Gastroenterology 2023;164:1108–1118

ESOPHAGUS
The Efficacy of Peroral Endoscopic Myotomy vs Pneumatic
Dilation as Treatment for Patients With Achalasia Suffering From
Persistent or Recurrent Symptoms After Laparoscopic Heller
Myotomy: A Randomized Clinical Trial
ESOPHAGUS

Caroline M. G. Saleh,1 Pietro Familiari,2 Barbara A. J. Bastiaansen,1 Paul Fockens,1 Jan Tack,3
Guy Boeckxstaens,3 Raf Bisschops,3 Aaltje Lei,1 Marlies P. Schijven,1
Jan Guido Costamagna,2 and Albert J. Bredenoord1
1
Department of Gastroenterology and Hepatology, Amsterdam UMC, University of Amsterdam, Amsterdam, The Netherlands;
2
Digestive Endoscopy Unit, Agostino Gemelli University Hospital, Università Cattolica del Sacro Cuore, Rome, Italy; and
3
Department of Gastroenterology and Hepatology, University Hospitals Leuven, KU Leuven, Leuven, Belgium

the esophageal body and insufficient relaxation of the lower

See editorial on page 1053. esophageal sphincter (LES). Treatment options for patients
with achalasia include the traditional pneumatic dilation (PD),
BACKGROUND & AIMS: For patients with achalasia experi- laparoscopic Heller myotomy (LHM), and per-oral endoscopic
encing persistent or recurrent symptoms after laparoscopic myotomy (POEM). POEM and LHM provide longer-lasting
Heller myotomy (LHM), pneumatic dilation (PD) is the most symptomatic responses in treatment-naïve patients with
frequently used treatment. Per-oral endoscopic myotomy achalasia than a single series of PDs.1,2 Despite the high effi-
(POEM) is increasingly being investigated as rescue therapy. cacy rates of POEM and LHM, persistent or recurrent symp-
This study aimed to determine the efficacy of POEM vs PD for toms occur in 10%–20% of patients treated with LHM.3,4
patients with persistent or recurrent symptoms after LHM. Until now, treatment of persistent or recurrent symp-
METHODS: This randomized multicenter controlled trial toms after LHM remained controversial. Rescue therapy
included patients after LHM with an Eckardt score >3 and includes PD, revisional LHM, POEM, and as a last resort,
substantial stasis (
2 cm) on timed barium esophagogram and esophagectomy.3,5–11 Studies investigating the efficacy of PD
randomized to POEM or PD. The primary outcome was treat-
for recurrent or persistent symptoms after LHM show a
ment success, defined as an Eckardt score of 3 and without
variable success rate ranging from 57%–96%.12–15 A
unscheduled re-treatment. Secondary outcomes included the
retrospective study conducted in our center showed a long-
presence of reflux esophagitis, high-resolution manometry, and
term success rate of 57%. Revisional LHM has a reported
timed barium esophagogram findings. Follow-up duration was
1 year after initial treatment. RESULTS: Ninety patients were efficacy of 71%–90%.16–19 However, this procedure re-
included. POEM had a higher success rate (28 of 45 patients quires extensive expertise, as it is invasive and with more
[62.2%]) than PD (12 of 45 patients [26.7%]; absolute differ- difficulties due to adhesions, fibrosis, and the loss of tissue
ence, 35.6%; 95% CI, 16.4%–54.7%; P ¼ .001; odds ratio, 0.22; planes in the area of the gastric esophageal junction caused
95% CI, 0.09–0.54; relative risk for success, 2.33; 95% CI, 1.37– by the original operation.10,19 The latter may support a
3.99). Reflux esophagitis was not significantly different be- preference to initially treat persistent or recurrent symp-
tween POEM (12 of 35 [34.3%]) and PD (6 of 40 [15%]). Basal toms with PD before moving on to a revisional LHM,
lower esophageal sphincter pressure and integrated relaxation although multiple PDs may be required.10
pressure (IRP-4) were significantly lower in the POEM group Over the past decade, POEM gained acceptance as an
(P ¼ .034; P ¼ .002). Barium column height after 2 and 5 endoscopic alternative to LHM for the primary treatment of
minutes was significantly less in patients treated with POEM achalasia. Considering the technique and its high efficacy
(P ¼ .005; P ¼ .015). CONCLUSIONS: Among patients with rates, POEM is increasingly being investigated as a rescue
achalasia experiencing persistent or recurrent symptoms after therapy for patients with persistent or recurrent symptoms
LHM, POEM resulted in a significantly higher success rate than after LHM. Case series suggest that POEM has an efficacy
PD, with a numerically higher incidence of grade A–B reflux
esophagitis. Netherlands Trial Registry: NL4361 (NTR4501),
Abbreviations used in this paper: ADSQoL, achalasia disease-specific
https://trialsearch.who.int/Trial2.aspx?TrialID ¼ NTR4501. quality of life questionnaire; AE, adverse event; LES, lower esophageal
sphincter; LHM, laparoscopic Heller myotomy; PD, pneumatic dilation;
Keywords: Per-Oral Endoscopic Myotomy; Pneumatic Dilation; POEM, per-oral endoscopic myotomy; PPI, proton pump inhibitor; SAE,
serious adverse event; SF-36, 36-Item-Short Form Health Survey.
Laparoscopic Heller Myotomy; Eckardt Score; High-Resolution
Manometry. Most current article
© 2023 The Author(s). Published by Elsevier Inc. on behalf of the AGA
Institute. This is an open access article under the CC BY license (http://

A
creativecommons.org/licenses/by/4.0/).
chalasia is a rare esophageal motility disorder 0016-5085
characterized by dysfunctional or absent motility of https://doi.org/10.1053/j.gastro.2023.02.048
June 2023 POEM vs PD After Failed Laparoscopic Heller Myotomy 1109

ranging from 81%–100% when applied as rescue therapy

WHAT YOU NEED TO KNOW
after LHM.10,20 Unfortunately, these studies represent small
series with relatively short follow-up times and retrospec- BACKGROUND AND CONTEXT
tively collected data. More importantly, no studies have Patients with achalasia experiencing persistent or
compared POEM with other performed rescue therapies, recurrent symptoms after laparoscopic Heller myotomy
such as PD, the most widely used. are most frequently treated with pneumatic dilation. Per-
Therefore, this study aimed to determine and compare oral endoscopic myotomy is being investigated
the efficacy of POEM vs PD for patients with persistent or increasingly as rescue therapy; this study aimed to
recurrent symptoms after LHM. determine the efficacy of per-oral endoscopic myotomy
vs pneumatic dilation for patients with persistent or
recurrent symptoms after laparoscopic Heller myotomy.
Methods

