Kamal A.N., Clarke J.O., Oors J.M., Smout A.J., Bredenoord A.J. The Role of Symptom Association Analysis in Gastroesophageal Reflux Testing, The American Journal of Gastroenterology: Dec 2020. V.115. Issue 12. P. 1950-1959.


: Kamal A.N. / Clarke J.O. / Oors J.M. / Smout A.J.P.M. / Bredenoord A.J.


The Role of Symptom Association Analysis in Gastroesophageal Reflux Testing

Afrin N. Kamal, MD1, John O. Clarke, MD1, Jac M. Oors, RN2,
André J. Smout, MD, PhD2 and Albert J. Bredenoord, MD2


1 Division of Gastroenterology and Hepatology, Stanford University School of Medicine, Stanford, California, USA;
2 Department of Gastroenterology and Hepatology, Amsterdam University Medical Centre, Amsterdam, the Netherlands.

Correspondence: Albert J. Bredenoord. E-mail: a.j.bredenoord@ amc.uva.nl.

The American Journal of Gastroenterology: December 2020 - Volume 115 - Issue 12 - p 1950-1959, doi: 10.14309/ajg.0000000000000754

Abstract
Gastroesophageal reflux disease is characterized by the reflux of gastric contents into the esophagus with an estimated worldwide prevalence of 8%33%. The current paradigm in gastroesophageal reflux disease diagnosis relies on recognition of symptoms and/or the presence of mucosal disease at the time of esophagogastroduodenoscopy. Recognition of symptoms, however, can arise with challenges, particularly when patients complain of less typical symptoms. Since first reported in 1969 by Spencer et al., the application of prolonged intraesophageal pH monitoring to identify pathologic reflux has evolved considerably. Utility of pH monitoring aims to investigate the degree of acid burden and frequency of reflux episode, and the relationship between symptoms and acid reflux events. This relationship is represented by either the Symptom Index, Symptom Sensitivity Index, Symptom Association Probability, or Ghillebert Probability Estimate. This article reviews symptom-association analysis during esophageal reflux testing, covering the literature on current methods of reflux testing, interpretation of symptom association, and practical issues that can arise during symptom analysis.

INTRODUCTION

Gastroesophageal reflux disease (GERD) is characterized by symptoms or mucosal injury due to reflux of gastric contents into the esophagus or beyond and has an estimated worldwide prevalence of 8%33%. Symptom presentation can vary, ranging between typical to atypical or extraesophageal symptoms (1). Typical symptoms consist of acid regurgitation or heartburn, whereas atypical presentations can consist of chest pain, chronic cough, asthma, or laryngitis (2).

The current paradigm in GERD diagnosis relies on recognition of symptoms and/or the presence of mucosal disease at the time of esophagogastroduodenoscopy. Recognition of symptoms, however, can be challenging, particularly when patients complain of less typical symptoms that manifest with ear, nose, throat, pulmonary, or cardiac complaints without the classic symptoms of heartburn and/or regurgitation (3). In addition, although the classic endoscopic appearance of esophageal mucosal injury can help guide the diagnosis for symptomatic reflux disease, the conclusion from normal esophageal mucosa in the setting of symptomatic disease (nonerosive reflux disease) may not always be intuitive (4). To add to the complexity of accurately diagnosing reflux, patients with reflux hypersensitivity encompass a lower threshold for symptom perception and may experience reflux symptoms within a physiologic (i.e., normal) acid exposure environment. Therefore, recognizing these diagnostic challenges has led to the emergence of esophageal reflux monitoring to quantitate acid exposure times by pH and importantly assess the reflux-symptom relationship (58). Acid exposure times have limitations. Alone, a diagnosis of pathologic acid exposure does not relay. if patients' symptoms are in fact caused by acid reflux episodes. To overcome this limitation, reflux-symptom association indices developed to quantitate a temporal relationship between the onset of a patients' symptom (i.e., heartburn or chest pain) and actual reflux episode as seen by pH monitoring. This article aims to review on reflux-symptom association analysis during esophageal reflux testing, covering the literature on the current methods of reflux testing, interpretation of symptom association, and practical issues that can arise during symptom analysis.

AMBULATORY ESOPHAGEAL REFLUX TESTING

Catheter-based 24-hour pH monitoring
Since first reported in 1969 by Spencer et al., the application of prolonged intraesophageal pH monitoring to identify pathologic reflux has evolved considerably. The value of prolonged measurement of esophageal pH is supported by the high sensitivity (77%100%) and specificity (85%100%) of detecting excessive esophageal acid exposure in patients with endoscopically proven esophagitis when compared with normal individuals (9,10).

Ambulatory esophageal reflux testing began with a catheter-based pH monitor using a pH electrode (glass, antimony, or ion sensitive field effect transistor) attached to a flexible catheter. After calibration, the catheter is inserted transnasally and the pH sensor is positioned approximately 5 cm above the proximal border of the lower esophageal sphincter (LES). On positioning, recording of pH begins and data sampling occurs at every 45 seconds, recorded over a 24-hour period. The aim of this measurement is to capture changes in esophageal intraluminal pH during a regular day to day basis that may trigger symptoms over a 24-hour window. Patients are therefore encouraged to continue with their routine daily behaviors over the 24-hour period (11).
Wireless pH monitoring
Challenges arose with 24-hour pH monitoring because of patient complaints of discomfort and restriction to daily activities and diets, leading to the U.S. Food and Drug Administration (FDA) approval of a catheter-free pH monitoring system for clinical use. A radiotelemetry capsule fitted with an antimony pH electrode attaches endoscopically onto the esophageal mucosa, approximately 6 cm above the LES. The device subsequently transmits data to an external receiver by radiofrequency telemetry (433 MHz). Unlike the traditional catheter-based pH system, this wireless pH monitoring device allows patients to resume normal daily activities without the uncomfortable presence of a transnasal catheter. Furthermore, esophageal pH can be measured over a more prolonged period of time (i.e., 4896 hours) compared with the standard 24 hours (12,13). Despite these advantages, wireless pH measurements come with higher cost and lack of impedance measurements the primary reason why this technique has not replaced catheter-based reflux monitoring.
Multichannel pH-impedance monitoring
A combined esophageal pH monitoring and impedance (pH-MII) catheter has more recently emerged as a sensitive tool to measure acidic and nonacidic reflux and has transitioned into the more preferential tool to assess GERD symptoms. pH-impedance measurement detects flow of reflux (liquid, gas, or mixed)regardless of the acidity. For example, as a liquid bolus passes along the electrodes, given the low electrical resistance of liquid, impedance is decreased. On the other hand, the low ion density of air results in higher resistance to electrical current and higher impedance during a belch. Using this knowledge, a pH-impedance catheter probe incorporates impedance electrodes to assess direction of bolus flow along the esophageal lumen and record frequency of antegrade movement after swallowing or retrograde movement with reflux (14).

