Introduction
Peroral endoscopic myotomy (POEM) was introduced by Inoue and colleagues nearly a
decade ago [1 ]. Since then, multiple studies have confirmed the safety and efficacy of POEM for
the management of achalasia [2 ]
[3 ]; however, initial studies concentrated mainly on the feasibility and efficacy of
POEM, and lacked a comprehensive evaluation of gastroesophageal reflux disease (GERD)
[1 ]
[4 ]
[5 ]. Recent studies incorporating an objective evaluation of GERD have indicated that
the incidence of GERD is high after POEM [6 ]. However, the existing literature is limited and divergent with regards to the incidence
and risk factors for GERD after POEM [6 ]
[7 ]
[8 ]
[9 ]
[10 ]
[11 ]
[12 ]. The observed heterogeneity in the published studies may be due to the selection
bias generated by the evaluation of selected patients. In addition, the response to
proton pump inhibitor (PPI) therapy has not been objectively evaluated in previous
studies.
In this study, we aimed to analyze the incidence and risk factors of post-POEM GERD
and its response to PPI therapy in consecutive patients with idiopathic achalasia.
Methods
The data on consecutive patients who underwent POEM (from December 2016 to January
2018) for idiopathic achalasia at a single tertiary center were analyzed from a prospectively
maintained database. Patients who underwent comprehensive evaluation of GERD with
data on all three parameters (symptoms, esophagogastroduodenoscopy [EGD] findings,
and 24-hour pH analysis) were included in the study. Children < 18 years of age, patients
with sigmoid achalasia or non-achalasia spastic esophageal motility disorders, and
those who did not agree to the evaluation of GERD were excluded from the study.
Multiple factors were analyzed to identify the variables that affect the incidence
of GERD after POEM ([Fig. 1 ]). The study was approved by the institutional review board committee.
Fig. 1 Flow diagram depicting the methodology of the current study. POEM, peroral endoscopic
myotomy; PPI, proton pump inhibitor; GERD, gastroesophageal reflux disease.
Pre-POEM evaluation
The pre-POEM evaluation included symptom analysis using the Eckardt score, timed barium
esophagogram, EGD, and high resolution esophageal manometry. The type of achalasia,
integrated relaxation pressure (IRP), and lower esophageal sphincter (LES) pressure
were recorded on esophageal manometry.
POEM technique
The standard technique of POEM has been described previously [3 ]
[13 ]. In brief, POEM was performed via an anterior (1 – 2 o’clock) or posterior (5 o’clock)
route by three experienced operators (Z.N., D.N.R., M.R.). Selective circular myotomy
was done in the upper part of the tunnel and full-thickness myotomy towards the lower
part of the tunnel. Myotomy was performed for a length of 2 – 3 cm beyond the gastroesophageal
junction (GEJ). The length of esophageal myotomy was left to the discretion of the
operating endoscopist.
Extension of the submucosal tunnel into the stomach was identified using a change
in vascular pattern; narrowing, followed by widening of submucosal space; and blanching
of the gastric mucosa visualized in retroversion ([Fig. 2a,b ]). The distal penetrating vessels were identified as the defining point for the end
of the submucosal tunnel on the gastric side. In addition, fluoroscopy was used to
confirm the extension of the tunnel for at least 2 cm beyond the GEJ ([Fig. 2c,d ]).
Fig. 2 Identification of the gastroesophageal junction (GEJ) and assessment of the gastric
myotomy: a endoscopically by identification of the narrow portion of the tunnel representing
the GEJ; b fluoroscopically by confirming the position of the endoscope at the GEJ; c with endoscopic view of the completed myotomy; d by fluoroscopic image confirming an adequate gastric myotomy.
Post-POEM management and follow-up
Follow-up was scheduled at 1, 3, 6, and 12 months after POEM. All the patients were
prescribed PPIs for a duration of 10 weeks after POEM. PPIs were stopped 2 weeks prior
to the first objective assessment of GERD during follow-up at 3 – 4 months. Post-POEM
evaluation at 3 months included symptom analysis, timed barium esophagogram, EGD,
24-hour pH study and esophageal manometry.
