Role of the Multichannel Intraluminal Impedance Technique in Infants and Children

Abstract

Recently, multichannel intraluminal impedance (MII) was added to the repertoire of tests available to study esophageal (patho)physiology in children. MII has since been used in 2 major areas. First, it has been used as a diagnostic test for gastroesophageal reflux disease (GERD). The main advantage over traditional pH monitoring is its ability to detect both acid and nonacid gastroesophageal reflux (GER) and to discern between liquid and gas GER.

Although feasible with multiple pH sensors, the MII technique routinely detects the proximal extent of a GER episode. When a pH sensor is added to the MII-catheter, important information about the acidity of a GER event can be gathered.

The second area in which the role of MII has been investigated is that of esophageal function testing. Manometry classically reveals information about esophageal pressure patterns and sphincter function, but does not inform us about bolus flow. MII not only detects the presence of esophageal flow but also adds information on the direction of flow, duration of bolus presence, completeness of bolus clearance, and composition of a bolus.

The combination of MII with manometry enables determination of the relationship between esophageal pressures and flow and, therefore, enhances evaluation of esophageal function in terms of assessment of mechanisms of esophageal volume clearance. In addition, this technique will improve our understanding of (patho)physiological mechanisms in pediatric GERD and other esophageal motility disorders.

Using manometry, esophageal motility patterns can be studied and normal motility is now well described in adults ⁽¹⁾. However, hardly any data are available on motility patterns in healthy children, with the exception of studies performed in premature infants ^(2–5). Gastroesophageal reflux disease (GERD) is by far the most common disease caused by abnormalities in esophageal motility. Abnormal motility patterns, however, are not always predictive of symptoms and patients with normal manometry findings can still experience severe symptoms ^(6,7). It is likely that not just the motor function of the esophagus, but rather the combination of motor events and the actual passage of a bolus (esophageal flow) may be important in the pathogenesis of these symptoms. Therefore, a test revealing both motility pattern and flow would be ideal for studying these disorders and may identify a relation between symptoms and objective abnormalities more often than with conventional manometry. Combined videofluoroscopy and manometry has been used for this purpose ⁽⁸⁾, but is hampered, especially in children, by the necessity for radiological exposure.

GASTROESOPHAGEAL REFLUX IN HEALTH AND DISEASE

Gastroesophageal reflux (GER) is defined as the retrograde flow of gastric contents into the esophagus and is universally present in infants and children as well as in adults. Its presence is especially noticeable during infancy, when occasional regurgitation is seen in nearly all children. When GER causes significant symptoms or complications, it is referred to as GER “disease” (GERD). This diagnosis is conventionally based on clinical history. However, diagnostic challenges arise in infants and children with less specific or even extra-esophageal symptoms (eg, irritability or apnea in infants, chronic cough or wheezing in older children). Therefore, an association between symptoms and GER is often sought.

The mechanism underlying most of the GER episodes in healthy volunteers and healthy premature infants as well as in adult and pediatric patients with GERD is transient lower esophageal sphincter relaxation (TLESR) ^(2,9–12). TLESRs not only depressurize the stomach following swallowing of air in association with food boluses but also allow reflux of liquid gastric contents to occur. It is now clear from adult studies that TLESRs are not more common in patients with GERD, but rather are associated with a higher proportion of liquid and acid GER ^(13,14). These findings have been confirmed in premature infants ⁽¹¹⁾.

Other factors than TLESRs contribute to the pathogenesis of GERD. Among these, abnormal esophageal peristalsis and deficient lower esophageal sphincter (LES) competence are thought to play a role. Although motility patterns in children with GERD have been described ^(12,15), the interpretation of these is difficult, because normative data do not exist. However, motility does appear to become impaired with increase severity of esophagitis ⁽¹²⁾.

