Study subjects and inclusion criteria

In this prospective observational case–control study, treated adult (≥18 years) achalasia patients visiting the outpatient clinic of the Gastroenterology and Hepatology Department of the Amsterdam UMC were approached to participate in the study. Patients were allocated into two groups depending on whether or not reflux symptoms were present. Treated achalasia patients with a total score of ≥8 on the Gastro-Oesophageal Reflux Disease Questionnaire (GORDQ) were classified as having reflux symptoms (RS+) and a score ≤8 as without reflux symptoms (RS−).16 17 The GORDQ is a widely used, validated, 6-item self-report questionnaire, evaluating reflux symptoms (heartburn, regurgitation and chest pain), sleep disturbance by reflux and antacid use, range per item 0–3 with a minimum total score of 0 and a maximum score of 18.16 17 The GORDQ was completed while the patients were off acid suppression. All included achalasia patients were ≥6 months post-treatment and in clinical remission for achalasia, defined as an Eckardt score ≤3. The Eckardt symptom score assesses the severity of achalasia symptoms by the sum of symptom frequency scores for dysphagia, regurgitation and chest pain (range 0–3: 0: absent; 1: occasionally; 2: daily; 3: every meal) combined with a weight loss score (range 0–3: 0: no weight loss; 1: <5 kg weight loss; 2: 5–10 kg weight loss; 3: >10 kg weight loss) resulting in a minimum score of 0 until a maximum score, indicating most pronounced symptoms, of 12.18 In all patients, the diagnosis of achalasia was previously confirmed by oesophageal manometry before treatment, showing absent peristalsis and impaired LOS relaxation. Treatment consisted of endoscopic pneumodilation, laparoscopic Heller’s myotomy with Dor fundoplication 180° and/or peroral endoscopic myotomy (POEM). Pneumodilations started with a 30 mm balloon, followed by a 35 mm balloon and in case of persistent symptoms a 40 mm balloon was used. Detailed eligibility criteria are provided in the study protocol (online supplementary 1 and 2).

The study was registered in the Dutch trial registry before the start of the study (NTR3838, trialregister.nl). Written informed consent was obtained from all patients before study participation. Normal values for acid sensitivity in healthy subjects were obtained in a previous study.19

Study protocol

Measurements were performed on two subsequent days after cessation of PPI, H2-receptor antagonists and/or prokinetic medication for 1 week (figure 1). Baseline data and questionnaires (Eckardt score, GORDQ, Reflux Disease Questionnaire (RDQ), Achalasia Disease-Specific Quality-of-Life questionnaire and Medical Outcomes Study 36-Item Short-Form Health Survey (SF-36)) were assessed before the measurements.20 21 One day before the measurements, patients were restricted to a liquid diet, followed by an overnight fast to minimise possible oesophageal stasis. Figure 1 displays the study protocol with the subsequent measurements that were performed during two study days including time intervals. On the first day, the distensibility of the OGJ was measured and oesophageal sensitivity for a mechanical stimulus was assessed (using EndoFLIP (Endo Functional Luminal Imaging Probe)), and stationary high-resolution manometry (HRM), acid perfusion test and a prolonged combined HRM and pH-impedance monitoring were performed. Thereafter the patients were dismissed, fitted with equipment for 24 hours ambulatory pH-impedance measurement. The next day, the 24 hours reflux monitoring was terminated and a timed barium oesophagogram was performed. Oesophagogastroduodenoscopy was not part of the protocol but in all patients performed off acid suppression as part of routine clinical practice before study participation. All measurements were analysed in a blinded fashion by the investigators.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Study protocol. EndoFLIP, Endo Functional Luminal Imaging Probe; HRM, high-resolution manometry; OGJ, oesophagogastric junction; PPI, proton pump inhibitors.

OGJ distensibility measurement and assessment of perception of oesophageal distension

To measure OGJ distensibility the EndoFLIP (Medtronic, Sunnyvale, California, USA) was used. By the use of impedance planimetry, EndoFLIP measures cross-sectional areas in the alimentary tract.22 Description of the EndoFLIP and the protocol to measure OGJ distensibility was previously described.23 The OGJ distensibility was determined at the 50 mL volume by dividing the median minimal cross-sectional area, reflecting the OGJ, by the median intrabag pressure during the complete recording period, expressed as mm²/mm Hg.

After measuring OGJ distensibility, the catheter was placed 10 cm above the OGJ to evaluate perception of oesophageal distension. The EndoFLIP bag was inflated from 20 to 70 mL, with a 10 mL increase of volume each minute. After each inflation, patients were asked to notify their first perception of the mechanical stimulus and to score the intensity of their perception on a Visual Analogue Scale (VAS), a horizontal 100 mm line marked with ‘no pain’ and ‘most extreme pain’.

