severity (scale of 0 to 5) of dysphagia. Participants who underwent LHC without fundoplication had a median decrease in dysphagia score of 9 points; participants who underwent LHC + Dor had a decrease of 8 points in dysphagia score. One participant randomised to the LHC + Dor group required redo Heller myotomy after 18 months. Simic 2010 used a scoring system (Grade 0 to 3) to record postoperative dysphagia. There was no statistical difference in the incidence of severe postoperative dysphagia between the two interventions, but two participants in the LHC + Dor arm required treatment with pneumatic dilatation. Grubnik 2017 reported dysphagia symptomatically by modified Mellow and Pinkas dysphagia scale and objectively by barium swallow. Complete relief of dysphagia, confirmed by barium swallow, was observed in 85% of participants from the LHC + Dor group, and in 95.5% of participants from the LHC without fundoplication group. There were no statistical differences in dysphagia score (0.8 versus 0.3). The follow‐up period in this study was 2 years, and has not reported the need for any further interventions for dysphagia in these patients.

We performed a meta‐analysis using the data from these three studies. We found that the evidence was uncertain (RR 3.00, 95% CI 0.34 to 26.33; I²= 0%, Z = 0.99, P = 0.32; 3 studies, 143 participants; low‐certainty evidence; Analysis 1.2); it is not possible to conclude if there was a difference in severe postoperative dysphagia between the two interventions.

1.2. Analysis.

Comparison 1: LHC + Dor versus LHC without fundoplication, Outcome 2: Severe postoperative dysphagia

Health‐related quality of life scores

Health‐related quality of life scores were reported in only one of the three studies in this group (Grubnik 2017). This study used the GERD‐HRQL score to assess quality of life following surgery and found no statistical difference between LHC without fundoplication (2.4) and LHC + Dor (1.9).

Major complications

Richards 2004 and Grubnik 2017 reported no major complication or mortality in their study. Simic 2010 reported two participants with mucosal perforation, three required splenectomy and two developed postoperative pneumonia out of a total of 84 participants in the three arms of their study. They did not mention complications in each of the individual arms, and hence we cannot use the information to compare the two interventions.

LOS pressure

In Richards 2004, LOS pressure dropped from their high preoperative levels after both interventions without any difference between them. In the LHC without fundoplication group, it dropped from a mean preoperative LOS pressure of 35 mm Hg (range 14 to 120) to a median postoperative LOS pressure of 13.7 mm Hg (range 6.4 to 25.9), while in the LHC + Dor group, it fell from mean preoperative LOS pressure of 33 mm Hg (range 20 to 120) to median postoperative LOS pressure of 13.9 mm Hg (range 5.8 to 26.8). Postoperative LOS pressures were measured at a median of six months (range three months to eight months) after surgery.

In Simic 2010, mean preoperative LOS pressure dropped from 34.2 +/‐ 5.1 mm Hg to 8.9 mm Hg in the LHC without fundoplication group and from 38.7 +/‐ 5.3 mm Hg to 9.6 mm Hg in the LHC + Dor group after three years. There was no difference in improvement in LOS pressure between the two interventions.

Grubnik 2017 did not report LOS pressure.

We did not perform a meta‐analysis of LOS pressure, as Richards 2004 presented preoperative data as mean values and postoperative data as median values. As the sample size was small and the outcome results were skewed, meta‐analysis would have given a misleading result.

DeMeester score

The DeMeester score was reported in the three studies. Richards 2004 reported that the median DeMeester score in the LHC without fundoplication group (31.5) was higher than that in the LHC + Dor group (6.5). Simic 2010 did not find any difference in postoperative DeMeester scores between the interventions after three years of follow‐up. They reported a mean DeMeester score of 12.6 following LHC without fundoplication and 13.5 after LHC + Dor. Grubnik 2017 found the mean DeMeester score was higher following LHC without fundoplication (42.1) compared to LHC + Dor (23.1), but the values did not reach statistical significance.

We did not perform a meta‐analysis as Richards 2004 presented median values of the data, while Simic 2010 presented mean values of their data. The data available from Grubnik 2017 were not sufficient to be used for meta‐analysis.

Length of hospital stay

Richards 2004 reported a median length of hospital stay of one day (range one to two days) for both interventions. Simic 2010 did not report length of hospital stay data for each individual intervention, but the mean length of hospital stay for the whole study was 5.2 days. Grubnik 2017 did not report any data regarding length of hospital stay in their study. It was not possible to compare the two procedures based on the data reported in these studies regarding this outcome.

Operating time

Simic 2010 did not report operating time for these two interventions. Both Richards 2004 and Grubnik 2017 reported shorter operating times with LHC without fundoplication. Richards 2004 reported a median operating time of 85 minutes (range 54 to 180) for LHC without fundoplication and 125 minutes (range 83 to 415) for LHC + Dor. Grubnik 2017 reported a mean operating time of 92 minutes with LHC without fundoplication and 130 minutes with LHC + Dor. We did not perform a meta‐analysis for this outcome as we did not have the required data.

LHC + Dor versus LHC + Toupet

Three studies compared these two interventions (Kumagai 2014; Rawlings 2012; Torres‐Villalobos 2018). These studies have a total of 174 participants (92 females and 82 males), with a mean age of 44.4 years. LHC + Dor was performed in 93 participants and 81 had LHC + Toupet. The mean follow‐up period in Rawlings 2012 and Kumagai 2014 was 12 months while that in Torres‐Villalobos 2018 was 24 months.

Postoperative pathological acid reflux

All three studies reported postoperative pathological acid reflux. Rawlings 2012 defined pathological acid reflux as DeMeester score > 14.73 or when the total time pH < 4 was > 4% in the 24‐hour study. Kumagai 2014 and Torres‐Villalobos 2018 defined pathological acid reflux when the total time pH < 4 was > 4% in the 24‐hour study; it did not consider the DeMeester score. Rawlings 2012 reported that in the Dor group, 10 of the 24 participants (41.7%) had pathological acid reflux compared to 4 of the 19 participants (21.1%) in the Toupet group. Kumagai 2014 reported that 2 of the 11 participants in the Dor group and 5 of the 13 participants in the Toupet group developed postoperative pathological acid reflux. Torres‐Villalobos 2018 reported that at six months, the number of participants with pathological acid reflux was lower in the Dor group (6.9%) than the Toupet group (34%), but no difference was found at 12 months (Dor 4.43% versus Toupet 9.5%) and at 24 months (Dor 10.5% versus Toupet 31.5%).

We performed a meta‐analysis of pathological acid reflux using data from these three studies and found that there was no evidence of a difference in the incidence of postoperative pathological acid reflux between these two procedures (RR 0.75, 95% CI 0.23 to 2.43; I² = 60%, Z = 0.47, P = 0.64; 3 studies, 105 participants; low‐certainty evidence; Analysis 2.1), but the level of certainty of the evidence is low.