The absence of a clear motor abnormality in GERD limits the utility of manometry as a diagnostic test for GERD. Until recently, the only other functional test was 24-hour esophageal pH monitoring, which is able to detect pH changes in the esophagus and thus acid GER. This test is often performed, but has major drawbacks. The results are highly reliant on the reflux index (the percentage of time with an esophageal pH of below 4), whereas its cutoff value for pathological GER is still a matter of debate ⁽¹⁶⁾. Furthermore, adult pH monitoring results have shown a significant day-to-day variability ⁽¹⁷⁾ and a poor correlation between pH monitoring results and symptoms or findings on endoscopy ^(18–21). In addition, pH monitoring also possesses intrinsic qualities, which limit sensitivity and specificity. First of all, GER episodes are indirectly detected by acidification of the lumen surrounding the pH sensor. For recognition of a GER episode, pH must drop by at least 1 point and to a value lower than 4. As a result, weakly acidic (4 < pH ≤ 7) and weakly alkaline (pH > 7) GER episodes are per definition not detected by standardized esophageal pH monitoring. This is especially problematic in infants, in whom frequent feeding and subsequent buffering of stomach contents causes the majority of GER episodes to have a pH above 4 ⁽²²⁾. Although precise pathophysiological mechanisms are yet to be determined, an increasing body of evidence exists suggesting a role for weakly acidic GER episodes in the development of several esophageal and extraesophageal symptoms ⁽²³⁾. Another consequence of this indirect measurement of GER is that superimposed acid GER (ie, acid GER during the period in which the acid of a previous GER episode is still being cleared) is not detected. Finally, the results of pH monitoring are highly influenced by dietary intake during the test, because acidic food and drinks will also cause a drop in pH in the surroundings of the pH sensor. Dietary restrictions are, therefore, commonly imposed on patients, resulting in test conditions that do not reflect daily routine.

Despite its limitations, esophageal pH monitoring is still performed in most centers due to the fact that it is readily available, analysis is automated, it is cost effective, and the result of the test can be interpreted in relation to established normative values for esophageal acid exposure time. In support of the continued use of 24-hour pH monitoring, existing evidence shows that increased acid exposure is a significant pathophysiological factor in GERD ⁽¹¹⁾. However, a test that could show all GER episodes would be preferable and could probably help clarify symptom genesis, especially in those patients not responding to acid suppressive therapy.

MULTICHANNEL INTRALUMINAL IMPEDANCE

Basic Principles

Electrical impedance is defined as the quotient of voltage and electrical current. Multiple electrodes can be assembled on an esophageal catheter to measure impedance between each pair of electrodes. This is now known as esophageal multichannel intraluminal impedance (MII). The technique is based on impedance being inversely proportional to ionic concentrations of luminal contents. As a result, bolus with relatively low ionic contents (eg, air) will result in higher impedance measurements compared with a bolus with relatively high ionic contents (eg, saline, refluxate). Hence, the passage of a bolus can be recorded and because of multiple measurements sites, bolus flow can be visualized (Fig. 1).

FIG. 1: Multichannel intraluminal impedance patterns of (liquid) gastroesophageal reflux and its proximal extent. Two tracings from a multichannel intraluminal impedance recording. Lines indicate position of impedance segments in the esophagus. Arrows indicate a pattern of retrograde impedance drops (ie, gastroesophageal reflux) extending into the proximal esophagus in the right panel but not in the left. GER = gastroesophageal reflux.

Since its introduction, MII has been used in 2 major areas: first, as a new diagnostic tool for GERD and second, in combination with manometry to enable determination of the relation between esophageal pressures and esophageal bolus flow, thereby enhancing the ability to evaluate esophageal function in terms of assessment of mechanisms of esophageal volume clearance. Furthermore, it gives the opportunity to study motor patterns underlying GER.

MII as a Diagnostic Tool for GERD

Because MII detects esophageal flow directly, it is possible to detect all GER episodes and in combination with a pH sensor classify these into acidic, weakly acidic, and weakly alkaline GER. Due to the multiple measuring sites, the direction of flow can be determined. Hence, GER can easily be discerned from swallowed material, making dietary restrictions unnecessary. Furthermore, this makes it possible to study the mechanisms of bolus and acid clearance and provides information on the proximal extent of a GER episode (Fig. 1), which can be helpful for determining a relation between GER and extraesophageal symptoms ⁽²⁴⁾.