ESOPHAGUS
NEW FINDINGS
Study Design Per-oral endoscopic myotomy resulted in a significantly
This study was designed as a multicenter randomized higher success rate than pneumatic dilation (62.2% and
controlled trial. Inclusion occurred in 3 achalasia expert centers 26.7%, respectively) in patients with achalasia
in the Netherlands, Belgium, and Italy, from January 2014 to experiencing recurrent or persistent symptoms after
June 2020. The Institutional Medical Ethics Board approved the laparoscopic Heller myotomy.
study protocol in each hospital.
LIMITATIONS
Patients were enrolled in the study after obtaining written
informed consent. The primary end point was measured at 1- Primary and secondary outcomes were assessed at 1-
year follow-up. Follow-up of patients took place at 3 months year follow-up, meaning no conclusions can be drawn
for longer-term treatment success, which is important,
and 1 year after initial treatment.
given that achalasia is a lifelong chronic disease.
A data and safety monitoring board consisting of a method- Furthermore, like most endoscopic or surgical studies
ologist, surgeon, and gastroenterologist was installed to monitor that evaluate new interventional techniques, patients
the safety and efficacy of treatment groups. Moreover, the study and caregivers were not blinded to treatment allocation.
underwent an extensive randomly assigned internal quality audit Lastly, multiple pneumatic dilation sessions might form
in May 2017. Study sites were monitored by a research nurse, in a potential bias in the comparison with 1 treatment
which the case report forms and source data were checked. intervention; however, this was done deliberately to
This study was not classified as single-blind; to minimize optimally reflect routine clinical care in these patients.
bias several interventions were implemented. First of all, CLINICAL RESEARCH RELEVANCE
questionnaires were filled in by patients without the presence
of research personnel. Diagnostic measurements were evalu- Per-oral endoscopic myotomy can be considered as the
initial treatment option for patients with achalasia
ated by an observer unaware of the patients’ treatment. The
experiencing persistent or recurrent symptoms after
interpretation whether an unscheduled treatment was indi- laparoscopic Heller myotomy.
cated was solely based on a previously set cutoff.
All authors had access to the study data and reviewed and BASIC RESEARCH RELEVANCE
approved the final manuscript. This randomized controlled trial found that per-oral
endoscopic myotomy results in significantly lower
Patients and Inclusion and Exclusion Criteria Eckardt scores than pneumatic dilation for patients with
Adult patients aged 18–80 years were eligible for enroll- achalasia experiencing persistent or recurrent symptoms
ment if they had persistent or recurrent symptoms after LHM, after laparoscopic Heller myotomy.
defined as having an Eckardt symptom score >3 in combina-
tion with significant stasis (
2 cm) seen on timed barium fluoroscopic guidance, the balloon was positioned at the gastric
esophagogram after 2 minutes. esophageal junction and dilated at a pressure of 5 psi for 1
Exclusion criteria included previous PDs after LHM, previous minute, followed by 7 psi for 1 minute.21 A graded distension
attempt at POEM, previous surgery to the stomach or esophagus protocol was implemented; initial PD was performed using a
(except for LHM), known coagulopathy, presence of liver cirrhosis 30-mm balloon and 1–3 weeks later a subsequent 35-mm
and/or esophageal varices, eosinophilic esophagitis, stricture of balloon dilation was performed. In case of symptom persis-
the esophagus, (pre)malignant esophageal lesions, 1 or more tence or recurrence within 3 months, a PD with a 40-mm
esophageal diverticula, and pregnancy at time of treatment. balloon was performed. Patients presenting with symptom
recurrence between 3 and 12 months after inclusion were
Randomization offered additional PD treatment with a 35- and 40-mm balloon.
Randomization was done using a web-based program If the treating physician judged that repeating a 30-mm PD was
(ALEA Clinical B.V.) that assigned patients to POEM or PD in a required before performing the 35-mm PD, this was allowed.
1:1 ratio, stratified according to the research site. Local study PD was considered a failure in case of symptom persistence or
staff enrolled the patients. The number of patients treated with recurrence after this additional round of PDs. All further re-
POEM or PD was similar for each center. treatments with PD were considered unscheduled re-
treatments. Patients undergoing unscheduled PD re-
Interventions treatments were considered failures at 1-year follow-up
Pneumatic dilation. For PDs, a series of dilations with regardless of their Eckardt score at 1-year follow-up. Thus, all
Rigiflex balloons (Boston Scientific) was performed. Under patients randomized to the PD treatment arm received at least
 1110 Saleh et al Gastroenterology Vol. 164, Iss. 7

2 dilations, with the last dilation at least 35 mm. PDs were Questionnaire, SF-36, and ADSQoL questionnaires. High-
performed by experienced endoscopists who completed more resolution manometry was performed to confirm the reoccur-
than 20 PDs independently. rence of achalasia.23 Upper endoscopy and timed barium
Preprocedural instructions consisted of a liquid diet for 3 esophagogram were performed to quantify esophageal stasis by
days before PD, which included a clear liquid diet 24 hours measuring barium column height at 0, 1, 2, and 5 minutes on
before PD and nil per mouth 8 hours before PD. Post PD, pa- radiographic images after ingestion of 100–200 mL of low-
tients were prescribed a proton pump inhibitor (PPI) once per density barium sulfate suspension during a time window of
day for 2 weeks after each dilation. 30–60 seconds.24
Per-oral endoscopic myotomy. POEM was performed Symptoms (Eckardt score) and questionnaires were
under general anesthesia, including endotracheal intubation, assessed at 3-month and 1-year follow-up. High-resolution
with the patient in a supine position. The POEM procedure was manometry and timed barium esophagogram were obtained
ESOPHAGUS