Combined esophageal pH monitoring and impedance adds the advantage of characterizing the type of reflux a feature unavailable with a standard pH probe. Reflux can be characterized into (i) acidic, (ii) weakly acidic, and (iii) alkaline reflux. Although acidic reflux is defined by a reflux event associated with a parallel drop in esophageal pH to a value <4, weakly acidic occurs at a pH between 4 and 7 and alkaline reflux occurs in the setting of an impedance reflux event without the pH dropping below 7 (13). Several studies demonstrate the importance of adding esophageal impedance monitoring to identify reflux as the cause of symptoms (1517).

ESOPHAGEAL pH AND IMPEDANCE DATA INTERPRETATION

Esophageal pH monitoring
The utility of both catheter-based and wireless pH monitoring aims to answer 2 important questions: (i) Is the quantity of (distal) esophageal acid exposure in excess to that of a normal subject? and (ii) Are the symptoms described by patients correlated with the occurrence of acid reflux? Answering these questions requires a further understanding of what is truly considered normal and what abnormal. Normally, esophageal luminal pH is slightly below 7.0, and acidic reflux episodes can be recognized as sudden drops of intraesophageal pH to a value <4.0. Beyond identifying reflux episodes, pH monitoring can quantify the severity of reflux (time with intraesophageal pH <4.0), number of episodes (i.e., drop in pH <4.0), duration of episodes, and the number of episodes longer than 5 minutes (11). It is generally agreed on that the primary outcome of pH testing and the most reproducible is the esophageal acid exposure time (time with pH <4). The cutoff level for this parameter was recently agreed on by the Lyon GERD Consensus. The group proposed that an acid exposure time <4% is definitively physiologic (i.e., normal), whereas an acid exposure time >6% is definitively abnormal (18).
Impedance monitoring
The Lyon GERD Consensus determined a cutoff of reflux episode numbers to determine what would be considered nonphysiologic. The caveat however is that reliable reflux episode measurement requires the application of impedance measurement using a combined pH-impedance catheter therefore not available with standard 24-hour or wireless pH monitoring. Furthermore, reflux episodes are often over detected by an automated analysis, and reliable measurement of the number of reflux episodes requires manual review and interpretation. Ultimately, the group proposed that >80 reflux episodes per 24-hour would be considered abnormal, whereas <40 episodes was physiologic. However, the clinical relevance of measuring the number of reflux episodes is somewhat unclear and at this point the parameter is considered adjunctive data when acid exposure time is inconclusive (i.e,. between 4% and 6%) (18). It should be recognized that there is no justification for a strict cutoff value, and the mere presence of slightly increased acid exposure does not prove that a patient's symptoms are due to reflux.

ESOPHAGEAL SYMPTOM ASSOCIATION

Limitations of acid exposure time exist and are in part related to high variability of normative values, false negative results seen in up to 20% of patients with documented reflux esophagitis (19), and the inability to provide evidence that symptoms occurred due to acid reflux. To overcome these limitations, multiple symptom association indices have been developed in an effort to describe the cause-and-effect relationship between reflux episodes and symptom development (20).

When accounting for a reflux-symptom association, the most bothersome or dominant symptom should be taken into account. Although pH allows a diagnosis of atypical GERD manifestations, only symptoms reasonably be reflux-related, such as heartburn, cough, or chest pain, should be considered for symptom analysis. Symptoms lacking a discrete onset and that are chronically present, such as sore throat or globus, should not be used for reflux-symptom association measurements (21).

Patients presenting with persistent cough can introduce challenge in determining a reflux-cough association. Etiologies for chronic cough can include postnasal drip and nonasthmatic bronchitis, in addition to GERD. Patients often arrive to a gastroenterology clinic with concern for chronic cough symptoms are induced by underlying reflux. To investigate causality of the temporal relationship, assessment encourages the use of ambulatory esophageal manometrya pH impedance catheter with combined pressure sensors to detect pressure changes in both the esophageal and gastric lumen. This application allows one to distinguish between cough induced by reflux (reflux-cough sequence) from cough contributing to reflux (cough-reflux sequence) (21,22). Therefore, when cough presents as a patient's dominant symptom, relying entirely on reflux-symptom association analysis may not accurately define the underlying etiology.
Defining measurements time window
Reflux-symptom association indices have a strong impact on decision-making as clinicians develop a treatment strategy. In the process of measuring reflux-symptom association, an accurate definition of the criteria is imperative. First, determination of a reflux episode depends on the type of pH measurement used. For example, when pH testing is used, a reflux episode is defined by a drop in pH <4.0, whereas impedance methods indicate reflux by a retrograde 50% drop in impedance from baseline (21,23).