Evaluation of GERD
GERD was evaluated using symptoms (heartburn and regurgitation) at each of the follow-up
visits (1, 3, 6, and 12 months). In addition to symptom analysis, EGD and 24-hour
pH impedance study were also performed at 3 months. The 24-hour pH study was performed
as follows: the pH probe was placed transnasally, which was connected to a pH data
acquisition device (ZepHr pH monitor with ComforTEC disposable catheters; Sandhill
Scientific, Highlands Ranch, Colorado, USA). A DeMeester score > 14.72 was considered
to be indicative of GERD [14 ].
An EGD was repeated at 1 year to document an objective assessment of the response
to PPIs and to detect new cases, if any, of reflux esophagitis. The severity of reflux
esophagitis was classified as per the Los Angeles grading for esophagitis (grade A
to D) [15 ].
The following variables were analyzed for their association with GERD: age, sex, body
mass index (BMI), type of achalasia, length of myotomy, history of previous treatment,
and manometric findings, including IRP and LES pressure.
Management of post-POEM GERD
All patients with evidence of reflux esophagitis, with or without symptoms, were prescribed
a PPI (equivalent to 40 mg pantoprazole or 20 mg rabeprazole) until the next objective
evaluation at 1 year after POEM. PPIs were not prescribed in patients with GERD on
24-hour pH study in the absence of symptoms and erosive esophagitis. In patients with
either persistence of symptoms or erosive esophagitis on subsequent follow-up (i. e.
6 months or 1 year), double-dose PPIs were prescribed.
Lifestyle interventions, including avoidance of late-night meals, cessation of smoking
and alcohol, and head-end elevation of the bed, were advised to all the patients.
Definitions
Treatment failure
We defined prior treatment failure as patients who had relapsed with symptoms along
with objective evidence of stasis on timed barium esophagogram after previous pneumatic
dilation (single or multiple sessions) or Heller’s myotomy with or without fundoplication.
GERD
We defined GERD as the presence of typical symptoms, including heartburn and regurgitation,
in conjunction with objective evidence of GERD in form of a positive pH impedance
study or erosive esophagitis.
Statistical analysis
The data were analyzed and compared between the groups of patients with and without
reflux after the POEM procedure. The data were presented as median (range) or mean
(standard deviation [SD]). A Student’s paired t test was used to analyze continuous variables and a chi-squared test for categorical
variables. P values of < 0.05 were considered to be statistically significant. Multiple logistic
regression was performed using the stepwise method irrespective of the significance
on univariate analysis to avoid errors of multiple comparisons and identify important
predictors of outcome that could have been missed while analyzing each outcome separately.
The data were analyzed using MedCalc for Windows, version 12.2.1.0 (MedCalc Software,
Ostend, Belgium).
Results
A total of 209 adult patients underwent POEM for esophageal motility disorders during
the study period (December 2016 to January 2018). Of these, 12 patients had sigmoid
achalasia, 18 did not agree to complete evaluation of GERD, three had non-achalasia
spastic esophageal disorders, and nine were lost to follow-up. Objective evaluation
of GERD was available at 3 months on 167 patients with idiopathic non-sigmoid achalasia
(86.1 %) ([Fig. 3 ]).
Fig. 3 Flow diagram depicting the analysis and outcomes of gastroesophageal reflux disease
(GERD) after peroral endoscopic myotomy (POEM). EGD, esophagogastroduodenoscopy; PPI,
proton pump inhibitor.
Demographics of study patients
Overall, 167 patients (52.7 % men; mean (SD) age 41 (14.42) years) underwent a complete
evaluation for GERD. The majority of the patients were treatment naïve (118 [70.7 %])
and had type II (64.7 %) or type I (25.8 %) achalasia. Pneumatic dilation, used in
40/49 patients (81.6 %) was the main modality of management in patients with a history
of prior treatment. Other baseline characteristics including Eckardt score and pre-POEM
esophageal manometry parameters have been outlined in [Table 1 ].