High-resolution manometry

HRM was performed using a 22-channel water-perfused catheter (Laborie, Williston, Vermont, USA) with an incorporated infusion channel until the end of the catheter. The proximal, first 15 channels were spaced at 2 cm intervals, followed by 6 channels at 1 cm intervals for measuring the LOS and the last channel at the end of the catheter to measure gastric pressure. The catheter was introduced transnasally and positioned to measure from hypopharynx to stomach. Following a standardised protocol, patients were placed in supine position (20°) and received 10 boluses of 5 mL water with an interval of 20 s. Prior and subsequently to the swallows, a period of 30 s not swallowing was assessed for baseline measures. Manometric signals were recorded with a frequency of 20 Hz. The HRM studies were analysed by dedicated software (Laborie, Williston, Vermont, USA), according to the Chicago classification V.3, adjusted for water-perfused values.24 25 The following key oesophageal pressure topography metrics were assessed: OGJ basal pressure at end-expiration, the 4 s integrated relaxation pressure (IRP), distal contractile integral, distal latency, peristaltic integrity using the 20 mm Hg isobaric contour and intrabolus pressure pattern with ≥30 mm Hg isobaric contour.24 26 OGJ/LOS pressure was referenced to gastric pressure and oesophageal contraction metrics to atmospheric pressure.24 26

Acid perfusion test

An acid perfusion test was performed according to a previously described protocol.19 The water-perfused HRM catheter was used with the incorporated infusion channel 6 cm above the OGJ. Patients were in semirecumbent position. After an adaptation period of 10 min, perfusion with a neutral solution (saline, NaCl 0.9% at pH 6.5) was performed for 10 min, followed by an acidic solution (0.1 n HCl at pH 1.1) for 30 min. The perfusion rate was 8 mL/min. Patients were blinded for the nature of the solutions and unaware of the switch to acid perfusion. The time to first perception of heartburn and time to discomfort were noted. Symptom severity was scored every 2 min on a VAS, a horizontal 100 mm line marked with ‘no pain’ and ‘most extreme pain’ labelled at the beginning and end of the line. Perfusion was stopped when symptoms were intolerable. A perfusion sensitivity score was calculated as follows: ((total perfusion time – lag time to perception) x maximum VAS), conform previously described methods.19 27 Patients with a first perception of heartburn within 20 min after acid perfusion were considered increased hypersensitive to acid.19

Postprandial stationary HRM and pH-impedance measurement

Combined HRM and pH-impedance monitoring was performed after the acid perfusion test. The pH-impedance catheter consisted of 6 impedance segments and 1 ISFET pH electrode (Unisensor AG, Attikon, Switzerland) and was placed next to the HRM catheter, with the pH electrode 5 cm above the upper border of the LOS. The impedance segments were located at 2–4, 4–6, 6–8, 8–10, 14–16 and 16–18 cm above the upper border of the LOS. A second pH catheter without impedance electrodes was placed 10 cm above the upper border of the LOS to detect the proximal extent of reflux or acidification beside the impedance measurement. Low distal baseline impedance tracings could prevent adequate detection of proximal acid exposure. Impedance (50 Hz), pH and pressure (20 Hz) signals were recorded and stored on a computer with dedicated software (Laborie, Williston, Vermont, USA). After an adaptation period of 30 min, intragastric infusion of a standardised high-caloric liquid meal, 250 mL nutrient drink (600 kcal, 18 g protein, Nutridrink Compact Protein, Nutricia, Zoetermeer, The Netherlands) diluted by 150 mL water, was started. For the perfusion the incorporated infusion channel of the HRM catheter was used with a perfusion rate of 13 mL/min during 30 min. This was followed by a postprandial measurement during 120 min. Impedance and pH tracings were analysed for acid patterns (see definitions below), acid exposure time (percentage of time pH <4), occurrence of reflux episodes according to previously described criteria and discriminating reflux from fermentation.28 29 Combined HRM and pH-impedance monitoring was used to detect mechanisms of acid exposure (eg, swallow induced, transient LOS relaxation) and clearance.