Symptom indices are usually calculated by a system software and rarely manually by a technician or clinician; therefore, it becomes important to understand the concept of a time window for symptoms after a reflux to be associated. The time window defines a period of time after a reflux episode, for which the symptom is associated with that reflux event. In determination of the optimal time window, Lam et al. compared time windows of various onsets and durations among patients with noncardiac chest pain, evaluating symptom associations with reflux and dysmotility. Applying the Poisson theory, symptom indices demonstrated a predictable gradual increase when windows of increasing length were used. A window beginning at 2 minutes before the onset of chest pain and ending at the onset of pain seemed to be optimal. This led to the current standard of 2 minutes as the optimal time window between a reflux episode and development of an associated symptom (Figure 1). Symptoms occurring outside this 2-minute window are therefore not considered (Figure 2). However, one must recognize this study was carried out in patients with noncardiac chest pain, and therefore, this optimum time window may not be universal among patients with other reflux symptoms, typical or atypical (24).


Figure 1.: Example of an acidic reflux episode (solid arrow) on ambulatory 24-hour pH monitoring represented by a drop in impedance in at least 2 channels with an associated drop in pH (dashed arrow). Reflux episode occurs within the 2-minute window, depicted as a colored rectangular box preceding patient's reported symptom (circle). Given reflux falls within the 2-minute window, the computer will mark as positive for reflux-symptom association.



Figure 2.: Example of an acidic reflux episode (solid arrow) on ambulatory 24-hour pH monitoring with an associated drop in pH (dashed arrow). Reflux episode occurs outside the 2-minute window, depicted as a colored rectangular box preceding patient's reported symptom (circle). Given this set time-window, the computer will mark as negative for reflux-symptom association.
Symptom index and symptom sensitivity index
First described by Wiener et al. (25), the Symptom Index (SI) reflects the percentage of symptom events occurring immediately before a drop in esophageal pH <4.0, calculated as:


This index is calculated separately for each symptom (i.e., heartburn or cough). Therefore, if a patient has 10 reported symptom events with 7 of them occurring shortly after a reflux event during ambulatory pH or impedance monitoring, the SI would be calculated as 70%. Symptom events not connected with a reflux would be omitted from the numerator calculation (Figure 3). Applying a receiver operating characteristic curve, optimum sensitivity and specificity determined a threshold value of 50% for a positive SI for heartburn. Thus, a positive index is reached when at least half of the patient-reported symptoms occurred shortly after the intraesophageal pH dropped below 4.0 (11,26). As described, a 2-minute window is used in the analysis. A limitation of the SI, however, is not taking into account the total number of reflux episodes. For example, if a person has frequent reflux events and only a few symptom episodes, the likelihood that a positive SI is reached by chance is high.



Figure 3.: An example of a 2-minute window preceding patient's report of pain (circle), depicted as a colored rectangular box. This symptom event lacks the presence of an acidic or nonacidic reflux episode within the 2-minute window, therefore interpreted as a symptom of pain without a reflux association.

Symptom association probability
Recognizing that the probability of a positive SI increases with higher numbers of reflux episodes, Weusten et al. (27) proposed the Symptom Association Probability (SAP) a complex probability approach to determine whether the pattern of reflux and symptoms occurred by chance. Twenty-four-hour data are divided into consecutive 2-minute segments, with each segment assessed for the presence of reflux to determine the total number of 2-minute segments with (total R+) and without (total R−) reflux (28). Subsequent calculation includes constructing a 2 × 2 contingency table in which the number of 2-minute segments with and without symptoms and with and without reflux are arranged. Subsequently, a Fisher exact test measures the probability that the temporal association between reflux episodes and symptoms occurred by chance. SAP is calculated as (1-probability) × 100% (Figure 4). A positive SAP score is ≥95%, interpreted as the probability of an association occurring by chance is <5% (11,26). Importantly, one must recognize that an SAP is a statistical parameter and that a statistically significant association does not necessarily imply a causal relationship (28). On the other hand, an elevated acid exposure time or SI provide even less certain evidence of a causal relationship between reflux and symptoms.

Figure 4.: Two by 2 table of the calculation of the SAP. ; R, acid reflux; S, symptom; SAP, symptom association probability.


Interpretation of SI and SAP are complementary and provide different data SI representing an effect size of a relationship, whereas SAP measuring probability of a relationship (18). When both values suggest a positive association, together these reflux-symptom association indices suggest the strong evidence of an association between reflux and symptoms. Often this is not the case, and one test remains positive, whereas the other is negative. Suggested by the Porto consensus, this requires the reader to rely on other parameters including the acid exposure time, number of reflux episodes, and clinical symptoms to make a judgement on a reflux-symptom association (21). In clinical practice, however, SAP is regarded with stronger statistical validity for reflux-symptom correlation given it is a calculation of probability therefore, some argue it is the superior symptom measurement (13).

