Accepted Manuscript

Gastroesophageal reflux disease after per-oral endoscopic myotomy as compared

with Heller’s myotomy with fundoplication: a systematic review with meta-analysis

Alessandro Repici, MD, Lorenzo Fuccio, MD, Roberta Maselli, MD, Fabrizio Mazza,
MD, Loredana Correale, PhD, Daniele Mandolesi, MD, Cristina Bellisario, PhD,
Amrita Sethi, MD, Mouen Kashab, MD, Thomas Rösch, MD, Cesare Hassan, MD,
PhD.

PII: S0016-5107(17)32419-7
DOI: 10.1016/j.gie.2017.10.022
Reference: YMGE 10800

To appear in: Gastrointestinal Endoscopy

Received Date: 20 August 2017

Accepted Date: 13 October 2017

Please cite this article as: Repici A, Fuccio L, Maselli R, Mazza F, Correale L, Mandolesi D, Bellisario
C, Sethi A, Kashab M, Rösch T, Hassan C, Gastroesophageal reflux disease after per-oral endoscopic
myotomy as compared with Heller’s myotomy with fundoplication: a systematic review with meta-
analysis, Gastrointestinal Endoscopy (2017), doi: 10.1016/j.gie.2017.10.022.

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.
 ACCEPTED MANUSCRIPT

Gastroesophageal reflux disease after per-oral endoscopic myotomy as compared with

Heller’s myotomy with fundoplication: a systematic review with meta-analysis

Alessandro Repici1, MD, Lorenzo Fuccio2, MD, Roberta Maselli1, MD, Fabrizio Mazza1,MD,
Loredana Correale3, PhD, Daniele Mandolesi2, MD, Cristina Bellisario4, PhD, Amrita Sethi5, MD,
Mouen Kashab6, MD, Thomas Rösch7, MD, and Cesare Hassan8, MD, PhD.

PT
RI
Institutions:
1
Digestive Endoscopy Unit, Division of Gastroenterology, Humanitas Research Hospital,

SC
Humanitas University, Milan, Italy.
2
Gastroenterology Unit, Department of Medical and Surgical Sciences, S.Orsola-Malpighi
University Hospital, Bologna, Italy.

U
3
Im3D Medical Imaging Lab, Turin, Italy.
AN
4
Department of Cancer Screening, Centre for Epidemiology and Prevention in Oncology (CPO),
University Hospital Città della Salute e della Scienza di Torino, Turin, Italy.
5
Division of Digestive and Liver Diseases, Department of Medicine, Columbia University Medical
M

Center, New York, New York, USA.

6
Division of Gastroenterology, Johns Hopkins Hospital, Baltimore, Maryland, USA.
D

7
University Medical Center Hamburg-Eppendorf, Hamburg, Germany.
TE

8
Digestive Endoscopy Unit, Nuovo Regina Margherita Hospital, Rome, Italy.
EP

Short Title: GERD post-POEM and Heller’s myotomy with fundoplication

Words count
C
AC

Abstract: 250 words; Full text: 3780 words; References: 75; Figures: 3; Appendix: 4.

Specific author contributions: Concept and design: Repici A, Fuccio L, Rösch T, Hassan C;
analysis and interpretation of the data: Repici A, Fuccio L, Correale L, Hassan C; drafting of the
article: Repici A, Fuccio L, Hassan C; critical revision of the article for important intellectual
content: Rösch T, Maselli M, Mazza F, Sethi A, Kashab M; final approval of the article: all the
Authors; statistical expertise: Correale L; collection and assembly of data: Fuccio L, Mandolesi D,
and Bellisario C.

Financial support: None; Grant support: None; Potential competing interests: None.
 ACCEPTED MANUSCRIPT

Keywords: Achalasia; Per-Oral Endoscopy Myotomy; Heller’s Myotomy; Interventional

Endoscopy.

Corresponding author:

Prof. Alessandro Repici

PT
Digestive Endoscopy Unit
Humanitas Research Hospital,
Via Manzoni 56,

RI
20089 Rozzano (Milan), Italy

SC
Fax: +39 (0)282242579
e-mail: [email protected]

U
AN
M
D
TE
C EP
AC
 ACCEPTED MANUSCRIPT
ABSTRACT
Background and Aims: Per-oral endoscopic myotomy (POEM) represents a less invasive
alternative, as compared with conventional laparoscopic Heller’s myotomy (LHM), for achalasia
patients. It cannot be excluded, however, that the lack of fundoplication after POEM may result in a
higher incidence of reflux disease, as compared with LHM. The aim was to conduct a systematic
review of prospective studies reporting the incidence of reflux disease developed after POEM and
LHM.

PT
Methods: Literature search with electronic databases was performed (up to February 2017) to

RI
identify full articles on the incidence of gastro-esophageal reflux symptoms, endoscopic- and pH-
monitoring-findings after POEM and LHM (with fundoplication). Proportions and rates were

SC
pooled by means of a random or fixed effects models, according to the level of heterogeneity
between studies.

U
Results: After applying the selection criteria, 17 and 28 studies, including 1,542 and 2,581 subjects
AN
who underwent POEM and LHM, respectively, were included. Pooled rate of postprocedure
symptoms was 19.0% (95% CI, 15.7%-22.8%) after POEM, and 8.8% (95% CI, 5.3%-14.1%) after
LHM, respectively. Pooled rate estimate of abnormal acid exposure at pH-monitoring was 39.0%
M

(95% CI, 24.5%–55.8%) after POEM, and 16.8% (95% CI, 10.2%–26.4%) after LHM, respectively.
Rate of post-POEM esophagitis was 29.4% (95% CI, 18.5%-43.3%) after POEM, and 7.6% (95%
D

CI, 4.1%-13.7%) after LHM. At meta-regression, heterogeneity was partly explained by POEM
TE

approach and study population.

Conclusions: Incidence of reflux-disease appears to be significantly more frequent after POEM

EP

than after LHM with fundoplication. pH-monitoring and appropriate treatment after POEM should
be considered in order to prevent long-term reflux-related adverse events.
C
AC

Introduction
Achalasia and other spastic esophageal disorders represent a wide variety of esophageal
motility disorders with different pathophysiology and manometric pattern according to the Chicago

1
 ACCEPTED MANUSCRIPT
Classification. [1] Despite these differences, however, the presenting symptoms are quite similar
including dysphagia, chest pain, heartburn, and/or regurgitation.
As no curative treatment is available, current therapeutic options aim to reduce the pressure
of the lower esophageal sphincter (LES), in order to facilitate the transport of the bolus into the
stomach, resulting into symptomatic improvement. Until recently, disruption of the LES was best
accomplished by pneumatic dilation or laparoscopic Heller’s myotomy (LHM) and, less effectively,

PT
by pharmacological agents, such as botulinum toxin injection or calcium channel blockers. [2-9]
Per-oral endoscopic myotomy (POEM) is a relatively novel minimally invasive technique

RI
that is used to treat achalasia and other spastic esophageal disorders. It may be considered as the
endoscopic equivalent of surgical myotomy with the substantial advantage for POEM to adapt the

SC
length of the myotomy according to the motility disorder. A systematic review on the efficacy of
POEM for achalasia reported technical and clinical success of 97% and 93%, respectively. [10]
One of the main POEM drawbacks may be represented by the development of gastro-

U
esophageal reflux, when considering that – differently from LHM – no anti-reflux intervention is
AN
performed. However, differently from LHM, no alteration of the diaphragmatic and gastro-
esophageal anatomy occurs with POEM, potentially mitigating the risk of reflux.
M

Aim of our study was to systematically assess the incidence of reflux disease after POEM
and LHM, when taking into consideration symptoms, endoscopic and pH-monitoring findings.
D

Methods
TE

The systematic review was carried out in according with the guidelines of the preferred
reported items for systematic review and meta-analyses (PRISMA-P). [11] The following methods
EP

are reported in Appendix 1: data sources and search strategy, the selection process, data extraction,
and the quality assessment.
C
AC

Eligibility criteria
To be included in the systematic review, we retrieved only prospective studies including
>10 adult patients (>18 years old) with a diagnosis of achalasia or other spastic esophageal disorder
(as defined by revised Chicago classification [1]) treated by POEM or LHM (with fundoplication),
and providing the incidence of gastro-esophageal reflux disease after at least 2 months of follow-up.

2
 ACCEPTED MANUSCRIPT
In case of mixed population of different ages (ie, pediatric and adult population) and results not
accordingly stratified, the study was considered only if the pediatric population represented ≤10%
of the entire study population. Only articles published as full-text and in the English language were
considered.
The presence of gastro-esophageal reflux disease was assessed according to:
● Symptoms: the presence of symptomatic reflux (heartburn and/or regurgitation) with or

PT
without a specific score system.
● Endoscopic findings: the presence of reflux esophagitis according to the Los Angeles

RI
classification (LA). [12] Where reported, we included the degree of severity of the
esophagitis (Grade A-D).

SC
● pH-monitoring findings: altered esophageal acid exposure defined by a % total reflux time
(TRT, esophageal pH<4) greater than 5% evaluated by 24 hours esophageal pH monitoring
and/or DeMeester score greater than 14.72 on the 24-hour periods.

U
AN
Outcomes
The primary outcome was to evaluate the incidence of post-POEM and post-LHM gastro-
esophageal reflux disease according to 3 subcategories: symptoms, endoscopic and pH-monitoring
M

findings in patients with esophageal motility disorders.

Secondary outcome was the assessment of factors influencing the risk of developing
D

postprocedure gastroesophageal reflux disease.

TE

Statistical analysis
Clinical patient data from individual studies were analyzed to obtain summary statistics.
EP

Continuous data were reported as the median (inter-quartile range [IQR]) or mean (standard
deviation [SD]), whereas categorical data were reported as the count and percentage. Event rates for
C

postprocedure reflux according to the 3 subcategories for each primary outcome were defined as the
AC

proportion of events among the number of patients who were available for follow-up. Standard
errors (SEs) and 95% confidence intervals (CIs) for a single proportion were derived.
Heterogeneity among studies was assessed by using the I2 statistic, which describes the
percentage of total variation attributable to between-study heterogeneity as opposed to random error
or chance. Percentages of 25% (I2 = 25), 50% (I2 = 50), and 75% (I2 = 75) were considered as low,

3
 ACCEPTED MANUSCRIPT
moderate, and high degree of heterogeneity, respectively. In the presence of substantial
heterogeneity (I2 > 50%), a random effect model was used as the pooling method; otherwise, a fixed
effect model was adopted as the pooling method. Prediction intervals for size effects of a new study
were also calculated according to Higgins et al. [13] in cases where, due to marked heterogeneity, it
was not possible to have interpretable overall effects. With regard to outcomes, when significant
heterogeneity existed across studies, random effects univariate and multivariate meta-regressions

PT
were performed to assess the potentially important covariates: socio-demographic attributes (eg,
geographic origin of the studies) and sex, as well as methodology-related attributes (eg, study

RI
design, year of publication, and study sample size), in addition to clinical and technical parameters.
The proportion of total between-study variances was explained by the models and reported as R2.

