Article Text
PDF
Abstract
Background Oesophageal intraluminal impedance is currently used for assessment of reflux in gastro-oesophageal reflux disease (GORD). Oesophageal mucosa integrity may have a key role in heartburn perception in non-erosive reflux disease (NERD). Severe erosive oesophagitis is associated with low impedance baseline. We hypothesised that impedance baseline measurements could be used to evaluate changes in oesophageal mucosa integrity in man.
Methods We measured oesophageal impedance baseline before, during and after acid perfusion in rabbits and healthy subjects. Transepithelial resistance (TER) was determined and dilated intercellular spaces (DIS) were assessed in isolated rabbit oesophageal mucosa. Impedance baseline was measured retrospectively at different levels of the oesophagus in impedance-pH recordings from asymptomatic volunteers and patients with GORD.
Results In healthy subjects and rabbits, impedance baseline dropped dramatically during perfusion of control solution (pH 7.2) but after perfusion, impedance recovered. In rabbits, after perfusion with saline pH 1.5 and 1.0 impedance values remained a 39.1±7.0% and 63.9±6.5% (p<0.05) lower respectively. There was a positive correlation between in vivo basal impedance and in vitro TER values (r=0.72, p=0.0021). Tissue showed no erosions but both acidic solutions induced DIS. In healthy subjects, after perfusion with saline pH 2.0 and 1.0 the impedance baseline remained lower a 21.9±6.5% and 52.7±5.0%, (p<0.0001) respectively. Patients with GORD have a lower impedance baseline than healthy volunteers at the distal oesophagus.
Conclusions Impedance baseline measurements might be used to evaluate the status of the oesophageal mucosa and to study the role of the impaired mucosal integrity in acid-induced heartburn in healthy volunteers and in patients with GORD.
- Impedance baseline
- impaired integrity
- gastro-oesophageal reflux disease
- non-erosive reflux disease
- healthy volunteers
- acid
- epithelial barrier
- gastro-oesophageal reflux disease
- oesophageal reflux
Statistics from Altmetric.com
- Impedance baseline
- impaired integrity
- gastro-oesophageal reflux disease
- non-erosive reflux disease
- healthy volunteers
- acid
- epithelial barrier
- gastro-oesophageal reflux disease
- oesophageal reflux
Significance of this study
What is already known about this subject?
Oesophageal multichannel intraluminal impedance (MII) is a widely spread technique used for the diagnosis of gastro-oesophageal reflux disease (GORD).
The status of the oesophageal mucosa integrity can be assessed in vitro functionally by measuring transepithelial resistance and morphologically by transmission electron microscopy assessing the presence of dilated intercellular spaces.
Transmucosal potential difference was proposed to be a useful in vivo technique for evaluating functionally the status of the oesophageal mucosa. Nevertheless, this technique is not used in clinical practice, is available only in a few hospitals and its utility is doubtful for this purpose.
Clinical practice shows that patients with severe oesophagitis have low baseline impedance.
What are the new findings?
We demonstrated in a rabbit perfused model that impedance baseline measured in vivo correlates with transepithelial resistance assessed in vitro.
In healthy volunteers, lower impedance baseline values were observed after oesophageal perfusion with acidic solutions.
Patients with oesophagitis but also GORD patients without erosions have lower impedance baseline values in the more distal oesophagus compared with healthy volunteers.
How might it impact on clinical practice in the foreseeable future?
Impedance baseline measurements might be used to evaluate the status of the oesophageal mucosa and to study the role of the impaired mucosal integrity in acid-induced heartburn in healthy volunteers and in patients with GORD. Normalisation of impedance baseline may determine symptom relief and a low impedance baseline may predict which patients will respond to treatment.