Table 1
Baseline demographics, esophageal manometry findings, and details of the peroral endoscopic
myotomy (POEM) procedure in the 167 study subjects.
Age, mean (SD), years
41 (14.42)
Sex, female, n (%)
79 (47.3 %)
Body mass index, mean (SD), kg/m2
22.2 (3.89)
Type of Achalasia, n (%)
43 (25.8 %)
108 (64.7 %)
11 (6.6 %)
5 (3.0 %)
Previous therapy, n (%)
118 (70.7 %)
40 (24.0 %)
2 (1.2 %)
1 (0.6 %)
6 (3.6 %)
Baseline Eckardt score, mean (SD)
6.93 (1.50)
Pre-POEM manometry, mean (SD)
26.4 (12.99)
35.42 (14.73)
Characteristics of POEM procedure
Length of myotomy, mean (SD), cm
10.58 (3.07)
7.62 (2.99)
2.94 (0.455)
Orientation of myotomy, n (%)
106 (63.5 %)
61 (36.5 %)
SD, standard deviation; LES, lower esophageal sphincter.
Intraprocedural variables
POEM was performed via an anterior and posterior route in 106 (63.5 %) and 61 patients
(36.5 %), respectively. The mean (SD) length of esophageal myotomy was 7.62 (2.99)
cm and gastric myotomy 2.94 (0.45) cm ([Table 1 ]).
Analysis of GERD
Erosive esophagitis and reflux on 24-hour pH study were found in 41.9 % and 47.9 %
of patients, respectively. The majority of the subjects (92.9 %) had mild erosive
esophagitis (LA grade A or B). The incidence of erosive esophagitis was not significantly
different in patients with high (> 14.72) or normal (< 14.72) DeMeester scores (48.8 %
vs. 35.6 %; P = 0.12). Symptomatic GERD was detected in 29.3 % of patients. There was no significant
difference in the incidence of symptoms in patients with high and normal DeMeester
scores (33.8 % vs. 25.3 %; P = 0.24) ([Table 2 ]).
Table 2
Comparison of patient- and technique-related factors in patients with and without
gastroesophageal reflux after undergoing peroral endoscopic myotomy (POEM).
No reflux (normal DeMeester score) (n = 87)
Reflux present (high DeMeester score) (n = 80)
P value
Age, mean (SD), years
42.2 (14.45)
40.01 (14.38)
0.31
Sex, female, n (%)
37 (42.5)
42 (52.5)
0.63
BMI, mean (SD), kg/m2
22.8 (3.83)
21.5 (3.86)
0.07
Type of achalasia
23 (26.4 %)
20 (25.0 %)
56 (63.4 %)
52 (65.0 %)
5 (5.7 %)
6 (7.5 %)
Previous therapy, n (%)
23 (26.4 %)
26 (32.5 %)
Pre-POEM manometry, mean (SD)
27.42 (15.52)
24.68 (9.33)
0.68
36.45 (14.47)
34.67 (14.93)
0.50
Length and orientation of myotomy
Length of myotomy, mean (SD), cm
10.43 (3.28)
10.74 (2.83)
0.27
7.52 (3.20)
7.74 (2.76)
0.36
2.89 (0.45)
3.0 (0.45)
0.36
Orientation of myotomy, n (%)
57 (64.77 %)
49 (62.03 %)
31 (35.23 %)
30 (37.97 %)
Post-POEM findings
Post-POEM Eckardt score, mean (SD)
0.61 (0.733)
0.70 (0.80)
0.53
Post POEM manometry, mean (SD)
14.33 (7.15)
15.02 (7.74)
0.68
8.90 (4.35)
9.19 (4.78)
0.79
Symptoms of reflux, n (%)
22 (25.3 %)
27 (33.8 %)
0.24
Erosive esophagitis, n (%)
31 (35.6 %)
39 (48.8 %)
0.12
16 (18.4 %)
18 (22.5 %)
13 (14.9 %)
18 (22.5 %)
2 (2.3 %)
3 (3.8 %)
0
0
No esophagitis, n (%)
56 (64.4 %)
41 (51.3 %)
0.12
SD, standard deviation; IRP, integrated relaxation pressure; LES, lower esophageal
sphincter; PPI, proton pump inhibitor.
