Automated impedance-manometry analysis detects esophageal motor dysfunction in patients who have non-obstructive dysphagia with normal manometry

Authors


Address for Correspondence
Nam Q. Nguyen, Department of Gastroenterology and Hepatology, Royal Adelaide Hospital, North Terrace, Adelaide, SA 5000, Australia.
Tel: +61 8 8222 5207; fax: +61 8 8222 5885;
e-mail: quoc.nguyen@health.sa.gov.au

Abstract

Background  Automated integrated analysis of impedance and pressure signals has been reported to identify patients at risk of developing dysphagia post fundoplication. This study aimed to investigate this analysis in the evaluation of patients with non-obstructive dysphagia (NOD) and normal manometry (NOD/NM).

Methods  Combined impedance-manometry was performed in 42 patients (27F : 15M; 56.2 ± 5.1 years) and compared with that of 24 healthy subjects (8F : 16M; 48.2 ± 2.9 years). Both liquid and viscous boluses were tested. MATLAB-based algorithms defined the median intrabolus pressure (IBP), IBP slope, peak pressure (PP), and timing of bolus flow relative to peak pressure (TNadImp-PP). An index of pressure and flow (PFI) in the distal esophagus was derived from these variables.

Key Results  Diagnoses based on conventional manometric assessment: diffuse spasm (= 5), non-specific motor disorders (= 19), and normal (= 11). Patients with achalasia (= 7) were excluded from automated impedance-manometry (AIM) analysis. Only 2/11 (18%) patients with NOD/NM had evidence of flow abnormality on conventional impedance analysis. Several variables derived by integrated impedance-pressure analysis were significantly different in patients as compared with healthy: higher PNadImp (< 0.01), IBP (< 0.01) and IBP slope (< 0.05), and shorter TNadImp_PP (= 0.01). The PFI of NOD/NM patients was significantly higher than that in healthy (liquid: 6.7 vs 1.2, = 0.02; viscous: 27.1 vs 5.7, < 0.001) and 9/11 NOD/NM patients had abnormal PFI. Overall, the addition of AIM analysis provided diagnoses and/or a plausible explanation in 95% (40/42) of patients who presented with NOD.

Conclusions & Inferences  Compared with conventional pressure-impedance assessment, integrated analysis is more sensitive in detecting subtle abnormalities in esophageal function in patients with NOD and normal manometry.

Introduction

Non-obstructive dysphagia (NOD) is defined as the sensation of impaired esophageal transit that occurs in the absence of an identifiable obstruction of the esophageal lumen on endoscopy and/or radiology. In most cases, it indicates the presence of impaired bolus transport and motor function of the esophagus.1 Effective esophageal emptying is influenced primarily by the interplay between peristaltic propulsive force and the outflow resistance at the esophago-gastric junction (EGJ).2,3 Esophageal manometry is therefore often considered as the investigation of choice for NOD and can identify motility abnormalities such as achalasia, ineffective esophageal motility (IEM), and diffuse esophageal spasm (DES).1,4,5 However, a significant proportion of patients with NOD have esophageal motility and EGJ pressures that are within normal limits.6,7 Based on conventional combined impedance and pressure measurements, the majority of these patients have been shown to have effective bolus transit and clearance.6

Automated impedance-manometry (AIM) analysis is a novel method that allows automated concurrent analyses of both intrabolus pressure and bolus flow during swallowing.8,9 It uses the impedance nadir recorded during bolus flow to define the pressure within the bolus. Furthermore, the time from nadir impedance to peak pressure allows the movement of the bolus to be tracked in relation to following peristaltic clearing wave. Together, AIM analysis provides an objective assessment of esophageal motor function by measuring time from bolus movement to peak, intrabolus and intrabolus ramp pressure. This novel analytic approach has recently been reported to be useful in identifying subtle pre-operative esophageal dysfunction in patients who developed post fundoplication dysphagia.10 AIM analysis has high intrarater and interrater reproducibility.11

The value of AIM analysis in patients with NOD and normal manometry has not been previously evaluated. We hypothesized that AIM analysis might identify subtle esophageal dysfunction in patients who have NOD and normal manometry (NOD/NM). The aim of the current study, therefore, was to investigate the performance of AIM analysis in the evaluation of patients who have NOD with a normal manometric study, in comparison with conventional analysis of manometric and impedance recordings.