Introduction
Gastro-oesophageal reflux disease (GORD) is very common, with approximately 20% of the adult Western population complaining of symptoms once per week,1 leading to a large financial burden on society.2 Even though the impact of GORD is evident, the underlying pathophysiology is incompletely elucidated. Non-erosive reflux disease (NERD) is the most common phenotype of GORD. Patients with NERD have typical reflux symptoms caused by the intra-oesophageal reflux of gastric contents (acid, bile acids, and others) but have no visible oesophageal mucosal injury at routine endoscopy.3 4 This is in contrast to patients with reflux oesophagitis, who have obvious oesophageal mucosal injury on endoscopy in the absence of medical therapy. In patients with NERD as well as those with erosive oesophagitis, dilated intercellular spaces (DIS) within the squamous epithelium are typically present. This ultrastructural abnormality is readily identified on transmission electron microscopy (TEM) and is considered a histopathological feature most indicative of GORD.5
Traditionally, the clinical manifestations of GORD are believed to result from prolonged contact of acid with the epithelium. The literature provides strong evidence that the acidity of the reflux event is a major determinant of signs (erosions) of tissue injury and also symptom generation (especially heartburn). Acidic reflux episodes are associated with symptoms6 7 in patients with GORD and, in addition, oesophageal acid perfusion triggers heartburn in these patients as well as in healthy volunteers.8–11 In animal studies, prolonged exposure of the oesophageal epithelium to luminal acid affects mucosal integrity, resulting functionally in decreased transepithelial resistance, increased paracellular permeability and structurally in DIS.12–14 Taken together, these results indicate that DIS may be a marker of altered oesophageal mucosal integrity in patients with NERD, although this hypothesis was not convincingly proven.15 This is at least partly caused by the lack of an easy method to study oesophageal mucosal integrity in man, as assessment of DIS requires sophisticated morphological analysis on biopsies obtained at endoscopy.16–18
It has previously been suggested that measurement of the oesophageal transmucosal electrical potential difference (PD) in vivo provides useful information about oesophageal mucosal integrity.19 Nevertheless, according to Ohm's law (PD=I×R), the PD does not only reflect the integrity of the tissue, as changes in PD may also occur when changes in secretion occur. In an experimental study on the rabbit oesophagus, the transepithelial resistance measured in vivo20 was a more sensitive marker of early damage induced by acid, bile or trypsin than PD.
Although the measurement of oesophageal resistance in man has never been reported, there is emerging evidence that the multichannel intraluminal impedance (MII) technique may provide a potentially useful tool. This technique measures changes in conductivity to an alternating electrical current by a pair of metallic rings mounted on a catheter. The electrical impedance is expressed in ohms and is the equivalent to the resistance to the direct electrical current. In an empty tubular organ such as the oesophagus the metallic rings are in close contact with the mucosa.21 Interestingly, simulation studies show that the status of the mucosa has the largest effect on oesophageal impedance baseline suggesting that it has the potential to reflect mucosal integrity or damage. Practical experience shows that the range of the impedance baseline measured by oesophageal MII in the normal subject is in the range of thousands of ohms, while in the distal oesophagus of patients with severe oesophagitis, MII values are in the range of only a few hundreds of ohms.22 23 While these observations support the use of MII to evaluate oesophageal mucosal integrity, this has never been systematically investigated. Our main aims therefore were (1) to assess whether oesophageal intraluminal impedance is a useful tool to evaluate oesophageal mucosa integrity in an animal model, (2) to evaluate whether this can be confirmed in humans in vivo, and (3) to study whether patients with GORD have altered oesophageal mucosa integrity.
Materials and methods
Animals
Fifteen adult New Zealand male rabbits weighing 2.5–4 kg were anaesthetised with xylazine (10 mg/kg) intramuscularly 10–20 min in advance to the administration of intramuscular ketamine chloride (40 mg/kg) (Parke-Davis, NV Warner-Lambert, Zaventem, Belgium).