The achalasia subtypes, pre-procedural patient-related variables (age, sex, and BMI),
parameters of esophageal manometry (LES pressure and IRP), intraprocedural variables
(length and orientation of myotomy) were not significantly different between patients
in the high and low DeMeester groups ([Table 2 ]).
Logistic regression analysis of factors effecting GERD
On logistic regression analysis, the patient-related factors including BMI, age, sex,
and type of achalasia did not significantly impact on the prevalence of GERD after
POEM. The technique of POEM, including the length of myotomy (esophageal or gastric)
and orientation of myotomy (anterior vs. posterior), also did not affect the incidence
of post-POEM GERD ([Table 3 ]).
Table 3
Logistic regression analysis of risk factors for gastroesophageal reflux after peroral
endoscopic myotomy (POEM).
Univariate analysis
Multivariate analysis
OR
95 %CI
P value
Adjusted OR (for age and sex)
95 %CI
P value
Age
1.31
0.71 – 2.40
0.38
0.74
0.35 – 1.55
0.43
Male
1.00
Reference
1.00
Female
0.63
0.34 – 1.17
0.15
0.67
0.42 – 1.49
0.71
Median BMI (< 22.2 vs. > 22.2 kg/m2 )
1.31
0.71 – 2.42
0.37
0.67
0.32 – 1.41
0.29
Type of achalasia
0.66
0.32 – 1.42
0.31
Type II
1
Reference
1
Reference
Previous therapy
1.09
0.56 – 2.14
0.78
1.21
0.53 – 2.76
0.64
Pre-POEM manometry
Median IRP (> 24.6 vs. < 24.6 mmHg)
1.16
0.61 – 2.19
0.64
1.14
0.51 – 2.55
0.74
LES pressure, mmHg
0.96
0.44 – 2.09
0.93
Median length and orientation of myotomy
Esophageal (> 7.0 vs. < 7.0 cm)
1.31
0.71 – 2.46
0.38
1.12
0.51 – 2.44
0.77
Gastric (> 3.0 vs. < 3.0 cm)
2.32
0.68 – 8.18
0.17
1.67
0.52 – 5.37
0.39
Anterior myotomy
0.89
0.47 – 1.66
0.71
1.14
0.51 – 2.44
0.74
Posterior myotomy
1.00
Reference
1.00
Reference
Post-POEM Eckardt score (> 1 vs. ≤ 1)
1.21
0.65 – 2.22
0.38
1.35
0.64 – 2.86
0.43
Post-POEM manometry
Median IRP (> 8.4 vs. < 8.4 mmHg)
1.07
0.57 – 2.01
0.82
0.79
0.35 – 1.80
0.59
LES pressure, mmHg
1.37
0.60 – 3.12
0.44
CI, confidence interval; BMI, body mass index; IRP, integrated relaxation pressure;
LES, lower esophageal sphincter.
Response to antireflux therapy
At 1 year, 139 patients (83.2 %) were available for follow-up and 106 of them (63.5 %)
underwent EGD, including 43 of the patients (61.4 %) with reflux esophagitis and 63
of those (65.0 %) without evidence of esophagitis at 3 months. Complete resolution
of esophagitis was documented in 35 patients (81.4 %) after PPI therapy. Persistent
esophagitis was found in eight patients (18.6 %). Of these, two patients were not
compliant with PPI therapy, while six had persistent or residual erosive esophagitis
(grade A 4; grade B 2) on regular once a day PPI therapy. Among the 63 patients without
reflux esophagitis on initial EGD at 3 months, reflux esophagitis was diagnosed in
an additional 12 patients (grade A 9; grade B 3) at 1 year. A high DeMeester score
was previously documented at 3 months in eight of these patients (grade A 5; grade
B 3) ([Fig. 3 ]).