MII has its limitations, too. First, normal values do not exist in the pediatric age range and although efforts to establish these have been made ⁽²²⁾ and are ongoing, it is unlikely that truly normative data will ever become available because of ethical considerations. Even though MII can detect both acid and weakly acidic GER, it cannot discriminate between the two. This is easily resolved by adding a pH sensor to the MII catheter (Fig. 2). These combined MII-pH catheters are now commercially available for use with infants and children. Other issues limiting its clinical use at present are the costs of consumables, which are 4 to 5 times higher than a pH study, and the time required for analysis, which ranges from 30 minutes to 4 hours depending on the frequency of reflux and the experience of the analyst. The recent development of displaying impedance patterns in spatio-temporal color plots, enhances easier recognition of GER and will probably improve sensitivity of GER detection and reduce analysis time (Fig. 3). Although available and continuously improving, the sensitivity and specificity of the software currently used for automated recognition of MII GER patterns requires further optimization and is unlikely to obviate the need for some degree of manual review of the tracing.

FIG. 2: Examples of infant catheters. (A) Catheter for combined MII and manometry, incorporating a feeding tube. (B) Catheter for combined MII and pH-metry. Configurations change with age. Most commercially available catheters have 6 impedance channels (7 electrodes). Imp = impedance segment; LES = Lower esophageal sphincter.

FIG. 3: Conventional line plot and novel spatio-temporal color plot of the same gastroesophageal reflux event. In the top panel a conventional line plot is shown where the x-axis represents time, and the position in the esophagus is depicted on the y-axis (with the most proximal channel on top and the most distal channel on the bottom). Each line represents a single impedance segment. This Gastroesophageal reflux event is a mixed episode, with gas (rise in impedance) preceding the liquid (drop in impedance). The drop in impedance, however, is hard to recognize due to the preceding gas. In the bottom panel the same reflux event is seen. As in the line plot, x-axis represents time and y-axis represents the position along the catheter. However, the pressure scale is color-coded, with blue representing baseline impedance levels, black representing high impedance levels (gas), and yellow and red representing low impedance levels (liquid). It can be appreciated that the mixed nature of this gastroesophageal reflux episode is much easier to recognize than on the conventional line plot.

Pediatric Studies Using MII For GER Detection

Skopnik et al ⁽²⁵⁾ were the first to describe a study using MII in the pediatric population. They evaluated 17 infants with symptoms of GER with both pH-metry and MII and found that the majority of postprandial GER episodes are not accompanied by a drop in esophageal pH. Furthermore, it became clear that almost 40% (71/185) of all episodes detected by pH-metry were actually associated by an antegrade pattern of impedance drops, suggesting it was part of the clearance of a preceding acid GER episode, rather than a new episode. In subsequent studies, pH drops to below 4 were seen without any change in impedance ⁽²⁶⁾. Theoretically, these could be bolus too small to be detected by MII, but it was shown that even small bolus could in fact be detected by MII ⁽²⁷⁾. It is still largely unclear what these “pH-only” events represent. Most experienced analysts of impedance tracings would, however, recognize some degree of subthreshold impedance change with most pH-only GER. These patterns become increasingly more recognizable using spatio-temporal plotting as mentioned above. Irrespective of how they are caused and what they represent, the most relevant question with respect to pH-only GER is whether they are important from a diagnostic perspective. Further studies are needed to elucidate this.

Since the paper by Skopnik ⁽²⁵⁾, multiple studies were performed with MII and pH-metry in children with symptoms of GER and healthy premature infants ^{(22,24,26,28–45)}. The results of these studies are summarized in Table 1.

TABLE 1: Prospective studies using MII and pH for the detection of GER

Recently, normal values in preterm infants were proposed ⁽²²⁾. These are helpful for the interpretation of MII/pH studies, but some drawbacks of this study need to be kept in mind. First, all infants received tube feeding at the time of the study. This means that they were in fact not completely healthy, because their gastrointestinal tract was unable to tolerate full oral feeding. Second, a nasogastric tube crossing the esophagogastric junction influences the number of GER episodes ⁽⁴⁶⁾. Where MII/pH studies could normally be performed with the catheter above the esophagogastric junction, Lopez-Alonso et al had to position the catheter in the stomach to allow for tube feeding ⁽²²⁾.

All other studies were performed in symptomatic children, using different inclusion criteria and different study protocols in different age and patient groups. Therefore, data cannot simply be pooled. However, from these studies it becomes clear that a significant proportion of all GER episodes is weakly acidic and that weakly acidic GER is more prevalent than acid GER in infants with symptoms ^{(22,25,26,28,29,34,40)}. This is probably because of the frequent feeding and subsequent buffering of gastric contents ⁽⁴⁷⁾. In older symptomatic children acid and weakly acidic GER seem to occur approximately at the same rate ^(31,35–38) as in adults with GERD ⁽⁴⁸⁾. On the contrary, in a group of children with chronic respiratory complaints, hardly any weakly acidic GER was shown ⁽²⁴⁾.