2.1. Analysis.

Comparison 2: LHC + Dor versus LHC + Toupet, Outcome 1: Postoperative pathological acid reflux

Severe postoperative dysphagia

Rawlings 2012 used a 5‐point Likert scale (0 to 4) to assess dysphagia, and found improvements in postoperative symptom scores for solid and liquid dysphagia. There was no statistical difference between the two interventions in terms of postoperative symptom severity or in the degree of improvement over preoperative symptoms. Three participants (11.5%) in the Dor group and one participant (6.7%) in the Toupet group required endoscopic dilatation postoperatively for symptomatic dysphagia.

Kumagai 2014 used the Eckardt score to assess symptomatic improvement in response to the interventions. They found an improvement with both interventions without any statistical difference between them. According to the Eckardt score, 1 out of 11 (9.1%) participants in the Dor group and 2 out of 14 (14.3%) participants in the Toupet group experienced recurrence or treatment failure 12 months after the procedures. They also used timed barium oesophagogram, which showed improvement in oesophageal emptying with Toupet fundoplication but not with a Dor fundoplication. However, there was no difference in the magnitude of improvement between the two groups.

Torres‐Villalobos 2018 reported Eckardt scores of < 3 in 100% of the Dor group and 90% of the Toupet participants after 24 months of follow‐up. There was no difference between the two interventions and none of the participants required any re‐intervention or other treatments.

We performed a meta‐analysis of severe postoperative dysphagia using data from these three studies and found no difference in severe postoperative dysphagia leading to treatment failure or further interventions after LHC + Dor (RR 0.78, 95% CI 0.19 to 3.15; I² = 0%, Z = 0.35, P = 0.73; 3 studies, 123 participants; low‐certainty evidence; Analysis 2.2), but again the certainty of the evidence is low.

2.2. Analysis.

Comparison 2: LHC + Dor versus LHC + Toupet, Outcome 2: Severe postoperative dysphagia

Health‐related quality of life scores

Rawlings 2012 used the SF‐36 questionnaire (Ware 1992) to compare the interventions. They found improvements in 5 out of the 10 health‐related domains in the Dor group and 7 out of the 10 domains in the Toupet group without any differences between them after 12 months of follow‐up. Kumagai 2014 used EORTC QLQ‐OES 18 (Blazeby 2003) and GERD‐HRQL (Velanivich 1996) questionnaires and found improvement in quality of life following both interventions, without any difference between them. Torres‐Villalobos 2018 also used the GERD‐HRQL questionnaire and found improvement in symptoms following both interventions, without any obvious difference between them.

Major complications

Rawlings 2012 reported two perforations (gastric/oesophageal) in each group which were repaired primarily. Kumagai 2014 reported no complication in the Dor group, but one patient in the Toupet group developed a leak from a perforation in the myotomy which was treated with a covered stent. Torres‐Villalobos 2018 reported three complications in their study. In the Dor group, one patient had gastric perforation which was repaired primarily and another patient required re‐intervention for bleeding from the trocar site. One patient in the Toupet group developed portal thrombosis 30 days after the surgery. None of the studies found any difference in the occurrence of major complications between the two interventions.

We performed a meta‐analysis of major complications using the data from the three studies; we did not find any statistical difference between the two procedures (RR 0.84, 95% CI 0.22 to 3.18; I² = 0%, Z = 0.26, P = 0.79; 3 studies, 174 participants; low‐certainty evidence; Analysis 2.3), but the level of certainty of the evidence is low.

2.3. Analysis.

Comparison 2: LHC + Dor versus LHC + Toupet, Outcome 3: Major complications

LOS pressure

Kumagai 2014 did not report postoperative LOS pressure. Rawlings 2012 reported that both the basal and residual LOS pressures decreased in both groups, and there was no difference between the two groups. Torres‐Villalobos 2018 used high‐resolution manometry to compare pre‐ and postoperative LOS pressures between the two groups. They reported no difference between the basal and integrated relaxation pressures of LOS between the Dor and Toupet groups.

We calculated a meta‐analysis of postoperative basal LOS pressure following LHC + Dor and LHC + Toupet using the data from the two studies and found that there is probably no difference between the two interventions (MD ‐2.11, 95% CI ‐5.04 to 0.81; I² = 17%, Z= 1.42, P= 0.16; 2 studies, 93 participants; moderate‐certainty evidence; Analysis 2.4).

2.4. Analysis.

Comparison 2: LHC + Dor versus LHC + Toupet, Outcome 4: LOS Pressures

DeMeester score

Kumagai 2014 did not report the DeMeester score of participants in their study. Rawlings 2012 reported that the median postoperative DeMeester score in the Dor group was 5.6 (range 0 to 131) and in the Toupet group was 3 (range 0 to 107). There was no statistical difference in the DeMeester scores between the two interventions. Torres‐Villalobos 2018 also did not find any difference in postoperative DeMeester scores between the two interventions at 24 months of follow‐up. We could not perform a meta‐analysis for this outcome as we did not have sufficient data.

Length of hospital stay

Kumagai 2014 reported a median hospital stay of two days in both groups. Torres‐Villalobos 2018 reported a mean length of stay of 2.54 (+/‐ 2.56) days in the Dor group and 2.54 (+/‐ 2.76) days in the Toupet group. Rawlings 2012 did not report the length of hospital stay in their study.

Operating time

All three studies reported operating times for the procedures and did not find any difference in the length of the two procedures. Rawlings 2012 reported a mean operating time of 158.9 +/‐ 45.5 minutes for LHC + Dor and 160.1 +/‐ 39.6 minutes for LHC + Toupet. Kumagai 2014 reported a median operating time of 103 (range 83 to 151) minutes for LHC + Dor and 95 (range 83 to 135) minutes for LHC + Toupet. Torres‐Villalobos 2018 reported a mean operating time of 114.9 +/‐ 24.6 minutes for LHC + Dor and 124.77 +/‐ 25.01 minutes for LHC + Toupet.

We calculated a meta‐analysis of the operating times for the two procedures using data from Rawlings 2012 and Torres‐Villalobos 2018. We found that there is probably no difference between the two procedures (MD ‐7.94 minutes, 95% CI ‐18.03 to 2.15; I² = 0%, Z = 1.54, P = 0.12; 2 studies, 133 participants; moderate‐certainty evidence; Analysis 2.5), but the large 95% CI makes the evidence uncertain.

2.5. Analysis.

Comparison 2: LHC + Dor versus LHC + Toupet, Outcome 5: Operating time

LHC + Dor versus LHC + Nissen

Only one of the included studies, Rebecchi 2008, compared these two interventions. The study included 144 participants; half (72 participants) underwent LHC + Dor and the other half, LHC + Nissen. The mean age of the participants was 49 years and there were 69 female participants in the study. The mean follow‐up period was 60 months.