Impedance baseline measurements and oesophageal perfusion in rabbits
Oesophageal intraluminal impedance was recorded with a series of cylindrical electrodes, each 4 mm in axial length, spaced at 3 cm intervals and incorporated into a 1.5 mm diameter polyvinyl assembly (prototype developed by Professor Jiri Silny). The assembly used had three recording segments over a distance of 14 cm. Each pair of electrodes was connected to an impedance transducer which delivered a measuring current of less than 6 mA at a frequency of 1 kHz. Impedance was recorded continuously, digitised at 100 Hz, and stored in a computer for subsequent analysis. A polyvinyl tube attached to the impedance catheter was passed by mouth into the oesophagus and the first sensor was located at 2 cm above the lower oesophageal sphincter (LES). Perfusion was performed at 1 ml/min allowing perfusion of the lower two impedance recording segments (10 cm above the LES) for 30 min (Braun Perfusor ED, Braun Melsungen AG, Melsungen, Germany). During the perfusions, rabbits were maintained under sedation with ketamine and in a 20° anti-Trendelenburg position. All solutions were perfused at room temperature. The configuration of the impedance catheter with the polyvinyl tube allows the animal to swallow. After the completion of the perfusion period impedance monitoring continued for 30 min and then animals were sacrificed by a blow on the neck followed immediately by exsanguination. The complete oesophagus was excised, opened, and stripped of its muscle layers in a paraffin tray containing carbogenated Krebs–Henseleit bicarbonate buffer (KHBB, pH 7.4 containing (in mM): 118 NaCl; 4.7 KCl; 1.2 CaCl2; 1.2 MgSO4; 1.2 NaH2PO4; 25 NaHCO3 and 11 glucose).
Solutions used were the following: (1) neutral solution: NaCl 0.9% at pH 7.2; (2) acidic solution at pH 1.5: NaCl 0.9% at pH 1.5; and (3) acidic solution at pH 1.0: NaCl 0.9% at pH 1.0.
Transepithelial electrical resistance
Oesophageal mucosal sections were cut and fixed in Ussing chambers with an aperture of 0.5 cm2. The tissue was incubated in 37°C carbogenated KHBB and transmucosal potential difference was continuously monitored with Ag/AgCl electrodes. Basal transepithelial electrical resistance (TER) was calculated according to Ohm's law from the voltage deflections induced by bipolar constant current pulses of 50 μA (every 60 s) with duration of 200 ms applied through platinum wires (Mussler Scientific Instruments, Aachen, Germany). A correction for fluid resistance was made.
Morphological studies
Following the transepithelial electrical resistance experiments in Ussing chambers, tissues were examined using both light microscopy and TEM to assess the presence of erosions and diameter of intercellular spaces. Tissues were fixed in neutral buffered 4% (w/v) paraformaldehyde for light microscopy and in 2.5% (v/v) glutaraldehyde and 2% (v/v) paraformaldehyde (EM grade, Sigma) for TEM. Paraffin blocks were sectioned (5 μm) and were stained using haematoxylin–eosin methods. For TEM, tissues were then post-fixed in 1% (w/v) osmium tetroxide containing 0.8% (w/v) of potassium hexacyanoferrate(III) (Sigma/RBI, Bornem, Belgium) and infiltrated in Epon's resin and polymerised at 60°C. Ultrathin sections were mounted in copper grids, contrasted and observed in a TEM Hitachi H-7000 at 75 kV equipped with a camera MegaView III (Soft Imaging System).
Four TEM photos/per animal were taken (×4000 magnification) and analysed using custom-written image analysis software in IGOR Pro (WaveMetrics Inc., Oregon, USA). Intercellular spaces were delineated between five and ten epithelial cells from the basal and lower prickle layers in each microphotograph. The intercellular spaces area was measured and related to the perimeter of the corresponding cells to obtain a relative measure of DIS.24 25 The morphological evaluations were performed by one of the investigators who was blinded to the type of mucosal exposure (test solution).
Subjects
Eight healthy volunteers (five men, age 23 (19–32) years) without upper gastrointestinal symptoms were enrolled in the study. All subjects gave written informed consent before inclusion.
Experimental protocol
Participants underwent intra-oesophageal perfusion sessions in a randomised order with at least a 4 week period in between. In the first session, manometry was performed to locate the proximal margin of the LES. A single lumen perfusion catheter and an intraluminal impedance catheter (Sandhill Scientific, Highlands Ranch, CO, USA) were positioned transnasally into the oesophagus so that the perfusion point was located at 6 cm proximal to the upper margin of the LES and the impedance sensors were located 1–3–5–7–11–13 cm above the LES with the pH sensor situated 3 cm above the LES.