SC
We used meta-regression models to test between-subgroup interactions, and a P value ≤0.05
indicated a significant difference. Subgroup analyses were performed according to the source of
heterogeneity or using covariates if possible. Sensitivity analysis was performed by sequentially

U
omitting each study to test the influence of each individual study on pooled data. [14] The Egger’s
AN
test and the trim and fill method of Duval and Tweedie were used to test and adjust for publication
bias. [15, 16] All statistical analyses were conducted in R. In particular, we used the metafor
M

libraries in R to carry out the meta-analysis and meta-regression. [17]

Secondary analyses
D

The studies included in our meta-analysis had multiple outcome variables (because of the
TE

three different methods used to evaluate reflux disease: symptoms, endoscopic and pH-monitoring
findings) for the same sample of participants which are dependent outcomes. A multivariate random
EP

effect model was used to meta-analyze the different outcomes simultaneously, accounting for
dependence in the outcome estimates within studies. The model has a random part that adds random
effect for each outcome within each trial to the standard random effect model. Therefore, outcomes
C

within the same study receive the same random effect, whereas outcomes across different studies
AC

are assumed to be independent. The model also includes an unstructured variance-covariance matrix
that allows the random effects to have different variances for each outcome and to be correlated (ie,
unstructured variance-covariance matrix). Finally, the model includes study-level covariates (eg, the
method of disease evaluation). These covariates were considered to have fixed-effects and,
therefore, did not impact the specification of correlation structure. Data were presented as odds
4
 ACCEPTED MANUSCRIPT
ratios (ORs) and 95% CIs. The OR is interpreted as the ratio of the odds of detecting reflux disease
for one group (eg, symptoms evaluation) compared with the odds for another (eg, PH findings).
Meta-analysis via multivariate random-effect model was performed by using R [17].

RESULTS

PT
POEM

After removing duplicates, a total of 296 records were identified through databases searches;

RI
271 records were excluded on the basis of titles and abstracts as clearly not relevant; 25 were
judged potentially relevant and acquired in full text and after applying the selection criteria, 17

SC
studies were finally included in the systematic review.[18-34] The flow-chart of the selection
process and the main characteristics of the included studies are reported in Appendix 2. Altogether

U
these 17 studies included 1542 subjects. Except for 3 studies [20, 24, 29], which reported data on
treatment of naive patients, most (n=12) studies included patients who were both treatment naive
AN
and previously treated patients; there was no reported data on study population in 2 trials [21, 26].
Twelve studies reported data on the use of postoperative proton-pump inhibitor (PPI). Across these
M

12 trials, the rates of PPIs use ranged from 2.6% to 27.8% (pooled estimate, 10.6%; 95% CI, 6.5%-
17.3%) with a pooled estimated success rate of 91.5% (95% CI, 80.9%-96.5%). According to the
D

modified Newcastle–Ottawa Scale, the mean score was 3.4 (range 2-4).
TE

LHM with fundoplication

After removing duplicates, a total of 1379 records were identified through databases
EP

searches; 1313 records were excluded on the basis of titles and abstracts as clearly not relevant; 66
were judged potentially relevant and acquired in full text and after applying the selection criteria, 28
C

studies were finally included in the systematic review [22, 35-61]. The flow-chart of the selection
AC

process and the main characteristics of the included studies are provided in Appendix 2. Altogether
these 28 studies included 2581 subjects. Most (n=15) trials were in mixed populations (including
either näive patients or patients with a prior treatment), followed by 9 trials in naive patients, and 1
trial included only patients with a prior treatment; there was no reported data on study population in
3 trials (n=888 patients). All but one of the study reported data about the type of fundoplication

5
 ACCEPTED MANUSCRIPT
after LHM (Appendix 2). According to the modified Newcastle–Ottawa Scale, the mean score was
2.4 (range 1-4).

OUTOCOMES

Rates of Postprocedure Reflux According to Symptoms Evaluation.

PT
POEM
Follow-up data were provided in all 17 studies for a total of 1275 (82.6%) patients. Meta-
analysis of all 17 studies yield an overall pooled rate of 19.0% (95% CI, 15.7%-22.8%) with a

RI
moderate degree of between-study heterogeneity (I2 =43.3%; p=0.024) (Figure 1A). Sensitivity
analysis revealed that no single study has substantial influence on the outcome. There was no

SC
significant publication bias (Egger test: z = -1.8413, p = 0.070).
On univariate meta-regression, the heterogeneity between studies was explained by the country

U
of the study, by the ratio between female and male and by the prevalence of Achalasia type III.
AN
Details provided in Table A in Appendix 3. However, in multivariable analysis, the study country
was found to be only marginally significant related to outcome, whereas the effect of female/male
ratio and type of achalasia were no longer significant.
M
D

LHM with fundoplication

TE

Follow-up data were provided in 20 studies for a total of 1136 (44%). Meta-analysis of all 20
studies yield an overall pooled rate of 8.8% (95% CI, 5.3%-14.1%) with a high degree of between-
study heterogeneity (I2 =80%) (Figure 2A). Sensitivity analysis revealed that no single study has
EP

substantial influence on the outcome. There was evidence for significant publication bias (Egger
test: z = -4.0852, p < .0001). After the trim fill analysis, we found 6 studies missing on the bottom
C

right side of the funnel plot using the random effect model (Supplementary Figure A in Appendix
AC

4). After adjusting for these studies in the forest plot, the pooled symptoms rate was 14.4% (95%
CI, 8.4%-23.5%). However, the heterogeneity was still evident in the simulated meta-analysis
(I2=87%, p<.001).
On univariate meta-regression, the heterogeneity between studies was explained by multiple
factors (Supplementary Table A in Appendix 4). A significant linear relationship between the

6
 ACCEPTED MANUSCRIPT
sample size of the study and the rate of reflux symptoms was observed. The rate decreased as the
sample size increased (regression coefficient, log odds: -0.011; 95% CI, -0.020 to -0.001)
(Supplementary Figure B in Appendix 4). The same linear relationship was found for the number
of patients who were actually included in the follow-up analysis (regression coefficient, -0.013;
95% CI, -0.023 to -0.003). The type of fundoplication itself accounted for 33.7% of the total
amount of heterogeneity (test for residual heterogeneity, QE (df = 16) = 62.4218, p-value <.0001),

PT
and it significantly influenced the occurrence of reflux symptoms: an increasing number of Nissen
and Toupet procedures were both inversely associated with the outcome (Nissen, -0.033; 95% CI, -

RI
0.063 to -0.004; Toupet, -0.010; 95% CI, -0.019 to -0.0001; P=0.048). In the analysis, was also
found an association between follow-up duration and outcome. The rates tended to be lower among

SC
studies with longer follow-up duration (minimum follow-up, -0.092; 95% CI, -0.166 to -0.0193).
Multivariate analysis showed that the only independent factors impacting on symptoms rate
were the minimum follow-up duration (log odds, -0.086; 95% CI, -0.155 to -0.018; p=0.013) and

U
the number of Nissen procedures (log odds, 95% CI, -0.0565 to 0.0042; p=0.092). In detail, the
AN
occurrence of reflux symptoms among studies with a minimum follow-up duration of <12 months
was more common than among studies with a minimum follow-up of 12 months or longer (12%;
M

95% CI, 6%-22% vs 4.8%; 95% CI, 2.5%-9.0%). In the later subgroup, moderate heterogeneity was
observed (Supplementary Figure C in Appendix 4).
D

Rates of Postprocedure Reflux According to Ph Monitoring.

TE

POEM
Clinical follow-up evaluation by pH monitoring was available for 289 patients from 5 trials [19,
EP

21, 23, 30, 31]. A forest plot of the rates of the patients detected with abnormal acid exposure is
shown in Figure 1B. Rates ranged from 12.9% to 57.7%, with a random effects pooled rate
C

estimate of 39.0% (95% CI, 24.5%–55.8%). A significant level of heterogeneity was found I2 =
AC

85.0% (p < .0001). The heterogeneity largely disappeared after elimination of the Shiwahu et al.
study (pooled rate estimate, 47.1%; 95% CI, 39.7%-54.9%; I2 = 20.5%; p=0.223) [23]. Of note, the
study by Shiwahu et al was the only one study conducted in Asia. There was no significant
publication bias (Egger test: z = -0.1578, p = 0.8746).

7
 ACCEPTED MANUSCRIPT
LHM with fundoplication
Clinical follow-up evaluation by pH monitoring was available for 1022 patients from 14 trials
mean period of follow-up was 22.8 months (range: 2-83 months). A forest plot of the rates of the
patients detected with abnormal acid exposure is shown in Figure 2B. Rates ranged from 0% to
57%, with a random effects pooled rate estimate of 16.8% (95% CI, 10.2%–26.4%). A significant
level of heterogeneity was found (I2=86.0%). A sensitivity analysis suggested that the data were not

PT
affected by the sequential exclusion of any particular trial from the pooled analysis. There was
evidence for significant publication bias (Egger test: z = -2.8092, p = 0.005). After the trim fill

RI
analysis, we found 3 studies missing on the bottom right side of the funnel plot using the random
effect model (Supplementary Figure D in Appendix 4). After adjusting for these studies in the

SC
forest plot, the pooled symptoms rate was 22.2% (95% CI, 12.0%-36.0%). However, the
heterogeneity was still evident in the simulated meta-analysis (I2=91%, p<.001).
In univariate meta-regression (Supplementary Table B in Appendix 4), the heterogeneity

U
between studies was partly explained by the type of fundoplication. This factor accounted for 76.4%
AN
of the total amount of heterogeneity (test for residual heterogeneity, Q(df=11) = 30.253, p<.0014)
and it was significantly related to the outcome: the rate of abnormal events decreased significantly
M

with an increasing number of Nissen procedures (log odds, -0.0183; 95% CI, -0.034 to -0.003;
p=0.022). Conversely, the rate tended to be higher with higher numbers of Toupet procedures (log
D

odds, 0.0130; 95% CI, -0.0014 to 0.0273), although this did not reach statistical significance
(p=0.076). In a further analysis (see Supplementary Figure E in Appendix 4), the point estimate
TE

for studies using either Toupet or Dor procedures was significantly higher than that for studies
using only Dor procedures (log odds, 1.352; 95% CI, 0.561-2.144) and that for studies using either
EP

Nissen or Dor procedures (log odds, 2.70; 95% CI, 1.45- 3.95).
The relationship of the study country (dichotomized into USA/Brazil vs. Europe) to the
outcome was also significant (log odds, 0.35; 95% CI, 0.35-2.42). However, with only this factor,
C

the heterogeneity remained significantly high (amount of heterogeneity accounted for, 43.0%; test
AC

for residual heterogeneity, Q=46.940; p<.001). By multivariate analysis, the number of Nissen
procedures remained significantly associated (log odds, -0.016; 95% CI, -0.031 to 0.0003;
p=0.054).