Symptom Association

Several authors have focused on the association between GER and symptoms. A clear temporal association between GER episodes and symptoms seems to provide convincing evidence of one causing the other. With the ability of MII to detect all GER, this is a promising means of diagnosing such an association.

However, several difficulties arise. First, not all patients experience symptoms during a 24-hour study period. Second, a clear definition of temporal association is lacking. In the case of apnea, several studies have used different time intervals to investigate an association ranging from anywhere between 20 seconds and 5 minutes ^{(24,28,29,31,34,36,37,40,44,49)}. Also, some authors have only included the time period before a symptom, whereas others have suggested using only those GER episodes that reach the proximal esophagus, especially in patients with respiratory symptoms ⁽²⁴⁾. Third, it is unclear how to measure a temporal association over a 24-hour (or other) period.

Three measures have been described: the symptom index (SI), the symptom sensitivity index (SSI) and the symptom association probability score (SAP). The SI is the percentage of symptoms related to a GER episode and is considered positive when above 50% ⁽⁵⁰⁾. The SSI is defined as the percentage of GER episodes that is associated with symptoms and is considered abnormal when higher than 10% ⁽⁵¹⁾. These 2 tests, however, have cutoff points that were arbitrarily chosen. Moreover, they do not take into account the total number of GER episodes and total number of symptoms, respectively. Therefore, the SI can be positive simply because of a high number of GER episodes, whereas the SSI can be false positive because of a high number of symptoms. The SAP was developed to overcome these problems. It is a statistical means (Fisher exact test) of calculating the probability that the symptoms and GER episodes found are unrelated. The P value of this test is then subtracted from 100% to reveal the SAP ⁽⁵²⁾. However, it is unclear whether these measures truly reveal associations between GER and symptoms because these are based on statistical probability and assumptions. Nevertheless, they are the best available and have been used extensively in the adult literature, particularly with respect to heartburn. No gold standard exists to validate these measures against, and therefore prospective intervention studies are needed to objectify the value of each of these tests. Such studies should incorporate all 3 measures and look for associations and (ideally) different epochs to answer the question: which of these measures most closely reflects reality. It may well be that such studies will show that different cutoff values or epochs are appropriate for different symptoms or age groups. It could also show that none of these estimates of association are good enough and novel statistical means need to be developed. Nevertheless, several studies in children have been performed that assessed the association between GER and symptoms. Results from these studies are outlined in Table 2 and further underline the problems mentioned above. Multiple symptoms are studied using different means of association and different time intervals. This makes a comparison hardly possible. The symptom most commonly studied is apnea in premature infants. Even for a symptom, which is much more objective than symptoms subjectively reported by patients or caregivers, the association varies widely in the different studies (Table 2). This is not only due to the imperfect measures of association but also to largely varying inclusion criteria, different definitions of apnea and low numbers of patients. Despite the difficulties, most studies fail to show a temporal link between apnea and GER and when an association is found, apnea usually precedes GER. Other symptoms commonly thought to be GER related are only studied with subjective measures (eg, reported by parents), but may be worth studying when an objective means of establishing the presence of the symptom can be used.

TABLE 2: Studies using MII to study the association between GER and symptoms

Even though symptom association testing in children is far from perfect, it can be helpful in selected cases, as long as the limitations mentioned above are taken into consideration when the results are interpreted by the clinician.

By virtue of the fact that it records all GER episodes irrespective of pH, pH-impedance monitoring has been shown in adults to improve the yield of symptom–reflux associations in GERD patients on and off therapy ^(53–55). As such it is a far better clinical tool than pH monitoring alone. The high cost of consumables, however, makes it essential to gain a clear understanding of when pH-MII testing is warranted and when pH-only testing will suffice.