After positioning of the catheter, the subjects remained in bed in a semi-recumbent position. Prior to perfusion, oesophageal impedance recordings were performed for 30 min in basal conditions. Next, oesophageal perfusions were performed during 30 min at a flow rate of 2 ml/min. The three first impedance channels were directly in contact with the different solutions. The following solutions were used: (1) neutral solution: NaCl 0.9% at pH 7.2; (2) acidic solution at pH 2.0: NaCl 0.9% at pH 2.0 plus 0.5 mg/ml of pepsin; and (3) acidic solution at pH 1.0: NaCl 0.9% at pH 1.0. After the perfusion had ended, oesophageal impedance–pH recordings continued for 120 min. Impedance–pH recordings were downloaded into a personal computer and analysed using the designated software (Bioview; Sandhill Scientific), impedance baseline values before perfusion for the three most distal MII segments were extracted and compared with values after perfusion.
Symptom score
Subjects scored the intensity of heartburn or pain every 5 min during the perfusion period, using a 10 cm visual analogue scale (VAS) ranged from 0 to 10 (0=no sensation to 10=maximum intensity).
Basal impedance measurements in GORD
We retrospectively analysed 41 consecutive oesophageal impedance-pH recordings from 15 asymptomatic volunteers and 26 patients with GORD (18 NERD and eight oesophagitis grade A–B patients). Seven out of 18 patients with NERD had pathological acid exposure (>5%). All patients had endoscopy within 6 months of the impedance study. All patients were studied ‘off’ treatment with proton pump inhibitors. Impedance baseline values were evaluated during 1 h period at 3, 5 and 7 cm proximal to the LES, in fasting conditions, 30 min after placing the impedance-pH catheter and excluding swallow and reflux induced changes (Sandhill Scientific).
Statistics
All data are expressed as mean±SEM. Single comparisons were performed by paired or unpaired Student t test when appropriate. The effect of the different solutions on time–permeability curves was analysed using two-way repeated measures ANOVA. When the ANOVA test was significant, Dunnett's test was used to determine the times with statistical significant difference. For calculating TER four tissues averaged per animal were taken and for impedance baseline measurements the average of the two (rabbits) or three (humans) recording segments that were in contact with the solutions were considered. Correlations were tested using the Spearman and Pearson tests when appropriate. Significance was declared at p<0.05.
Results
Animal experiments
Effect of acid perfusion on impedance baseline measurements and on in vitro transepithelial resistance
Impedance baseline values before perfusion with the three different solutions did not differ significantly (p=0.607). Perfusion with the solution at pH 7.2 did not induce any changes in impedance (4092±539 Ω at baseline vs 3443±278 Ω, N=5, p=0.153). Impedance baseline before and after perfusion with the solution at pH 1.5 was 4788±580 Ω; and 2840±396 Ω, respectively (decrease of 39.1±7.0%, N=5, p=0.017). Before perfusion with the solution at pH 1.0 the impedance baseline was 4492±281 Ω and after perfusion it was 1552±174 Ω (decreased of 63.9±6.5%, N=5). The decrease in impedance was significantly higher at pH 1.0 compared to pH 1.5 (p=0.0177) (figure 1). Figure 1A illustrates the difference in impedance baseline after perfusion with pH 7.2 and pH 1.0. A quick drop in impedance occurs at the beginning of the perfusion because of the presence of liquid in the oesophagus.
(A) Representative tracings showing the effect of saline pH 7.2 (upper panel) and pH 1.0 (lower panel) on impedance baseline in rabbit oesophagus. (B) Summary of the effect of different solutions on impedance baseline measurements in rabbit oesophagus after perfusion with the different solutions. *p<0.05 versus pH 7.2, ***p<0.001 versus pH 7.2, #p<0.05 versus pH 1.5.