Rates of postprocedure reflux according to endoscopic findings.

8
 ACCEPTED MANUSCRIPT
POEM
Data on the findings after a postoperative upper endoscopy were available in 12 studies for a
total of 1056 patients. The average mean period of follow-up was 9.3 months (range: 2-30 months).
Of the 1056 included patients, 449 were diagnosed with esophagitis of any severity (grade A to D,
Los Angel Classification) corresponding to a pooled overall rate of 29.4% (95% CI, 18.5%-43.3%).
The rates of esophagitis of any severity for all studies were given in Figure 1C. Although there was

PT
heterogeneity across the trials (I2=93% 95% CI, 89.6%-95.1%), a sensitivity analysis suggested that
no obvious changes were observed for the pooled rate when single study was excluded respectively.

RI
There was also significant publication bias (Egger test: z = -2.6626, p = 0.0078). After adjusting for
missing trials (n=2), the overall rate of esophagitis was 35.3% (95% CI, 22.5%-50.7%). On

SC
multivariable analysis, the achalasia type III prevalence and the posterior approach orientation
remained significantly associated with the outcome. According to LA classification, most of the
esophagitis was classified as mild grade, accounting for 92.0% (54.3%, 244 of 449 findings in LA-

U
A and 37.6%, 169 of 449 findings in LA-B). Considering moderate to severe grade (LA-C and LA-
AN
D) as a clinically relevant disease, the overall pooled rate of reflux disease was 4.47% (95% CI,
3.27%-6.07%). Low heterogeneity was found for this outcome (I2 = 0%; 95% CI, 0%-53.3%; p=
M

0.54).
D

LHM with fundoplication

Data on the findings after a postoperative upper endoscopy were available in 5 studies for a
TE

total of 752 patients. The average mean period of follow-up was 26.6 months (range, 12-49
months). Of the 752 included patients, 46 were diagnosed with esophagitis of any severity (grade A
EP

to D, Los Angeles Classification) corresponding to a pooled overall rate of 7.6% (95% CI: 4.1%-
13.7%). The rates of esophagitis of any severity for all studies were given in Figure 2C. Test for
heterogeneity was significant (I2 = 62%; p=0.03). After excluding the trial [36] that contributed to
C

the heterogeneity, the mean rate was 6.0% (95% CI, 5.0%-9.0%) (Q (df = 3) = 5.485, p-value =
AC

0.140). There was no evidence for publication bias (Egger test: z = -1.3754, p = 0.1690). According
to LA classification, most of the esophagitis were classified as mild grade, accounting for 48.5%
(39.4%, 13 of 43 findings in LA-A and 9.1%, 3 of 33 findings in LA-B). Considering moderate to
severe grade (LA-C and LA-D) as a clinically relevant disease, the overall pooled rate of reflux

9
 ACCEPTED MANUSCRIPT
disease was 1.0% (95% CI, 0.4%-.3.0%). Test for heterogeneity was not significant for this
outcome [Q (3) =5.08; p=0.1675; I2 =40.7%].

Comparison between POEM and LHM

According to the results from multivariate model [Appendix 3], in the LHM with
fundoplication, the rate of reflux disease was 8.6% (95% CI, 5.4%-13.3%), 14.9% (95% CI, 9.4-

PT
23.0) and 8.3% (95% CI, 3.8%-17.0%) for disease determined by symptoms evaluation, esophageal
pH test and endoscopic findings, respectively. The corresponding rates in POEM cohorts were

RI
18.1% (95% CI, 13.1%-24.4%), 39.3% (95% CI, 25.2%-55.4%) and 30.7% (95% CI, 22.2%-
40.5%) [Figure 3].

SC
DISCUSSION

U
According to our findings, POEM is associated with a 2 to 3 folds increased risk of post-
AN
intervention reflux when compared with LHM with fundoplication, the result being consistent
across the three main parameters assessed, namely reflux-symptoms, abnormal pH-monitoring or
M

endoscopic diagnosis of esophagitis.

Our findings are relevant for the following reasons. First, we may estimate that the number
D

of patients needed to be treated with LHM over POEM in order to prevent the incidence of reflux at
pH-monitoring is 2.8, suggesting an intrinsic superiority in the anti-reflux barrier of the surgical
TE

procedure. This was not unexpected when considering that the lack of any protection by post-
POEM reflux that is comparable with the very effective fundoplication after LHM. However, the
EP

clinical implications of such higher incidence of reflux appeared to be somewhat less relevant. For
instance, as many as 9 patients should be treated with LHM over POEM to prevent the incidence of
C

symptomatic reflux disease. In addition, despite the incidence of esophagitis was significantly more
frequent after POEM than LHM, the gradient in severe esophagitis was quite low, so that
AC

approximately 30 patients should be treated with LHM over POEM to prevent one case of
postprocedure severe esophagitis. This appears to be indirectly confirmed by the low rate of
prolonged PPI therapy shown in our analysis after POEM. This lower rate of post-POEM
symptomatic GERD or severe esophagitis may be related with the preservation of all the anatomic
structures that contribute to the gastroesophageal barrier, representing a main advantage on LHM,
10
 ACCEPTED MANUSCRIPT
where a partial fundoplication is required to compensate for the lack of integrity of such anatomic
structures. Secondly, our comparative meta-analysis shows a similar degree of dissociation between
the high rate of abnormal acid exposure and the occurrence of symptoms in both POEM and LHM
settings. Such difference may be explained by the persistence of a suboptimal esophageal motility
after LES disruption, resulting in an abnormal stasis of reflux material after the reflux itself. It is
unclear at this stage what the clinical implications of such asymptomatic reflux documented by pH

PT
monitoring are, and in particular what therapeutic strategy must be put on in those asymptomatic
cases with abnormal exposure. Third, the presence of between-study heterogeneity was investigated

RI
by meta-regression and subgroups analyses suggesting interesting associations. For POEM, we
showed that geographical setting and gender seems to be associated with the incidence of post-

SC
POEM symptomatic reflux, whereas the type of fundoplication was associated with heterogeneity in
LHM estimate. The incidence of post-POEM reflux tended to be lower in Asian as compared to
non-Asian studies, occurring in 16.0% and 22.8% of the cases, respectively. This may be at least

U
partially related with a lower prevalence of GERD, in general, and of GERD-pathogenetic factors in
AN
Eastern as compared with Western populations, such as different lifestyles, eating habits and
increased incidence of Helicobacter pylori-related chronic atrophic gastritis.[62] Similarly, the
M

association between male sex and a higher incidence of post-POEM symptomatic GERD may be
explained by the well-known pathogenetic role of male sex in the development of GERD in
D

general.[63] Of note, there was a publication bias in two out of the three estimates of post-LHM
reflux, namely symptoms and pH-monitoring. When adjusting for such bias, the difference between
TE

the two techniques appeared to be substantially reduced.

There are limitations to the present analysis. Regarding post-POEM/LHM incidence of
EP

symptomatic GERD, definition of reflux symptoms was not standardized across the various studies.
Only some studies adopted questionnaires with predefined scores on the duration, frequency and
intensity of symptoms. However, there was only moderate heterogeneity across the studies for this
C

primary endpoint. There was a high degree of heterogeneity in the estimates of pH and endoscopic
AC

findings. However, we were partially able to explain it at meta-regression, and the exploratory
results of such analysis may be useful for driving future research in this setting. Publication bias
might partly be explained by the restriction of our search only to studies published in English and
by the exclusion of the proceedings of the international congresses due to incomplete reporting.

11
 ACCEPTED MANUSCRIPT
Because a not negligible proportion of patients after POEM present abnormal pH-
monitoring findings without typical symptoms of GERD, patients should follow a long-term yearly
surveillance program, including ph-monitoring, endoscopic and symptoms assessment. Long-term
effects of abnormal ph-monitoring in asymptomatic subjects are not well known. In particular it is
not clear whether asymptomatic abnormal ph-exposure significantly increases the risk of
metaplastic changes, therefore strong recommendation to treat this subgroup of patients cannot be

PT
provided. Most probably an empirical short course of PPIs (eg, 3 months) should be routinely
suggested after POEM, whereas a more tailored approach based on symptoms and tests appears to

RI
be more appropriate in the long-term. In selected cases of nonresponders to PPI treatment or
patients refusing chronic PPI treatment, fundoplication after POEM should be strongly considered

SC
[76].
In our view, the decision between POEM and Heller’s myotomy with fundoplication for the
treatment of achalasia should be based on several factors (eg, age, comorbidities, local expertise,

U
type of achalasia according to Chicago classification, previous treatments, risk of short- and long-
AN
term major adverse events) and consider the risk of postprocedure GERD only as a secondary
factor, readily treatable in almost all the cases with standard PPI dose.
M

In conclusion, incidence of reflux-disease appears to be significantly more frequent after

POEM than after LHM with fundoplication. pH-monitoring and appropriate treatment after POEM
D

should be considered in order to prevent serious long-term reflux-related adverse events.

TE

Figure Legend
EP

Figure 1: (A) Random-effect meta-analysis of persistent reflux symptoms after POEM shown in a
forest plot. Between-study variability is also provided showing moderately large (I2=43.3%)
heterogeneity. (B) Random-effect meta-analysis of abnormal acid exposure after POEM shown in a
C

forest plot. Between-study variability is also provided showing high heterogeneity (I2=85%). (C)
AC

Random-effect meta-analysis of reflux events according to endoscopic findings after POEM.

Between-study variability is also provided showing high (I2=93%) heterogeneity. The squares
represent individual studies, and the size of the square represents the weight given to each study in
the meta-analysis. Error bars represent 95% confidence intervals (CI). The diamond represents the
combined results.
12
 ACCEPTED MANUSCRIPT

Figure 2: (A) Random-effect meta-analysis of persistent reflux symptoms after Heller’s myotomy
shown in a forest plot. Between-study variability is also provided showing high heterogeneity
(I2=80%). (B) Random-effect meta-analysis of abnormal rate of abnormal esophageal acid exposure
using prolonged pH monitoring after Heller’s myotomy. Between-study variability is also provided
showing high (I2=86%) heterogeneity. (C) Random-effect meta-analysis of reflux events according

PT
to endoscopic findings for an average follow-up of 26.6 months after Heller’s myotomy shown in a
forest plot. Between-study variability is also provided showing significant (I2=62%) heterogeneity.

RI
The squares represent individual studies, and the size of the square represents the weight given to
each study in the meta-analysis. Error bars represent 95% confidence intervals [CI]. The diamond

SC
represents the combined results.

Figure 3. Comparisons of rates of reflux disease after POEM and LHM with fundoplication

U
according to symptoms evaluation, pH assessment and endoscopic findings.
AN
M

REFERENCES

[1] Kahrilas PJ, Bredenoord AJ, Fox M, et al. The Chicago Classification of esophageal motility
D

disorders, v3.0. Neurogastroenterology and motility: the official journal of the European
TE

Gastrointestinal Motility Society. 2015;27:160-74.