Postoperative pathological acid reflux

This study defined pathological acid reflux as DeMeester score > 18 for 24 consecutive hours and area under H+ (AUH+) parameter > 114.1 mmoL/L (Rebecchi 2002). Postoperative pathological acid reflux was found in two participants (2.8%) in the LHC + Dor group and none in the LHC + Nissen group. We analysed their data to find any difference in pathological acid reflux between the two interventions (RR 4.72, 95% CI 0.23 to 96.59; Z = 1.81, P = 0.31; 1 study, 138 participants; low‐certainty evidence; Analysis 3.1), but given the wide 95% CI and low number of events, certainty in the evidence is low.

3.1. Analysis.

Comparison 3: LHC + Dor versus LHC + Nissen, Outcome 1: Postoperative pathological acid reflux

Severe postoperative dysphagia

Rebecchi 2008 used Modified DeMeester Scoring System to assess clinical dysphagia. Postoperative clinical dysphagia was higher in the LHC + Nissen group (10 participants) than in the LHC + Dor group (2 participants). In the LHC + Nissen group, four participants required endoscopic pneumatic dilatation for dysphagia whereas none of the participants in the LHC + Dor required any further intervention. We analysed their data, which showed increased dysphagia after LHC + Nissen compared to LHC + Dor (RR 0.19, 95% CI 0.04 to 0.83, Z = 2.21, P = 0.03; 1 study, 138 participants; high‐certainty evidence; Analysis 3.2).

3.2. Analysis.

Comparison 3: LHC + Dor versus LHC + Nissen, Outcome 2: Severe postoperative dysphagia

Health‐related quality of life score

Rebecchi 2008 did not report quality of life scores in the study.

Major complications

There were no major postoperative complications in this study.

LOS pressure

The mean LOS pressure improved following both procedures compared to their preoperative values. The postoperative LOS pressure was considerably more in the LHC + Nissen group (22.6 mm Hg) than in the LHC + Dor group (12.8 mm Hg). The study did not report these results in sufficient detail for us to carry out further analysis.

DeMeester score

The mean postoperative DeMeester score after 60 months of follow‐up is reported as 16 (range 0 to 34) for the LHC + Dor group and 15 (range 0 to 18) for the LHC + Nissen group.

Length of hospital stay

The mean hospital stay in the Dor group was 3.2 days (range 2 to 6) and in the Nissen group was 3.6 days (range 2 to 5).

Operating time

The mean operating time for LHC + Dor was 100 minutes (range 86 to 132) and for LHC + Nissen was 106 minutes (range 90 to 135).

LHC + Dor versus LHC + angle of His accentuation

Gupta 2018 compared these two interventions in their study with 110 participants. LHC + Dor was performed on 54 participants and LHC + angle of His accentuation was performed on 56 participants. Mean follow‐up for the whole group was 36 months (range 12 to 72). The study included 55 female participants with a mean age of 33.5 years.

Postoperative pathological acid reflux

Participants in this study did not undergo postoperative pH studies and so this outcome was not reported. There was no difference in symptomatic acid reflux between the two interventions.

Severe postoperative dysphagia

Clinical assessment of postoperative dysphagia was performed using a modified Mellow‐Pinkas score (Mellow 1985). Clinically‐significant dysphagia (dysphagia score > 2) at the last follow‐up was present in three (5%) participants in the LHC + angle of His accentuation group and two (4%) participants in the LHC + Dor group. One participant in the LHC + Dor group required re‐intervention in the form of endoscopic dilatation and none in the LHC + angle of His accentuation group needed any further intervention. We analysed their reported data and found that there is probably no difference in the incidence of severe postoperative dysphagia following the two interventions (RR 1.56, 95% CI 0.27 to 8.95, Z = 0.49, P = 0.62; 1 study, 110 participants; moderate‐certainty evidence; Analysis 4.1), but given the wide 95% CI, the evidence is of moderate certainty.

4.1. Analysis.

Comparison 4: LHC + Dor versus LHC + angle of His accentuation, Outcome 1: Severe postoperative dysphagia

Health‐related quality of life scores

This study used an achalasia‐specific quality of life (ASQOL) questionnaire (Urbach 2005) to assess quality of life pre‐ and postoperatively. ASQOL showed improvement in all aspects following surgery in both groups. The mean ASQOL score improved from 58.2 (+/‐ 7.3) to 14.1 (+/‐ 14.5) in the LHC + angle of His accentuation group and from 58.6 (+/‐ 8.7) to 13.3 (+/‐ 16.4) in the LHC + Dor group. We used their data to analyse health‐related quality of life scores and found that there is probably no difference between the two interventions (MD ‐0.80, 95% CI ‐6.59 to 4.99, Z = 0.27, P = 0.79; 1 study, 110 participants; moderate‐certainty evidence; Analysis 4.2), but the evidence is uncertain.

4.2. Analysis.

Comparison 4: LHC + Dor versus LHC + angle of His accentuation, Outcome 2: Quality of life (ASQOL) score

Major complications

No major complication (Clavien‐Dindo Grade III and above) was reported in this study.

LOS pressure

This outcome was not reported in this study as the patient did not undergo postoperative manometry.

DeMeester score

A postoperative DeMeester score was not reported in this study as the participants did not undergo postoperative pH studies.

Length of hospital stay

Median length of hospital stay was 3 days in both groups with a range of 2 days to 8 days in the LHC + angle of His accentuation group and 2 days to 12 days in the LHC + Dor group.

Operating time

Median operating time was shorter in the LHC + angle of His accentuation group (128 minutes, range: 110 to 180) than the LHC + Dor group (144 minutes, range: 100 to 165).

Discussion

Summary of main results

In this systematic review, we compared the outcomes of LHC in patients with achalasia with or without fundoplication. We also compared the advantages and disadvantages of various types of fundoplication when used in association with LHC. We included eight RCTs in this review with a total of 571 participants who underwent LHC. Of these, 298 participants had Dor fundoplication, 81 had Toupet fundoplication, 72 had Nissen fundoplication, 55 had angle of His accentuation and 65 participants had no fundoplication. We grouped these studies according to the interventions they compared: LHC + Dor versus LHC without fundoplication (3 RCTs); LHC + Dor versus LHC + Toupet (3 RCTs); LHC + Dor versus LHC + Nissen (1 RCT); and LHC + Dor versus LHC + angle of His accentuation (1 RCT).

When comparing LHC + Dor and LHC without fundoplication, we performed meta‐analysis to study the difference in the outcomes of postoperative acid reflux, severe postoperative dysphagia, LOS pressures and DeMeester scores. Based on the results of the meta‐analysis, it is very uncertain whether or not there was any difference in postoperative pathological acid reflux between the two interventions. It is also uncertain if there was any difference in postoperative severe dysphagia, between the two procedures. The studies did not report any difference in health‐related quality of life scores, major postoperative complications and length of hospital stay. All studies reported that the operating time was shorter for LHC without fundoplication than for LHC + Dor.