Methods

Subjects

Combined pressure-impedance recordings from 42 consecutive patients (15 M : 27 F; 56.2 ± 5.1 years), who were referred to the Department of Gastroenterology at the Royal Adelaide Hospital, Adelaide, Australia (= 22), and the Gastrointestinal Motility Research Unit of University Medical Center, Utrecht, The Netherlands (= 20) for assessment of NOD were reviewed and analyzed. Manometric and standard impedance data from 38/42 patients from this cohort have been reported previously.6 None of the NOD patients had evidence of esophageal strictures or other luminal lesions on upper endoscopy. Thirty-one (84%) patients had esophageal biopsies taken and none had changes consistent with eosinophilic esophagitis. In addition, there was no evidence of active esophagitis on endoscopy and no patient was known to have gastro-esophageal reflux disease (GERD) on the basis of pH-monitoring or symptom response to PPI therapy. Patients with hiatus hernia ≥3 cm were excluded. The data were compared with those from 24 healthy subjects from Adelaide (16 M : 8 F; 48.2 ± 2.9 years), who were free of gastrointestinal symptoms, had no evidence of acute or chronic illness and were not taking medications known to influence esophageal motor function. The protocol was approved by the institutional ethics committees of both centers and all subjects gave written informed consent.

Esophageal manometry and impedance recording

The esophageal motor function of all subjects was assessed using combined perfused multilumen manometry (DentSleeve Pty Ltd, Wayville, SA, Australia) and intraluminal impedance measurement (Sandhill Scientific, Highland Ranch, CO, USA) techniques.3 The technical details of the catheter’s configuration, positioning, perfusion rate, impedance frequency, and current setting have been previously reported.3,12 Specifically, the catheter included a 6-cm long, reverse-perfused sleeve sensor to record LES pressure; and four side holes at 2, 7, 12, and 17 cm proximal to the LES to assess esophageal body motor function. In the current study, the most distal esophageal site (i.e., 2 cm above the LES) was labeled as recorded site 4 and the most proximal recorded site (i.e., 17 cm above the LES) as site 1.

Manometric and impedance signals were recorded by a computerized combined manometry-impedance recording system (Insight Acquisition; Sandhill Scientific) and displayed and stored on a personal computer for subsequent display and analysis.

Protocol

All studied subjects were fasted for at least 4 h before the study. The assembly was passed into the esophagus through a nostril anesthetized with lignocaine spray and positioned such that the sleeve sensor straddled the EGJ high-pressure zone and the distal side hole/impedance segment was situated about 2 cm above the LES. The subjects were then positioned in the right lateral position and allowed to accommodate to the assembly for 5–10 min. The subjects were then asked to swallow ten 5-mL boluses of normal saline (0.9%) followed by ten 5-mL boluses of a low impedance viscous solution (Sandhill Scientific). The swallows were spaced at least 20 s apart.

Data analysis

All manometric and impedance measurements were reviewed manually and analyzed by NQN. The clinical manometric diagnoses were independently reviewed and confirmed by NQN and RHH. Raw pressure-impedance traces from all subjects were imported into the AIM analysis software program and the analyses were performed by TO, who was blinded from the manometric diagnosis.

Manometric measurements  The manometric recordings were analyzed manually to determine: (i) esophageal pressure wave amplitude and velocity, (ii) peristaltic success, and (iii) basal and nadir LES pressure. The definition of normal and various abnormal esophageal motility conditions have been previously reported.3,6,13

Impedance measurements  The recordings were analyzed using the impedance analysis software (Bioview Analysis; Sandhill Scientific). Esophageal bolus clearance was assessed by measurement of two variables: bolus presence time (BPT) and total bolus transit time (TBTT).12 Normal values of these two variables for both liquid and viscous swallows had been previously published.12 Complete bolus transit for a swallow was defined as one for which BPT at all sites and TBTT values were within normal limits.12 An individual was deemed to have normal esophageal bolus clearance if complete bolus transit was achieved in eight or more liquid swallows and seven or more viscous swallows.12