After impedance measurements, in vitro TER was tested in Ussing chambers. Tissues perfused with the neutral solution reached a value of 2079.0 ±105.8 Ω/cm2 (N=5) after the equilibration period. These experiments show that solution pH 1.5 and pH 1.0 reduced TER by approximately 20% (N=5, p<0.05) and 65% (N=5, p<0.0001) respectively (figure 2). The decrease in TER was significantly higher at pH 1.0 compared to pH 1.5 (726.6±125.8 Ω/cm2 and 1710.1±117.1 Ω/cm2, respectively, p<0.001). A significant positive correlation was found between impedance values after perfusion in vivo and the values of TER obtained in vitro (r=0.72, p=0.0021, N=15, figure 3A).
(A) Representative tracings showing the effect of saline pH 7.2 (upper panel) and pH 1.0 (lower panel) on transepithelial resistance (TER) in rabbit oesophageal mucosa after an equilibration period of 2 h. (B) Summary of the effect of different solutions on transepithelial resistance evaluated in vitro. *p<0.05 versus pH 7.2, ***p<0.001 versus pH 7.2, ##p<0.01 versus pH 1.5.
Correlations between (A) in vivo intraluminal impedance baseline measurements and in vitro transepithelial resistance (TER) and (B) in vitro transepithelial resistance and intercellular spaces after perfusion with the different solutions in rabbits.
Histological studies
In animals perfused with solution pH 7.2, pH 1.5 and pH 1.0, gross inspection revealed no erosions, ulcerations or haemorrhage. Light microscopy confirmed the absence of erosions and showed that epithelial thickness was similar in all groups (161.8±13.4 μm, 204.3 ±35.7 μm and 188.8±19.0 μm, respectively, N=5 per group). Despite the absence of changes at light microscopy, perfusion with solution at pH 1.5 and pH 1.0 induced a significant increase in intercellular spaces to 0.12±0.02 μm (p<0.0001) and 0.28±0.06 μm (p=0.0006) respectively compared to the effect of the neutral solution pH 7.2 (0.01±0.002 μm, N=3) (figure 4). The magnitude of the dilation was higher at pH 1.0 as compared with pH 1.5 (p<0.05). A significant negative correlation was found (r=−0.78, p=0.012 and N=9) between the size of the intercellular spaces and the TER values (figure 3B).
Ultrastructural examination of the rabbit oesophageal mucosa after in vivo perfusion with saline at (A) pH 7.2, (B) pH 1.5 and (C) pH 1.0. Observe the presence of dilated intercellular spaces after exposure with acidic solutions.
Human experiments
Effect of acid perfusion on baseline impedance
Impedance baseline at the distal oesophagus (1–5 cm above the LES) before perfusion was 2673±217 Ω, after perfusion with the solution at pH 7.2 an increase by 25.8±7.3% occurred, to reach 3289±246 Ω (p=0.0069). Impedance baseline before perfusion with the solution at pH 2.0 was 2960±244 Ω, after perfusion this value decreased by 21.9±6.5% to 2244±202 Ω (p=0.0011). As in the experiments in rabbits, perfusion with a solution at pH 1.0 resulted in the largest decrease in impedance baseline. Before perfusion the value was 3256±412 Ω and after 1378±103 Ω (decrease of 52.7±5.0%) (figure 5). The decrease in impedance was significantly higher at pH 1.0 compared to pH 2.0 (p=0.0019). Impedance baseline values before perfusion with the three different solutions did not differ (p=0.413).
(A) Representative impedance-pH tracings showing the effect of saline pH 7.2 and (B) pH 1.0 on impedance baseline in human oesophagus. Lower tracing depicts the oesophageal pH at 3 cm above the lower oesophageal sphincter (LES), after perfusion with the acidic solution pH recovers gradually. (C) Summary of the effect of perfusion of different solution on impedance measurements in human oesophagus. ***p<0.001 versus pH 7.2, #p<0.05 versus pH 2.0.