A caveat to interpreting reflux-symptom associations is recognizing the importance on the total number of reported symptoms. Suggested in the Lyon consensus on recommendations of ambulatory reflux monitoring, reliable interpretation of a reflux-symptom association requires at least 3 events per symptom. If the total number of symptoms experienced by a patient for a specific symptom (i.e., heartburn) occurs <3, a reliable assessment cannot be made (21).
Ghillebert probability estimate
An additional method to assess the temporal correlation between reflux episodes and symptom events is the Ghillebert Probability Estimate (GPE). This method consists of a calculation of the sum of partial probabilities for all reflux-associated symptoms among the total number of symptoms, considering total pH-impedance study duration and total exposure time. In contrast to standard measurements for SAP, calculated after knowledge of reflux and symptom data for each of the 2-minute intervals, the GPE use post hoc summary parameters after completion of ambulatory pH monitoring. Similar to the SAP, the GPE is considered positive when the estimated chance that reflux episodes and symptom events are not related is <5% (29,30). Concordance between SAP and GPE was measured by Kushnir et al. after a review of 772 symptomatic subjects with ambulatory pH data. Standard SAP and GPE were positive in 42.7% and 39.3%, with major discordance found in only 2.8%. Furthermore, when compared with SAP, sensitivity and specificity returned at 0.95 and 0.91 for GPE. The authors therefore concluded that GPE can be used interchangeably with SAP when evaluating for symptom correlation (5). In addition, Kushnir et al. measured the value of GPE, in addition to SI and acid exposure time, assessing whether GPE alone or in combination with SI best predicted a successful response to antireflux therapy. Among 98 subjects with noncardiac chest pain who had underwent esophageal pH monitoring, 80.6% experienced symptoms during the procedure, but only 53% demonstrated an elevated acid exposure time. Furthermore, when authors focused on symptom scores, similar symptom scores were seen with GPE and SI (26.5% and 25.5%, respectively). All patients eventually were treated with proton pump inhibitors (PPI) and within the group, and 24 underwent fundoplication. At time of follow-up (2.8 0.9 years), 59.2% achieved sustained response with the best outcomes among individuals with a combination of a positive SI, GPE, and elevated acid exposure time (sensitivity 24%, specificity 98%, positive predictive value 15%, and negative predictive value 85%) (31).
Length of time and reflux-symptom measurements
The length of pH monitoring can affect reflux-symptom association values. With the introduction of wireless pH monitoring, recording time could be extended from 24 hours to 48 or 96 hours. In extended recording times as seen in wireless pH monitoring, SI and SAP calculate separately at 24 hours and for the total study period. Although extended recording periods of 4896 hours may increase the yield for detecting pathologic esophageal acid exposure, does the calculation affect reflux-symptom association indices? Slaughter et al. determined in a cross-sectional study of 254 patients undergoing either wireless pH or impedance/pH monitoring a longer monitoring period affected SAP values. The authors determined as study length increased from 24 to 48 hours, SAP scores switched from negative to positive associations as reflux rates ranged between 5% and 10%. SI values were independent to measurement length. Furthermore, the authors noted significant day-to-day variability with SAP values while maintaining reflux and symptom probabilities constant (9). Prakash and Clouse similarly aimed to assess for changes in reflux-symptom association with extended wireless pH monitoring. The authors noted a 5.2% gain in the number of reported symptoms contributing to an increase in available symptoms measured in a reflux-symptom association (29). Although extended pH monitoring can provide advantages in increasing patient reports on the number of symptoms, one must recognize the possible change in SAP assessment as measurement length increase.
How to apply symptom indices reliability and accuracy
To apply reflux-symptom association analysis in clinical practice, it is imperative to first test whether these diagnostic tests are reproducible. Reproducibility, or reliability, represents the degree of precision a study maintains with repeated measures.

Aanen et al. investigated the reproducibility of these indices when measured in the same individuals at 2 separate time points. Among 21 patients with typical reflux symptoms who underwent 24-hour combined pH-impedance recording off antisecretory therapy on 2 separate days, separated by 14 weeks, SAP and Symptom Sensitivity Index had the highest reproducibility (Kendall coefficients of concordance (W-value) = 0.90, P = 0.01, and W = 0.86, P < 0.05). On the other hand, the number of symptom episodes was not found to be very similar on the 2 separate days, contributing to a less reproducible SI (W = 0.73, P = 0.09). The authors concluded that the SAP, as being the most objective parameter by eliminating chance occurrence, is a reliable measurement to express the relationship between reflux and symptoms and the symptom association parameters are actually more reproducible than the parameters that express reflux severity (32).

To measure the reliability of reflux-symptom associations beyond expected by chance alone, Abdul-Hussein et al. calculated the measures of agreement applying a kappa coefficient (k). Among 1,471 patients testing positive for GERD by combined impedance-pH reflux studies, authors separated patients by on and off acid suppression therapy and reviewed SAP and SI for typical and atypical symptoms including heartburn, regurgitation, chest pain, cough, indigestion, and throat clearing. Authors demonstrated substantial inter-rater agreement between SAP and SI with patients on PPI therapy for regurgitation (k = 0.68) and indigestion (k = 0.64), whereas moderate agreement for heartburn (k = 0.48) and chest pain (k = 0.51), and fair agreement for cough (k = 0.33) and throat clearing (k = 0.29). Lower level of agreement was seen with patients off acid-suppressive therapy. Therefore, the authors demonstrated that reliability of SAP and SI measurement increases with typical symptoms and while on acid suppression therapy (33).

To assess the extent of reflux-symptom indices capturing a true association, or accuracy, Taghavi et al. aimed to validate SI and SAP with changes in symptom after omeprazole. The authors studied 38 patients with predominant heartburn symptoms who underwent 24-hour esophageal pH monitoring and treated with high-dose omeprazole (40 mg morning and 20 mg evening). Applying symptom response with PPI therapy, the authors calculated a modest sensitivity and specificity for SI and SAP (34.8% and 80% and 65.2% and 73.3%, respectively). Interpretation of this study emphasizes that although symptom indices hold some value in predicting reflux treatment outcome, at a stand-alone, one cannot make a complete interpretation (34).
How to interpret the results of reflux-symptom association analysis.
After understanding the method of reflux-symptom analysis calculation and the overall reliability and accuracy of testing, it becomes clinically relevant to recognize how SI and SAP compliment the interpretation of esophageal AET. As proposed by the Lyon consensus, AET is the most reproducible and the primary outcome applying pH monitoring. However, although pH monitoring can diagnosis pathologic acid exposure, a positive reflux-symptom association not only supplements interpretation when AET return inconclusive (4%6%) but additionally determines if and which symptoms are indeed caused by acid reflux. This information can suggest if symptoms are more likely to respond after medical (30,35) or surgical therapy (36). For example, Patel et al. revealed in patients undergoing medical therapy for reflux symptoms, GPE scores predicted a linear relationship with improvements in global symptom severity (30). The same author differentiated phenotypes of patients based on strength of reflux, ranging between strong (abnormal AET, positive SAP), good (abnormal AET, negative SAP), and equivocal evidence for GERD and aimed to compare the outcomes to predict symptom response. The authors applied a symptom survey generated a priori and a global symptom severity score on a 100-mm visual analogue scale. The study results demonstrated patients with strong or good evidence for GERD, who receive treatment by medical or surgical therapy, predict to have the greatest symptomatic responsecompared with those with equivocal reflux evidence. In addition, symptom response is most pronounced when symptoms fit typical (i.e., heartburn and regurgitation) GERD symptoms (37).