performed as described by Ponds et al21; however, the mucosal after 3-month and 1-year follow-up, whereas upper endoscopy
incision, tunnel, and myotomy were slightly more toward the was only performed after 1-year follow-up. The severity of
posterior orientation of the esophagus to stay away from the reflux esophagitis was scored according to the Los Angeles
original myotomy scar. Patients randomized to undergo POEM Classification.25 PPI use was documented and was prescribed
received the same preprocedural instructions as those who for patients who experienced reflux symptoms independent of
underwent PD. Admission took place on the same day as follow-up time or when reflux esophagitis was observed during
treatment or the day before, depending on the travel distance upper endoscopy.
to the hospital; patients were discharged at least 1 day after
POEM. Before treatment, patients were administered prophy-
lactic antibiotics according to local hospital recommendations Re-treatment After Unsuccessful Treatments
and a double-dose PPI intravenously. Post-discharge patients Patients randomized to the PD treatment arm were initially
were advised to adhere to a liquid diet for 7 days, followed by a treated with a 30- and 35-mm balloon. A 3-week follow-up was
soft diet for 1 more week and were prescribed a single-dose PPI set to assess symptom severity; in case of an Eckardt score >3,
for 2 weeks. patients were treated with a 40-mm PD. If symptoms recurred
within 1 year, patients were treated with additional PDs, up to a
maximum diameter of 40 mm. Patients were offered POEM if
Outcomes symptoms persisted or recurred after 1 year or if they refused
The primary outcome was treatment success after 1-year additional or unscheduled re-treatment with PDs within 1 year
follow-up, which was defined as an Eckardt score of
3 from initial PD treatment.
without any unscheduled re-treatment. For patients random- Patients who failed after POEM treatment were offered
ized to the PD treatment arm, this meant dilation with a 30-mm unscheduled re-treatment consisting of PDs, according to the
and 35-mm balloon and possibly PDs up to 40 mm; for patients graded distension protocol described above.
randomized to the POEM arm, this meant undergoing POEM Follow-up after re-treatment was continued according to
without any PDs or other unscheduled re-treatments. the initial treatment protocol.
Secondary outcomes were assessed at baseline and 3-
month and 1-year follow-up. The quality of life and the
achalasia-specific quality of life were measured using the 36- Statistical Methods
Item-Short Form Health Survey (SF-36) and the achalasia Sample size calculation was based on the reported long-
disease-specific quality of life questionnaire (ADSQoL). The SF- term success rates of PD after Heller myotomy (50%–67%)
36 measured general quality of life by scoring mental and and the reported short-term success rates of POEM after Heller
physical aspects, ranging from 0 to 100, with higher scores myotomy (91%–100%).1,26,27 One study in previously non-
indicating a better quality of life.22 The presence of reflux surgically treated patients reported a success rate of 82% after
symptoms and reflux esophagitis was assessed using the 12-month follow-up.28 Therefore, we assumed long-term suc-
Gastroesophageal Reflux Disease Questionnaire and upper cess rates of 58% for PD and 85% for POEM after Heller
endoscopy; use of acid suppressant drugs was also docu- myotomy. With these success rates, we estimated that with 43
mented. Esophageal stasis, as seen on the timed barium patients in each group, the study would have 80% power to
esophagogram, was measured. detect a significant difference in success rate between PD and
All adverse events (AEs) and serious adverse events (SAEs) POEM, with a 2-sided a level of .05. To cope with an estimated
were documented. Treatment complications were defined as 5% loss to follow-up, we aimed to enroll 90 patients.
any AEs that arose after the treatment or secondary to the
treatment. AEs were classified as “severe” when they resulted
in (prolonged) admission of more than 24 hours, medium or Primary Analysis
intensive care unit admission, additional endoscopic proced- An intention-to-treat analysis was performed containing all
ures, blood transfusion, or death. Other complications were patients as randomized to their treatment group. According to
classified as “mild.” distribution, continuous data are presented as mean (SD) or
median (interquartile range). Categorical data are presented as
percentages.
Clinical Assessment and Follow-Up The primary outcome included treatment effectivity based
The clinical assessment started at baseline and included a on Eckardt score at 1-year follow-up without re-treatment.
medical history, physical examination, and routine laboratory Fisher exact test was used to calculate the odds ratio and
tests. Patients completed the Gastroesophageal Reflux Disease relative risk for treatment outcome and treatment-related SAEs.
June 2023 POEM vs PD After Failed Laparoscopic Heller Myotomy 1111

The secondary outcomes were analyzed using Mann- In the patients randomized to receive POEM, 3 patients
Whitney U test for continuous data or Fisher exact test for did not undergo a complete POEM; 2 patients did not
categorical data. receive POEM because fibrotic submucosa prohibited the
Absolute differences of comparative results were calculated creation of a submucosal tunnel and performance of the
by subtracting percentages, means, or medians of the groups endoscopic myotomy, and 1 patient was lost to follow-up
and calculating the 95% CIs of the difference. after randomization (Figure 1).
In the patients randomized to PD, 1 patient underwent
Post-Hoc Sensitivity Analysis only a single 30-mm PD with a good response and refused
further dilation with a 35-mm balloon. The other patients
A sensitivity analysis was performed to increase the credi-
bility of the results.29 For the primary outcome, a per-protocol received dilations with 30- and 35-mm balloons (n ¼ 19) or
up to 40-mm balloon (n ¼ 25). Waist obliteration was ob-