AIM analysis  Raw manometric and impedance data for each test bolus were visualized over a 30-s window and exported from the recording system in ASCII text format, then analyzed using MATLAB (version 7.9.0.529 R2009b; The MathWorks Inc, Natick, MA, USA). Pressure and impedance data were smoothed by a cubic interpolation method in which temporal data were doubled and spatial data increased by a factor of 10, achieving a virtual increase in data sampling from one value per 5 cm sampled at 30 Hz to 10 values per 5 cm sampled at 60 Hz.

As reported previously,10 five pressure-flow variables were derived from the automated analyses (see Fig. 1A): (i) pressure at nadir impedance (PNadImp, mmHg), (ii) peak pressure (PP, mmHg), (iii) intrabolus pressure (IBP, mmHg), (iv) time interval between nadir esophageal impedance and peak esophageal pressure (TNadImp-PP, sec), and (v) IBP slope (IBP slope, mmHg s−1). These variables were derived at each recording side hole from the proximal (site 1) to distal (site 4) esophagus. As previously discussed,10 the most relevant site was deemed to be in the distal esophagus and the data reported in this study are from site 4.

Figure 1.

 Esophageal impedance and pressure signals and derivation of AIM analysis metrics. (A) The AIM metrics TNadImp to PP, PNadImp, IBP, and Slope IBP are measured using a software algorithm, which is guided by the timing of landmarks NadImp and PP. (B) An esophageal pressure topography plot showing pressure (iso-contours) and impedance (horizontal lines) changes during swallowing of a 5-mL viscous bolus in a control subject. Circles and lines indicate the timing of NadImp and PeakP. (C) A plot as for B, however, in a patient with NOD considered having a normal manometry based on standard criteria. Note, when compared with B, the timing of NadImp is shifted to the right and travels closer to PP.

Based on our recent work in patients who have pharyngeal dysphagia8,9 and post fundoplication dysphagia,10 three AIM variables were found to significantly associate with dysphagia: IBP, IBP slope, and TNadImp-PeakP. To amplify the differences in these variables between patients with and without dysphagia, the three variables were combined to form an index of pressure and flow (PFI) in the distal esophagus. The index was calculated using the following formula: PFI = (IBP4 * distal _IBP4 slope)/(TNadImp-PeakP4). Overall, the PFI reflects the relationship between peristaltic strength and flow resistance in the distal esophagus. The PFI has previously been shown to be a predictor of dysphagia risk in patients undergoing antireflux surgery.10

Normal values for the PFI for liquid and viscous swallows were derived from the 24 healthy subjects, with a PFI of less than the 90th percentile being defined as normal.

Statistical analysis

Data for all liquid or viscous bolus swallows for each subject were averaged and compared. The differences between the groups were compared using Fisher’s exact test for categorical data and the Mann–Whitney test for continuous data. Liquid bolus swallows were compared with viscous within each subject group using the signed rank test. Significance was accepted at a P < 0.05.

Results

Manometric and impedancometric diagnosis

Based on conventional manometric assessment, achalasia was found in seven patients, DES in five patients, non-specific esophageal motility disorder (NSEMD) in 19 patients, and ‘normal’ motility in 11 patients. All healthy volunteers had normal manometric assessment. Based on published criteria for standard impedance signal assessment, 2/11 NOD patients with normal manometry had impaired esophageal bolus clearance.

AIM analysis in NOD patients with known esophageal motor disorders and healthy subjects

Given AIM analysis could not be performed reliably in patients with achalasia because of the absence of a propulsive wave and intrabolus pressure, these patients were excluded from the analysis.