Intervention Studies

MII has been used as an objective outcome measure in intervention studies. Two of these studies looked at the effect of feed thickeners ^(30,39). In both studies, infants with symptoms suggestive of GER were subjected to a crossover design, thereby avoiding the problem of the absence of normative data. Interestingly, Wenzl et al found a small but statistically significant reduction in the number of weakly acidic GER episodes and proximal extent of GER in full-term infants fed with formula with and without bean gum ⁽³⁰⁾. In contrast, Corvaglia et al found no difference at all in preterm infants who were given human milk with and without precooked starch ⁽³⁹⁾. Although different age and/or feeding regimes may play a role in the effectiveness of thickening, results of these studies may also be influenced by the small number of patients studied (5 and 14, respectively) ^(30,39).

Del Buono et al showed that Gaviscon Infant does not reduce the number of GER episodes and only marginally reduces the proximal extent of GER, raising questions regarding the clinical benefit of Gaviscon Infant ⁽³³⁾.

Recently, the first study looking at the effect of proton pump inhibitors on GER in preterm infants and neonates was presented showing that esomeprazole does not change the total number of GER episodes, but merely changes acid GER into weakly acidic GER ⁽⁵⁶⁾. This is expected because proton pump inhibitors do not target TLESR. It, therefore, reaffirms the need for more research into pathophysiological mechanisms underlying GER, which can eventually lead to reflux-inhibiting medication that targets triggering TLESRs.

MII in Studying Physiology of Pediatric GER

The influence of nasogastric tubes on GER in children is a matter of debate ⁽⁵⁷⁾. Using MII, it was convincingly shown that more GER episodes were triggered with the catheter placed in the stomach than when it was placed in the esophagus ⁽⁴⁶⁾. This suggests that gastric contents are not simply leaking into the esophagus because of the catheter being present, but that this catheter may actually be triggering more GER episodes.

Using MII/pH, it was shown that even though the esophageal emptying (volume clearance) after a GER episode is equal during feeding and fasting, the neutralization of acid takes up to 2.4 times as long during fasting ⁽⁵⁸⁾. The authors suggest that acid-clearance mechanisms are less efficient during periods of fasting.

Esophageal MII/Manometry

In adults, the combination of MII and manometry has become a widely used technique and has been useful in studying pathophysiology of motility disorders. Patients are given liquid and viscous bolus to determine motor patterns and the effectiveness of esophageal bolus transport. As such, it has been labeled esophageal function testing and is used in patients with any suspected esophageal motility disorders ⁽⁵⁹⁾. Other than that, MII/manometry also allows for the detection of motor events underlying GER episodes and can, therefore, be helpful in determining pathophysiological mechanisms ⁽¹³⁾.

In children, the number of studies is limited, mainly because of the ethical problem that performing a study with both techniques means that the child needs to be intubated with 2 catheters. For research purposes, this is unacceptable in the pediatric age range. Therefore, catheters have been developed combining both techniques and also incorporating a feeding tube, without increasing their size (Fig. 2) ⁽⁶⁰⁾. Using such a catheter, we have shown that positioning healthy infants in the left lateral position slows down gastric emptying, but paradoxically increases the number of TLESRs and GER episodes ⁽⁶⁰⁾. Subsequently, we have shown that placing an infant in the right lateral position for the first postprandial hour and in the left lateral position thereafter dramatically reduces TLESRs and GER in the late postprandial period (when GER is acidic), whereas gastric emptying is enhanced ⁽⁶¹⁾. Even though these results need to be confirmed in infants with GERD in a large clinical trial, they are promising and may result in a conservative therapy that actually targets the underlying mechanism of GER.

Although adult data show promise, pediatric studies using MII/manometry for esophageal function testing are lacking but may prove to become useful in selected patient groups, such as patients undergoing surgery for esophageal motility disorders. It is even conceivable that in the future patients will be selected for surgery on the basis of esophageal function test findings.

CONCLUSIONS

MII is a technique that allows for detection of flow throughout the esophagus. It is, therefore, able to detect all GER, whether acidic, weakly acidic, or weakly alkaline. Although the lack of normative data will prevent it from becoming an absolute gold standard in the diagnosis of GERD, MII in combination with pH-metry will replace pH-metry alone. It not only detects all GER but also enables us to more accurately reveal associations between GER and symptoms. Considering these advantages, it is now a matter of time before the hurdles preventing MII from becoming a standard clinical tool will be overcome.

In combination with manometry, MII is a powerful tool to clarify (patho)physiological mechanisms underlying esophageal motility disorders, including GERD. In the future, esophageal function testing could prove to be useful, for example, as a preoperative assessment for children in whom antireflux surgery is considered.