ESOPHAGUS
analysis was used. For the secondary outcomes, the post-hoc
analysis included the use of linear mixed models and general- tained in all PDs.
ized linear models to account for missing values and to adjust
for repeated effects or possible confounders. Specifically, linear
Secondary Outcomes
mixed models were used to analyze the effect of treatment type
Reflux esophagitis, proton pump inhibitor use,
on continuous secondary outcome parameters with fixed ef-
and reflux symptoms (Gastroesophageal Reflux Dis-
fects for time and treatment. An unstructured covariance ease Questionnaire). At 1-year follow-up, a numerically
structure was used when running the linear mixed models. The
higher incidence of reflux esophagitis was observed in pa-
generalized linear models were used to analyze the association
tients treated with POEM (12 of 35 [34.3%]) than PD (6 of
between treatment on binary outcome parameters, such as the
40 [15%]), but this was not statistically significant. Further
presence of reflux esophagitis or PPI use. The generalized linear
specified, for the patients randomized to POEM, 11 of 12
models used a binomial distribution and logit link function.
(91.7%) were assigned grade A–B and 1 (8.3%) grade C, and
for patients randomized to PD, 5 of 6 (83.3%) were assigned
Results grade A–B and 1 (16.7%) grade C. Reflux symptoms and
daily use of PPI did not differ between treatment groups
Enrollment and Patient Characteristics (Table 3 and Figure 3 and Supplementary Table 1).
Between January 2014 and June 2020, ninety patients Eckardt score, high-resolution manometry, timed
with achalasia and experiencing persistent or recurrent barium esophagogram, and quality of life (achalasia
symptoms after LHM were randomized; 45 were randomly disease-specific quality of life questionnaire and
assigned to receive POEM and 45 were assigned to receive Medical Outcomes Study 36-Item-Short Form Health
PD (Figure 1). All patients were treated with LHM and a Dor Survey). This study found a significantly lower Eckardt
fundoplication. One patient randomized to POEM never score was measured in the patients treated with POEM vs
received treatment. In 2 patients, the myotomy as part of those treated with PD (P ¼ .016) (Figure 4 and
POEM was not possible because the submucosal tunnel Supplementary Table 1). Basal LES pressure and integrated
could not be created due to submucosal fibrosis (Figure 1). relaxation pressure (IRP-4) were significantly lower at 1-year
A protocol deviation occurred related to the PD treatment, follow-up for patients treated with POEM vs patients treated
as 1 patient received a single 30-mm PD and refused further with PD (P ¼ .034; P ¼ .002). A significant difference was
treatment because of a significant reduction of symptoms. found between POEM and PD for barium column height after
The final date of the 1-year follow-up period of the last 2 and 5 minutes, with less stasis observed in the POEM group
patient was June 2021. Baseline characteristics were similar (P ¼ .005; P ¼ .015). There was no significant difference
between groups (Table 1). between POEM and PD when evaluating the maximum
esophageal width measured during timed barium esophago-
gram (P ¼ .121) (Figure 3 and Supplementary Table 1).
Primary Outcome With regard to the baseline measurements, this study
Analysis of the primary outcome showed higher treat- found no significant differences in median Eckardt score
ment success at 1-year follow-up in the patients treated (P ¼ .920), basal LES pressure (P ¼ .109), IRP-4 (P ¼ .631),
with POEM (28 of 45 patients [62.2%]) compared with the achalasia subtype (P ¼ .927), and barium column height
patients treated with PD (12 of 45 patients [26.7%]; abso- after 2 minutes (P ¼ .282) and 5 minutes (P ¼ .830) be-
lute difference, 35.6%; 95% CI, 16.4%–54.7%; P ¼ .001; tween unsuccessfully and successfully treated patients. The
odds ratio, 0.22; 95% CI, 0.09–0.54; relative risk for success, same applied when performing subgroup analysis within
2.33; 95% CI, 1.37–3.99) (Figure 2 and Table 2). A total of 5 the treatment groups; for patients treated with POEM and
missing values were observed, which were assumed failures PD, there were no significant differences in median Eckardt
according to the intention-to-treat principle (Figure 2 and score (P ¼ .910; P ¼ .699), basal LES pressure (P ¼ 1.0; P ¼
Supplementary Figure 1). Single imputations were used for .501), IRP-4 (P ¼ .756; P ¼ .926), achalasia subtype (P ¼
3 missing values by logically inferencing; 2 patients were .765; P ¼ .843), and barium column height after 2 minutes
considered successful at 1 year, as they were successfully (P ¼ .597; P ¼ .669) and 5 minutes (P ¼ .597; P ¼ .830)
treated at 3-month and 2-year follow-up (without any re- between unsuccessfully and successfully treated patients.
treatments), and 1 patient was deemed a failure, as this Importantly, our study found a significantly lower mean
patient was a failure at 3-month follow-up. ADSQoL score in the POEM group. The overall quality of life
 1112 Saleh et al Gastroenterology Vol. 164, Iss. 7
ESOPHAGUS

Figure 1. Flowchart, randomization, and follow-up according to intention-to-treat analysis. *1This patient underwent a PD with
a 30-mm balloon only because adequate symptom control (Eckardt score <3) was achieved after a single PD, the patient
refused PD with a 35-mm balloon.

was measured using the SF-36 score, which is composed of consisted of chronic severe reflux symptoms after PD and
8 sections. There was a significant difference between POEM was treated with a Toupet fundoplication. Both patients
and PD, favoring POEM for Physical Functioning, Emotional continued in the study. Detailed information on SAEs inde-
Well-Being, and Social Functioning. For the components pendent of the study interventions is provided in the
General Health, Limitations Due to Physical Health, Limita- Supplementary Material.
tions Due to Emotional Problems, Energy/Fatigue, and Pain, AEs were more common after POEM (14 of 45 patients
we found no difference (Table 3). [31.1%]) vs after PD (9 of 45 [20%]). AEs in the POEM
group were related to candida esophagitis (n ¼ 1), Heli-
cobacter pylori infection (n ¼ 3), periprocedural mucosal
Serious Adverse Events and Adverse Events bleeding (n ¼ 2), gastric perforations (1 caused by the spray
Eight SAEs occurred during the study, 2 were related to catheter that was managed conservatively and 1 that was
treatment and 6 occurred independently of the study treated by placement of 3 clips) (n ¼ 2), food impaction
intervention. One microperforation occurred after a POEM, (n ¼ 1), and several not-upper-gastrointestinal–related AEs
which required admission and treatment with antibiotics for (n ¼ 5).
2 days with subsequent discharge; this patient was initially In the PD group, reported AEs were retrosternal pain
randomized and treated with PD and failed. Another SAE after PD (n ¼ 2), pneumoperitoneum and subcutaneous
June 2023 POEM vs PD After Failed Laparoscopic Heller Myotomy 1113

Table 1.Baseline Characteristics of Included Patients

Presented per Treatment Group (n ¼ 90)

Characteristic POEM (n ¼ 45) PD (n ¼ 45)

Sex, female, n (% male) 27 (40) 29 (35.6)

Age, y, median (IQR) 53 (29–77) 52 (25–79)
a
Eckardt score, median (IQR) 6 (4–8) 6(4–6)
Achalasia subtypes, n (%)
I (n ¼ 19) 9 (47) 10 (53)
II (n ¼ 30) 15 (50) 15 (50)