We performed meta‐analysis on the outcomes of postoperative acid reflux, severe postoperative dysphagia, major complications, LOS pressures and Demeester score when comparing the two partial fundoplications Dor and Toupet in association with LHC. We did not find any evidence suggesting a difference in the outcomes between the two interventions, however the evidence was uncertain.

In the one study comparing LHC + Dor to LHC + Nissen (Rebecchi 2008), we found high‐certainty evidence that LHC + Nissen caused an increase in severe postoperative dysphagia. The study did not report any difference in postoperative acid reflux, but the certainty of evidence was low. The study also showed that postoperative LOS pressure was considerably higher after LHC + Nissen, but DeMeester scores, major complications, length of hospital stay and operating time were similar following the two procedures.

Gupta 2018 compared LHC + Dor to LHC + angle of His accentuation and did not report any symptomatic difference in acid reflux, dysphagia, health‐related quality of life scores, major complications or length of hospital stay. We performed meta‐analysis using their data on severe postoperative dysphagia and health‐related quality of life scores. We found that the outcomes were probably similar following the procedures, but the evidence was not certain. This study did not report postoperative pathological acid reflux following the procedures, as the participants were not followed up with 24‐hour pH studies.

Overall completeness and applicability of evidence

The participants of the studies included in this review were adults who underwent laparoscopic Heller's cardiomyotomy for achalasia. Patients who had undergone unsuccessful non‐surgical treatments like botulinum toxin injections or endoscopic dilatation previously were included in these studies. The findings of this review are not applicable to the paediatric age group, patients with recurrent achalasia following previous surgery, those undergoing open surgery or where LHC was performed with circumferential hiatal dissection. Two studies in this review excluded participants with sigmoid shaped oesophagus (Gupta 2018; Rebecchi 2008). One study included in this review excluded participants with previous oesophagogastric surgeries (Rawlings 2012). The study findings also do not apply to patients with known gastro‐oesophageal cancers, oesophageal strictures, large hiatus hernias (> 5 cm) or other causes of oesophageal dysmotility, like Chagas disease or neuromuscular disorders.

Certainty of the evidence

All the studies included in this review were RCTs. Grubnik 2017 has only been published as an abstract, and we were not successful in our attempts to contact the authors. We have very little information from this study and did not use any of its data for analysis. Apart from Grubnik 2017, the remaining seven studies described robust selection and randomisation methods with low risk of allocation bias. The information on blinding of participants and assessors has not been provided in many of these studies. The follow‐up period varied from six months in Richards 2004 to 125 months in Rebecchi 2008 and there was heterogeneity in the methods of assessing participants during the follow‐up. Three of the studies had high numbers of dropouts (Rawlings 2012; Simic 2010; Torres‐Villalobos 2018), which is likely to have caused increased attrition bias. We did not find any evidence of reporting bias or other potential bias in these studies.

With achalasia being a rare disease, it is not surprising that the studies had small numbers of participants with a low number of events and wide confidence intervals. We downgraded the certainty of evidence by one level in all our assessments due to serious risk of imprecision. Some of the included studies had different parameters of defining pathological acid reflux and used different clinical scoring systems for dysphagia. When we encountered this while comparing different studies, we reduced the certainty of evidence by one level to reduce risk due to indirectness of evidence. Similarly, we also reduced the certainty of evidence by one level where analysis of studies showed substantial heterogeneity (50% to 90%) leading to inconsistency of results. We did not find any evidence of publication bias in the included studies.

Potential biases in the review process

Our two primary outcomes were severe dysphagia and pathological acid reflux in patients who had undergone LHC. The studies reported these two outcomes with different clinical scoring systems, making it difficult for us to compare. We used the outcome postoperative pathological acid reflux, which is obtained from the results of pH studies to assess gastro‐oesophageal acid reflux. But the studies included here used different parameters to define pathological acid reflux, which is likely to have caused some bias. We considered dysphagia where the symptoms were severe (Eckardt score of > 3), or where medical or surgical intervention was required to relieve symptoms.

There is substantial statistical heterogeneity between studies comparing LHC + Dor and LHC without fundoplication when assessing reflux and DeMeester scores (I² = 56%). We also observed considerable heterogeneity (I² = 66%) when assessing reflux in the studies comparing LHC + Dor and LHC + Toupet. Heterogeneity did not change when we changed the effect from risk ratio (RR) to risk difference (RD) or changed the model from fixed‐effect to random‐effects.

The 95% CI was quite large in most of the analysis, suggesting imprecision and uncertainty regarding the certainty of the evidence.

We could not exclude publication bias, as the number of included studies was low (< 10).

As we restricted our inclusion criteria to RCTs, only eight studies were eligible for our review. If we had widened our criteria to include non‐randomised studies we could have increased the number of studies and participants, but it would have introduced a lot more bias, and the certainty of evidence would have suffered. We also decided to exclude open surgical procedures as they are hardly performed these days, and the results would not be relevant for our current practice.

Agreements and disagreements with other studies or reviews

Achalasia is a rare disease and there are very few RCTs comparing its surgical management in the literature. The available RCTs have small sample sizes and short follow‐up periods. There are quite a few systematic reviews in the literature, most of which have looked into a combination of non‐randomised and randomised studies to come to a conclusion.

In the studies comparing LHC + Dor to LHC without fundoplication, Richards 2004 showed the addition of a Dor fundoplication reduced the incidence of postoperative pathological reflux without increasing the risk of dysphagia. The follow‐up period of this study was only six months. The other two RCTs comparing these two interventions had longer follow‐up times (two to three years) and did not report any difference in postoperative reflux and dysphagia (Grubnik 2017; Simic 2010). Broman 2018 managed to contact 66% (21/41) of the participants included in the Richards 2004 RCT, and found that the long‐term (11.8 years) patient‐reported outcomes after LHC + Dor and LHC without fundoplication were comparable. The Lyass 2003 systematic review reported that postoperative reflux after LHC appeared to increase over time and is not necessarily eliminated with the addition of a fundoplication. Based on their data, the authors could not find any evidence that adding a fundoplication to LHC, improved the postoperative reflux symptoms. Wei 2013, in another systematic review, found there was no difference between LHC + Dor and LHC without fundoplication. On the other hand, two systematic reviews found the addition of a fundoplication to LHC reduced postoperative reflux symptoms (Campos 2009; Mayo 2012). The SAGES guidelines strongly recommend the addition of a fundoplication with LHC to prevent postoperative reflux and minimise treatment failure (Stefanidis 2012). It also suggests that partial fundoplication (Dor or Toupet) should be favoured over total (Nissen) fundoplication due to reduced dysphagia and similar reflux control.