Discussion
In this study, we found that the incidence of GERD is high (48 %) after POEM in Indian
patients with idiopathic achalasia. There are no intraprocedural or patient-related
factors that influence the occurrence of GERD after POEM.
The efficacy of POEM has been unequivocally proven in multiple studies with short-
and medium-term follow-up [2 ]
[3 ]
[16 ]. However, recent reports indicate that the incidence of GERD may be higher after
POEM compared with Heller’s myotomy and pneumatic dilation [17 ]
[18 ]
[19 ]
[20 ]. The emergence of reports of peptic stricture and Barrett’s esophagus after POEM
highlight the importance of evaluation and management of GERD in these patients [12 ]
[21 ].
The current literature depicts contrasting results with regards to the incidence and
risk factors for GERD after POEM, possibly due to selection bias ([Table 4 ]). In addition, the response of post-POEM GERD to PPI therapy has not been previously
evaluated. In one of the largest multicenter studies [6 ], objective evaluation of post-POEM GERD (by pH study) was performed in only 60 %
of the subjects. In another study, only 68/112 patients were analyzed for GERD after
POEM. Of note, the symptoms of GERD were significantly less severe in those who did
not return for objective evaluation of GERD [12 ]. This suggests that symptomatic patients are more likely to be evaluated for GERD;
therefore, the true incidence of post-POEM GERD cannot be quantified by analyzing
a selected group of patients. We attempted to address this concern and analyzed consecutive
patients who underwent POEM for achalasia.
Table 4
Studies depicting the objective evaluation of gastroesophageal reflux after peroral
endoscopic myotomy (POEM).
n
Study design/ country
Percentage of subjects with GERD on basis of:
Predictors of GERD
symptoms
reflux esophagitis (percentage undergoing endoscopy)
24-hour pH study (percentage undergoing pH testing)
Sharata et al. 2015 [11 ]
100
Retrospective, USA
Heartburn 8 %;
27.4 % (73 %)
38.2 % (68 %)
Not reported
Hungness et al. 2016 [12 ]
115
Retrospective,
28 % (GERD-Q > 7)
25 % (61 %)
45 % (20 %)
Hiatus hernia, BMI > 35 kg/m2
Shiwaku et al. 2016 [9 ]
105
Prospective, Japan
Not reported
60 %
13 %
IRP
Familiari et al. 2016 [7 ]
103
Retrospective, Italy
Heartburn 18.4 %
20.4 %
50.5 %
IRP
Wang et al. 2016 [10 ]
56
Retrospective,
23.2 %
21.4 %
44.6 %
Low IRP, full-thickness myotomy
Jones et al. 2016 [8 ]
43
Retrospective, USA
28 % (GERD-HRQL, GERSS)
Not reported
58 % (60 %)
Not reported
Kumbhari et al. 2017 [6 ]
282
Retrospective, multicenter
39.9 %
23 % (83 %)
57.8 %
Female sex
Current study
167
Retrospective, India
29.3 %
41.9 %
47.9 %
No factors
GERD, gastroesophageal reflux disease; GERD-Q, GERD questionnaire; BMI, body mass
index; IRP, integrated relaxation pressure; GERD-HRQL, GERD health-related quality
of life; GERSS, GERD symptom scale.
In the current study, the majority of the patients (86 %) underwent objective evaluation
for GERD using EGD and 24-hour pH impedance analysis. About half of the patients had
GERD (high DeMeester score) on pH study. Erosive esophagitis (mostly grade A or B)
was detected in 42 % of patients; however, symptoms of reflux were evident in fewer
patients (29 %). Only one-third of the patients with a high DeMeester score were symptomatic
for GERD, suggesting that the majority of the patients with reflux were detected incidentally.
Our results are in concordance with a recent multicenter study by Kumbhari et al.
[6 ], where a high DeMeester score, reflux esophagitis, and asymptomatic GERD were found
in 57.8 %, 23.2 %, and 60.1 % of patients, respectively. In other studies, the incidence
of symptomatic GERD and reflux esophagitis have been found to be 18 % – 40 % and 20 % – 60 %,
respectively ([Table 4 ]).