ESOPHAGUS
III (n ¼ 7) 4 (57) 3 (43)
Basal LES pressure, 22.7 (17.9–27.5) 25.4 (19.3–31.5)
mmHg, mean (95% CI)
Basal IRP-4, mmHg, 17.2 (13.6–20.8) 21.3 (16.1–26.6)
mean (95% CI)
Barium esophagogram,
median (IQR)
Column height 4.7 (1.9–7.5) 4.3 (1–7.6)
T ¼ 2 min, cm
Column height 3.4 (0.3–6.5) 4.0 (1.2–6.8)
T ¼ 5 min, cm
Maximum diameter, cm 3.5 (2.6–4.4) 3.3 (2.1–4.7) Figure 2. Primary outcome for POEM and PD (absolute
numbers).
ADSQoL score,b 25 (25–27) 26 (23–28)
median (IQR)
SF-36 score,c
median (IQR)
patient was lost to follow-up after treatment with POEM,
General Health 50 (35–70) 45 (35–60) and 1 was lost to follow-up after randomization. Within the
Physical Functioning 80 (65–93.8) 82.5 (57.5–95) POEM group, 27 of 41 patients (65.8%) vs 12 of 47 (25.5%]
Limitations Due to 25 (0–100) 50 (0–100) in the PD group were successfully treated at 1-year follow-
Physical Health up (absolute difference, 40.3%; 95% CI, 21.2%–59.5%;
Limitations Due to 100 (41.7–100) 66.7 (0–100) relative risk, 2.6; 95% CI, 1.51–4.41).
Emotional Problems
In the PD group, 14 patients received re-treatment with
Energy/Fatigue 50 (35–65) 42.5 (30–60)
Emotional Well-Being 72 (53–88) 70 (37–84) POEM; 6 of 14 (42.9%) were successfully treated at 1-year
Social Functioning 75 (50–87.5) 56.3 (28.2–75) follow-up. Within the POEM group, 2 patients received re-
Pain 57.5 (45–80) 45 (32.5–67.5) treatment with PD and both failed.
Secondary outcomes. Linear mixed models were
used to determine the difference in treatment effect on
IQR, interquartile range. secondary outcomes. The differences are represented as
a
Eckardt score ranges from 0 to 12, with a higher score
indicating more severe symptoms.
parameter estimates (Supplementary Table 1 and Table 3).
b
ADSQoL score ranges from 10 to 33, with a lower score Linear mixed models adjusted for time showed a significant
indicating a better quality of life. difference in Eckardt score, basal LES pressure, IRP-4, and
c
SF-36 score consisted of a Physical Component Summary barium contrast height at T ¼ 2 minutes and 5 minutes, at
score and Mental Component Summary score, each ranging 1-year follow-up in favor of POEM. This study did not find
from 0 to 100, with higher scores indicating better quality of significant differences in the maximum esophageal width
life. measured during timed barium esophagogram between
emphysema after PD (n ¼ 1), mild bleeding during PD POEM and PD using linear mixed models. With regard to the
managed conservatively (n ¼ 1), an allergic reaction after ADSQoL and SF-36 scores, linear mixed models showed
endoscopy (n ¼ 1), and not-upper-gastrointestinal–related similar results as the classical statistical analysis (Table 3).
AEs (n ¼ 4). By using generalized linear models, the association be-
tween treatment and binary outcomes, such as the occur-
rence of reflux esophagitis and reflux symptoms, could be
Post-Hoc Sensitivity Analysis determined. The strength of this association is represented
Primary outcome. Post-hoc sensitivity analysis of the as a b-coefficient. With the generalized linear model, it was
primary outcome was performed by looking at the data with also possible to adjust for certain confounding factors, such
the “per-protocol” principle. Within the POEM group, 2 pa- as PPI use within the first year of follow-up. Both with and
tients received PD as the primary treatment after random- without adjustment, there was no significant association
ization to POEM because of fibrotic mucosa, which between treatment and occurrence of reflux esophagitis and
prohibited the performance of POEM. Furthermore, 1 reflux symptoms (Supplementary Table 1). The same
 1114 Saleh et al Gastroenterology Vol. 164, Iss. 7

Table 2.Primary Outcome of Patients With Achalasia at 1-Year Follow-Up After POEM or PD as Intention-to-Treat Analysis

Unadjusted
1-y follow-up POEM, n (%) PD, n (%) Odds Relative absolute difference,
primary end point (n ¼ 45) (n ¼ 45) P Value ratio (95% CI) risk (95% CI) % (95% CI)

Overall treatment 28 (62.2) 12 (26.7) .001 0.22 (0.09–0.54)a 2.33 (1.37–3.99)b 35.6 (16.4–54.7)
success

a
POEM is less likely to result in failure than PD.
b
Relative risk for success, success was 2.33 times more likely in patients randomized to receive POEM.
ESOPHAGUS

applies to the use of a PPI within the first year of follow-up and barium height at T ¼ 2 minutes and 5 minutes, in
(Supplementary Table 1). These results fall in line with the favor of POEM at 1-year follow-up. Importantly, no statis-
classical statistical analysis presented above and thereby tically significant differences between groups were
showed consistent results. measured for occurrence of reflux esophagitis, PPI use, and
reflux symptoms. When looking at treatment effect on the
quality of life, a significant difference in ADSQOL score was
Discussion found, again favoring POEM. However, for quality of life
This randomized controlled clinical study demonstrated measured by the SF-36, significant differences were
that POEM is more efficacious than PD as rescue therapy for observed for only 3 of 8 components, that is, Physical
patients with achalasia who experience persistent or Functioning, Emotional Well-Being, and Social Functioning.
recurrent symptoms after LHM. With respect to safety, there were 2 treatment-related
Regarding the secondary outcomes parameters, this SAEs, including a microperforation caused by POEM,
study found significant differences in LES pressure, IRP-4, which was treated with antibiotics and 2 days of admission,

Table 3.Secondary Objective Outcomes After 1-Year Follow-Up After POEM or PD

POEM vs PD

Parameter
Variable POEM, mean (SD) PD, mean (SD) P Valuea estimateb SE P Valuec 95% CI

Eckardt score 2.95 (1.44) 3.77 (1.78) .016 –0.788 0.361 .031 –1.505 to –0.071
LES pressure, mmHg 14.81 (7.37) 19.97 (9.99) .034 –4.95 2.41 .043 –9.73 to –0.160
IRP-4, seconds 9.64 (4.96) 15.62 (9.08) .002 –5.998 1.727 .001 –9.425 to –2.571
Barium height T ¼ 2 min 2.97 (1.74) 4.64 (2.90) .005 –1.658 0.592 .006 –2.833 to –0.483
Barium height T ¼ 5 min 2.47 (1.77) 4.02 (2.89) .015 –1.558 0.592 .01 –2.732 to –0.384
Maximum width, cm 3.23 (1.25) 3.65 (1.36) .121 –0.363 0.283 .203 –0.925 to 0.199

Variable POEM, n (%) PD, n (%) P Valuea b-coefficientd SE P Valuee 95% CI

Reflux 12/35 (34.3) 6/40 (15) .062 –0.770 (1.022f) 0.658 (1.398f) .242 (.465f) –2.06 to 0.530
esophagitis (n ¼ 75) (–3.762 to 1.718 e)
Grade A 7/35 (20) 4/40 (10) NA NA NA NA NA
Grade B 4/35 (11.4) 1/40 (2.5) NA NA NA NA NA
Grade C 1/35 (2.9) 1/40 (2.5) NA NA NA NA NA
Grade D 0 (0) 0 (0) NA NA NA NA NA
PPI use (n ¼ 87) 29 (69) 26 (57.8) 0.374 –0.489 0.45 0.277 –1.37 to 0.393

a
P value for the difference in outcome of continuous data analyzed using Mann-Whitney U test and categorical data using X 2
test between treatment groups at 1-year follow-up.
b
Parameter estimates represent the difference in outcome of continuous data between treatment groups at 1-year follow-up,
adjusted for repeated measurements over 1 year time; measured by linear mixed models with PD as the reference treatment.
c
P value for parameter estimates as measured by linear mixed models with PD as the reference treatment.
d
b-coefficients represent the association between categorical data at 1-year follow-up and the treatment groups; measured by
generalized linear models using PD as the reference treatment.
e
P value for b-coefficients as measured by generalized linear models with PD as the reference treatment.
f
Results of generalized linear models adjusted for PPI use during 1-year follow-up.
June 2023 POEM vs PD After Failed Laparoscopic Heller Myotomy 1115