Two hours after perfusion with pH 2.0 the drop in impedance was only 9.4±7.0% and this was not significantly different from the value before perfusion, indicating a recovery of the effect. In contrast, after perfusion with pH 1.0 no recovery was observed (reduction of 47.6±5.5% after 2 h). After perfusion of both pH 2.0 and pH 1.0 solutions the pH recovered to values before perfusion, indicating that the oesophagus was emptied.
Symptom score
Only one of the subjects perfused with neutral solution reported minimal discomfort. In spite of reduced basal impedance, only two subjects reported symptoms during perfusion with the acidic solution at pH 2.0. Overall, symptom scores were similar to those obtained during perfusion of neutral solution (p=0.39) with a median score of 0. In contrast, the perfusion with the solution at pH 1.0 provoked discomfort/pain in six out of eight healthy volunteers with a median VAS score of 4.5 (p<0.05 vs pH 7.2). A significant correlation was found between VAS score and percentage of change in impedance baseline values (r=−0.51, p=0.009). When only symptomatic healthy volunteers were considered, the correlation improved (r=−0.86, p=0.004).
Impedance baseline measurements in GORD
Impedance baseline in healthy volunteers was measured at 3, 5 and 7 cm above the LES. The value at 5 cm was lower compared to 3 and 7 cm (p<0.05). Compared to controls, patients with grade A–B oesophagitis had significantly lower impedance baseline values at 3 and 5 cm, while patients with NERD had only significantly lower impedance baseline values at 3 cm above the LES (p<0.05). The impedance value at 3 cm was significantly higher in NERD compared to oesophagitis (p<0.05) (figure 6).
Impedance baseline measurements at 3, 5 and 7 cm above the lower oesophageal sphincter (LES) in patients with erosive and non-erosive reflux disease (NERD). *p<0.05 versus healthy; **p<0.01 versus healthy #p<0.05 versus NERD. Healthy volunteers N=15, NERD N=18 and oesophagitis grade A–B N=8.
Discussion
We tested the hypothesis that multichannel intraluminal impedance (MII) may be a suitable tool for the assessment of oesophageal mucosal integrity, by performing in vivo experiments of acid perfusion in rabbits and humans, and by retrospectively analysing impedance-pH tracings from patients with GORD. We showed that: (1) impedance baseline values reflect the status of the oesophageal mucosa both in an animal model and in healthy volunteers, indicating that MII is a useful tool to evaluate the oesophageal mucosa integrity; (2) patients with erosive oesophagitis have lower impedance baseline at 3 cm and 5 cm above the LES compared to healthy volunteers; and (3) patients with NERD have lower impedance baseline only at 3 cm above the LES compared to healthy volunteers. These results indicate the presence of impaired mucosa integrity which may help to explain why patients with erosive oesophagitis and, more interestingly, those with NERD experience heartburn.
The oesophagus has a thick (20–30 cell layers) and tight non-keratinised squamous epithelium compared with the rest of the mucosa of the gastrointestinal tract as is shown by a high transepithelial resistance.26 This property avoids the passage of acid through the mucosa in normal conditions.27 When erosions are present it is likely that acid can pass through and reach sensory nerve endings to provoke heartburn. In patients with NERD, visceral hypersensitivity and sustained oesophageal contractions appear to be an important pathogenic mechanism for the development of heartburn. An impaired mucosal integrity (abnormal transepithelial resistance) may underlie the occurrence of oesophageal visceral hypersensitivity and sustained oesophageal contraction,28 suggesting that the status of the oesophageal mucosa may have a key role in heartburn perception in patients with GORD without any erosions.
Experimental in vivo oesophageal acid perfusion in rabbits results in a reduction of transepithelial resistance and an increase in paracellular permeability to small molecules assessed in vitro. These alterations are not linked with erosions but are associated ultrastructurally with the presence of dilated intercellular spaces (DIS).12 13 DIS has been established as a morphological feature in GORD, both in non-erosive and erosive types. Taken together, these results suggest that the presence of DIS in patients with NERD, mainly in the basal cell layer of the epithelium, may indicate an impaired mucosal integrity. Nevertheless, functional data of altered mucosal integrity in NERD patients are not available.