The second major finding of our study was a lack of association between patient-related
factors, like age, sex, BMI, type of achalasia, and the incidence of GERD. Similarly,
the technique of POEM (anterior vs. posterior), length of myotomy, and post-POEM esophageal
manometry parameters, such as IRP and LES pressure, did not influence the occurrence
of post-POEM GERD. In order to substantiate the findings in the current study and
reduce the margin of error, we performed a multivariate analysis, even though no significant
factors were identified on univariate analysis. Several randomized trials have confirmed
that the orientation of the myotomy does not influence the incidence of GERD after
POEM [22 ]
[23 ]
[24 ]. Similarly, a high BMI was not associated with an increased incidence of post-POEM
GERD in a recent study [25 ]. More recently, Tanaka and colleagues evaluated a novel technique for POEM to preserve
the oblique muscle fibers and prevent GERD [26 ]. Reflux esophagitis of grade B and higher severity was significantly less frequent
in the group treated with the new technique (31.3 % vs. 58.1 %; P = 0.02). In our study, we did not specifically record and analyze the technique of
gastric myotomy and its impact on GERD.
In addition to the incidence and risk factors for post-POEM GERD, we also analyzed
the outcome of PPI therapy at 1 year after POEM. Although, the resolution of reflux
esophagitis was documented in the majority of the patients, several new cases of erosive
esophagitis were detected. Most of these patients (67 %) had mild esophagitis (grade
A and B) and had had a high DeMeester score at 3 months. Therefore, increased esophageal
acid exposure in the absence of endoscopic evidence of esophagitis may not be innocuous
and close follow-up is required. In addition, non-compliance with PPI therapy was
documented in nearly a quarter of patients. Because most of the patients with post-POEM
GERD (70 %) are asymptomatic, ensuring compliance with PPI therapy may be especially
challenging in these patients.
There are several implications of our study. First, in contrast to the prevalence
of GERD in the general population, post-POEM GERD is highly prevalent in Asian patients,
similar to that reported in Western studies [27 ]. Therefore, it is important to convey the risk of GERD and discuss alternative treatment
options with patients before contemplating POEM. Second, poor correlation between
symptoms and the presence of GERD signifies the importance of universal screening
for GERD, irrespective of the presence of symptoms. Third, the occurrence of GERD
is not dependent on the technique of POEM. Consequently, novel strategies need to
be explored for the prevention of post-POEM GERD in future studies [28 ]
[29 ]
[30 ]
[31 ]. Last, the response to PPIs appears adequate in the majority of the patients who
are compliant with them. However, surveillance endoscopy may be required, especially
in patients with increased esophageal acid exposure, to look for new cases of reflux
esophagitis. Long-term follow-up studies are required to enlighten our knowledge in
this group of patients.
The strengths of our study are its large sample size, the objective evaluation of
GERD, and a reduced likelihood of selection bias owing to the inclusion of consecutive
patients who underwent POEM. The objective documentation of response to PPIs and the
clinical significance of increased esophageal acid exposure in the absence of symptoms
have not been discussed in previous studies. However, certain drawbacks are noteworthy.
About one-third of the patients could not be objectively evaluated for GERD at 1-year
follow-up. As a consequence, the response to PPI therapy may have been marginally
underestimated or overestimated. We did not use the double-scope method to confirm
the gastric extent of the myotomy. An excessive myotomy (> 4 cm) on the gastric side
has been proposed as one of the risk factors for post-POEM reflux [32 ]. Instead, we used the “two penetrating vessels” technique and fluoroscopy to define
the gastric extent of the myotomy. These techniques have been previously evaluated
for estimating the adequacy of gastric myotomy [33 ]
[34 ]. Although, a marginal error is possible using these methods, they are unlikely to
grossly over- or underestimate the length of the gastric myotomy.
In conclusion, the incidence of GERD is high after POEM and this risk should be conveyed
to the patients. Patient characteristics, orientation of the myotomy, and post-POEM
IRP do not correlate with the development of GERD. The majority of patients develop
mild erosive esophagitis and respond well to PPI therapy.