When patients demonstrate a positive reflux-symptom association score in the setting of physiologic (i.e., normal) reflux acid exposure times, one must consider a diagnosis of reflux hypersensitivitya disease of heightened esophageal visceral sensitivity and not because of pathologic acid disease. Under the umbrella of functional gastrointestinal disorders per Rome IV (38), the diagnostic term reflux hypersensitivity describes patients with greater affinity to symptoms with reflux in the presence of physiologic acid exposure and lack of endoscopic mucosal disease. Reflux hypersensitivity accounts for a large proportion of patients presenting with typical GERD-like symptoms refractory to high-dose medical treatment. Because these patients present with typical GERD-like symptom, distinguishing reflux hypersensitivity from acid reflux disease improved after the introduction of ambulatory reflux monitoring with reflux-symptom association analysis. Now patients are appropriately being categorized as reflux hypersensitivity and management focuses less on purely acid suppression therapy and more on concomitant neuromodulators including tricyclic antidepressants and selected serotonin reuptake inhibitors (39,40).

Patients with functional heartburn similarly experience typical GERD-like symptoms in the presence of physiologic esophageal acid exposure and normal endoscopic mucosal findings. In contrast with reflux hypersensitivity, this cohort of patients lack a symptom correlation with gastroesophageal reflux eventsdistinguishing functional heartburn from reflux hypersensitivity. Per the Rome IV criteria, diagnostic criteria for functional heartburn requires all 4 qualities: (i) burning retrosternal discomfort or pain, (ii) no symptom relief despite optimal antisecretory therapy, (iii) absence of GERD or eosinophilic esophagitis, and (iv) absence of major esophageal motor disorders. The underlying pathophysiology of functional heartburn is not entirely understood, but current treatment strategies focus with pain modulators such as tricyclic antidepressants or selected serotonin reuptake inhibitors in addition to psychological interventions including hypnotherapy (41,42).

How well reflux-symptom association indices distinguish between reflux hypersensitivity and functional heartburn has been questioned by Choksi et al. The authors aimed to measure the reliability of reflux-symptom association analysis among 205 patients with PPI refractory GERD. The study separated patients into 3 groups: (i) pH-negative without esophagitis, (ii) pH-positive without esophagitis, and (iii) with esophagitis. All patients underwent SAP calculation scores. Differences in reflux-symptom association were found to be small among the 3 groups, with a difference of only 0.48% leading to a diagnosis of reflux hypersensitivity in the pH-negative without esophagitis group. Furthermore, significant variability in SAP was seen between day 1 and day 2 of pH testing. Therefore, the authors suggested that in a PPI-refractory GERD population, the SAP may not be as accurate to distinguish reflux hypersensitivity from functional heartburn (43).

One may question although reflux-symptom association analysis allow providers to differentiate reflux hypersensitivity from functional heartburn in the setting of physiologic esophageal acid exposure, is there truly any need to differentiate the 2 functional esophageal conditions? It is actually recognized patients with reflux hypersensitivity may respond to medical, even surgical, acid reductionwhile not the case in functional heartburn. Watson et al. measured symptomatic response and SF-36 quality of life among patients with reflux hypersensitivity treated with omeprazole 20 mg twice daily × 4 weeks and revealed decrease symptom frequency, severity, and intake of antacids. In the comparator with functional heartburn (i.e., negative SI), the authors noted only one patient improved (44). Broeders et al. additionally compared outcomes of long-term antireflux surgery among patients with reflux hypersensitivity and pathologic acid exposure, revealing both arms demonstrated similar benefits after fundoplication. Therefore, although providers may treat both conditions with similar neuromodulator, emerging data suggest some role of acid suppression therapy specific to reflux hypersensitivity.

Last, emerging evidence suggest reflux-symptom association in regurgitation may indicate rumination syndrome. Rumination describes the behavioral process of recurrent regurgitation of undigested food and frequently misdiagnosed as GERD. Recognized by the Rome IV criteria, rumination syndrome is defined by (i) persistent or recurrent regurgitation of recently ingested food and subsequent spitting or mastication and swallowing and (ii) regurgitation not preceded by retching for the past 3 months (38). Although more common to be a pediatric condition, rumination syndrome presents in up to 20% of adults with medically refractory GERD-like symptoms. Aiming to identify distinct patterns on impedance-pH monitoring among children with rumination syndrome, Nikaki et al. revealed significant higher SAP scores (regurgitation and/or reflux and/or vomiting) in those with rumination syndrome compared with controls (P < 0.0001). In addition, the authors determined a large number of proximal reflux episodes and high frequency of postprandial nonacidic reflux episodes highly suggesting rumination syndrome (45). Patients with conditions possibly mimicking GERD, including rumination syndrome, should therefore undergo appropriate testing before drawing conclusions after reflux-symptom association analysis.