[2] Boeckxstaens GE, Annese V, des Varannes SB, et al. Pneumatic dilation versus laparoscopic
Heller's myotomy for idiopathic achalasia. The New England journal of medicine. 2011;364:1807-
EP

16.
[3] Spechler SJ. Pneumatic dilation and laparoscopic Heller's myotomy equally effective for
C

achalasia. The New England journal of medicine. 2011;364:1868-70.

AC

[4] Campos GM, Vittinghoff E, Rabl C, et al. Endoscopic and surgical treatments for achalasia: a
systematic review and meta-analysis. Annals of surgery. 2009;249:45-57.
[5] Yaghoobi M, Mayrand S, Martel M, et al. Laparoscopic Heller's myotomy versus pneumatic
dilation in the treatment of idiopathic achalasia: a meta-analysis of randomized, controlled trials.
Gastrointestinal endoscopy. 2013;78:468-75.
13
 ACCEPTED MANUSCRIPT
[6] Schoenberg MB, Marx S, Kersten JF, et al. Laparoscopic Heller myotomy versus endoscopic
balloon dilatation for the treatment of achalasia: a network meta-analysis. Annals of surgery.
2013;258:943-52.
[7] D'Onofrio V, Annese V, Miletto P, et al. Long-term follow-up of achalasic patients treated with
botulinum toxin. Dis Esophagus. 2000;13:96-101; discussion 2-3.
[8] Annese V, Bassotti G, Coccia G, et al. A multicentre randomised study of intrasphincteric

PT
botulinum toxin in patients with oesophageal achalasia. GISMAD Achalasia Study Group. Gut.
2000;46:597-600.

RI
[9] Leyden JE, Moss AC, MacMathuna P. Endoscopic pneumatic dilation versus botulinum toxin
injection in the management of primary achalasia. Cochrane database of systematic reviews.

SC
2006:CD005046.
[10] Barbieri LA, Hassan C, Rosati R, et al. Systematic review and meta-analysis: Efficacy and
safety of POEM for achalasia. United European Gastroenterol J. 2015;3:325-34.

U
[11] Shamseer L, Moher D, Clarke M, et al. Preferred reporting items for systematic review and
AN
meta-analysis protocols (PRISMA-P) 2015: elaboration and explanation. Bmj. 2015;349:g7647.
[12] Lundell LR, Dent J, Bennett JR, et al. Endoscopic assessment of oesophagitis: clinical and
M

functional correlates and further validation of the Los Angeles classification. Gut. 1999;45:172-80.
[13] Higgins J, Version 5.0.2 [updated September 2009] - The Cochrane Collaboration. [accessed
D

April 18, 2008]. Available at http://www.cochrane-handbook.org.

[14] Viechtbauer W, Cheung MWL. Outlier and influence diagnostics for meta-analysis. Res Synth
TE

Methods 2010;1:112-25.
[15] Egger M, Davey Smith G, Schneider M, et al. Bias in meta-analysis detected by a simple,
EP

graphical test. Bmj. 1997;315:629-34.

[16] Duval S, Tweedie R. Trim and fill: A simple funnel-plot-based method of testing and adjusting
for publication bias in meta-analysis. Biometrics. 2000;56:455-63.
C

[17] Viechtbauer W. Conducting meta-analyses in R with the metafor package. J Stat Softw
AC

2010;36:1-48.
[18] Li CJ, Tan YY, Wang XH, et al. Peroral endoscopic myotomy for achalasia in patients aged
>/= 65 years. World journal of gastroenterology: WJG. 2015;21:9175-81.

14
 ACCEPTED MANUSCRIPT
[19] Hungness ES, Sternbach JM, Teitelbaum EN,et al. Per-oral Endoscopic Myotomy (POEM)
After the Learning Curve: Durable Long-term Results With a Low Complication Rate. Annals of
surgery. 2016;264:508-17.
[20] Hungness ES, Teitelbaum EN, Santos BF, et al. Comparison of perioperative outcomes
between peroral esophageal myotomy (POEM) and laparoscopic Heller myotomy. J Gastrointest
Surg. 2013;17:228-35.

PT
[21] Jones EL, Meara MP, Schwartz JS, et al. Gastroesophageal reflux symptoms do not correlate
with objective pH testing after peroral endoscopic myotomy. Surgical endoscopy. 2016;30:947-52.

RI
[22] Ujiki MB, Yetasook AK, Zapf M, et al. Peroral endoscopic myotomy: A short-term
comparison with the standard laparoscopic approach. Surgery. 2013;154:893-7; discussion 7-900.

SC
[23] Shiwaku H, Inoue H, Sasaki T, et al. A prospective analysis of GERD after POEM on anterior
myotomy. Surg Endosc. 2016 Jun;30(6):2496-504.

U
[24] Von Renteln D, Fuchs KH, Fockens P, et al. Peroral endoscopic myotomy for the treatment of
achalasia: an international prospective multicenter study. Gastroenterology. 2013;145:309-11 e1-3.
AN
[25] von Renteln D, Inoue H, Minami H, et al. Peroral endoscopic myotomy for the treatment of
achalasia: a prospective single center study. The American journal of gastroenterology.
M

2012;107:411-7.
[26] Chan SM, Wu JC, Teoh AY, et al. Comparison of early outcomes and quality of life after
D

laparoscopic Heller's cardiomyotomy to peroral endoscopic myotomy for treatment of achalasia.

TE

Digestive endoscopy : official journal of the Japan Gastroenterological Endoscopy Society.

2016;28:27-32.
[27] Ramchandani M, Nageshwar Reddy D, Darisetty S, et al. Peroral endoscopic myotomy for
EP

achalasia cardia: Treatment analysis and follow up of over 200 consecutive patients at a single
center. Digestive endoscopy: official journal of the Japan Gastroenterological Endoscopy Society.
C

2016;28:19-26.
AC

[28] Minami H, Isomoto H, Yamaguchi N, et al. Peroral endoscopic myotomy for esophageal
achalasia: clinical impact of 28 cases. Digestive endoscopy: official journal of the Japan
Gastroenterological Endoscopy Society. 2014;26:43-51.

15
 ACCEPTED MANUSCRIPT
[29] Ling TS, Guo HM, Yang T, et al. Effectiveness of peroral endoscopic myotomy in the
treatment of achalasia: a pilot trial in Chinese Han population with a minimum of one-year follow-
up. J Dig Dis. 2014;15:352-8.
[30] Sharata AM, Dunst CM, Pescarus R, et al. Peroral endoscopic myotomy (POEM) for
esophageal primary motility disorders: analysis of 100 consecutive patients. J Gastrointest Surg.
2015;19:161-70; discussion 70.

PT
[31] Familiari P, Greco S, Gigante G, et al. Gastroesophageal reflux disease after peroral
endoscopic myotomy: Analysis of clinical, procedural and functional factors, associated with

RI
gastroesophageal reflux disease and esophagitis. Digestive endoscopy : official journal of the Japan
Gastroenterological Endoscopy Society. 2016;28:33-41.

SC
[32] Hu JW, Li QL, Zhou PH, et al. Peroral endoscopic myotomy for advanced achalasia with
sigmoid-shaped esophagus: long-term outcomes from a prospective, single-center study. Surgical
endoscopy. 2015;29:2841-50.

U
[33] Kumagai K, Tsai JA, Thorell A, et al. Per-oral endoscopic myotomy for achalasia. Are results
AN
comparable to laparoscopic Heller myotomy? Scandinavian journal of gastroenterology.
2015;50:505-12.
M

[34] Inoue H, Sato H, Ikeda H, et al. Per-Oral Endoscopic Myotomy: A Series of 500 Patients.
Journal of the American College of Surgeons. 2015;221:256-64.
D

[35] Bhayani NH, Kurian AA, Dunst CM, et al. A comparative study on comprehensive, objective
outcomes of laparoscopic Heller myotomy with per-oral endoscopic myotomy (POEM) for
TE

achalasia. Annals of surgery. 2014;259:1098-103.

[36] Boeckxstaens GE, Annese V, des Varannes SB, et al. Pneumatic dilation versus laparoscopic
EP

Heller's myotomy for idiopathic achalasia. The New England journal of medicine. 2011;364:1807-
16.
[37] Di Martino N, Brillantino A, Monaco L, et al. Laparoscopic calibrated total vs partial
C

fundoplication following Heller myotomy for oesophageal achalasia. World journal of

AC

gastroenterology : WJG. 2011;17:3431-40.

[38] Gulpinar K, Celasin H, Sozener U, et al. The therapeutic results after laparoscopic hellers'
myotomy and partial fundoplication for achalasia. The Turkish journal of gastroenterology : the
official journal of Turkish Society of Gastroenterology. 2014;25 Suppl 1:54-8.

16
 ACCEPTED MANUSCRIPT
[39] Hamdy E, El Nakeeb A, El Hanfy E, et al. Comparative Study Between Laparoscopic Heller
Myotomy Versus Pneumatic Dilatation for Treatment of Early Achalasia: A Prospective
Randomized Study. Journal of laparoendoscopic & advanced surgical techniques Part A.
2015;25:460-4.
[40] Kostic S, Kjellin A, Ruth M, et al. Pneumatic dilatation or laparoscopic cardiomyotomy in the
management of newly diagnosed idiopathic achalasia. Results of a randomized controlled trial.

PT
World journal of surgery. 2007;31:470-8.
[41] Novais PA, Lemme EM. 24-h pH monitoring patterns and clinical response after achalasia

RI
treatment with pneumatic dilation or laparoscopic Heller myotomy. Alimentary pharmacology &
therapeutics. 2010;32:1257-65.

SC
[42] Rawlings A, Soper NJ, Oelschlager B, et al. Laparoscopic Dor versus Toupet fundoplication
following Heller myotomy for achalasia: results of a multicenter, prospective, randomized-
controlled trial. Surgical endoscopy. 2012;26:18-26.

U
[43] Rebecchi F, Giaccone C, Farinella E, et al. Randomized controlled trial of laparoscopic Heller
AN
myotomy plus Dor fundoplication versus Nissen fundoplication for achalasia: long-term results.
Annals of surgery. 2008;248:1023-30.
M

[44] Richards WO, Torquati A, Holzman MD, et al. Heller myotomy versus Heller myotomy with
Dor fundoplication for achalasia: a prospective randomized double-blind clinical trial. Annals of
D

surgery. 2004;240:405-12; discussion 12-5.