ESOPHAGUS
Figure 3. Mean basal LES pressure, IRP-4, barium column height at 2 and 5 minutes over 1 year, and presence of reflux
esophagitis at 1-year follow-up for POEM and PD.

and extreme reflux symptoms as a result of PD, which were clinical success rates ranged between 81% and
treated with a Toupet fundoplication. In contrast to studies 96%.9,20,26,29–33 This discrepancy cannot be attributed to a
comparing POEM and PD for treatment-naïve patients with difference in the definition of success, as these studies also
achalasia, this study did not find a statistically significant defined clinical success as an Eckardt score 3. However,
difference in development of reflux esophagitis, experience most of these studies included small samples, had a
of reflux symptoms, and use of PPI between patients treated retrospective design in which inclusion in the cohort was
with POEM and patients treated with PD.21 determined afterward, and presented shorter follow-up
To our knowledge, this is the first randomized times, which could explain higher success rates.8,9,20 This
controlled trial that compared POEM with PD as treatment study’s observed success rate of 62.2% at 1 year should be
in patients with achalasia experiencing persistent or considered a medium-term outcome. Longer follow-up
recurrent symptoms after LHM. The efficacy rate of POEM data will help provide information about the duration of
observed in our study (62.2%) was lower compared with the treatment effect. Moreover, our data confirmed the
uncontrolled prospective and retrospective studies, where low-risk nature of POEM.

Figure 4. Mean Eckardt score over 1 year for POEM and PD.
 1116 Saleh et al Gastroenterology Vol. 164, Iss. 7

As for PD, this study observed an efficacy rate of There were also limitations identified. Firstly, primary
26.7%, which is also lower than the efficacy reported by and secondary outcomes were assessed at 1-year follow-up.
published case series, where efficacy ranged from 57% to Consequently, no conclusions can be drawn for longer-term
96%.9,12,15,17,34 One reason for this discrepancy could be the treatment success, which is important, given that achalasia
heterogeneity of PD protocols used and the retrospective is a lifelong chronic disease. Secondly, like most endoscopic
nature of most published reports. In this study, patients or surgical studies that evaluate new interventional tech-
were initially treated with 30-mm and 35-mm PD (except niques, patients and caregivers were not blinded for treat-
for 1 patient), and in case of persistent symptoms after 3 ment allocation. Although a blinded study would have been
weeks, an additional 40-mm PD was performed. If symp- very challenging—requiring general anesthesia and admis-
toms recurred within 1 year, patients were treated with sion for the PD group and undergoing several sham PDs in
another round of PDs. Repeat series of PD is an accepted the POEM group—bias was minimized to the greatest extent
ESOPHAGUS

clinical strategy and reflects daily practice. Still, patients possible by blinding observers of diagnostic measurements
may experience another series of PDs as failed treatment. to the patients’ treatment; questionnaires were filled in by
Indeed, in this study, a few patients randomized to the PD patients without the presence of research personnel; and
arm refused additional rounds of PD when they experienced indication for an unscheduled treatment was based solely
persistent or recurrent symptoms after their first round of on the previously set cutoff, that is, Eckardt score >3. Lastly,
PD. These patients were considered failures, and some multiple PD sessions might form a potential bias in the
received POEM in consultation with their physician; 14 comparison with 1 treatment intervention; however, as
patients were re-treated with POEM after failed PDs within stated before, this was done deliberately to optimally reflect
their first-year of follow-up. Furthermore, in this study, routine clinical care in these patients.
pressurization of balloons started at 5 psi for 1 minute, In conclusion, among patients with achalasia experi-
followed by 7 psi for another minute. Although this might encing persistent or recurrent symptoms after LHM,
differ from other protocols, it is important to realize waist treatment with POEM resulted in a significantly higher
obliteration was obtained in all PDs, most of these already success rate compared with PD, with a numerically (not
occurring with 5 psi. Therefore, it was considered unlikely statistically significant) higher incidence of grade A–B
that the difference in pressurizations used played a role in reflux esophagitis. These findings support the consider-
the high degree of PD failure. ation of POEM as the initial treatment option for patients
Although POEM is more invasive and requires more with achalasia experiencing persistent or recurrent symp-
technical endoscopic skills, the risk of severe complications toms after LHM.
was not higher than was seen with PD. Data from this study
suggest that in previously treated patients with achalasia,
POEM did lead to more grade A–B reflux esophagitis,
Supplementary Material
Note: To access the supplementary material accompanying
although this was not statistically significant. This was
this article, visit the online version of Gastroenterology at
most likely the result of the small number of events in this
www.gastrojournal.org, and at https://doi.org/10.1053/
subgroup analysis. However, POEM did not conduce more
j.gastro.2023.02.048.
reflux symptoms or PPI use than PD. Interestingly, it was
after a series of PD that 1 patient experienced severe reflux
symptoms and required a Toupet fundoplication. References
Taking into account the efficacy rate, occurrence of
1. Campos GM, Vittinghoff E, Rabl C, et al. Endoscopic and
complications, and presence of reflux esophagitis and reflux surgical treatments for achalasia. Ann Surg 2009;
symptoms within the clinical context, it seems reasonable to 249:45–57.
offer POEM as the primary treatment option for patients 2. Oude Nijhuis RAB, Prins LI, Mostafavi N, et al. Factors
with achalasia experiencing persistent or recurrent symp- associated with achalasia treatment outcomes: system-
toms after LHM. atic review and meta-analysis. Clin Gastroenterol Hep-
atol 2020;18:1442–1453.
Strengths and Limitations 3. Ortiz A, de Haro LFM, Parrilla P, et al. Very long-term
The strengths of this randomized controlled trial are objective evaluation of Heller myotomy plus posterior
the substantial number of patients included, particularly partial fundoplication in patients with achalasia of the
given the rare nature of this disorder, and the stratifica- cardia. Ann Surg 2008;247:258–264.
tion of the randomization by center. In addition, this 4. Moonen A, Annese V, Belmans A, et al. Long-term results
study used objective measures at baseline to determine of the European Achalasia Trial: a multicentre randomized
the nature of the persistent or recurrent symptoms and controlled trial comparing pneumatic dilation versus
the eligibility for the trial. The objective measures were laparoscopic Heller myotomy. Gut 2016;65:732–739.
also used to analyze treatment effect and esophageal 5. Rakita S, Villadolid D, Kalipersad C, et al. Outcomes
function. Concerning the statistical methods: primary data promote reoperative Heller myotomy for symptoms of
analysis was performed according to the intention-to-treat achalasia. Surg Endosc 2007;21:1709–1714.
principle. Nonetheless, to increase the credibility and 6. Devaney EJ, Iannettoni MD, Orringer MB, et al. Esoph-
strength of the forthcoming conclusions, a sensitivity agectomy for achalasia: patient selection and clinical
analysis was implemented.29 experience. Ann Thorac Surg 2001;72:854–858.
June 2023 POEM vs PD After Failed Laparoscopic Heller Myotomy 1117