Furthermore, patients with GERD commonly may report excessive belching along with reflux symptoms. Gastric belch describes the accumulation of gastric air, triggering a reflex response of LES relaxation and retrograde movement of air into the esophagus and through the oral cavity (46). Supragastric belching, on the other hand, describes a distinct belch pattern of rapid air entry by sucking air into the esophagus from the mouth followed by immediate expulsion of air through the mouth (47). Liquid reflux commonly can accompany gastric or supragastric belch and therefore may increase confusion among patients and their report on a symptom diary. These mixed liquid-gas relationships have been defined applying impedance analysis: (i) belch preceding liquid reflux, (ii) belch occurring during liquid reflux, and (iii) liquid reflux induced by belch (48). However, unlike impedance analysis, patients are not as skilled to characterize these unique patterns of belch and liquid reflux. Instead patients are left to rely on perceived symptoms and on occasion may describe regurgitation a typical symptom of GERD instead in a symptom diary. Therefore, we encourage readers to review tracings for gastric and supragastric belch patterns before making conclusions of GERD with positive reflux-symptom association because treatment remain dramatically different.
Practical issues of symptom analysis and controversy.
It has been put forward that, in clinical practice, one should be careful not to overinterpret the symptom association indices. In a cross-sectional study of patients with PPI-refractory disease, Slaughter et al. aimed to assess the effectiveness of both SI and SAP. Patients underwent (i) wireless pH or impedance/pH monitoring off PPI (ii) impedance/pH monitoring off PPI therapy, or (iii) impedance/pH monitoring on twice-daily PPI therapy. SI and SAP values were subjected to Monte Carlo simulation to measure variability, demonstrating that only 33% of patients had positive SI or SAP scores and that abnormal indices required high frequencies of reflux episodes. When reflux rates dropped to values <10%, positive SI and SAP values occurred more by chance than representing a true symptom and reflux relationship. The authors therefore argue that SI and SAP could be overinterpreted, unless patients demonstrate a high frequency of reflux (9).

Assessing the temporal relationship between the patient's symptoms (i.e., heartburn, cough, or chest pain) and reflux episodes relies on patients pressing an event marker button on feeling their symptoms. Therefore, an important step ensuring accuracy of symptom association measurement is providing patients with the appropriate instruction on documenting each event and patient compliance. When patients forget to document symptoms (Figure 5) or only document heartburn, but not cough, they may have a positive symptom association for one and not the other or vice versa. For example, in pediatric patients, reflux-symptom association testing has met with challenges because many children do not express themselves in a similar manner as adults. The parent or caregiver is left to be responsible for indicating onset and type of symptoms, resulting in challenges with interpretation of the analysis. Furthermore, during pH assessment patients are provided instructions to eat, drink, work, and exercise as they would normally do (28). Patient compliance, however, is an emerging topic in health care, and practice issues arise due to patient discomfort or embarrassment. Therefore, patients may reduce normal daily activities or food consumption behaviors (i.e., eating out at a restaurant), which may lead to an inaccurate symptom assessment. Musser et al. assessed the impact of 24-hour catheter-based esophageal pH monitoring among 82 participants using a short questionnaire in addition to a review of each participant's food diary. Surprisingly, the authors found that despite mild discomfort, most participants did not experience a significance alteration in their typical routines by the presence of the pH catheter and many returned to the work with the catheter in place (49). In a randomized trial by Wong et al. comparing feasibility and tolerability of a wireless pH capsule (Bravo pH monitoring system) vs traditional 24-hour esophageal pH monitoring, the authors reported less nose pain (P = 0.047), less interference with daily activities (P = 0.001), eating (P = 0.003) and sleep (P = 0.025), and overall higher satisfaction with the testing (P = 0.023) in patients receiving the wireless pH capsule (50). Wireless pH monitoring, however, does not go without complaints either. Sensation of a foreign body leading to esophageal discomfort or pain has been reported in up to 35% of patients, with endoscopic removal occurring in approximately 3%5% of cases (51). Of course the extended measurement time compared with catheter-based techniques allows to register more symptoms, and this increases the sample size and thus the statistical power of tests such as the SAP.



Figure 5.: An example of 2 acidic reflux episodes (solid and dashed arrow) on ambulatory 24-hour pH monitoring occurring without an associated symptom, therefore interpreted as asymptomatic reflux episodes.


The results of ambulatory reflux monitoring and reflux-symptom association analysis are critically dependent on the interpretation of the data, both with pH and with impedance monitoring. An important aspect of the analysis of pH and impedance signals is the visual inspection, which is required to exclude artifacts not identified by the computer software. Artifacts in pH signals can occur because of intake of acidic foods or beverages during monitoring by technical problems in the monitoring circuit or by too distal placement of the pH probe so it will be positioned in the stomach during (part of) the measurement. Agrawal et al. found acidic foods such as carbonated beverages, coffee, ketchup, tea, orange or apple juice, lemonade, red or white wine, and strawberry juice produced a decrease in intraesophageal pH <4, therefore likely to produce artifacts that mimic reflux at the time of pH monitoring (52). There are 2 ways to overcome this: one is to prohibit consumption of acidic foods during the measurement and the other, more often used approach, is to register meal consumptions times during the measurement period and exclude these from the analysis. Mathus-Vliegen et al. investigated the effect of manual correction for food/drink-associated pseudoreflux among 216 patients. The authors found a mean acid exposure (% total time with pH <4) was 1.28 with raw unprocessed data; however, after manual correction, this decreased to 0.41%. Similarly, total refluxes numbers decreased from 33.6 to 12.9 after manual correction. Last, changes in body position, swallowing, or phonation may induce proximal displacement or accidental migration of the pH probe into the stomach, all resulting to erroneous pH assessment. Traditionally, the pH sensor of the monitoring catheter is positioned at 5 cm proximal to the LES, with the LES location based on previous esophageal manometry. This location largely prevents slipping of the electrode into the stomach during esophageal shortening (53). Unfortunately, current software programming does not caution readers to abnormal acid exposure due to pH probe migration; therefore, common practice is to manually assess for possible erroneous measurements at the beginning of interpretation. Overall, these corroborate the importance of manually reviewing the computer-generated data from pH and impedance monitoring prior to making the assessment (54).

CONCLUSION

In conclusion, the emergence of intraluminal esophageal pH and impedance monitoring has introduced the ability to detect abnormal intraluminal esophageal acid exposure and excessive nonacid reflux events in addition to measure reflux-symptom association analysis as a cause-and-effect relationship. Several indices have been developed; however, SAP continues to be regarded with the strongest statistical validity and highest reproducibility, therefore the preferred measurement. Recognizing the importance of assessment of reflux-symptom associations, their analysis should form part of ambulatory pH-impedance study evaluations. Symptom association analysis strengthens the overall information provided by gastroesophageal reflux monitoring.