In the three RCTs comparing LHC + Dor to LHC + Toupet (Kumagai 2014; Rawlings 2012; Torres‐Villalobos 2018), none of the studies reported any difference in the control of postoperative pathological acid reflux. All three studies reported no difference in postoperative dysphagia between the procedures, but Kumagai 2014 reported improved oesophageal emptying after LHC + Toupet. Aiolfi 2020, a systematic review using data from these three RCTs, concluded that LHC + Dor and LHC + Toupet were comparable to each other in terms of postoperative pathological acid reflux and dysphagia. In another systematic review, Kurian 2013 reported that LHC + Toupet was associated with lower rates of re‐intervention for postoperative dysphagia than LHC + Dor, while providing similar benefits in controlling postoperative reflux. Siddaiah‐Subramanya 2019, another systematic review and meta‐analysis, showed that LHC + Toupet was superior to LHC + Dor in terms of length of hospital stay and quality of life scores, but there was no difference in terms of postoperative reflux and dysphagia.

Falkenback 2003, in an RCT comparing Nissen fundoplication to no fundoplication in association with open Heller's procedure, reported that Heller's procedure induces pathological acid reflux, which is eliminated by adding a floppy Nissen fundoplication without increasing the risk of dysphagia. Cuttitta 2011 (a non‐randomised study), reported higher incidence of dysphagia with LHC + Nissen when compared with LHC + Dor, without any advantage in control of reflux. Di Martino 2011 (another non‐randomised study) reported that LHC + Nissen calibrated intraoperatively with endoscopy and manometry was more effective in controlling acid reflux without increasing the risk of postoperative dysphagia. The SAGES guidelines favour partial fundoplication over total fundoplication due to reduced dysphagia and similar reflux control (Stefanidis 2012).

Ramacciato 2005 (a non‐randomised study) reported that angle of His reconstruction was a safe and effective alternative to anterior (Dor) fundoplication to reduce the incidence of recurrent achalasia following LHC.

Authors' conclusions

Implications for practice.

Our study is limited by a small number of RCTs with small sample sizes, short periods of follow‐up, and varying methodology. When LHC was performed with minimal hiatal dissection, we were very uncertain whether the addition of a Dor fundoplication made a difference in controlling postoperative reflux, and we were uncertain if it increased the risk of severe postoperative dysphagia. We did not find any difference in the outcomes of postoperative reflux and severe postoperative dysphagia, when comparing Dor and Toupet fundoplication in association with LHC, but the certainty of the evidence was low. Nissen (total) fundoplication used in combination with LHC for achalasia increased the risk of severe postoperative dysphagia. The angle of His accentuation and Dor fundoplication had similar effects when combined with LHC on severe postoperative dysphagia, but there was no objective evidence to compare their effects on postoperative pathological acid reflux.

Implications for research.

We would recommend further RCTs with robust methodology and longer follow‐up periods, especially to compare LHC without fundoplication and LHC + Dor. We would also recommend that studies compare the efficacy of long‐term use of proton pump inhibitors to control postoperative reflux after LHC compared to Dor or Toupet fundoplication. A network meta‐analysis comparing the different types of fundoplications may also reveal helpful evidence.

History

Protocol first published: Issue 7, 2019

Appendices

Appendix 1. Glossary of terms

1. Achalasia: disease where the valve at the lower end of the food pipe is in a permanent state of contraction, causing difficulty in swallowing

2. Aetiology: cause

3. Anterior: front

4. Aspiration: food or fluids entering the airway

5. Barium swallow: an investigation to assess swallowing by x‐ray imaging when the patient is swallowing a radio‐opaque dye

6. Botox: botulinum toxin which causes long‐term relaxation of smooth muscles

7. Cardiomyotomy: division of the muscles forming a valve at the junction of the gullet and stomach

8. Cardiospasm: sustained contraction of the muscles at the lower end of the food pipe resulting in difficulty in swallowing

9. Crus: elongated attachment of the diaphragm to the spine

10. Circumferentital: encircling

11. Diaphragm: flat muscle separating the chest and abdominal cavities

12. Dissection: surgical act of cutting and separating tissues

13. Distally: away from the point of attachment

14. Dysphagia: difficulty in swallowing

15. Endoscopy: examination of the gullet, stomach and intestine using a flexible camera

16. Fundoplication: wrapping the upper part of stomach around the lower end of the gullet to form an artificial valve

17. Ganglion cells: a collection of nerve cells

18. Gastro: related to stomach

19. Gastro‐oesophageal junction: where the stomach and oesophagus join

20. General anaesthesia: a state of medically‐induced unconsciousness during surgery

21. Heartburn: burning sensation in the chest due to regurgitation of stomach acid in the gullet

22. Hiatus: an opening (in the diaphragm) through which the oesophagus enters the abdominal cavity from the chest

23. Hypertensive: higher than normal pressure

24. Inhibitory neurons: nerve cells which slow or stop the excitatory electrical signals generated in response to a stimulus

25. Intraoperative: occurring during surgery

26. Laparoscopy: keyhole surgery

27. Lower oesophageal sphincter (LOS): muscle valve at the lower end of the gullet

28. Manometry: investigation to measure pressure

29. Morbidity: illness or disease

30. Motility: the contraction of muscles of the digestive tract which mixes and propels the contents forward

31. Myenteric: related to muscle layers in the intestine

32. Myotomy: the surgical division of a muscle

33. Nitric oxide: a colourless gas made up of nitrogen and oxygen with the chemical formula NO. It relaxes the smooth muscles in the body, causing dilation of blood vessels and reducing contraction of the intestine

34. Nocturnal: relating to night time

35. Oesophagogastroduodenoscopy: examination of the oesophagus, stomach and duodenum (first part of the intestine) with a flexible camera

36. Oesophagus: gullet or the tubular structure that carries food from the mouth to the stomach

37. Peristalsis: muscular contraction of the gullet or intestine which help to push food forward

38. Peroral: by mouth

39. Pharmacotherapy: treatment with medicines

40. Phreno‐oesophageal ligament: thin membranous attachment between the oesophagus (food pipe) and the diaphragm (muscle separating the chest and abdominal cavities

41. Pneumatic dilation: stretching using gas or air under high pressure

42. Pneumothorax: presence of air in the chest cavity outside the lung, causing partial or complete collapse of the lung on that side

43. Posterior: back

44. Postoperative: occurring after surgery

45. Ports: small surgical openings used in keyhole surgery to insert instruments in the body cavity

46. Proximally: close to the point of attachment

47. Pseudoachalasia: dysphagia similar to achalasia, caused by other diseases like tumours of the gullet

48. Reflux: regurgitation of food and acid from the stomach into the gullet and mouth

49. Regurgitation: bringing up undigested food from the stomach into the mouth

50. Retractor: a surgical instrument used to expose the site of operation

51. Sphincter: a ring of muscle that acts like a valve

52. Sutures: surgical stitches

53. Synthesis: producing a new substance by combining different simpler substances

54. Thoracic: related to chest cavity

55. Vasoactive intestinal peptide: a hormone found in the digestive tract which relaxes the smooth muscles of the intestine and dilates the blood vessels.