Automated impedance-manometry analysis was performed in 35 NOD patients. Data related to comparisons of AIM variables in the distal esophagus (PNadImp, IBP, PeakP, TNadImp-PP, IBP slope, and PFI) between all NOD, DES, NSEMD, NOD patients with normal manometry and healthy subjects for liquid and viscous swallows are summarized in Figs 2 and 3. Compared with healthy subjects, NOD patients overall had higher PNadImp, shorter TNadImp_PP, and a higher PFI for both liquid and viscous swallows (Figs 2 and 3). In general, the differences in the AIM variables between healthy subjects and patients with NSEMD were minimal, except for a significantly lower peak pressure in the patients. In contrast, NOD patients with DES had higher PNadImp, IBP, IBP slope and PFI for both liquid and viscous swallows than those of healthy subjects (Figs 2 and 3).

Figure 2.

 Comparison of AIM variables (PNadImp, IBP, PP, TNadImp_PP, and Slope IBP) and Index of pressure and flow in the distal esophagus (PFI) for liquid swallows between patients with NOD and healthy subjects (controls). Overall, NOD patients had higher PNadImp, lower PP, shorter TNadImp_PP, and higher PFI. When variables were stratified by manometric diagnosis, abnormality was most commonly observed in patients with DES and normal manometry (NOD_NM). *< 0.05, vs healthy subjects. #= 0.06, vs healthy subjects.

Figure 3.

 Comparison of AIM variables (PNadImp, IBP, PP, TNadImp_PP, and Slope IBP) and index of pressure and flow (PFI) in the distal esophagus for viscous swallows between patients with NOD and healthy subjects. In contrast with liquid swallows, abnormalities in the AIM variables (PNadImp, IBP, TNadImp_PP, and Slope IBP) and PFI were only observed in patients with DES and normal manometry (NOD_NM). *< 0.05, vs healthy subjects. **< 0.01, vs healthy subjects. #= 0.07, vs healthy subjects.

AIM analysis in NOD patients with normal manometry and healthy subjects

There was no difference in age (49.8 ± 4.6 years vs 48.2 ± 2.9 years, > 0.05) or gender (7 M : 4 F vs 16 M : 8 F; > 0.05) between NOD patients with normal manometry and healthy subjects. The differences in distal esophageal PNadImp, IBP, PP, TNadImp-PP, IBP slope, and PFI between NOD patients with normal manometry and healthy subjects for liquid and viscous swallowed are summarized in Figs 2 and 3. Example of esophageal pressure topography plots showing the differences in the timing of NadImp and PeakP between the healthy subjects and NOD patients with normal manometry are depicted in Fig. 1B,C. The PNadImp, IBP, and IBP slope were significantly higher in NOD patients with normal manometry for both liquid and viscous swallows than those of healthy subjects. The TNadImp_PP of both liquid and viscous swallows in these patients was significantly shorter than those of healthy subjects. The peak pressures (PP) were similar between the NOD patients with normal manometry and healthy subjects. For both liquid and viscous swallows, the PFI of NOD patients with normal manometry was significantly higher than those of healthy subjects (liquid: 6.7 [2.4–9.3] vs 1.2 [0.4–2.2], = 0.02; viscous: 27.1 [18.9–58.2] vs 5.7 [4.7–10.4], < 0.001).

Distal esophageal pressure-flow index (PFI) in NOD patients and healthy subjects

Normality for PFI was defined as within the 90th percentile range with values of ≤5.4 for liquid swallows and ≤19.4 for viscous swallows. Based on these cutoff values, only one healthy subject had an abnormal PFI for both liquid and viscous swallows. There were two healthy subjects who had an abnormal PFI for liquids, but normal for viscous swallows and vice versa, two healthy subjects had an abnormal PFI for viscous swallows, but normal for liquids.

Compared with healthy subjects, a significantly greater proportion of all NOD patients had abnormal PFI for both liquid (55%vs 13%, < 0.05) and viscous (82%vs 13%, < 0.05) swallows; the highest PFI scores were observed in NOD patients with DES (< 0.05) and NOD patients with normal manometry (< 0.01) (Fig. 4). Significantly more NOD patients with normal manometry were found to have subtle esophageal motor dysfunction on AIM analysis (i.e., abnormal PFI) (9/11 vs 2/11; = 0.009) than evident from standard combined pressure-impedance assessments.