CONFLICTS OF INTEREST

Guarantor of the article: Albert J. Bredenoord, MD.

Specific author contributions: A.N.K. drafted and revised the manuscript. A.J.B. revised and reviewed the manuscript. A.N.K., J.O.C., J.M.O., A.J.S., and A.J.B. have approved the final draft submitted. J.O.C. is a consultant for Isothrive, Medtronic, Pfizer, and Regeneron/Sanofi, and the principal investigator for Ironwood and Impleo Medical.

Financial support: None to report.

Potential competing interests: A.J.B. received research funding from Nutricia, Norgine and Bayer and received speaker and/or consulting fees from Laborie, EsoCap, Diversatek, Medtronic, Dr Falk Pharma, Calypso Biotech, Thelial, Robarts, Reckitt Benckiser, Regeneron, Celgene, Bayer, Norgine, AstraZeneca, Almirall, and Allergan.

REFERENCES

  1. El-Serag HB, Sweet S, Winchester CC, et al. Update on the epidemiology of gastro-oesophageal reflux disease: A systematic review. Gut 2014;63: 87180.
  2. Saritas Yuksel E, Vaezi MF. New developments in extraesophageal reflux disease. Gastroenterol Hepatol 2012;8:5909.
  3. Vaezi MF. Atypical manifestations of gastroesophageal reflux disease. Medscape Gen Med 2005;7:25.
  4. Modlin IM, Hunt RH, Malfertheiner P, et al. Diagnosis and management of non-erosive reflux diseasethe Vevey NERD Consensus Group. Digestion 2009;80:7488.
  5. Kushnir VM, Sathyamurthy A, Drapekin J, et al. Assessment of concordance of symptom reflux association tests in ambulatory pH monitoring. Aliment Pharmacol Ther 2012;35:10807.
  6. Poh CH, Gasiorowska A, Navarro-Rodriguez T, et al. Upper GI tract findings in patients with heartburn in whom proton pump inhibitor treatment failed versus those not receiving antireflux treatment. Gastrointest Endosc 2010;71:2834.
  7. Forootan M, Zojaji H, Ehsani MJ, et al. Advances in the diagnosis of GERD using the esophageal pH monitoring, gastro-esophageal impedance-pH monitoring, and pitfalls. Open Access Maced J Med Sci 2018;6:193440.
  8. Prakash Gyawali C. Esophageal hypersensitivity. Gastroenterol Hepatol 2010;6:497500.
  9. Slaughter JC, Goutte M, Rymer JA, et al. Caution about overinterpretation of symptom indexes in reflux monitoring for refractory gastroesophageal reflux disease. Clin Gastroenterol Hepatol 2011;9:86874.
  10. Kahrilas PJ, Quigley EM. Clinical esophageal pH recording: A technical review for practice guideline development. Gastroenterology 1996;110: 198296.
  11. Tutuian R, Castell DO. Gastroesophageal reflux monitoring: pH and impedance. Aliment Pharmacol Ther 2006;24:2737.
  12. Bechtold ML, Holly JSL, Thaler K, et al. Bravo (wireless) ambulatory esophageal pH monitoring: How do day 1 and day 2 results compare? World J Gastroenterol 2007;13:40915.
  13. Hirano I, Richter JE. ACG practice guidelines: Esophageal reflux testing. Am J Gastroenterol 2007;102:66885.
  14. Prakash C, Jonnalagadda S. Esophageal impedance testing: Unraveling the mysteries of gastroesophageal reflux. Gastroenterology 2006;131: 3223.
  15. Zerbib F, Roman S, Ropert A, et al. Esophageal pH-impedance monitoring and symptom analysis in GERD: A study in patients off and on therapy. Am J Gastroenterol 2006;101:195663.
  16. Hemmink GJM, Bredenoord AJ, Weusten BL, et al. Esophageal pHimpedance monitoring in patients with therapy-resistant reflux symptoms: on or off proton pump inhibitor? Am J Gastroenterol 2008; 103:244653.
  17. Bredenoord AJ, Weusten BL, Timmer R, et al. Addition of esophageal impedance monitoring to pH monitoring increases the yield of symptom association analysis in patients off PPI therapy. Am J Gastroenterol 2006; 101:4539.
  18. Gyawali CP, Kahrilas PJ, Savarino E, et al. Modern diagnosis of GERD: The Lyon consensus. Gut 2018;67:135162.
  19. Savarino E, Tutuian R, Zentilin P, et al. Characteristics of reflux episodes and symptom association in patients with erosive esophagitis and nonerosive reflux disease: Study using combined impedance-pH off therapy. Am J Gastroenterol 2010;105:105361.
  20. Frazzoni L, Frazzoni M, de Bortoli N, et al. Critical appraisal of Rome IV criteria: Hypersensitive esophagus does belong to gastroesophageal reflux disease spectrum. Ann Gastroenterol 2018;31:17.
  21. Roman S, Gyawali CP, Savarino E, et al. Ambulatory reflux monitoring for diagnosis of gastro-esophageal reflux disease: Update of the Porto consensus and recommendations from an international consensus group. Neurogastroenterol Motil 2017;29:115.
  22. Sifrim D, Dupont L, Blondeau K, et al. Weakly acidic reflux in patients with chronic unexplained cough during 24 hour pressure, pH, and impedance monitoring. Gut 2005;54:44954.
  23. Kim GH. How to interpret ambulatory 24 hr esophageal pH monitoring. J Neurogastroenterol Motil 2010;16:20710.
  24. Lam HG, Breumelhof R, Roelofs JM, et al. What is the optimal time window in symptom analysis of 24-hour esophageal pressure and pH data? Dig Dis Sci 1994;39:4029.
  25. Wiener GJ, Richter JE, Copper JB, et al. The symptom index: A clinically important parameter of ambulatory 24-hour esophageal pH monitoring. Am J Gastroenterol 1988;83:35861.
  26. Vaezi MF. Use of symptom indices in the management of GERD. Gastroenterol Hepatol (N Y) 2012;8:1857.