[45] Salvador R, Pesenti E, Gobbi L, et al. Postoperative Gastroesophageal Reflux After
TE

Laparoscopic Heller-Dor for Achalasia: True Incidence with an Objective Evaluation. J Gastrointest
Surg. 2017;21:17-22.
EP

[46] Khajanchee YS, Kanneganti S, Leatherwood AE, et al. Laparoscopic Heller myotomy with
Toupet fundoplication: outcomes predictors in 121 consecutive patients. Archives of surgery
(Chicago, Ill : 1960). 2005;140:827-33; discussion 33-4.
C

[47] Oelschlager BK, Chang L, Pellegrini CA. Improved outcome after extended gastric myotomy
AC

for achalasia. Archives of surgery (Chicago, Ill : 1960). 2003;138:490-5; discussion 5-7.
[48] Pechlivanides G, Chrysos E, Athanasakis E, et al. Laparoscopic Heller cardiomyotomy and
Dor fundoplication for esophageal achalasia: possible factors predicting outcome. Archives of
surgery (Chicago, Ill : 1960). 2001;136:1240-3.

17
 ACCEPTED MANUSCRIPT
[49] Ramacciato G, D'Angelo FA, Aurello P, et al. Laparoscopic Heller myotomy with or without
partial fundoplication: a matter of debate. World journal of gastroenterology : WJG. 2005;11:1558-
61.
[50] Rossetti G, Brusciano L, Amato G, et al. A total fundoplication is not an obstacle to esophageal
emptying after heller myotomy for achalasia: results of a long-term follow up. Annals of surgery.
2005;241:614-21.

PT
[51] Smith CD, Stival A, Howell DL, et al. Endoscopic therapy for achalasia before Heller
myotomy results in worse outcomes than heller myotomy alone. Annals of surgery. 2006;243:579-

RI
84; discussion 84-6.
[52] Vogt D, Curet M, Pitcher D, et al. Successful treatment of esophageal achalasia with

SC
laparoscopic Heller myotomy and Toupet fundoplication. American journal of surgery.
1997;174:709-14.
[53] Perrone JM, Frisella MM, Desai KM,et al. Results of laparoscopic Heller-Toupet operation for

U
achalasia. Surgical endoscopy. 2004;18:1565-71.
AN
[54] Douard R, Gaudric M, Chaussade S, et al. Functional results after laparoscopic Heller
myotomy for achalasia: A comparative study to open surgery. Surgery. 2004;136:16-24.
M

[55] Wills VL, Hunt DR. Functional outcome after Heller myotomy and fundoplication for
achalasia. J Gastrointest Surg. 2001;5:408-13.
D

[56] Ackroyd R, Watson DI, Devitt PG, et al. Laparoscopic cardiomyotomy and anterior partial
fundoplication for achalasia. Surgical endoscopy. 2001;15:683-6.
TE

[57] Teitelbaum EN, Rajeswaran S, Zhang R, et al. Peroral esophageal myotomy (POEM) and
laparoscopic Heller myotomy produce a similar short-term anatomic and functional effect. Surgery.
EP

2013;154:885-91; discussion 91-2.

[58] Kumagai K, Kjellin A, Tsai JA, et al. Toupet versus Dor as a procedure to prevent reflux after
cardiomyotomy for achalasia: results of a randomised clinical trial. International journal of surgery
C

(London, England). 2014;12:673-80.

AC

[59] Balakrishna P, Parshad R, Rohila J, et al. Symptomatic outcome following laparoscopic

Heller's cardiomyotomy with Dor fundoplication versus laparoscopic Heller's cardiomyotomy with
angle of His accentuation: results of a randomized controlled trial. Surgical endoscopy.
2015;29:2344-51.

18
 ACCEPTED MANUSCRIPT
[60] Gholoum S, Feldman LS, Andrew CG, et al. Relationship between subjective and objective
outcome measures after Heller myotomy and Dor fundoplication for achalasia. Surgical endoscopy.
2006;20:214-9.
[61] Simic AP, Radovanovic NS, Skrobic OM, et al. Significance of limited hiatal dissection in
surgery for achalasia. J Gastrointest Surg. 2010;14:587-93.
[62] Sharma P, Wani S, Romero Y, et al. Racial and geographic issues in gastroesophageal reflux

PT
disease. The American journal of gastroenterology. 2008;103:2669-80.
[63] Dent J, El-Serag HB, Wallander MA, et al. Epidemiology of gastro-oesophageal reflux disease:

RI
a systematic review. Gut. 2005;54:710-7.
[64] Li CJ, Tan YY, Wang XH, Liu DL. Peroral endoscopic myotomy for achalasia in patients aged

SC
≥ 65 years. World J Gastroentero. 2015;21:9175-81.
[65] Hungness ES, Sternbach JM, Teitelbaum EN, et al. Per-oral endoscopic myotomy (POEM)
after the learning curve: Durable long-term results with a low complication rate. Annals of surgery.
2016;264:508-15.
U
AN
[66] Jones EL, Meara MP, Schwartz JS, et al. Gastroesophageal reflux symptoms do not correlate
with objective pH testing after peroral endoscopic myotomy. Surgical Endoscopy and Other
M

Interventional Techniques. 2016;30:947-52.

[67] Shiwaku H, Inoue H, Yamashita K, et al. A prospective analysis of GERD after POEM on
D

anterior myotomy. Gastrointestinal endoscopy. 2015;81:AB487.

[68] Von Renteln D, Inoue H, Minami H, et al. Peroral endoscopic myotomy for the treatment of
TE

achalasia: A prospective single center study. American Journal of Gastroenterology. 2012;107:411-

7.
EP

[69] Chan SM, Wu JCY, Teoh AYB, et al. Comparison of early outcomes and quality of life after
laparoscopic Heller's cardiomyotomy to peroral endoscopic myotomy for treatment of achalasia.
C

Digest Endosc. 2016;28:27-32.

[70] Ramchandani M, Reddy DN, Darisetty S, et al. Peroral endoscopic myotomy for achalasia
AC

cardia: Treatment analysis and follow up of over 200 consecutive patients at a single center. Digest
Endosc. 2016;28:19-26.
[71] Minami H, Isomoto H, Yamaguchi N, et al. Peroral endoscopic myotomy for esophageal
achalasia: Clinical impact of 28 cases. Digest Endosc. 2014;26:43-51.

19
 ACCEPTED MANUSCRIPT
[72] Ling TS, Guo HM, Yang T, et al. Effectiveness of peroral endoscopic myotomy in the
treatment of achalasia: A pilot trial in Chinese Han population with a minimum of one-year follow-
up. J Digest Dis. 2014;15:352-8.
[73] Sharata AM, Dunst CM, Pescarus R, et al. Peroral endoscopic myotomy (POEM) for
esophageal primary motility disorders: analysis of 100 consecutive patients. Journal of
gastrointestinal surgery: official journal of the Society for Surgery of the Alimentary Tract.

PT
2015;19:161-70.
[74] Kumagai K, Tsai JA, Thorell A, et al. Per-oral endoscopic myotomy for achalasia. Are results

RI
comparable to laparoscopic Heller myotomy? Scandinavian journal of gastroenterology.
2015;50:505-12.

SC
[75] Inoue H, Sato H, Ikeda H, et al. Per-Oral Endoscopic Myotomy: A Series of 500 Patients.
Journal of the American College of Surgeons. 2015;221:256-64.

U
[76] Zak Y, Meireles OR, Rattner DW Laparoscopic Toupet fundoplication for GERD after poem
AN
Surgical Endoscopy and Other Interventional Techniques 2015 29 SUPPL. 1 (S366).
M
D
TE
C EP
AC

20
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC
 ACCEPTED MANUSCRIPT

Abbreviations used: POEM = Per-Oral Endoscopic Myotomy; LHM = Heller’s myotomy; LES=
Lower Esophageal Sphincter; TRT = total reflux time; PPI = proton-pump inhibitor.

PT
RI
U SC
AN
M
D
TE
C EP
AC
 ACCEPTED MANUSCRIPT

Appendix 1

Data sources and search strategy

We performed a comprehensive literature search by using PubMed, EMBASE, SCOPUS,
Google Scholar and the Cochrane Central Register of Clinical Trials (up to February 28th 2017) to

PT
identify full articles evaluating the incidence of gastro-esophageal reflux symptoms, endoscopic
findings and pH-monitoring alterations following both POEM and Heller’s myotomy. In order not

RI
to miss any relevant article, we preferred to perform two different systematic search of the
literature, namely one for POEM and one for LHM with fundoplication. Studies with LHM without

SC
anti-reflux procedures were excluded. The search was restricted to articles published in English.
Electronic searches were supplemented by manual searches of references of included studies and
review articles.

U
AN
Search strategy
M

POEM
We identified studies using the following medical subject headings (MeSH) and keywords
including: “esophagus”, “POEM” and “peroral endoscopic myotomy”.
D
TE

In detail:
• The PubMed search strategy was: POEM [all fields] or peroral endoscopic myotomy [all
EP

fields]
• The EMBASE search strategy was: (POEM:ab,ti OR 'peroral endoscopic myotomy':ab,ti OR
C

'esophagus myotomy'/exp) AND ('gastroesophageal reflux'/exp OR GERD:ab,ti OR

reflux:ab,ti OR 'esophagitis'/exp OR (pH:ab,ti AND impedance:ab,ti AND monitoring:ab,ti)
AC

OR 'pH-metry':ab,ti).

LHM
In detail:
• The PubMed search strategy was: myotomy[All Fields] and esophagus[MesH]
• The EMBASE search strategy was: ('cardioesophagomyotomy'/exp OR 'Heller
myotomy':ab,ti) AND ('gastroesophageal reflux'/exp OR GERD:ab,ti OR reflux:ab,ti OR
 ACCEPTED MANUSCRIPT
'esophagitis'/exp OR (pH:ab,ti AND impedance:ab,ti AND monitoring:ab,ti) OR 'pH-
metry':ab,ti)

Selection process
Three review authors (LF; DM; FM) independently screened the titles and abstracts yielded
by the search against the inclusion criteria. Full reports were obtained for all titles that appeared to

PT
meet the inclusion criteria or where there was any uncertainty. Review author pairs then screened
the full text and decided whether these met the inclusion criteria. Disagreements were resolved

RI
through discussion of all the authors. The reasons for excluding trials were recorded. When there
were multiple articles for a single study, we used the latest publication and supplemented it, if

SC
necessary, with data from the more complete version. Abstracts published in the proceedings of
international congresses were not included, because they did not provide enough information and
details on the methods and/or results. Case series with less than ten patients were also excluded in

U
order to reduce the connected bias.
AN
Data extraction
M

Using standardized forms, two reviewers (LF, DM) extracted data independently and in
duplicate from each eligible study. Reviewers resolved disagreements by discussion, and the
D

arbitrators (CH and AR) reconciled disagreements. The following data were extracted for each
study including the publication status, year of publication, study sample size and design
TE

(prospective, retrospective), the geographic origin of the study (i.e., hospital, city and country), the
number of centres involved, the enrolment period, patient characteristics (mean age of the study
EP

population and sex), indication for POEM/LHM and classification of esophageal motility disorders
according to the Chicago classification (type I, type II and type III), previous treatment of the
esophageal motility disorder (i.e. endoscopic treatments, including balloon dilation and botox
C

injection, and surgical treatments), length of myotomy and esophageal wall in which POEM was
AC

performed (i.e. anterior or posterior wall), type of fundoplication intervention (i.e.Toupet’s,

Nissen’s, Dor’s fundoplication), number of patients in follow up, mean period of follow up
(months), number of patients developing gastro-esophageal reflux disease based on symptoms of
reflux disease, endoscopic findings (i.e. esophagitis and its severity classified according to Los
Angeles classification), abnormal pH-metry evaluation, number of patients needing of proton pump
inhibitor (PPI) after the procedure because of the new onset of GERD symptoms.
 ACCEPTED MANUSCRIPT

Quality assessment
Quality was assessed by the tool developed by Moga et al [A]. Two reviewers (LF, DM)
assessed quality measures for included studies and discrepancies were adjudicated by collegial

PT
discussion.