The transepithelial potential difference (PD) is principally generated through active transport of sodium ions from lumen to blood and a passive and slower diffusion of chloride ions in the same direction resulting in a measurable PD. Available data suggest that oesophageal mucosal integrity can be assessed by measuring in vivo transepithelial PD.19 An important finding for this assumption is that transepithelial PD measured in vivo in rabbits strongly correlate with TER measured in vitro with Ussing chambers.12 Controversial results have been published with this direct functional measurement in GORD patients. Although several studies reported abnormal PD values in patients with erosions29–31 compared to healthy volunteers, others have failed to confirm these findings.15 In contrast, in patients with NERD, basal PD values do not differ from controls.
In the present study we showed for the first time that oesophageal intraluminal impedance may be a tool used for estimating the status of the oesophageal mucosa in addition to the use of this technique for the detection of reflux.6 We validated the technique in a rabbit model and we confirmed part of these findings in human volunteers. We found that changes in TER provoke by different acidic solutions were also detected in vivo by measuring impedance baseline. Moreover, we showed a good correlation between impedance baseline values and TER, a validated tool for testing the status of the mucosa. Transepithelial resistance changes induced by the different solutions in rabbit oesophagus did not provoke erosions or apparent damage, only the presence of dilated intercellular spaces as was already reported.13 Interestingly, we found similar results in human volunteers perfused with similar acidic solutions. Acidic solutions at pH 2.0 and 1.0 also provoked a reduction in impedance baseline after perfusion that was comparable to the effect observed in rabbits. Although we did not assess the presence of DIS in healthy volunteers in the presence study, we already published using the same methodology and solutions that after perfusion with the solution at pH 2.0 intercellular spaces were significantly wider compared with the solution at pH 7.2 (0.98 μm vs 0.61 μm).16 We did not assess the intercellular spaces after perfusion of pH 1.0 but considering the higher concentration of H+ in this solution and the effect on rabbit oesophageal mucosa, it is perfectly plausible to accept that it may provoke DIS also in humans. Moreover, other investigators, using a similar protocol, have previously shown that an acidic solution at pH 1.0 provokes widening of the intercellular spaces in human oesophageal mucosa.32
A limitation of the present study is that we did not measure TER in human tissues as we performed in rabbit because of the difficulty to evaluate this on biopsies. Because similar findings were observed in rabbit – lower impedance baseline and the presence of dilated intercellular spaces after acidic solutions – we assume that transepithelial resistance may be also reduced in humans after perfusions. All these findings together suggest us that MII technique could be a tool to detect not only alterations in the oesophageal mucosa from patients with GORD with erosion, but also non-erosive damage as may occur in patients with NERD. For this purpose we analysed, retrospectively, impedance-pH tracings from patients with GORD and healthy volunteers and we calculated the impedance baseline at different levels above the LES. We found that patients with oesophagitis grade A–B have lower basal impedance at 3 and 5 cm indicating impaired tissue integrity. However, at 7 cm above the LES, the impedance baseline in erosive disease did not differ from healthy subjects. This is not so surprising, since most reflux-related lesions are found in the most distal few centimetres of the oesophagus, close to the oesophago-gastric junction. Previous studies reported that an abnormal oesophageal potential difference was found over an area of the oesophageal mucosa with an erosive or ulcerative lesion but not in areas without lesions.15 30 Interestingly, as we expected, patients with NERD have lower impedance baseline at 3 cm above the LES but not in upper levels. Nevertheless the magnitude of the difference is lower compared to patients with oesophagitis. These results indicate that patients with NERD may not have only a structural (presence of DIS) but also a functional alteration in the oesophageal mucosa. As a consequence, the increased sensitivity to acidic solutions (pH 1.0) described in patients with NERD11 may reflect a facilitated passage of acid through the impaired epithelium, leading to activation of sensory nerve endings.28 While the morphological alteration of the intercellular spaces appears along the distal and proximal oesophageal epithelium in patients with NERD,33 lower impedance baseline values were observed only at the very distal oesophagus. An