7. Fernández-Ananín S, Fernández A, Balagué C, et al. 23. Kahrilas PJ, Bredenoord AJ, Fox M, et al. The Chicago
What to do when Heller’s myotomy fails? Pneumatic Classification of Esophageal Motility Disorders, v3.0.
dilatation, laparoscopic remyotomy or peroral endo- Neurogastroenterol Motil 2015;27:160–174.
scopic myotomy: a systematic review. J Minim Access 24. de Oliveira JM, Birgisson S, Doinoff C, et al. Timed
Surg 2018;14:177. barium swallow: a simple technique for evaluating
8. Milito P, Siboni S, Lovece A, et al. Revisional therapy for esophageal emptying in patients with achalasia. AJR Am
recurrent symptoms after Heller myotomy for achalasia. J Roentgenol 1997;169:473–449.
J Gastrointest Surg 2022;26:64–69. 25. Lundell LR, Dent J, Bennett JR, et al. Endoscopic
9. Weche M, Saad AR, Richter JE, et al. Revisional pro- assessment of oesophagitis: clinical and functional cor-
cedures for recurrent symptoms after Heller myotomy relates and further validation of the Los Angeles Classi-
and per-oral endoscopic myotomy. J Laparoendosc Adv fication. Gut 1999;45:172–180.

ESOPHAGUS
Surg Tech A 2020;30:110–116. 26. Zhou P, Li Q, Yao L, et al. Peroral endoscopic remyot-
10. Felix VN, Murayama KM, Bonavina L, et al. Achalasia: omy for failed Heller myotomy: a prospective single-
what to do in the face of failures of Heller myotomy. Ann center study. Endoscopy 2013;45:161–166.
N Y Acad Sci 2020;1481:236–246. 27. Onimaru M, Inoue H, Ikeda H, et al. Peroral endoscopic
11. Zaninotto G, Costantini M, Portale G, et al. Etiology, myotomy is a viable option for failed surgical esoph-
diagnosis, and treatment of failures after laparoscopic agocardiomyotomy instead of redo surgical Heller
Heller myotomy for achalasia. Ann Surg 2002; myotomy: a single center prospective study. J Am Coll
235:186–192. Surg 2013;217:598–605.
12. Saleh CMG, Ponds FAM, Schijven MP, et al. Efficacy of 28. von Renteln D, Inoue H, Minami H, et al. Peroral endo-
pneumodilation in achalasia after failed Heller myotomy. scopic myotomy for the treatment of achalasia: a pro-
Neurogastroenterol Motil 2016;28:1741–1746. spective single center study. Am J Gastroenterol 2012;
13. Kumbhari V, Behary J, Szczesniak M, et al. Efficacy and 107:411–417.
safety of pneumatic dilatation for achalasia in the treat- 29. Thabane L, Mbuagbaw L, Zhang S, et al. A tutorial on
ment of post-myotomy symptom relapse. Am J Gastro- sensitivity analyses in clinical trials: the what, why, when
enterol 2013;108:1076–1081. and how. BMC Med Res Methodol 2013;13:92.
14. Legros L, Ropert A, Brochard C, et al. Long-term results 30. Vigneswaran Y, Yetasook AK, Zhao JC, et al. Peroral
of pneumatic dilatation for relapsing symptoms of endoscopic myotomy (POEM): feasible as reoperation
achalasia after Heller myotomy. Neurogastroenterol Motil following Heller myotomy. J Gastrointest Surg 2014;
2014;26:1248–1255. 18:1071–1076.
15. Amani M, Fazlollahi N, Shirani S, et al. Assessment of 31. Zhang X, Modayil RJ, Friedel D, et al. Per-oral endo-
pneumatic balloon dilation in patients with symptomatic scopic myotomy in patients with or without prior Heller’s
relapse after failed Heller myotomy: a single center myotomy: comparing long-term outcomes in a large U.S.
experience. Middle East J Dig Dis 2015;8:57–62. single-center cohort (with videos). Gastrointest Endosc
16. Iqbal A, Tierney B, Haider M, et al. Laparoscopic re- 2018;87:972–985.
operation for failed Heller myotomy. Dis Esophagus 32. Ngamruengphong S, Inoue H, Ujiki MB, et al. Efficacy
2006;19:193–199. and safety of peroral endoscopic myotomy for treatment
17. Ellis FH. Failure after esophagomyotomy for esophageal of achalasia after failed Heller myotomy. Clin Gastro-
motor disorders. Causes, prevention, and management. enterol Hepatol 2017;15:1531–1537.e3.
Chest Surg Clin North Am 1997;7:477–487; ; discussion 488. 33. Akimoto S, Yano F, Omura N, et al. Redo laparoscopic
18. Gockel I. Persistent and recurrent achalasia after Heller Heller myotomy and Dor fundoplication versus rescue
myotomy. Arch Surg 2007;142:1093. peroral endoscopic myotomy for esophageal achalasia
19. Santes O, Coss-Adame E, Valdovinos MA, et al. Does after failed Heller myotomy: a single-institution experi-
laparoscopic reoperation yield symptomatic improve- ence. Surg Today 2022;52:401–407.
ments similar to those of primary laparoscopic Heller 34. Guardino JM, Vela MF, Connor JT, et al. Pneumatic
myotomy in achalasia patients? Surg Endosc 2021; dilation for the treatment of achalasia in untreated pa-
35:4991–5000. tients and patients with failed Heller myotomy. J Clin
20. Huang Z, Cui Y, Li Y, et al. Peroral endoscopic myotomy Gastroenterol 2004;38:855–860.
for patients with achalasia with previous Heller myotomy:
a systematic review and meta-analysis. Gastrointest
Endosc 2021;93:47–56.e5. Received October 4, 2022. Accepted February 18, 2023.
21. Ponds FA, Fockens P, Lei A, et al. Effect of peroral
Correspondence
endoscopic myotomy vs pneumatic dilation on symptom Address correspondence to: Albert J. Bredenoord, MD, PhD, Department of
severity and treatment outcomes among treatment-naive Gastroenterology and Hepatology, Academic University Medical Center, PO
patients with achalasia: a randomized clinical trial. JAMA Box 22660, 1100 DD Amsterdam, The Netherlands. e-mail:
[email protected].
2019;322:134–144.
22. Brazier JE, Harper R, Jones NM, et al. Validating the SF- CRediT Authorship Contributions
Caroline M. G. Saleh, MSc (Conceptualization: Equal; Formal analysis: Lead;
36 health survey questionnaire: new outcome measure Investigation: Equal; Methodology: Equal; Supervision: Equal; Validation:
for primary care. BMJ 1992;305(6846):160–164. Equal; Writing – original draft: Lead; Writing – review & editing: Equal).
 1118 Saleh et al Gastroenterology Vol. 164, Iss. 7

Pietro Familiari, MD (Formal analysis: Supporting; Methodology: Supporting; Albert J. Bredenoord, MD, PhD (Conceptualization: Equal; Formal analysis:
Resources: Supporting; Supervision: Supporting; Writing – review & editing: Supporting; Funding acquisition: Lead; Investigation: Equal; Methodology:
Supporting). Equal; Supervision: Equal; Writing – review & editing: Equal).
Barbara A. J. Bastiaansen, MD, PhD (Formal analysis: Supporting;
Methodology: Supporting; Resources: Supporting). Conflicts of interest
Paul Fockens, MD, PhD (Formal analysis: Supporting; Methodology: These authors disclose the following: Albert J. Bredenoord received research
Supporting; Resources: Supporting; Supervision: Supporting; Writing – funding from Nutricia, Norgine, DrFalkPharma, Gossamer, Thelial, SST, and
review & editing: Supporting). Bayer, and received speaker and/or consulting fees from Laborie, Arena,
Jan Tack, MD, PhD (Formal analysis: Supporting; Investigation: Supporting; EsoCap, Medtronic, Dr. Falk Pharma, Calypso Biotech, Alimentiv, Sanofi/
Methodology: Supporting; Resources: Supporting; Supervision: Supporting). Regeneron, Reckett, and AstraZeneca. Pietro Familiari consults for Olympus
Guy Boeckxstaens, MD, PhD (Formal analysis: Supporting; Investigation: and Cook Medical. The remaining authors disclose no conflicts.
Supporting; Resources: Supporting; Supervision: Supporting).
Raf Bisschops, MD, PhD (Formal analysis: Supporting; Investigation: Funding
Supporting; Methodology: Supporting; Resources: Supporting; Supervision: This study was made possible with financial support from Fonds NutsOhra
Supporting). (FNO grant 1402-002).
ESOPHAGUS

Aaltje Lei, BSc (Formal analysis: Supporting; Investigation: Equal;

Methodology: Supporting; Resources: Equal). Data Availability
Marlies P. Schijven, MD, PhD (Writing – review & editing: Supporting). Data, analytic methods, and study materials are available to other researchers
Jan Guido Costamagna, MD, PhD (Formal analysis: Supporting; upon request. The study protocol is included as Supplementary Material
Investigation: Supporting; Methodology: Supporting; Resources: Supporting). available with the online version of this article.
June 2023 POEM vs PD After Failed Laparoscopic Heller Myotomy 1118.e1

Supplementary Material liver cirrhosis and ischemic cardiomyopathy, which was

identified shortly after receiving POEM; this is the only SAE
not related to a treatment that resulted in early discontin-
Serious Adverse Events Independent of Study uation of the study.
Intervention For the group of patients treated with PD, there were 2
Within the group of patients treated with POEM, there SAEs not related to treatment. One patient was diagnosed
were 3 SAEs not related to the intervention. One patient was with a viral pericarditis, for which they were admitted and
hospitalized because of worsened dysphagia symptoms, treated with nonsteroidal anti-inflammatory drugs. The
which required placement of a feeding tube; symptoms second patient was diagnosed with a, probably viral,
resolved spontaneously and the feeding tube could be pneumonia; because this SAE occurred shortly after PD, it
removed swiftly. In another patient, an atrium septum was doubtful whether it was or was not related to the
defect type II (birth defect) was discovered during the trial, intervention. The patient required admission and was
which was treated operatively. Lastly, 1 patient died due to treated with antibiotics and antivirals.
1118.e2 Saleh et al Gastroenterology Vol. 164, Iss. 7

Supplementary Figure 1. Eckardt score range from baseline to follow-up year 1 for POEM and PD.
June 2023 POEM vs PD After Failed Laparoscopic Heller Myotomy 1118.e3

Supplementary Table 1.Secondary Subjective Outcomes After 1-Year Follow-Up After POEM or PD

POEM vs PD

POEM, PD, Parameter

Variable median (IQR) median (IQR) P valuea estimateb SE P valuec 95% CI

ADSQoL score 18 (15–21.25) 21 (16.5–24) .023 –2.535 1.046 .017 –4.164 to –0.455
SF-36
General Health 65 (36.25–80) 52.5 (45–75) .636 3.258 4.902 .508 –6.504 to 13.021
Physical Functioning 95 (81.25–100) 80 (55 – 95) .002 10.945 5.320 .043 0.359 to 21.531
Limitations Due to 100 (0–100) 50 (0–100) .217 12.493 8.964 .167 –5.333 to 30.319
Physical Health
Limitations Due to 100 (33.3–100) 100 (0–100) .110 12.838 8.703 .144 –4.465 to 30.141
Emotional Problems
Energy/Fatigue 62.5 (45–82.5) 55 (42.5–72.5) .334 5.597 4.938 .260 –4.216 to 15.409
Emotional Well-Being 84 (76–92) 76 (48–84) .007 12.458 4.294 .005 –3.926 to 20.991
Social Functioning 87.5 (75–100) 75 (56.25–87.5) .005 16.186 5.729 .006 4.803 to 27.569
Pain 77.5 (60–90) 67.5 (45–90) .096 9.459 5.263 .076 –1.001 to 19.918
GERDQ 7 (6–9.75) 8 (7–10) .395 –0.500 0.587 .396 –1.665 to 0.644

Variable POEM, n (%) PD, n (%) P value b-coefficientd SE P valuee 95% CI

GERDQ
8(n ¼ 79) 18/40 (45) 22/39 (56.4) .371 –0.458 (–0.649 f) 0.453 (1.012 f) .312 (.521f) –1.347 to 0.430
(–2.633 to 1.334f)

GERDQ, Gastroesophageal Reflux Disease Questionnaire; IQR, interquartile range.

a
P value for the difference in outcome of continuous data analyzed using Mann-Whitney U test and categorical data using X 2
test between treatment groups at 1-year follow-up.
b
Parameter estimates represent the difference in outcome of continuous data between treatment groups at 1-year follow-up,
adjusted for repeated measurements over 1 year; measured by linear mixed models with PD as the reference treatment.
c
P value for parameter estimates as measured by linear mixed models with PD as the reference treatment.
d
b-coefficients represent the association between categorical data at 1-year follow-up and the treatment groups; measured by
generalized linear models using PD as the reference treatment.
e
P value for b-coefficients as measured by generalized linear models with PD as the reference treatment.
f
Results of generalized linear model adjusted for PPI use during 1-year follow-up.