Appendix 2. CENTRAL search strategy (via OvidSP)

1. exp Esophageal Achalasia/
2. ((oesophag* or esophag* or cardia*) and achalasia*).tw,kw.
3. (cardiospasm* or (cardia adj3 spasm*) or (cardia adj2 spastic contraction*) or megaesophag* or megaoesophag* or mega oesophag* or mega esophag*).tw,kw.
4.((esophag* or oesophag*) adj5 aperistalsis).tw,kw.
5. or/1‐4
6. exp Laparoscopy/
7. laparoscop*.tw,kw.
8. myotom*.tw,kw.
9. (esophagomyotom* or oesophagomyotom* or oesophagocardiomyotom* or esophagocardiomyotom*).tw,kw.
10. cardiomyotom*.tw,kw.
11. (Heller* or LHM or LHC).tw,kw.
12. or/6‐11
13. 5 and 12

Appendix 3. MEDLINE search strategy (via OvidSP)

1. exp Esophageal Achalasia/

2. achalasia*.tw,kw.

3. (cardiospasm* or (cardia adj3 spasm*) or (cardia adj2 spastic contraction*) or megaesophag* or megaoesophag* or mega oesophag* or mega esophag*).tw,kw.

4. ((esophag* or oesophag*) adj5 aperistalsis).tw,kw.

5. or/1‐4

6. exp Laparoscopy/

7. laparoscop*.tw,kw.

8. myotom*.tw,kw.

9. (esophagomyotom* or oesophagomyotom* or oesophagocardiomyotom* or esophagocardiomyotom*).tw,kw.

10. (cardiomyotom* or fundoplication).tw,kw.

11. (Heller* or LHM or LHC).tw,kw.

12. or/6‐11

13. 5 and 12

14. randomized controlled trial.pt.

15. controlled clinical trial.pt.

16. random*.ab.

17. placebo.ab.

18. trial.ab.

19. groups.ab.

20. or/14‐19

21. exp animals/ not humans.sh.

22. 20 not 21

23. 13 and 22

Note: Lines14‐22. RCT filter: “Cochrane Highly Sensitive Search Strategy for identifying randomized trials in MEDLINE: sensitivity‐maximizing version (2008 revision); Ovid format”. We made the following minor revision: we used “random*.ab.” instead of “randomized.ab” or “randomly.ab.” to capture word variations such as “randomised, randomization, random”

Appendix 4. Embase search strategy (via OvidSP)

1. esophagus achalasia/

2. achalasia*.tw,kw.

3. (cardiospasm* or (cardia adj3 spasm*) or (cardia adj2 spastic contraction*) or megaesophag* or megaoesophag* or mega oesophag* or mega esophag*).tw,kw.

4. ((esophag* or oesophag*) adj5 aperistalsis).tw,kw.

5. or/1‐4

6. exp Laparoscopy/

7. laparoscop*.tw,kw.

8. exp myotomy/ or myotom*.tw,kw.

9. (esophagomyotom* or oesophagomyotom* or oesophagocardiomyotom* or esophagocardiomyotom*).tw,kw.

10. (cardiomyotom* or fundoplication).tw,kw.

11. (Heller* or LHM or LHC).tw,kw.

12. or/6‐11

13. 5 and 12

14. random:.tw.

15. placebo:.mp.

16. double‐blind:.tw.

17. or/14‐16

18. exp animal/ not human/

19. 17 not 18

20. 13 and 19

Note: Lines 14‐17. RCT filter. Hedge Best balance of sensitivity and specificity filter for identifying randomized trials in Embase. https://hiru.mcmaster.ca/hiru/HIRU_Hedges_EMBASE_Strategies.aspx

Data and analyses

Comparison 1. LHC + Dor versus LHC without fundoplication.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1.1 Postoperative pathological acid reflux	2	97	Risk Ratio (M‐H, Random, 95% CI)	0.37 [0.07, 1.89]
1.2 Severe postoperative dysphagia	3	143	Risk Ratio (M‐H, Fixed, 95% CI)	3.00 [0.34, 26.33]

Comparison 2. LHC + Dor versus LHC + Toupet.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
2.1 Postoperative pathological acid reflux	3	105	Risk Ratio (M‐H, Random, 95% CI)	0.75 [0.23, 2.43]
2.2 Severe postoperative dysphagia	3	123	Risk Ratio (M‐H, Random, 95% CI)	0.78 [0.19, 3.15]
2.3 Major complications	3	174	Risk Ratio (M‐H, Random, 95% CI)	0.84 [0.22, 3.18]
2.4 LOS Pressures	2	93	Mean Difference (IV, Fixed, 95% CI)	‐2.11 [‐5.04, 0.81]
2.5 Operating time	2	133	Mean Difference (IV, Random, 95% CI)	‐7.94 [‐18.03, 2.15]

Comparison 3. LHC + Dor versus LHC + Nissen.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
3.1 Postoperative pathological acid reflux	1	138	Risk Ratio (M‐H, Random, 95% CI)	4.72 [0.23, 96.59]
3.2 Severe postoperative dysphagia	1	138	Risk Ratio (M‐H, Random, 95% CI)	0.19 [0.04, 0.83]

Comparison 4. LHC + Dor versus LHC + angle of His accentuation.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
4.1 Severe postoperative dysphagia	1	110	Risk Ratio (M‐H, Random, 95% CI)	1.56 [0.27, 8.95]
4.2 Quality of life (ASQOL) score	1	110	Mean Difference (IV, Random, 95% CI)	‐0.80 [‐6.59, 4.99]

Characteristics of studies

Characteristics of included studies [ordered by study ID]

Grubnik 2017.

Study characteristics
Methods	RCT
Participants	Country: Ukraine Study period: 2009 to 2014 Number of participants: 42 Follow‐up: 25.4 months (range 24‐28 months)
Interventions	Group I: LHC + Dor (n = 20) Group II: LHC (n = 22)
Outcomes	Primary outcome: recurrent dysphagia and gastroesophageal reflux Seconday outcome: operative time
Notes	This study was only published as an abstract. We contacted the authors for detailed results but did not receive any reply. There was no difference in recurrent dysphagia and reflux between the two interventions, but operative time was higher in LHC + Dor.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Details of randomisation are not available in the abstract.
Allocation concealment (selection bias)	Unclear risk	Details of allocation concealment are not available.
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	The text does not mention blinding of participants/personnel.
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	There is no information on whether the assessors were blinded.
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	There is no information on participant dropouts in follow‐up.
Selective reporting (reporting bias)	Low risk	The authors have reported on the important outcomes in the study.
Other bias	Unclear risk	There is no mention whether the authors had accepted funding from any source for this study.