  27. Weusten BL, Roelofs JM, Akkermans LM, et al. The symptom-association probability: An improved method for symptom analysis of 24-hour esophageal pH data. Gastroenterology 1994;107:17415.
  28. Bredenoord AJ, Weusten BL, Smout AJ. Symptom association analysis in ambulatory gastro-oesophageal reflux monitoring. Gut 2005;54: 18107.
  29. Prakash C, Clouse RE. Value of extended recording time with wireless pH monitoring in evaluating gastroesophageal reflux disease. Clin Gastroenterol Hepatol 2005;3:32934.
  30. Patel A, Sayuk GS, Gyawali CP. Acid-based parameters on pH-impedance testing predict symptom improvement with medical management better than impedance parameters. Am J Gastroenterol 2014;109:83644.
  31. Kushnir VM, Sayuk GS, Gyawali CP. Abnormal GERD parameters on ambulatory pH monitoring predict therapeutic success in noncardiac chest pain. Am J Gastroenterol 2010;105:10328.
  32. Aanen MC, Bredenoord AJ, Numans ME, et al. Reproducibility of symptom association analysis in ambulatory reflux monitoring. Am J Gastroenterol 2008;103:22008.
  33. Abdul-Hussein M, Zhang C, Castell D. Symptom index or symptom association probability?: A closer look at symptom association in suspected GERD patients. J Clin Gastroenterol 2018;52:e7e10.
  34. Taghavi SA, Ghasedi M, Saberi-Firoozi M, et al Symptom association probability and symptom sensitivity index: Preferable but still suboptimal predictors of response to high dose omeprazole. Gut 2005;54:106771.
  35. Aanen MC, Weusten BL, Numans ME, et al. Effect of proton-pump inhibitor treatment on symptoms and quality of life in GERD patients depends on the symptom-reflux association. J Clin Gastroenterol 2008;42: 4417.
  36. Díaz S. The symptom association probability (SAP) is superior to the symptom index (SI) for attributing symptoms to gastroesophageal reflux: Validation using outcome from laparoscopic antireflux surgery (LARS). Gastroenterology 2002;122:A75.
  37. Patel A, Sayuk GS, Kushnir VM, et al. GERD phenotypes from pHimpedance monitoring predict symptomatic outcome on prospective evaluation. Neurogastroenterol Motil 2016;28:51321.
  38. Stanghellini V, Chan FKL, Hasler WL, et al. Rome IV Gastroduodenal disorders. Gastroenterology 2016;150:138092.
  39. Viazis N, Keyoglou A, Kanellopoulos AK, et al. Selective serotonin reuptake inhibitors for the treatment of hypersensitive esophagus: A randomized, double-blind, placebo-controlled study. AmJ Gastroenterol 2012;107:16627.
  40. Yamasaki T, Fass R. Reflux hypersensitivity: A new functional esophageal disorder. J Neurogastroenterol Motil 2017;23:495503.
  41. Yamasaki T, ONeil J, Fass R. Update on functional heartburn. Gastroenterol Hepatol (N Y) 2017;13:72534.
  42. Riehl ME, Pandolfino JE, Palsson OS, et al. Feasibility and acceptability of esophageal-directed hypnotherapy for functional heartburn. Dis Esophagus 2016;29:4906.
  43. Choksi Y, Slaughter JC, Sharda R, et al. Symptom association probability does not reliably distinguish functional heartburn from reflux hypersensitivity. Aliment Pharmacol Ther 2018;47:95865.
  44. Watson RG, Tham TC, Johnston BT, et al. Double blind cross-over placebo controlled study of omeprazole in the treatment of patients with reflux symptoms and physiological levels of acid reflux the sensitive oesophagus. Gut 1997;40:58790.
  45. Nikaki K, Rybak A, Nakagawa K, et al. Rumination syndrome in children presenting with refractory gastroesophageal reflux symptoms. J Pediatr Gastroenterol Nutr 2020;70:3305.
  46. Koukias N, Woodland P, Yazaki E, et al. Supragastric belching: Prevalence and association with gastroesophageal reflux disease and esophageal hypomotility. J Neurogastroenterol Motil 2015;21:398403.
  47. Hemmink GJM, Bredenoord AJ, Weusten BLAM, et al. Supragastric belching in patients with reflux symptoms. Am J Gastroenterol 2009;104: 19927.
  48. Roman S, Keefer L, Imam H, et al. Majority of symptoms in esophageal reflux PPI non-responders are not related to reflux. Neurogastroenterol Motil 2015;27:166774.
  49. Musser J, Kelchner L, Neils-Strunjas J, et al. Patient compliance during 24-hour dual pH probe monitoring for extraesophageal reflux. ISRN Otolaryngol 2011;2011:703936.
  50. Wong WM, Bautista J, Dekel R, et al. Feasibility and tolerability of transnasal/per-oral placement of the wireless pH capsule vs. traditional 24-h oesophageal pH monitoring A randomized trial. Aliment Pharmacol Ther 2005;21:15563.
  51. Vaezi MF. Should we bravo? Gastroenterology 2006;130:22389.
  52. Agrawal A, Tutuian R, Hila A, et al. Ingestion of acidic foods mimics gastroesophageal reflux during pH monitoring. Dig Dis Sci 2005;50:191620.
  53. Fletcher J, Wirz A, Henry E, et al. Studies of acid exposure immediately above the gastro-oesophageal squamocolumnar junction: Evidence of short segment reflux. Gut 2004;53:16873.
  54. Mathus-Vliegen EMH, Smit CF, Devriese PP. Artefacts in 24-h pharyngeal and oesophageal pH monitoring: Is simplification of pH data analysis feasible? Scand J Gastroenterol 2004;39:149.


2020 by The American College of Gastroenterology

""

. : " ", " -",
" ", " " .

.

 -