explanation for this controversy can be that ultrastructural alterations in the oesophageal mucosa observed with EM may not be associated with functional changes of the epithelium measured by intraluminal impedance. Although less plausible, it may be also possible that impedance baseline measurements are not able to detect weak ultrastructural alterations in the mucosa. In healthy volunteers values of the intercellular spaces are 0.48 μm at the distal and 0.42 μm at the proximal oesophagus. In patients with NERD, the magnitude of the dilation is three times larger at the distal but only two times at the proximal oesophagus (1.5 μm and 0.82 μm, respectively). Current studies are being conducted in our laboratory to elucidate this discrepancy. The cause of the alterations in the oesophageal mucosa integrity observed in these patients is known. Acid can be the agent in a subgroup of patients since some of them have pathological acid oesophageal exposure.34 Nevertheless, in patients with normal acid exposure, luminal agents present in weakly acidic reflux events as bile acids and trypsin or endogenous factors as stress may play a role.16 24 25 35
Compared to PD measurements, basal impedance measurements offer several advantages in the evaluation of the status of the oesophageal mucosa. First, MII is already a widely spread technique used for the diagnosis of reflux and allows measurements at different levels of the oesophagus. Second, oesophageal transepithelial resistance in vivo20 is a more sensitive marker of early damage than PD measurements. Moreover, electrical impedance is the equivalent to the resistance of the direct electrical current, suggesting that MII may be a more sensitive technique for the evaluation of oesophageal mucosal integrity. Finally, changes in ion transport, which may alter PD measurements, are less likely to affect basal impedance measurements.
Another important finding of the present study is that perception of acid in human healthy volunteers is linked, at least in part, to the disruption of the oesophageal mucosa integrity. We found an excellent correlation (r=0.86) between the status of the mucosa and acid-induced heartburn when we considered only symptomatic volunteers. Oesophageal perfusion with solution at pH 2.0 provokes only a 20% of change in tissue integrity and, as expected, solution at pH 1.0 provoked a higher damage (60% drop of impedance baseline). It is important to remark that the difference in tissue integrity provoked by the two different acidic solutions not only concerns the magnitude of the effect but also the duration of the damage. Two hours after the oesophageal perfusion with the solution at pH 2.0 the mucosal integrity is recovered, but with the more acidic solution the impairment lasts for the entire 2 h study period (no recovery).
The fact that patients with NERD are hypersensitive to acid10 11 and that they may have impaired mucosa integrity as we show in the present paper, suggests that the breakdown in the epithelial barrier may play a key role in the pathophysiology of the disease. This hypothesis should be addressed in future studies evaluating whether normalisation of the altered impedance baseline occurs after medical or surgical treatment, and whether it affects clinical efficacy. It has been shown recently that proton pump inhibitory therapy normalises impedance baseline in infants with symptoms of GORD and low initial baseline impedance.36 This observation awaits further studies in adults to investigate whether normalisation of impedance determines symptom relief and whether a low impedance baseline predicts which patients will respond to proton pump inhibitors treatment.
In summary, this study indicates that impedance baseline measurements might be used to evaluate changes in the status of the oesophageal mucosa in man and allow us studying the role of the impaired mucosal integrity in acid-induced heartburn in healthy volunteers and in patients with GORD.
Acknowledgments
We would like to thank Onofre Castell and Alex Sànchez and the rest of the staff of Servei Microscopia of Universitat Autònoma de Barcelona (UAB) for their technical support.
References
Footnotes
RF and KB contributed equally to this paper.
Funding This work was supported by a ‘‘Geconcerteerde Onderzoeksactie’’ grant from the Catholic University of Leuven, Belgium and by Ministerio de Ciencia e Innovación (BFU2009-11118). RF and KB are postdoctoral fellows supported by FWO (Research Foundation, Flanders).
Competing interests None.
Ethics approval The procedure for the animal experiments was approved by the ethics committee for animal experiments of the Catholic University of Leuven, Belgium. The part of the study involving human subjects was approved by the Ethical Committee of the Catholic University Leuven and performed in accordance with the Declaration of Helsinki.
Provenance and peer review Not commissioned; externally peer reviewed.