RI
References

SC
A. Moga, C., Guo, B., Schopflocher, D., and Harstall, C. Development of a quality appraisal tool
for case series studies using a modified Delphi technique. (Available at

U
http://www.ihe.ca/advanced-search/development-of-a-quality-appraisal-tool-for-case-series-studies-
AN
using-a-modified-delphi-technique. Accessed September 14, 2015) Institute of Health Economics,
Edmonton; 2012
M
D
TE
C EP
AC
 ACCEPTED MANUSCRIPT
Appendix 2

Figure A) Flow chart of the study selection process for the POEM procedure. B) Flow
chart of the study selection process for the Heller’s procedure.

PT
RI
U SC
AN
M
D
TE
C EP
AC
 ACCEPTED MANUSCRIPT

Table 1 – POEM. Baseline characteristics of the study population. Only

prospective studies were included. Abbreviation used: uk = unknown.

PT
Study
Enrolment Male Mean Prior Balloon- Prior Heller Prior Botox
Reference Country Population
period (%) Age dilation (n) Myotomy (n) treatment (n)

RI
(N)
Li, et al [18] China 2011-2014 15 40% 67 4 1 uk

SC
Hungness, et al [19] USA 2010-2012 18 72% 38 uk uk uk
Hungness, et al [20] USA 2012-2015 112 61% 52.9 21 0 13

U
Jones, et al. [21] USA 2012-2014 43 63% 53.5 uk uk uk
Ujiki, et al [22] USA 2011-2013 18 72% 64.1 4 3 4

AN
Shiwahu, et al [23] Japan 2011-2014 105 59% 48.8 14 0 0
von Renteln, et al [25] Germany uk 16 75% 45 10 0 1

M
Germany;
Netherlands,
von Renteln, et al [24] Canada, uk 70 57% 45 0 0 0

D
Switzerland
Chan, et al [26] China 2010-2014 33 36% 48.2 uk uk uk
Ramchandani, et al [27]
Minami, et al [28]
India
Japan TE
2013-2014
2010-2012
220
28
53%
32%
39.3
52.2
77
1
10
0
1
0
EP
Ling, et al [29] China 2010-2012 87 46% 41.6 0 0 0
Sharata, et al [30] USA 2010-2013 100 51% 58 5 5 20
C

Familiari, et al [31] Italy 2011-2014 103 46% 46.6 20 3 uk

AC

Hu, et al [32] China 2010-2012 32 53% 43.6 14 3 3

Kumagai, et al [33] Sweden 2012-2014 42 64% 45.5 12 3 0
Inoue, et al [34] Japan 2008-2013 500 43% 43 179 10 6
 ACCEPTED MANUSCRIPT

Table 2 – POEM. Details on the type of achalasia (according to the Chicago

classification), POEM approach, length of myothomy and follow-up.
Abbreviation used: uk = unknown.
%Patients
% Achalasia % Achalasia % Achalasia
Duration with at %Patie

PT
Type I Type II Type III %Achalasia % Anterior Myotomy
Reference Follow Up least one on PP
Chicago Chicago Chicago Unclassified Approach length (cm)
(months) assessment post-PO
Classification Classification Classification
post-POEM

RI
Li, et al [18] uk uk uk uk 0% 9 18 100% 6.7%

SC
Hungness, et al [19] 33.3% 61.1% 5.5% 0% 100% 9 2 100% uk
Hungness, et al [20] 22.3% 51.8% 17.9% 8% 100% 10,7 24 100% 7.1%

U
Jones, et al. [21] uk uk uk uk 100% uk 6 60.5% 16.3%

AN
Ujiki, et al [22] uk uk uk uk 100% 11,2 3,6 100% 27.8%
Shiwahu, et al [23] 5.7% 52.4% 8.6% 0% 100% 12,6 3 66.7% 7.6%
von Renteln, et al [25] uk uk uk uk 100% 12 3 100% 6.3%

M
von Renteln, et al [24] uk uk uk uk 100% 13 3 100% 17.1%
Chan, et al [26]

D
uk uk uk uk 100% uk 6 100% 9.1%
Ramchandani, et al [27] 30.5% 65.9% 3.6% 2.3% 86.4% 12 13,4 46.4% uk

TE
Minami, et al [28] uk uk uk uk 100% 14,4 3 100% 21.4%
Ling, et al [29] uk uk uk uk 0% uk 3 100% 5.7%
EP
Sharata, et al [30] 30% 43% 2% 25% 100% 8 21,5 81% uk
Familiari, et al [31] 25.2% 44.7% 1.9% 28.1% 97% 12,9 7,6 100% uk
C

Hu, et al [32] uk uk uk uk 0% 10,3 30 96.8% 25%

AC

Kumagai, et al [33] uk uk uk uk 100% 8 12 100% uk

Inoue, et al [34] 11.8% 12.4% 5.4% 70.4% 96.4% 14 2 84.6% 2.6%
 ACCEPTED MANUSCRIPT

Table 3 – LHM. Baseline characteristics of the study population. Only

prospective studies were included. Abbreviation used: uk = unknown.
Study
Enrolment Male Mean Prior Balloon- Prior Heller Prior Botox
Reference Country Population
period (%) Age dilation (n) Myotomy (n) treatment (n)
(N)

PT
Bhayani, et al [35] USA 2007-2012 64 48% 57 0 0 0
Belgium;

RI
Boeckxstaens, et al [36] Italy; France; 2003-2008 106 54% 45.5 0 0 0
Spain; Netherlands
Di Martino, et al [37] Italy 2002-2007 56 46% 42.8 0 0 0

SC
Gulpinar, et al [38] Turkey 2002-2012 40 45% 41 31 0 0
Hamdi, et al [39] Egypt 2005-2010 25 32% 32 0 0 0

U
Kostic, et al [40] Sweden 2000-2005 25 44% 43 0 0 0

AN
Novais, et al [41] Brazil 2005-2009 47 38% 46.5 0 0 0
Rawlings, et al [42] USA 2003-2008 60 57% 48.7 uk 0 uk

M
Rebecchi, et al [43] Italy 1993-2002 144 52% 49 12 0 7
Richards, et al [44] USA 2000-2003 22 45% 50 uk 0 uk

D
Salvador, et al [45] Italy 1992-2015 806 54% 44 uk 0 uk

TE
Simic, et al [61] Serbia 2002-2005 26 uk uk 0 0 0
Ujiki, 2013[22] USA 2009-2013 21 57% 60.2 8 3 2
Khajancee, et al [46] USA 1996-2004 121 40% 46.4 10 0 8
EP
Oelschlager, et al [47] USA 1994-1998 110 uk uk 8 0 16
Pechlivanides, et al [48] Greece 1995-1998 29 41% 47.5 29 0 0
C

Ramacciato, et al [49] Italy 1997-2004 17 uk 42 0 0 0

AC

Rossetti, et al [50] Italy 1992-2003 195 47% 45.2 21 0 0

Smith, et al [51] USA 1994-2003 209 49% 47 100 0 33
Vogt, et al [52] USA 1993-1997 20 60% 42 3 0 3
Perrone, et al [53] USA 1996-2003 100 61% 47 24 0 25
Douard, et al [54] France 1993-2008 52 46% 37 28 0 3
 ACCEPTED MANUSCRIPT

Wills, et al [55] Australia 1992-1999 62 47% 47.4 39 0 1

Ackroyd, et al [56] Australia 1992-2000 82 57% 47 0 0 0
Teitelbaum, et al [57] USA 2004-2011 17 59% 51 uk 0 uk
Kumagai, et al
Sweden 2005-2011 41 46% 44.5 2 0 1

PT
[58_ENREF_37]
Balakrishna, et al [59] India 2010-2013 31 52% 35 10 0 0

RI
Ghouloum, et al [60] Canada 1999-2004 53 47% 51 9 0 4

U SC
AN
M
D
TE
C EP
AC
 ACCEPTED MANUSCRIPT

Table 4 – LHM. Details on the type of achalasia (according to the Chicago

classification), fundoplication type, length of myothomy and follow-up.
Abbreviation used: uk = unknown.
%Patients
% Achalasia % Achalasia % Achalasia
% Toupet’s % Dor’s % Nissen’s Duration with at %Pa
Type I Type II Type III Myotomy

PT
Reference fundoplicatio fundoplicatio fundoplicatio Follow Up least one on
Chicago Chicago Chicago length (cm)
n n n (months) assessment post-
Classification Classification Classification
post-LHM

RI
Bhayani, et al [35] uk uk uk 27 37 0 uk 6.8 59% u
Boeckxstaens, et al 0 106 0

SC
[36] uk uk uk 7.5 12 67% u
Di Martino, et al 0 30 26
uk 24 96% 7
[37] uk uk uk

U
Gulpinar, et al [38] 34 6 0
uk uk uk 5.5 6 100% 7

AN
Hamdi, et al [39] uk uk uk 0 25 0 8 48 100% u
Kostic, et al [40] uk uk uk 25 0 0 uk 12 100% u

M
Novais, et al [41] uk uk uk 0 47 0 8 1.8 91% u
Rawlings, et al [42] uk uk uk 24 36 0 8.7 7 100% u

D
Rebecchi, et al [43] uk uk uk 0 72 72 8.5 12 96% u
0 22 0

TE
Richards, et al [44] uk uk uk 6.5 6 100% u
Salvador, et al [45] 152 126 18 0 806 0 9 49 57% u
Simic, et al [61] 0 26 0
EP
uk uk uk 9.5 36 100% 23
Ujiki, 2013[22] uk uk uk 9 12 0 10 5.4 100% 19
Khajancee, et al [46] uk uk uk 121 0 0 7.5 9 49% u
C

Oelschlager, et al 58 52 0
AC

[47] uk uk uk uk 46 75% 17
Pechlivanides, et al 0 29 0
uk uk uk 6.5 12 72%
[48] u
Ramacciato, et al 0 17 0
8 12 100%
[49] uk uk uk u
Rossetti, et al [50] 0 0 195 7.5 83 8%
uk uk uk u
 ACCEPTED MANUSCRIPT

Smith, et al [51] 181 25 0 9 21 98%

uk uk uk u
Vogt, et al [52] 18 0 0 uk 12 90%
uk uk uk u
Perrone, et al [53] 94 6 0 9 26 92%
uk uk uk u

PT
Douard, et al [54] 0 52 0 8 50 100%
uk uk uk u
Wills, et al [55] uk uk uk 8.2 38 100%

RI
uk uk uk u
Ackroyd, et al [56] 0 78 0 6 24 95%
uk uk uk u

SC
Teitelbaum, et al 14 0 3
3 10 2 8.5 43 94%
[57] u
Kumagai, et al [58] 22 19 0 7 6.3 58%

U
uk uk uk u
Balakrishna, et al 0 31 0

AN
8 21 100% u
[59] uk uk Uk
Ghouloum, et al [60] uk uk uk 0 53 0 6.5 3 7% u

M
D
TE
C EP
AC
 ACCEPTED MANUSCRIPT

Table 5 – Quality assessment of case series studies evaluating the POEM

procedure; the 18-item quality assessment tool proposed by Moga et al were used
(see below for details).