Figure 4.

 Comparison of proportion of subjects with abnormal PFI after liquid and viscous swallows, with stratification according to manometric diagnosis. Compared with healthy subjects, a significantly greater proportion of NOD patients (< 0.05) had abnormal PFI for both liquid and viscous swallows, highest with those who have DES (< 0.05) and normal manometry (< 0.01).

Overall, the combination of manometry, impedancometry, and AIM analysis provided diagnoses and/or a plausible explanation in 95% (40/42) of our patients who presented with NOD, and achieved a significantly higher yield than manometric assessment alone (31/42 [74%], = 0.01), or standard combined pressure-impedance assessment (33/42 [79%], = 0.04; Fig. 5).

Figure 5.

 Comparison of diagnostic outcomes of combined manometry, impedancometry, and AIM analysis for patients who presented with NOD against conventional assessments using either manometry alone or combined manometry and impedance technique.

Discussion

Impaired bolus clearance related to disordered esophageal motor function is thought to be the primary cause of NOD. However, less than half of these patients have identifiable esophageal dysmotility on conventional manometry.6,7 In contrast with the gold standard videofluoroscopy, manometric assessment of esophageal bolus transport and clearance is only by inference1,4,5; and a recent study has indicated that, even with high-resolution manometry (HRM), this technique is insensitive for the prediction of bolus transit as the relationship between esophageal motility and transit is complex and only modest.14 The addition of intraluminal impedance monitoring allows for the evaluation of esophageal bolus transit without the need for radiation exposure, and has been shown to improve the identification of impaired bolus clearance in patients with NOD but with normal manometric studies.6 In contrast with the isolated analysis of either manometric or impedance data, AIM analysis provides a more objective assessment method of esophageal bolus transport by analyzing combined pressure and impedance data, taking into account both the driving forces and bolus transit. This approach may allow detection of more subtle abnormalities in esophageal functional variables related to peristaltic pressure and bolus movement that are not detected on conventional manometry.9,10 The findings of the current study are the first to show that AIM analysis is more sensitive for the detection of subtle abnormalities in esophageal motor function in patients who have NOD with no abnormality found on normal manometry. The addition of AIM analysis to conventional assessment of pressure-impedance signals was able to provide a diagnosis and/or a plausible explanation for dysphagia in 95% of patients who presented with NOD.

Our results indicate that AIM analysis is a complementary technique to the conventional manometric assessment in the evaluation of NOD as it is appears less useful in demonstrating abnormal motor function in known manometric diseases such as DES and NSEMD. Although the median PFI in patients with DES was significantly higher than that in healthy subjects, an abnormal PFI for either liquid or viscous swallows was only seen in 60% of these patients. The underlying reasons why AIM analysis revealed more abnormal PFI in NOD patients with normal manometry than in known esophageal motor disorders remain unclear. One possible explanation is that, as abnormal PFI reflects both peristaltic strength and abnormal flow resistance in the distal esophagus, sufficient esophageal motor contraction in the esophageal body is required to generate propulsive flow across the distal esophagus in order for abnormal flow resistance to be demonstrated. Given esophageal body motor function is universally poor in achalasia and NSEMD, AIM analysis may be less useful in these conditions characterized by overt dysmotility. Possibly, AIM analysis of high-resolution impedance and manometry recordings may yield better results. Thus, based on our findings, manometry should be used for the identification of defined esophageal motility disorders, and AIM analysis reserved for those patients who have normal manometry.

Automated impedance-manometry analysis is also able to identify subtle abnormalities in the pressure-flow relationship. As demonstrated in Fig. 1B and 1C, in NOD patients with normal manometry, the bolus appears to travel much closer to the peristaltic wave than in healthy subjects. We believe that there are two potential explanations for this pattern of bolus flow. First, this may reflect impaired wall compliance of the intrabolus domain that precedes the contraction wave, possibly due to an impairment of descending inhibition. Computational modeling of colonic wall contraction and flow dynamics has recently shown that perturbation of descending inhibition can lead to less efficient bolus transit and increased intraluminal pressures,15 a finding that is analogous to the observation of increased intrabolus pressures in the current study. We note that impaired descending inhibition has been previously characterized in dysphagia patients16 and it has been recently shown that multiple rapid swallowing, a test designed to uncover deficiencies in descending inhibition of the esophagus, has been found to be a predictor of post fundoplication dysphagia.17 Whereas multiple rapid swallowing may provide an indirect marker of inhibitory failure, AIM analysis, through measurement of elevated pressures within the bolus, may provide more direct markers of impaired descending inhibition. In support of this, we have recently shown that PNadImp and IBP are elevated pre-operatively in patients who develop dysphagia after fundoplication.10 Alternatively, there may be a deficiency in pharyngeal swallow, leading to the bolus being propelled a shorter distance into the esophageal lumen. If so then this would place a greater demand on esophageal transport, due to the bolus being less well dispersed. The role of descending inhibition and pharyngeal bolus propulsion in esophageal bolus transport warrants further evaluation and the AIM analysis technique has the potential to provide these new insights.18

The current study also suggests that viscous gel is the test medium of choice for AIM analysis as higher PFI scores were seen with viscous compared with liquid swallows. This is consistent with previous studies that have shown that a viscous test medium is more sensitive and more discriminatory for the detection of abnormal esophageal motor function19,20 given that the success of bolus clearance depends on both the strength of peristalsis as well as the viscosity of the test medium. Focal short-segment hypotensive responses have little impact on total bolus clearance, especially on liquids.3

Pressure-impedance measurements in the current study were made with a sleeve and 5-cm interval side-hole sensors. This might be viewed as a weakness compared with the more recent high-resolution techniques.21,22 Although studies using high-resolution impedance-manometry assessment have demonstrated that large isobaric contour breaks (>5 cm) were uniformly associated with incomplete bolus transit,23,24 a majority of swallows with contour breaks <5 cm are still associated with normal bolus clearance. In a recent study, only 16% swallows with smaller breaks (2–5 cm) were only associated with incomplete bolus transit.23 A side-hole spacing of at least 4 cm has been shown to compare favorably with HRM.25 Furthermore, we have recently demonstrated that small breaks that lead to failure of bolus transit reduce distal intrabolus pressures and PFI.18 Hence, as NOD patients with normal peristalsis still have an elevated PFI, it is unlikely that they have significant breaks that were missed by our low-resolution equipment. Together, we believe that it would be unlikely that our equipment has significant impact on our findings. Further studies with high-resolution impedance-manometry in a larger cohort of NOD are, however, warranted.

Another potential weakness of the current study is the lack of assessment on the relationship between the perception of dysphagia and esophageal motor function for each individual swallow. In patients with gastro-esophageal reflux disease, there is no agreement between objective measurements of esophageal function and subjective perception of bolus passage.26 Furthermore, a recent study indicates that stasis of barium boluses were highly prevalent and were not different between healthy subjects and patients with NOD, and can be regarded as physiological.14 The length of the transitional zone and distal contraction amplitudes were found to predict impaired bolus clearance. Unfortunately, this study did not assess the relationship between perception of dysphagia and bolus clearance on individual swallow basis. Together, these results suggest that perception of dysphagia in patients without mechanical obstruction might be due to esophageal hypersensitivity.

In conclusion, AIM analysis is more sensitive in detecting subtle abnormalities in esophageal functional variables related to peristaltic pressure and bolus movement in NOD patients with normal manometry than conventional assessment of pressure and impedance recordings. Clinically, the addition of AIM analysis to conventional manometry is able to provide an explanation for dysphagia in 95% of patients.

Acknowledgments

The authors would like to thank the technical support provided by Ms Katrina Ching and Mr Marcus Tippett.

Funding

None.

Conflict of interest

Dr. Omari is a technology consultant to Sandhill Scientific.

Author contribution

NQN and TIO conceived the study concept and design, and performed the studies and analyzed data; TIO developed the AIM analysis; NQN drafted the manuscript; NQN, RHH, AJM, and TIO contributed to critical revision of the manuscript and approval of the final version.

Ancillary