Twenty-four-hour ambulatory pH-impedance monitoring

After the combined HRM and pH-impedance monitoring, the HRM and single pH catheter were removed. The pH-impedance catheter stayed in place and was used for a 24-hour ambulatory measurement. The catheter was connected to a digital data logger (Laborie, Williston, Vermont, USA) to store pH and impedance signals at a frequency of 50 Hz. During the measurement, patients were instructed to consume meals and drinks at fixed times during the day and report symptoms in a diary. Analysing the ambulatory pH-impedance measurements, we distinguished five different acidification patterns: (A) acid reflux with normal clearance: rapid pH drop to below 4, drop rate ≥1 pH unit per second, lasting between 10 s and 5 minutes28 29; (B) acid reflux with delayed clearance: rapid pH drop to below 4, drop rate ≥1 pH unit per second, lasting longer than 5 min; (C) acid fermentation: slow pH drop to below 4, drop rate <1 pH unit per minute, lasting longer than 5 min; (D) stasis of recently ingested acidic food or drink: pH drop to below 4 during meal/drink, persisting longer than 5 min after meal/drink; (E) unclassified: pH drop to below 4 not meeting criteria for any of the acid patterns described above (figure 2). Low distal baseline impedance levels prevented the use of impedance for defining the observed acidification patterns. Impedance was used to identify prolonged acidification by further decrease of impedance levels (distal and proximal), clearance of acidification and air trapping. All acid episodes (pH <4) were analysed according to the predefined acidification patterns. Total acid exposure time, percentage of time pH <4, during the complete measurement and in upright and supine position were assessed. An acid exposure time >6% was considered pathological.29 The correlation between symptoms and acid patterns was analysed, with a positive correlation when symptoms were notified within 2 min from the start of the acid pattern. The Symptom Index (SI) was calculated by the number of symptoms associated with reflux as a percentage of the total number of symptoms. The optimal SI threshold was set at ≥50% of reported reflux symptoms.27 The symptom association probability was not calculated because in patients with achalasia the number of total acid reflux episodes cannot be determined reliably. Baseline impedance levels were measured every 2 hours in the proximal channel at 17 cm above the LOS and in the most distal channel at 3 cm above the LOS, as previously described.30 A 30 s time window was selected to calculate the baseline impedance by averaging the raw impedance values during this time period. The median values for proximal and distal impedance levels were calculated for the 24-hour measurement based on the 2 hours data.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Oesophageal acidification patterns observed during pH-impedance monitoring. (A) Acid reflux with normal clearance: rapid pH drop to below 4, drop rate ≥1 pH unit per second, lasting between 10 s and 5 min. (B) Acid reflux with delayed clearance: rapid pH drop to below 4, drop rate ≥1 pH unit per second, lasting longer than 5 min. (C) Acid fermentation: slow pH drop to below 4, drop rate <1 pH unit per minute, lasting longer than 5 min. (D) Stasis of recently ingested acidic food or drink: pH drop to below 4 during meal/drink, pH below 4 persisting longer than 5 min after meal/drink. The pink-coloured area indicates meal/drink ingestion. (E) Unclassified: pH drop to below 4 not meeting criteria for any of the acid patterns described above.

Timed barium oesophagogram and oesophagogastroduodenoscopy

A timed barium oesophagogram was performed on the second day, after the 24-hour pH-impedance monitoring. Patients were instructed to ingest a maximal tolerable amount of low density barium sulphate suspension up to 200 mL within 30–60 s in an upright, slight left posterior oblique position.31 Radiographs were taken at 0, 1, 2 and 5 min after ingestion of the suspension.31 To evaluate oesophageal emptying the barium column height at 5 min was measured from the OGJ to the top of the barium column in centimetres. Adequate oesophageal emptying was defined as ≤1 cm barium column height at 5 min. The maximal oesophageal diameter was measured by the oesophageal width at 5 min.

All patients had undergone an oesophagogastroduodenoscopy off PPI within 24 months before study participation. Severity of reflux oesophagitis was scored according to the Los Angeles classification.32

Statistical analysis

An assumed 45% difference in objectified reflux episodes, between reflux symptomatic and asymptomatic treated achalasia patients, was hypothesised for the purpose of sample size calculations. With 19 patients in each group, the study would have 80% power to detect significant differences in outcome parameters that could give insight in the underlying mechanism of reflux symptoms. To compensate for technical failures, the aim was to enrol 20 patients per group, with a 1:1 allocation per group based on treatment type and gender.

Continuous data are presented as mean (SD or 95% CI) or median (IQR), according to distribution. Categorical data are presented as percentages. Continuous data were compared by unpaired Student’s t-test and one-way analysis of variance or Mann-Whitney U test and Kruskal-Wallis, according to distribution. Categorical data were analysed by χ² and Fisher’s exact test. Relationships between parameters were analysed by linear regression analysis, (Pearson’s or Spearman’s correlation, r) or logistic regression analysis (OR with 95% CI). The time to perception during the acid perfusion test, was compared by the log rank test on Kaplan-Meier curves. A subgroup analysis of the outcome parameters of 24 hours pH-impedance, oesophagogastroduodenoscopy and provocation tests was also performed per treatment, laparoscopic Heller’s myotomy versus POEM. For both treatments, patients primarily treated by Heller’s myotomy or POEM were taken together with patients who failed on pneumodilation and retreated by Heller’s myotomy or POEM. Differences were considered statistically significant when p<0.05. All reported p values are two tailed. Statistical analysis was performed using IBM SPSS Statistics V.24 (IBM).

Patient and public involvement

Patient involvement started at study inclusion. Patients assessed the burden of the study and gave feedback for adjustments on the study design for further studies. Patients were not involved in development of the research question, outcome measures or study design. Patients’ personal results of the measurements were shared on request. After publication, the article will be disseminated to all study participants and shared on the Dutch achalasia patient Facebook page.