PT
Item
Reference 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

RI
Li, 2015[18] Y Y N Y Y N Y Y Y Y Y Y Y Y N Y Y Y
Hungness, 2013[19] Y Y N Y U Y Y Y Y Y Y Y Y Y N Y Y Y

SC
Hungness, 2016[20] Y Y N Y Y N Y Y Y Y Y Y Y Y N Y Y Y

U
Jones, 2015[21] Y Y N Y Y N Y Y Y Y Y Y Y Y N Y Y Y

AN
Ujiki, 2013[22] Y Y N Y U N Y Y Y Y Y Y Y Y N Y Y Y
Shiwahu, 2016[23] Y Y N Y U N Y Y Y Y Y Y Y Y N Y Y Y

M
von Renteln, 2012[25] Y Y N Y Y N Y Y Y Y Y Y Y Y N Y Y Y

D
von Renteln, 2013[24] Y Y Y Y Y Y Y Y Y Y Y Y Y Y N Y Y Y

TE
Chan, 2016[26] Y Y N Y U N Y Y Y Y Y Y Y Y N Y Y Y

Ramchandani, 2016[27] Y Y N Y Y N Y Y Y Y Y Y Y Y N Y Y Y
EP
Minami, 2014[28] Y Y N Y Y N Y Y Y Y Y Y Y Y N Y Y Y
C

Ling, 2014[29] Y Y N Y U Y Y Y Y Y Y Y Y Y N Y Y Y
AC

Sharata, 2014[30] Y Y N Y Y N Y Y Y Y Y Y Y Y N Y Y Y
Familiari, 2016[31] Y Y N Y Y N Y Y Y Y Y Y Y Y N Y Y Y
Hu, 2015[32] Y Y N Y Y N Y Y Y Y Y Y Y Y N Y Y Y
Kumagai, 2014[33] Y Y N Y Y N Y Y Y Y Y Y Y Y N Y Y Y
 ACCEPTED MANUSCRIPT

Inoue, 2015[34] Y Y N Y Y N Y Y Y Y Y Y Y Y N Y Y Y

Table 6 – Quality assessment of case series studies evaluating the Heller’s

myotomy with fundoplication; the 18-item quality assessment tool proposed by

PT
Moga et al were used (see below for details).

RI
U SC
AN
M
D
TE
C EP
AC
 ACCEPTED MANUSCRIPT Item

Reference 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

Bhayani, 2014[35] Y Y N Y Y Y Y Y Y Y Y Y Y Y N Y Y Y
Boeckxstaens, 2011[36] Y Y Y Y Y Y Y Y Y Y Y Y Y Y N Y Y Y
Di Martino, 2011[37] Y Y N Y U Y Y Y Y Y Y Y Y Y N Y Y N

PT
Gulpinar, 2014[38] Y Y N Y U N Y Y Y Y Y Y Y Y N Y Y Y

RI
Hamdi, 2015[39] Y Y N Y U Y Y Y Y Y Y Y Y Y N Y Y Y

SC
Kostic, 2007[40] Y Y N Y U Y Y Y Y Y Y Y Y Y N Y Y N
Novais, 2010[41] Y Y N Y U Y Y Y Y Y Y Y Y Y N Y Y Y

U
Rawlings, 2011[42] Y Y Y Y U U Y Y Y Y Y Y Y Y N Y Y Y

AN
Rebecchi, 2008[43] Y Y N Y U N Y Y Y Y Y Y Y Y N Y Y N
Richards, 2004[44] Y Y N Y U U Y Y Y Y Y Y Y Y N Y Y Y

M
Salvador, 2017[45] Y Y N Y Y U Y Y Y Y Y Y Y Y N Y Y N

D
Simic, 2010[61] Y Y N Y Y Y Y Y Y Y Y Y Y Y N Y Y Y

TE
Ujiki, 2013[22] Y Y N Y U N Y Y Y Y Y Y Y Y N Y Y N
Khajancee, 2005[46] Y Y N Y Y N Y Y Y Y Y Y Y Y N Y Y N
EP
Oelschlager, 2003[47] Y Y N Y Y N Y Y Y Y Y Y Y Y N Y Y N
C

Pechlivanides, 2001[48] Y Y N Y U N Y Y Y Y Y Y Y Y N Y Y N
AC

Ramacciato, 2005[49] Y Y N Y U Y Y Y Y Y Y Y Y Y N Y Y N
Rossetti, 2005[50] Y Y N Y U N Y Y Y Y Y Y Y Y N Y Y N
Smith, 2006[51] Y Y N Y U N Y Y Y Y Y Y Y Y N Y Y N
Vogt, 1997[52] Y Y N Y U N Y Y N N N N Y Y N Y Y N
 ACCEPTED MANUSCRIPT

Perrone, 2004[53] Y Y N Y Y N Y Y Y Y Y Y Y Y N Y Y N
Douard, 2004[54] Y Y N Y Y N Y Y Y Y Y Y Y Y N Y Y N
Wills, 2001[55] Y Y N Y Y N Y Y Y Y Y Y Y Y N Y Y N

PT
Ackroyd, 2001[56] Y Y N Y U Y Y Y Y Y Y Y Y Y N Y Y N
Y Y N Y U N Y Y Y Y Y Y Y Y N Y Y N

RI
Teitelbaum, 2013[57]
Kumagai, 2014[58] Y Y N Y U N Y Y Y Y Y Y Y Y N Y Y Y

SC
Balakrishna, 2014[59] Y Y N Y U N Y Y Y Y Y Y Y Y N Y Y Y
Ghouloum, 2006[60] Y Y N Y Y N Y Y Y Y Y Y Y Y N Y Y N

U
AN
Development Organization Major Components Judgment

M
D
TE
C EP
AC
 ACCEPTED MANUSCRIPT

Carmen Moga and his 1. Is the hypothesis/aim/objective of the study clearly stated? 1. Yes, Unclear, No
colleagues* 2. Are the characteristics of the participants included in the study described? 2. Yes, Partially reported, No
3. Were the cases collected in more than one centre? 3. Yes, Unclear, No
4. Are the eligibility criteria (i.e. inclusion and exclusion criteria) for entry into the study clearly 4. Yes, Partially reported, No
stated?

PT
5. Were participants recruited consecutively? 5. Yes, Unclear, No
6. Did participants enter the study at a similar point in the disease? 6. Yes, Unclear, No
7. Yes, Partially reported, No

RI
7. Was the intervention of interest clearly described?
8. Were additional interventions (co-interventions) reported in the study? 8. Yes, Unclear, No
9. Are the outcome measures established a priori? 9. Yes, Partially reported, No

SC
10. Were the relevant outcomes measured with appropriate objective and/or subjective 10. Yes, Unclear, No
methods? 11. Yes, Unclear, No
11. Were the relevant outcomes measured before and after the intervention?

U
12. Yes, Unclear, No
12. Were the statistical tests used to assess the relevant outcomes appropriate? 13. Yes, Unclear, No

AN
13. Was the length of follow-up reported? 14. Yes, Unclear, No
14. Was the loss to follow-up reported? 15. Yes, Unclear or partially reported,
15. Does the study provide estimates of the random variability in the data analysis of relevant No

M
outcomes? 16. Yes, Partially reported, No
16. Are the adverse events related with the intervention reported? 17. Yes, Partially reported, No

D
17. Are the conclusions of the study supported by results? 18. Yes, Partially reported, No
18. Are both competing interests and sources of support for the study reported? 19.

TE
C EP
AC
 ACCEPTED MANUSCRIPT

Appendix 3

Relationship among the different modality of reflux assessment within each group of
treatment according to the multivariate mixed-effects meta-regression model.

POEM

PT
In the POEM cohort, the odds of detecting reflux disease by using symptoms evaluation
(reference group) were significantly lower than were those by using ph-monitoring (ORs, 2.92; 95%

RI
CI: 1.37-6.22; P<0.001) and those by using endoscopic findings (ORs, 2.00; 95% CI: 1.12-3.56;
P=0.065), while differences in detecting disease between ph-monitoring and endoscopic findings

SC
were not significant (P=0.269). According to the multivariate model, the estimated rates of disease
by using symptoms, ph-monitoring and endoscopic findings were 18.1% (95% CI: 13.1-24.4%),

U
39.3% (95% CI: 25.2-55.4%) and 30.7% (95% CI: 22.2-40.5%). (see Figure 3).
AN
LHM

The odds of reflux disease by using symptoms evaluation (reference group) were comparable
M

with those by using endoscopic findings (ORs, 1.06; 95% CI: 0.58-2.10; P=0.852) and only
marginally lower than were those by using esophageal ph-monitoring (ph vs. GERD, ORs, 1.78;
D

95% CI:0.93-3.44; P=0.086). However, ph-monitoring was significantly associated with an

increased likelihood of reflux disease after LHM as compared with endoscopic findings (ORs, 1.67;
TE

95% CI: 1.40-2.00; P<0.001). (see Figure 3). According to the multivariate model, the estimated
rates of disease by using symptoms, ph-monitoring and endoscopic findings were 8.6% (95%CI:
EP

5.4-13-3%), 14.9% (95%CI: 9.4-23.0) and 8.3% (95%CI: 3.8-17.0%).

C
AC
 ACCEPTED MANUSCRIPT

Table A. Post-POEM reflux according to symptoms: univariate random-effects meta-regression.