Gupta 2018.

Study characteristics
Methods	RCT, single blind, single institution
Participants	Country: India Period of study: May 2010 to July 2016 Inclusion criteria: diagnosis of achalasia Exclusion criteria: unfit for general anaesthesia, did not consent to surgery, sigmoid oesophagus Number of participants:110 Lost to follow‐up: 1 Number of participants analysed for results = 109 Female: 53; male: 53 Mean age: 33.5 Follow‐up: 36 months (mean)
Interventions	Group I: LHC + AOH (n = 56) GroupII: LHC + Dor (n = 54)
Outcomes	Primary: relief of oesophageal symptoms (dysphagia, regurgitation, heartburn Secondary: new onset heartburn and ASQOL
Notes	Median operative time was higher in LHC + Dor (128 minutes versus 144 minutes; P < 0.01) There was no difference in improvement of oesophageal symptoms, ASQOL and new onset heartburn between the two interventions.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Block randomisation
Allocation concealment (selection bias)	Low risk	Variable block size using computer‐generated random number
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Participants were blinded (single blind)
Blinding of outcome assessment (detection bias) All outcomes	High risk	Assessors were not blinded
Incomplete outcome data (attrition bias) All outcomes	Low risk	Only 2 participants were lost to follow‐up (after 2 years)
Selective reporting (reporting bias)	Low risk	All important outcomes including the ones mentioned in the trial registry were reported.
Other bias	Low risk	No financial support or any other conflict of interest.

Kumagai 2014.

Study characteristics
Methods	RCT
Participants	Country: Sweden Inclusion criteria: participants with symptomatic achalasia (> 3 on Eckardt score) and proven manometry Exclusion criteria: severely unwell patients, pseudoachalasia, inability to retrieve data Number of eligible participants: 42 Excluded from analysis: 1 (did not receive the allocated treatment) Number of participants included in the analysis: 41 Female: 22; male: 19 Average age: 44.4 Follow‐up: 12 months
Interventions	Group 1: LHC + Dor (n = 20) Division of only anterior phrenoesophageal membrane with mobilisation of anterolateral oesophagus and hiatus Group 2: LHC + Toupet (n = 22) One of the participants had open Heller's procedure, but it has not been specified which group (either Dor or Toupet) he belonged to.
Outcomes	HRQL questionnaire and Eckardt score for symptoms 24‐h pH monitoring Timed barium oesophagogram Improvements in quality of life was measured by EORTC QLQ‐OES18 form
Notes	There was no difference between the procedures in terms of symptom improvement, failure of treatment, improvement in quality of life, timed barium oesophagogram and 24‐h pH manometry.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Block randomisation
Allocation concealment (selection bias)	Low risk	Computer‐generated permuted blocks of random sizes.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Participants and care providers were blinded.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Assessors were blinded. Only the principal investigator (BH) could access participants through codes.
Incomplete outcome data (attrition bias) All outcomes	High risk	All participants were followed‐up, but not all the participants were included in the analysis.
Selective reporting (reporting bias)	Low risk	All important outcomes, including the ones in the trial registry were addressed.
Other bias	Low risk	No other conflict of interest

Rawlings 2012.

Study characteristics
Methods	RCT, multicentric trial
Participants	Country: USA Inclusion criteria: age > 18, achalasia diagnosed by oesophagogram, manometry and endoscopy Exclusion criteria: other oesophageal motility disorder, previous gastroesophageal surgery, gastroesophageal cancer, Crohn's disease, hiatus hernia > 5 cm, cognitive impairment Participants eligible for the study: 85 Excluded from the analysis: 25 (did not undergo follow‐up evaluation) Participants analysed in the study: 60 Female: 26; male: 34 Average age: 48.8 Follow‐up: 12 months
Interventions	Group 1: LHC + Dor (n = 36) Group 2: LHC + Toupet (n = 24)
Outcomes	Oesophageal symptom questionnaire (for dysphagia, heartburn, regurgitation, chest pain, nausea/vomiting and bloating) SF‐36 quality of life assessment 24‐h pH testing Oesophageal manometry Barium swallow
Notes	Only 43 (71.7%) of the participants underwent 24‐h pH test. Conclusion: LHC improves symptoms of dysphagia and regurgitation in patients with achalasia regardless of the type of partial fundoplication.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Randomised by computer‐generated random number table.
Allocation concealment (selection bias)	Low risk	Random number table generated on Microsoft Excel. Multicentric study where allocation was most likely done centrally (offsite).
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Not specified in text
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Not specified in text
Incomplete outcome data (attrition bias) All outcomes	High risk	25 out of 85 initially randomised participants were excluded as they did not attend follow‐up. Only 71% of included participants underwent 24‐h pH test.
Selective reporting (reporting bias)	Low risk	All important outcomes, including the ones mentioned in the trial registry were reported.
Other bias	Low risk	The authors have not mentioned any financial or other support from anyone.

Rebecchi 2008.

Study characteristics
Methods	RCT
Participants	Country: Italy Inclusion criteria: achalasia diagnosed by manometry and oesophagogram Exclusion criteria: sigmoid shaped mega‐oesophagus, previous Heller's procedure, oesophagogastric malignancy, neuromuscular disorders, ASA score = 4, contraindications to laparoscopy Number of participants: 144 Number of participants analysed: 144 Female: 69, male: 75 Average age: 49 Follow‐up: 125 months
Interventions	Group 1: LHC + Dor (N = 72) Group 2: LHC + Nissen (N = 72)
Outcomes	Clinical and instrumental gastroesophageal reflux, dysphagia
Notes	Four participants were lost to long‐term follow‐up and 2 participants had died of unrelated causes. Conclusion: both techniques provided long‐term control of reflux, but recurrence of dysphagia was significantly higher in the Nissen fundoplication group.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Randomisation done the day before surgery with computer‐generated random numbers.
Allocation concealment (selection bias)	Low risk	Sealed envelopes
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Not specified in text
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Not specified in text
Incomplete outcome data (attrition bias) All outcomes	Low risk	Very low dropout rates even though the follow‐up was long
Selective reporting (reporting bias)	Low risk	All important outcomes including the ones mentioned in the trial registry were reported.
Other bias	Unclear risk	Not specified in text

Richards 2004.

Study characteristics
Methods	RCT
Participants	Country: USA Inclusion criteria: participants above 18 years of age diagnosed with achalasia Exclusion criteria: previous surgical treatment for achalasia, ongoing GORD, achalasia associated with gastroesophageal malignancy, pregnancy Number of participants: 43 Number of participants analysed 43 Female: 22, male: 21 Average age: 50 Follow‐up: 6 months
Interventions	Group 1: LHC (n = 21) Group 2: LHC + Dor fundoplication (n = 22)
Outcomes	Pathological gastroesophageal reflux (assessed by 24‐hour pH study), dysphagia score
Notes	Conclusion: LHC plus Dor was superior to LHC alone in regard to the incidence of postoperative gastroesophageal reflux.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Block randomisation using a random numbers table
Allocation concealment (selection bias)	Low risk	Sealed opaque envelopes prepared by a blinded investigators. The envelope was handed to the surgeon on the day of the procedure.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Participants and researchers were blinded.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Evaluators were blinded
Incomplete outcome data (attrition bias) All outcomes	Low risk	Intention‐to‐treat analysis
Selective reporting (reporting bias)	Low risk	All important outcomes including the ones mentioned in trial registry were reported
Other bias	Low risk	Authors have no conflict of interest and have not mentioned receiving any financial help from any source.

Simic 2010.

Study characteristics
Methods	RCT
Participants	Country: Serbia Inclusion criteria: achalasia diagnosed by manometry between May 2002 and May 2005 Exclusion criteria: none Total participants randomised: 102 Participants excluded: 18 (2 died of unrelated causes and 16 did not attend follow‐up) Total Participants analysed in the study: 84 Female: 48, males: 36 Average age: 49.6 Follow‐up: 36 months
Interventions	Group 1: LHC with full hiatal dissection + Dor fundoplication (N = 26) Group 2: LHC with limited hiatal and oesophageal dissection + Dor (N = 36) Group 3: LHC with limited hiatal dissection without fundoplication (N = 22) We included only Group 2 and Group 3 in our review, as Group 1 (full hiatal dissection) did not meet our inclusion criteria.
Outcomes	Amount of decrease in LOS pressures after surgery, postoperative DeMeester score, evidence of reflux oesophagitis and hiatus hernia on postoperative endoscopy, amplitude of oesophageal body contractions after surgery, postoperative dysphagia.
Notes	Conclusion: LHC with limited hiatal dissection obtains better reflux control compared to complete hiatal dissection, regardless of Dor fundoplication.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Computer‐aided randomisation
Allocation concealment (selection bias)	Low risk	Random Allocation Software (Edition 1) was used
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Not mentioned in text
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Not mentioned in text
Incomplete outcome data (attrition bias) All outcomes	High risk	16 out of the initial 102 were excluded as they did not attend follow‐up
Selective reporting (reporting bias)	Low risk	All the important outcomes including the ones mentioned in the trial registry were reported.
Other bias	Low risk	Authors have not mentioned receiving any financial or other help from any source.

Torres‐Villalobos 2018.

Study characteristics
Methods	RCT
Participants	Country: Mexico Inclusion criteria: adult participants (> 18 years) with diagnosis of achalasia confirmed with HRM, barium oesophagogram and endoscopy Exclusion criteria: Chagas disease, oesophageal stricture, gastric/oesophageal cancer, peptic stricture, other oesophageal motility disorders Total participants randomised: 73 Participants excluded: 0 Female: 44 (60.3%), male: 29 (39.6%) Average age: 40.5 Follow‐up: 24 months
Interventions	Group I: LHC + Dor (n = 38) Group II: LHC + Toupet (n =35)
Outcomes	Symptomatic reflux and dysphagia scores using Eckardt scores, EAT‐10 and GERD‐HRQL questionnaires. High‐resolution manometry and 24‐h pH monitoring
Notes	There was no difference in the outcomes studied between the two interventions.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Used Microsoft Excel to randomise
Allocation concealment (selection bias)	Low risk	Random number table generated in Microsoft Excel. The number was randomised as the diagnosis was confirmed.
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Not mentioned in text
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Not mentioned in text
Incomplete outcome data (attrition bias) All outcomes	High risk	There were considerable number of dropouts in the follow‐up and it is not mentioned what was done to reduce the attrition bias
Selective reporting (reporting bias)	Low risk	All important outcomes including the ones mentioned in the trial registry were reported
Other bias	Low risk	One of the authors received grants for speaking/teaching from EndoStim, but no financial support was accepted for the study

AOH: angle of His accentuation; ASA: American Society of Anaesthesiology; ASQOL: achalasia specific quality of life; EORTC QLQ‐OES18; European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire‐oesophageal cancer module (developed in 2003); EAT‐10: 10‐item Eating Assessment Tool; GORD: gastro‐oesophageal reflux disease; HRM: high‐resolution manometry; LHC: laparoscopic Heller's myotomy; RCT: randomised controlled trial; SF‐36: 36‐item short form.

Characteristics of excluded studies [ordered by study ID]

Study	Reason for exclusion
Broman 2018	The authors tried to contact participants of a previous RCT (Richards 2004) over phone and mail to collect patient‐reported outcomes. This study was not planned in the original RCT and only 27 out of the original 43 participants (62.7%) could be contacted. The methodology and outcomes measured were different from the original RCT.
Falkenback 2003	In this study all the participants had open Heller's myotomy (does not meet our inclusion criteria).
Torquati 2006	This is a cost‐utility analysis of a previously conducted RCT (Richards 2004) which we have already included), using a decision analysis model.

RCT: randomised controlled trial

Differences between protocol and review

1. In one of the included studies, Kumagai 2014, 1 out of the 41 participants had open surgery. It was not mentioned if this was planned as an open procedure or if it was a conversion of a planned laparoscopic procedure. Even though our protocol stated we would only include studies where Heller's cardiomyotomy was performed laparoscopically and exclude all studies where the procedure was performed by open surgery, we included this study because we decided that the impact of this one open procedure, on the results of the entire review would be very small.

2. Severe postoperative dysphagia is one of our primary outcomes. In the protocol, we stated that we would define this as severe dysphagia requiring medical or surgical intervention. But in our review we also included participants with severe symptomatic dysphagia (Eckardt score > 3) even if they did not have any intervention, as it is considered failure of treatment.

3. We could not carry out subgroup analyses comparing age groups, gender, ASA groups and severity of achalasia as planned in the protocol because the information required for these analyses was not available in the study reports.

4. We have only included the primary outcomes (postoperative pathological acid reflux and severe postoperative dysphagia) in the summary of findings table as these were the two factors which are important to clinical decision makers in choosing the type of fundoplication in LHC.

Contributions of authors

Conceiving and designing the review: SM, MWM

Co‐ordinating the review: SM, DG, MWM

Literature review, selecting studies and data collection: SM, DG

Analysis and interpretation of data: SM, DG

Writing the review: SM

Providing critical comments on the review: DG, MWM

Performing previous work that was the foundation of the current study: SM, MWM

Sources of support

Internal sources

• No sources of support provided

External sources

• No sources of support provided

Declarations of interest

SM: none

DG: none

MWM: none

New