This retrospective cohort study was conducted under the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines [16]. This study included consecutive patients who presented with heartburn symptoms despite standard-dose vonoprazan (20 mg) or standard-dose PPIs (lansoprazole of 30 mg, rabeprazole of 20 mg, and esomeprazole of 20 mg) for at least 8 weeks from September 2013 to November 2023, and with no evidence of erosive esophagitis and Barrett’s esophagus (> 1 cm of columnar-lined esophagus) on endoscopy (unproven GERD) and underwent MII–pH monitoring at Chiba University Hospital. Patients with PPI-refractory heartburn were defined as those who had not previously taken vonoprazan. On the other hand, patients with vonoprazan-refractory heartburn were not asked about their previous history of taking PPIs. Heartburn symptoms were confirmed with the frequency scale for the symptoms of GERD (FSSG) questionnaire and gastroesophageal reflux disease questionnaire (GERD-Q) [17, 18]. Patients were categorized into vonoprazan-refractory and PPI-refractory heartburn, and differences in their symptoms and test results were investigated. The Ethics Committee at Chiba University Hospital reviewed and approved this study, conducted in compliance with the Helsinki Declaration (approval number 3873). Informed consent was obtained from all patients to undergo the procedures involved.

All participants completed four questionnaires. The questionnaires were answered on the day of HRM and MII–pH testing, under vonoprazan/PPI discontinuation. The FSSG questionnaire consists of 12 items categorized into two domains. Each item was scored from 0 (never) to 4 (always). Dysmotility-like symptoms (dyspepsia score), the first domain, were calculated by summing the scores (range 0–28) for items 1, 4, 6, 7, 9, 10, and 12. Acid reflux-related symptoms (reflux score), the second domain, involved summing the scores (range 0–20) for items 2, 3, 5, 8, and 11 [17]. The GERD-Q is a self-reported questionnaire designed to diagnose GERD, comprising six items. Each item was rated on a 4-point scale. The scoring for items 1, 2, 5, and 6 is as follows: 0 points for no symptoms, 1 point for symptoms on 1 day, 2 points for 2–3 days, and 3 points for 4–7 days. The scoring for items 3 and 4 is reversed: 3 points for no symptoms, 2 points for 1 day, 1 point for 2–3 days, and 0 points for 4–7 days. Patients reported their symptoms over the previous week [18]. The Gastrointestinal Symptom Rating Scale (GSRS) questionnaire comprises 15 items and measures 5 categories of gastrointestinal symptoms, including reflux, abdominal pain, indigestion, diarrhea, and constipation. Values for each of the category scores were calculated as the mean of the respective items. The total GSRS score was calculated as the mean of all 15 items. The scoring is based on a Likert scale from 1 (minimal gastrointestinal symptoms) to 7 (most severe symptoms) points [19]. The Short Form-8 (SF-8) is a measure of health that generates physical and mental component summaries. Participants were asked eight questions about their health during the past 4 weeks to measure two scores, both between 0 and 100. The higher the score, the better the health state [20].

The MII–pH test was performed after discontinuing vonoprazan/PPI medication for > 7 days. After an overnight fast, the MII–pH catheter was inserted for 24 h. The catheters included Diversatek (formerly Sandhill Scientific, Boulder, Colorado, USA). MII–pH was used for diagnosis in accordance with the Lyon consensus [21, 22]. The MII–pH data were manually analyzed by three experienced investigators (M.S., T.M., and H.D.) using the BioView™ analysis software (Sandhill Scientific, Inc.). Patients were classified into three groups: true NERD, RH, and FH. AET > 6% was defined as true NERD, and AET of 4–6% with the presence of other impedance parameters that support GERD was also defined as true NERD. The other impedance parameters supporting GERD included excess total number of physiological reflux episodes (> 80), low post-reflux swallow-induced peristaltic wave (PSPW) index (≤ 50%), and low mean nocturnal baseline impedance (MNBI, ≤ 1500 Ω). RH was defined as AET ≤ 6%, not belonging to the above criteria, and a positive symptom index (> 50%) and/or positive symptom association probability (> 95%). FH was defined as AET ≤ 6%, not belonging to the above criteria, and negative for both symptom index and symptom association probability [21, 22]. The PSPW was defined as a swallow that occurs within 30 s of the end of a reflux episode that triggers an antegrade 50% drop in impedance relative to the pre-swallow baseline, originating in the most proximal impedance site and reaching all distal impedance sites [23]. The PSPW index was calculated by dividing the number of PSPWs by the number of total refluxes, TRs [24]. The MNBI was evaluated from the most distal impedance channel during night-time recumbence without reflux episodes, swallows, or pH drops [21, 25]. The mean of three measurements, each lasting at least 30 min, was manually calculated to obtain the MNBI [21].

All patients underwent esophagogastroduodenoscopy to confirm the absence of erosive esophagitis and Barrett’s esophagus. Erosive esophagitis was classified as A–D according to the Los Angeles classification [26], and Barrett’s esophagus was identified as the replacement of normal epithelium with columnar epithelium in the distal esophagus between the esophagogastric junction (EGJ) and the squamocolumnar junction (SCJ), as assessed by the Prague classification system [27]. The present study excluded patients with confirmed erosive esophagitis and Barrett’s esophagus (> 1 cm of columnar-lined esophagus) as patients with proven GERD.

The HRM systems used included the Diversatek system (Boulder, Colorado, USA) with 32 pressure sensors and 16 impedance sensors. EMD diagnosis followed the criteria set by the Chicago Classification System, version 4.0 [28]. Patients with disorders of EGJ outflow and hypercontractile esophagus were excluded. The esophagogastric junction contractile integral (EGJ-CI) serves as an index to evaluate the barrier function of the EGJ. It is calculated using the distal contractile integrals (DCI) box, comprising the lower esophageal sphincter and crural diaphragm over three respiratory cycles, thereby maintaining a threshold above gastric pressure. The resulting ‘DCI’ is then divided by the duration of these three respiratory cycles and expressed in mmHg cm units [21].

Categorical variables were expressed as n (%) and compared using the χ2 (chi-square) test; comparisons of True NERD, RH, and FH proportions were made by χ2 (chi-square) test with Bonferroni correction. Continuous variables were expressed as mean ± standard deviation or median and interquartile range (25th–75th percentile). Student’s t-test or Mann–Whitney U test were used to analyze these variables, respectively. All statistical analyses were performed using Statistical Package for the Social Sciences version 23 (SPSS, Inc., Chicago, IL). All authors had access to the study data and reviewed and approved the final manuscript.