Variable Coefficient 95%-CI P value
Lower Limit Upper Limit
Mean patients age (ya) 0.0012 -0.0395 0.0315 0.825

PT
Publication Year (ya) -0.0093 -0.2112 0.1926 0.928
Geographic location

RI
(Asia vs. North America/ Europe) -0.4537 -0.8225 -0.0849 0.016
Mean follow-up period (mo) 0.0120 -0.0127 0.0367 0.342

SC
Min Follow-up (mo) 0.0128 -0.0327 0.0583 0.581
Max Follow-up (mo) 0.0022 -0.0149 0.0193 0.802
Female:male ratio (%) 0.4545 -0.9803 0.0713 0.090

U
Sample Size (#) -0.0004 -0.0019 0.0012 0.641

AN
Study population (% of treatment-naive patients) -0.0022 -1.2091 1.240 0.9971
Study Design

M
(Retrospective vs. Prospective) 0.4197 -0.4784 1.3179 0.3597
Loss to follow-up (%) 0.2202 -0.8355 1.2759 0.6827

D
Myotomy Length (cm) 0.0045 -0.1184 0.1274 0.942

TE
Orientation approach (% of anterior approaches) 0.4697 -0.2078 1.1472 0.174
PPI Treatments (%) 2.5266 -0.6107 5.6639 0.115
Preoperative botox (%) 1.2705 -2.0682 4.6091 0.4558
EP
Preoperative balloon dilation (%) 0.0957 -1.4218 1.2304 0.887
Preoperative Heller myotomy (%) 0.2830 -5.9445 6.5105 0.9290
C

% of Achalasia Type III (continuous variable) 3.2351 -0.6256 7.0957 0.090

AC

Achalasia Type III (categorical variable)

Low vs. High -0.5754 -1.1624 0.0116 0.055
NA vs. High -0.4312 -1.0317 0.1692 0.1593
 ACCEPTED MANUSCRIPT

Table B. Post-POEM reflux according to pH monitoring: univariate random-effects meta-regression.

Variable Coefficient 95%-CI Pvalue

Lower Limit Upper Limit

PT
Mean patients age (ya) -0.0283 -0.2558 0.1993 0.807
Publication Year (ya) 0.1929 -0.9165 1.3023 0.733

RI
Geographic location
(Asia vs. North America/ Europe) -2.1470 3.5061 -0.7879 0.002

SC
Mean follow-up period (mo) -0.0224 -0.1176 0.0729 0.645
Min follow-up (mo) -0.1977 -0.6791 0.2838 0.421
Max follow-up (mo) 0.0099 -0.0523 0.0721 0.754

U
Female:male ratio (%) -2.3513 -4.6111 -0.0916 0.041

AN
Sample Size (#) -0.0233 -0.0456 -0.0010 0.040
Study population

M
(% of treatment-naive patients) -10.0993 -19.5457 -0.6529 0.036
Study Design

D
(Retrospective vs. Prospective) 1.8963 0.4161 3.3765 0.012
Loss to follow-up (%) 1.3512 -1.8227 4.5252 0.404
Myotomy Length (cm)
Orientation approach TE
-0.0077 -0.6427 0.6273 0.980
EP
(% of anterior approaches) -23.1064 -42.0856 4.1272 0.017
PPI Treatments (%) 8.8919 -0.0635 17.8473 0.051
C

Preoperative botox (%) 12.6328 -0.3170 25.5827 0.056

AC

Preoperative balloon dilation (%) 4.6321 -4.2031 13.467 0.304

Preoperative Heller myotomy (%) 21.4594 -0.0528 42.9715 0.051
% of Achalasia Type III -0.0144 -0.1572 0.1285 0.843
 ACCEPTED MANUSCRIPT

Table C. Post-POEM reflux according to endoscopic findings: univariate random-effects meta-regression.

Moderator Coefficient 95% Confidence Interval Pvalue
Lower Limit Upper Limit
Mean patients age (ya) 0.0541 -0.051 0.1596 0.315

PT
Publication Year (ya) 0.2436 -0.2303 0.7174 0.314
Geographic origin

RI
(North America/Europe vs. Asia) 0.0204 -1.1053 1.1462 0.972
Mean follow-up period (mo) -0.0226 -0.0790 0.0338 0.432

SC
Min follow-up (mo) -0.0559 -0.1452 0.0334 0.2200
Max follow-up (mo) -0.0212 -0.0560 0.0136 0.232
the female:male ratio 0.7710 -0.4911 2.0331 0.231

U
Sample Size (#) 0.0027 -0.0011 0.0064 0.164

AN
Study population
(% of treatment-naive pts) 0.6832 -2.0731 3.4395 0.627

M
Study Design
(retrospective vs. prospective) 1.0572 -1.1574 3.2718 0.350

D
Loss to follow-up (%) 1.1939 -1.9002 4.2879 0.450

TE
Myotomy Length (cm) 0.1477 -0.1157 0.4111 0.272
Orientation approach (% of anterior approaches) 1.5337 0.0553 3.0121 0.042
PPI Treatments (%) -2.1604 -10.3438 6.0230 0.605
EP
Preoperative botox (%) -1.7888 -11.3246 7.7470 0.713
Preoperative balloon dilation (%) -0.3444 -3.0188 2.3301 0.801
C

Preoperative Heller myotomy (%) -8.1727 -28.5238 12.1784 0.431

AC

% of Achalasia Type III (as continuous variable) 1.4337 0.7240 2.1435 <0.001
Achalasia Type III (as categorical variable)
Low vs. High -1.3987 -2.3209 -0.4765 0.003
NA vs. High -1.3987 -2.3209 -0.4765 0.002
 ACCEPTED MANUSCRIPT

APPENDIX 4

Figure A. Funnel plot for publication bias. Meta-analysis of reflux symptoms after Heller’s
myotomy. A funnel plot shows the observed rates of reflux symptoms on the x-axis against
standard error on the y-axis. Empty circles were added to represent studies imputed by trim and fill.

PT
RI
U SC
AN
M
D
TE
C EP
AC
 ACCEPTED MANUSCRIPT

Figure B. Heller’s myotomy. Pooled rate of post-procedure reflux symptoms grouped by sample
size (i.e, sample size of the study:N, <40 pts; N, 40-60 pts, N>60 pts. Rates of post-procedure
symptoms are shown with 95% confidence intervals [95% CIs].

PT
RI
U SC
AN
M
D
TE
C EP
AC
 ACCEPTED MANUSCRIPT
Figure C. Heller’s myotomy. Pooled rate of post-procedure reflux symptoms grouped by
minimum follow-up duration (i.e, N, <12 months; ≥12 months). Rates of post-procedure symptoms
are shown with 95% confidence intervals [95% CIs].

PT
RI
U SC
AN
M
D
TE
C EP
AC
 ACCEPTED MANUSCRIPT
Figure D. Heller’s myotomy. Funnel plot for publication bias. Meta-analysis of rate of abnormal
esophageal acid exposure using pH monitoring after Heller’s myotomy. A funnel plot shows the
observed (log) rates of abnormal events on the x-axis against standard error on the y-axis. Empty
circles were added to represent studies imputed by trim and fill.

PT
RI
U SC
AN
M
D
TE
C EP
AC
 ACCEPTED MANUSCRIPT
Figure E. Heller’s Myotomy. Pooled rate of abnormal esophageal acid exposure using pH
monitoring grouped by the type of fundoplication after heller myotomy. Rates of post-procedure
symptoms are shown with 95% confidence intervals [95% CIs].

PT
RI
U SC
AN
M
D
TE
C EP
AC
 ACCEPTED MANUSCRIPT
Table A. Univariate meta-regression analysis. Outcome: rate of reflux symptoms after Heller’s

myotomy.

Factor Coef. cilb ciub Pvalue

Publication Year
<2005 -2.546 -3.466 -1.627
2005-2010 -0.171 -1.450 1.108 0.793
>2010 0.843 -0.463 2.149 0.206
Sample Size (#) -0.011 -0.020 -0.001 0.026

PT
# pts at follow-up -0.013 -0.023 -0.003 0.010
Study Country
Europe -2.659 -3.584 -1.733
USA/Canada 0.142 -1.389 1.673 0.856

RI
Other countries 0.616 -0.592 1.824 0.318
Ratio of females to males 0.709 -0.786 2.204 0.353
Mean patient's age (ya) -0.008 -0.085 0.070 0.849

SC
Study population
Naive pts -2.409 -3.046 -1.772
mixed population -0.109 -1.295 1.077 0.857
Study Design

U
prospective studies -2.479 -2.985 -1.828
RCT 0.463 -1.055 1.982 0.550
AN
Prior endoscopic interventions (%)
<8% -2.515 -3.507 -1.523
8-30% 0.145 -1.363 1.653 0.850
>30% -0.129 -1.410 1.151 0.843
M

Type of Fundoplication
DOR (#) -0.008 -0.033 0.017 0.554
Nissen (#) -0.033 -0.063 -0.004 0.028
Toupet (#) -0.010 -0.019 -0.000 0.049
D

Myotomy length (m) 0.118 -0.469 0.705 0.694

Esophageal Diameter (m) -0.062 -1.090 0.966 0.906
TE

Stomach (m) 0.559 -0.609 1.727 0.348

Minimum Follow-up (months) -0.092 -0.166 -0.019 0.013
Max Follow-up (months) 0.010 -0.004 0.024 0.148
Mean Follow-up -0.168 -0.815 0.478 0.610
C EP
AC
 ACCEPTED MANUSCRIPT

Table B. Univariate meta-regression. Outcome: rate of abnormal esophageal acid exposure using
prolonged pH monitoring after Heller’s myotomy.

Factor Coef. cilb ciub Pvalue

Publication Year
<2005 -1.45 -2.93 0.03

PT
2005-2010 -0.86 -2.77 1.06 0.382
>2010 0.14 -1.64 1.91 0.881
Sample Size -0.001 -0.004 0.002 0.475
# pts at follow-up 0.003 -0.008 0.003 0.295

RI
Study Country
Europe -2.12 -2.77 -1.46
USA/Brazil 1.38 0.35 2.42 0.009

SC
Ratio of females to males -0.50 -13.39 12.39 0.939
Mean patient's age (ya) 0.127 -0.048 0.303 0.849
Study population
Naive pts -1.473 -2.609 -0.336
mixed population -0.212 -1.802 1.378 0.794

U
Study Design
prospective studies -1.579 -2.294 -0.865
AN
RCT -0.411 -2.271 1.448 0.665
% of prior endoscopic
interventions
botox (%) 8.418 -9.497 26.333 0.357
M

balloon (%) -0.779 -3.941 2.383 0.629

fundoplication procedures
Nissen (#) -0.018 -0.034 -0.003 0.022
D

Toupet (#) 0.018 0.002 0.033 0.025

Dor (#) -0.001 -0.004 0.002 0.511
Myotomy lenght (m) 0.118 -0.469 0.705 0.694
TE

Esophageal Diameter (m) -0.062 -1.090 0.966 0.906

Stomach (m) 0.559 -0.609 1.727 0.348
Minimum Follow-up (months) -0.018 -0.086 0.050 0.601
Max Follow-up (months) -0.001 -0.019 0.018 0.957
EP

Mean Follow-up -0.006 -0.041 0.028 0.728

C
AC
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC