2023 Impact Factor
2023 Impact Factor
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.Gastroesophageal reflux disease (GERD) is a chronic disease that affects up to 20% of the Western population.1-3 Although GERD is common, its diagnosis can be challenging as its symptoms are nonspecific, its clinical presentation is heterogeneous, and it overlaps with other gastrointestinal (GI) disorders.4,5 According to several authors, there is no gold standard in GERD diagnosis, and the diagnosis should be based on a combination of several factors, such as clinical presentation, therapeutic response, endoscopic evaluation, and prolonged monitoring of gastroesophageal reflux.6
A new metric assimilated in GERD diagnosis is the mean nocturnal baseline impedance (MNBI), evaluated during multichannel intraluminal impedance-pH monitoring (MII-pH). During rest at night, there are no refluxes or swallowing; therefore, the esophageal lumen collapses, leaving the metallic wipes in close contact with the mucosa. Thus, it is possible to estimate the electrical conductivity of the mucosa, which is considered the basal impedance.7,8
Previous studies have supported the idea that basal impedance decreases and remains low after the infusion of an acidic solution into the esophagus of healthy individuals.9 It is postulated that damage to the mucosa increases the dilation of intercellular spaces, with higher ionic concentration and, consequently, greater electrical conduction. If electric current is conducted more easily, the impedance decreases.9-11
The measurement of basal impedance using MNBI correlates with the integrity of the esophageal mucosa and is an interesting method to assess the effects of chronic exposure of the epithelium to reflux. Low baseline impedance values are associated with increased acid exposure and sensitivity, whereas normal values are observed in patients with functional heartburn.5,12-14 Low MNBI values (< 2292 Ω) predict, for example, a favorable response to anti-reflux therapy.15,16 However, for its measurement, it is necessary to perform MII-pH and, consequently, maintain the uncomfortable transnasal catheter for a long time.17 The development of new devices to assess the integrity of the mucosa is one of the main targets for the evolution in GERD diagnosis.
Yuksel et al18 developed a catheter to directly measure the conductivity of the esophageal epithelium during upper GI endoscopy (UGE). The catheter is introduced through the working channel of the device during the examination, and mucosal impedance (MI) is measured by direct contact.
Considering that this is a simple and promising method, the present study evaluates the reproducibility of this catheter for the direct measurement of esophageal MI during UGE in distinguishing patients with and without evidence of GERD on MII-pH.
This study comprised patients with typical GERD manifestations (heartburn and/or regurgitation) who were referred to our center for diagnostic testing for GERD. Individuals with Los Angeles (LA) grades C and D esophagitis, Barrett’s esophagus, hiatal hernia greater than 3 cm, or previous upper digestive tract surgeries were excluded from this study.
The patients were divided into 2 groups according to the acid exposure time (AET): group A (AET < 4%, GERD diagnosis excluded) and group B (AET ≥ 4%, possible GERD diagnosis, inconclusive or confirmed). All participants signed a consent form before participating in this study. This study was approved by the Research Ethics Committee of the Faculty of Medicine at the University of São Paulo (Approval No. 3.255.132).
The patients were instructed to discontinue proton pump inhibitors and histamine H2 blockers for at least 14 days before this study. Demographic data, medical histories, and symptoms were collected using standardized questionnaires. The Gastroesophageal Reflux Disease–Health-related Quality of Life (GERD-HRQL) questionnaire was used to assess the quality of life.19,20 After overnight fasting, the patients underwent high-resolution manometry and MII-pH. The following day, after removing the MII-pH catheter, the patients underwent UGE without prior knowledge of the results of the other tests. MI was measured directly during UGE with a modified impedance catheter introduced through the endoscope working channel.
High-resolution manometry was performed using a 24-channel water-perfused catheter (Alacer Biomédica, São Paulo, Brazil). Two experienced investigators evaluated all tracings according to the standard procedure recommended by the Chicago classification version 4.0.21 The following quantitative parameters were evaluated:21,22
Esophagogastric junction contractile integral (EGJ-CI) (mmHg∙cm): EGJ hypotonia was considered when values were < 25 mmHg∙cm.23
Integrated relaxation pressure (mmHg)
Distal contractile integral (mmHg∙s∙cm)
A catheter with 1 pH channel and 6 impedance channels (Alacer Biomédica) was used. The pH sensor was positioned 5 cm above the upper border of the lower esophageal sphincter and was determined manually. A diary was provided to the patients for activity registration (meals, lying down, sleeping, symptoms). The monitoring period was 24 hours, with pH values recorded every 4 seconds, that is 15 readings per minute. The tracing was exported to a computer and critically evaluated using the MII-pH analysis software version 1.1.2.1 (Alacer Biomédica) by 2 experienced researchers.
The following quantitative parameters were evaluated:24
Total AET %: according to the Lyon Consensus25
Number of distal refluxes
Number of proximal refluxes
MNBI (Ω)
Post-reflux swallow-induced peristaltic wave (PSPW) index (%)
For this study, a catheter was developed with Alacer Biomédica, based on Yuksel et al’s study.18 This device is originally 2 mm in diameter and has 2 circumferential ring sensors. Each of these sensors is 2 mm long, with the most distal sensor located 1 mm from the tip of the catheter, and the distance between them is 2 mm. This layout was adapted for stainless steel sensors 4 mm long in a polyvinyl catheter, with a sensor separation of 1 mm and a distal sensor mounted 2 mm away from the tip of the catheter (Fig. 1).
Figure 1. Schematic catheter representation for direct mucosal impedance measurement during upper gastrointestinal endoscopy. EGJ, esophagogastric junction.
The electrodes were connected to an impedance transducer located at the bedside using copper wires that ran internally along the entire length of the catheter. The voltage generated by the transducer was limited to a maximum current of 50 µA, whereas the frequency for the measurement circuit was 1 kHz. The impedance (I) is expressed in ohms (Ω) as the ratio between the voltage (T) and current (C), that is, I = T/C.
During UGE, an MI catheter was introduced through the endoscope’s working channel, and continuous measurements were taken for at least 10 seconds on the right lateral esophageal wall at 2, 5, 10, and 18 cm above the EGJ (Fig. 1). In the study by Yuksel et al,18 MI measurements were obtained continuously for 5 seconds at each location, and the measurements’ median for each location was used for analysis. In our research, we chose to extend this observation period, as we considered that one of the factors that most influenced the measurement was respiratory movements. Considering a minimum normal respiratory rate of 12, each respiratory movement will last about 5 seconds; therefore, 10 seconds would evaluate at least 2 complete respiratory cycles.
An adapter was used to connect the impedance device directly to a computer using a USB cable; thus, it was possible to record the impedance measurements in real-time using the MII-pH analysis software version 1.1.2.1 (Alacer Biomédica). Subsequently, 10 continuous seconds of the most stable tracing at each distance were identified to calculate the average. Figure 2 shows examples of the tracings obtained in this step.
Figure 2. Trace of mucosal impedance measurements at 2 cm above the esophagogastric junction in the analysis program. (A) The highlighted section represents the 10-second interval where the average impedance measurements (2190 Ω) were calculated. (B) The highlighted section represents the 10-second interval where the average of the measurements was calculated (5290 Ω) in another patient. Some artifacts are observed in this trace, whether due to esophageal contractility, respiratory movements, or heartbeats. This reveals the inefficiency of a single parameter measurement and the necessity to calculate the average of records.
Initially, the data are described with frequency and confidence interval for qualitative variables and with measures of central tendency (mean and median) and dispersion (standard deviation, interquartile range, minimum, and maximum) for quantitative data. Nonparametric tests were selected because the quantitative variables of interest did not present a normal distribution, a hypothesis tested using Kolmogorov–Smirnov statistics.
The existence of an association between the qualitative variables and AET (< 4% × ≥ 4%) was assessed using the chi-squared test. The Wilcoxon or Kruskal–Wallis evaluated the association between quantitative and qualitative variables. When necessary, in comparison between groups, post-hoc tests (chi-squared or Dunn test) with Bonferroni correction were used to locate the groups in which the statistical differences in the initial test were found. The diagnostic accuracy of the MI values at 2, 5, 10, and 18 cm above the EGJ was assessed using the receiver operating characteristic curve analysis. The correlation between MNBI and MI was tested with Spearman’s correlation. The strength of the association was classified according to the r-value as very weak (0.00 to 0.19), weak (0.20 to 0.39), moderate (0.40 to 0.69), strong (0.70 to 0.89), or very strong (0.90 to 1.00).
All statistical analyses considered a bidirectional α (P-value) of 0.05 and a confidence interval of 95%. All analyses were performed using R software or IBM SPSS Statistics for Windows version 25 (IBM Corp, Armonk, NY, USA).
Sixty patients were included in this study, of whom 28 were classified into group A (AET < 4%) and 32 into group B (AET ≥ 4%). Most participants were women, with a similar distribution in both groups (P > 0.999). Moreover, there was no difference between the groups regarding body mass index, comorbidities (hypertension and diabetes mellitus), and smoking. The distribution of symptoms was similar between the studied groups. A statistically significant difference was found only for eructation, which occurred in 42.8% of the participants in group A and 75.0% in group B (P = 0.023). The GERD-HRQL scores were also similar between groups (Table 1).
Table 1 . Demographic and Clinical Characteristics of the Patients, According to the Acid Exposure Time (< 4% vs ≥ 4%)
| Characteristics | Group A AET < 4% | Group B AET ≥ 4% | P-value |
|---|---|---|---|
| Age (yr) | 47.9 ± 10.3 | 46.2 ±12.3 | 0.500 |
| Women | 75.0 | 78.1 | > 0.999 |
| BMI (kg/m2) | 28.3 ± 4.6 | 25.8 ± 4.5 | 0.060 |
| Hypertension | 28.6 | 18.7 | 0.554 |
| Diabetes mellitus | 10.7 | 0.0 | 0.192 |
| Smoking | 7.1 | 6.2 | > 0.999 |
| Symptoms | |||
| Heartburn | 85.7 | 90.6 | 0.851 |
| Regurgitation | 82.1 | 84.4 | > 0.999 |
| Dysphagia | 32.1 | 40.6 | 0.681 |
| Chest pain | 32.1 | 28.1 | 0.955 |
| Epigastric pain | 35.7 | 37.5 | > 0.999 |
| Nausea | 42.9 | 37.5 | 0.874 |
| Bloating | 78.6 | 78.1 | > 0.999 |
| Eructation | 42.9 | 75.0 | 0.023a |
| Throat clearing | 42.9 | 53.1 | 0.593 |
| Chronic cough | 32.1 | 21.9 | 0.545 |
| Hoarseness | 10.7 | 25.0 | 0.275 |
| GERD-HRQL score | 11.14 ± 6.1 | 13.84 ± 6.2 | 0.100 |
aA P < 0.05 was considered statistically significant.
BMI, body mass index; GERD-HRQL, Health-related Quality of Life Scale for Gastroesophageal Reflux Disease.
Data are presented as mean ± SD or %.
The main manometric, endoscopic, and MII-pH characteristics are shown in Table 2.
Table 2 . Manometric, Endoscopic, and Multichannel Intraluminal Impedance pH-monitoring Parameters of the Patients, According to the Acid Exposure Time (< 4% vs ≥ 4%)
| Parameters | Group A AET < 4% | Group B AET ≥ 4% | P-value |
|---|---|---|---|
| Manometric parameters | |||
| EGJ-CI (mmHg∙cm) | 32.2 ± 15.7 | 30.8 ± 17.7 | 0.500 |
| EGJ hypotonia | 21.4 | 37.5 | 0.283 |
| IRP (mmHg) | 5.6 ± 3.1 | 5.9 ± 3.3 | > 0.999 |
| DCI (mmHg∙s∙cm) | 1699.7 ± 1195.5 | 1316.9 ± 831.5 | 0.300 |
| IEM (Chicago 3.0 criteria) | 17.9 | 28.1 | 0.527 |
| IEM (Chicago 4.0 criteria) | 10.7 | 18.7 | 0.612 |
| Endoscopic parameters | |||
| EE | 0.008 | ||
| Without EE | 23 (82.1) | 14 (43.8) | |
| EE grade A | 4 (14.3) | 11 (34.4) | |
| EE grade B | 1 (3.6) | 7 (21.9) | |
| Helicobacter pylori-positive | 5 (17.9) | 6 (18.8) | > 0.999 |
| MII-pH parameters | |||
| AET (%) | 1.7 ± 1.1 | 7.8 ± 4.2 | < 0.001 |
| Proximal reflux | 7.1 ± 7.6 | 21.2 ± 17.9 | < 0.001 |
| Distal reflux | 37.7 ± 23.9 | 78.0 ± 28.1 | < 0.001 |
| MNBI (Ω) | 5544.3 ± 2558.6 | 3702.2 ± 2264.9 | < 0.001 |
| PSPW index (%) | 59.9 ± 16.6 | 33.4 ± 21.4 | < 0.001 |
AET, acid exposure time; EGJ-CI, esophagogastric junction contractile integral; IRP, integrated relaxation pressure; DCI, distal contractile integral; IEM, ineffective esophageal motility; EE, erosive esophagitis; MII, multichannel intraluminal impedance-pH monitoring; MNBI, mean nocturnal baseline impedance; PSPW, post-reflux swallow-induced peristaltic wave.
Data are presented as mean ± SD, %, or n (%).
There were no significant differences in EGJ-CI, integrated relaxation pressure, and distal contractile integral values between the groups. Furthermore, although the proportion of EGJ hypotonia was greater in group B than in group A (37.5% vs 21.4% in group A), this difference was not statistically significant (P = 0.283). Similarly, the proportion of participants with ineffective esophageal motility was higher in group B than in group A when using the Chicago classification version 3.0 criteria (28.1% vs 17.9%) and when considering 4.0 (18.7% vs 10.7%), but without statistical significance.
Group B had a higher frequency of erosive esophagitis (34.4% of LA grade A esophagitis and 21.9% of LA grade B esophagitis) than group A (14.3% of grade A esophagitis and 3.6% grade B esophagitis), with statistical significance (P = 0.008). The only individual in group A with grade B erosive esophagitis had an AET of 2.9% in the MII-pH. The distribution of gastric Helicobacter pylori infection was similar between the 2 studied groups.
The mean AET in group A was 1.7% (± 1.1), whereas, in group B, it was 7.8% (± 4.2), with a statistical difference (P < 0.001). The proximal and distal refluxes amounts were also more significant in group B than in group A (P < 0.001). The MNBI was significantly lower in group B than in group A (3702.2 ± 2264.9 Ω vs 5544.3 ± 2558.5 Ω, P < 0.001), as was the PSPW index (33.4 ± 21.4% vs 59.9 ± 16.6%, P < 0.001).
The MI measurements evaluated during UGE are presented in Table 3 and Figure 3.
Figure 3. Box Plot of esophageal mucosal impedance evaluated at different positions above the esophagogastric junction according to the acid exposure time (AET; < 4% vs ≥ 4%).
Table 3 . Descriptive Statistics of Mucosal Impedance Measurements in Ohms Evaluated at Different Positions Above the Esophagogastric Junction, According to Acid Exposure Time (< 4% vs ≥ 4%)
| Position above the EGJ | Group A (AET < 4%) | Group B (AET ≥ 4%) | P-value | |||
|---|---|---|---|---|---|---|
| Mean ± SD | Median (IQR) | Mean ± SD | Median (IQR) | |||
| 2 cm | 4575.0 ± 1407.6 | 4210 (3695-5150) | 2264.4 ± 1099.0 | 2130 (1685-2535) | < 0.001a | |
| 5 cm | 5888.2 ± 2529.4 | 5485 (5115-6265) | 4221.2 ± 2623.7 | 3510 (2495-5265) | 0.001a | |
| 10 cm | 7181.4 ± 2656.2 | 6545 (5540-7860) | 7079.1 ± 3664.9 | 6325 (5365-8515) | 0.600 | |
| 18 cm | 8927.1 ± 2798.8 | 8260 (7220-9765) | 8464.7 ± 3709.3 | 8495 (6205-9870) | 0.600 | |
aA P < 0.05 was considered statistically significant.
EGJ, esophagogastric junction; AET, acid exposure time; IQR, interquartile range; MI, mucosal impedance.
MI was significantly lower in group B than in group A at positions 2 cm (2264.4 ± 1099.0 Ω vs 4575.0 ± 1407.6 Ω [group A], P < 0.001) and 5 cm above the EGJ (4221.2 ± 2623.7 Ω vs 5888.2 ± 2529.4 Ω [group A], P = 0.001). There was no significant difference in the MI between the 2 groups at 10 cm and 18 cm above the EGJ.
A subanalysis of group B was proposed to evaluate participants with AET between 4% and 6% and AET ≥ 6% (Table 4 and Fig. 4). There were no statistically significant differences in the MI measurements of the subgroups at any position. However, a statistically significant difference was noted when comparing AET < 4% × ≥ 6% (P < 0.001) and AET < 4% × 4-6% (P < 0.001) at the 2-cm position and when comparing AET < 4% × ≥ 6% (P = 0.004) at the 5-cm position.
Figure 4. Box Plot of esophageal mucosal impedance evaluated at different positions above the esophagogastric junction according to the acid exposure time (< 4% vs 4-6% vs ≥ 6%).
Table 4 . Mean ± Standard Deviation of Mucosal Impedance Measurements in Ohms Evaluated at Different Positions Above the Esophagogastric Junction, According to Acid Exposure Time (< 4% vs 4-6% vs ≥ 6%)
| Position above the EGJ AET | < 4% n = 28 | 4-6% n = 14 | ≥ 6% n = 18 | P-value |
|---|---|---|---|---|
| 2 cm | 4575.0 ± 1407.6 | 2479.3 ± 1531.3 | 2097.2 ± 584.6 | < 0.001ª |
| 5 cm | 5888.2 ± 2529.4 | 4064.3 ± 1906.3 | 4343.3 ± 3120.5 | 0.003b |
| 10 cm | 7181.4 ± 2656.2 | 6422.9 ± 2755.7 | 7589.4 ± 4249.1 | 0.800 |
| 18 cm | 8927.1 ± 2798.8 | 7817.1 ± 3415.9 | 8968.3 ± 3943.6 | 0.800 |
aPost-hoc: < 4% × ≥ 6%, P < 0.001; < 4 × 4-6%, P < 0.001; ≥ 6 × 4-6%, P > 0.999.
bPost-hoc: < 4% × ≥ 6%, P = 0.004; < 4 × 4-6%, P = 0.800; ≥ 6 × 4-6%, P > 0.999.
AET, acid exposure time; EGJ, esophagogastric junction; MI, mucosal impedance.
In Figure 5, it is possible to verify that MI at 2 and 5 cm above the EGJ presented a significantly higher area under the curve values. The MI value at 2 cm > 2970 Ω resulted in a sensitivity of 96.4% and a specificity of 87.5% to define that the participant had AET < 4% (exclude GERD). In the case of the MI at 5 cm, a value > 4825 Ω resulted in a sensitivity of 82.1% and a specificity of 71.9% to define AET < 4%.
Figure 5. Receiver operating characteristic (ROC) curves comparing mucosal impedance (MI) at 2, 5, 10, and 18 cm for prediction of acid exposure time (AET) ≥ 4%. The area under ROC curve (AUC) were respectively 0.937 (95% CI, 0.865-1.000), 0.748 (95% CI, 0.620-0.877), 0.541 (95% CI, 0.394-0.688), and 18 cm: 0.541 (95% CI, 0.393-0.689).
The correlation between MNBI and MI at 2 cm was significant (Fig. 6), with weak correlation strength (0.39).
Figure 6. Correlation between mean nocturnal baseline impedance (MNBI) measured during the multichannel intraluminal impedance-pH monitoring and the mucosal impedance (MI) at 2 cm measured during endoscopy. This correlation was significant, with weak correlation strength (0.39).
The division of groups was based on AET, the most crucial variable for the definitive diagnosis of GERD.25 The choice of the AET cutoff point for this classification considered the Lyon Consensus (2018), which indicated that an AET < 4% excluded the possibility of GERD diagnosis.25 In fact, a later validation study found that in healthy individuals, the median AET was 1.3%, with a 95th percentile of 4.6%.26
The study group comprised a homogeneous sample of demographic and clinical characteristics. Notably, most participants were female, which is justified by women seeking more healthcare services and complaining more about heartburn.27,28 According to a previous population study conducted in São Paulo, 12.5% of women had heartburn at least twice a week versus 7.7% of men (odds ratio, 1.41; 95% CI, 1.15-1.73).28
Dilated intercellular spaces (DIS) show radial and axial variations in the esophagus of patients with GERD. It was previously demonstrated that, in individuals with non-erosive GERD, there is a greater thickening of the basal cell layer (P = 0.011) and more DIS (P = 0.010) at the 3 o’clock position in the esophagus (equivalent to the right lateral wall) compared with the 9 o’clock position (left side wall). Furthermore, when comparing patients with non-erosive GERD and asymptomatic controls, there was a greater difference in the histological criteria for biopsies performed 1-2 cm above the EGJ than at the EGJ itself. During an endoscopic evaluation, erosion was predominant in the 3-hour quadrant.29 Given this non-homogeneous distribution of the DIS, MI measurements were performed on the right lateral wall at 2, 5, 10, and 18 cm above the EGJ.
The MI values increased from the distal (2 cm) to the proximal (18 cm) in both groups. Since gastroesophageal reflux occurs retrogradely in the esophagus, more tissue damage is expected in the distal esophagus. Therefore, lower MI values are observed closer to the EGJ. In addition, this pattern was identified even in group A (AET < 4.0%), as, although there were no pathological levels of reflux, the simple presence of physiological reflux can lead to greater dilation of intercellular spaces in the distal esophagus when compared with the proximal esophagus.30
There was no difference in MI between groups A and B in the middle and proximal esophagus (10 cm and 18 cm above the EGJ), where there is generally less tissue damage from reflux. In contrast, the MI was significantly lower in group B than in group A in the distal esophagus (2 and 5 cm above the EGJ). Furthermore, a subanalysis of group B revealed that MI values at 2 cm in individuals with AET between 4.0% and 6.0% were more similar to the AET ≥ 6.0% subgroup than to the AET < 4.0% group. Therefore, MI measurements seem more useful for ruling out GERD diagnosis.
An MI cutoff of 2970 Ω at 2 cm determined AET ≥ 4.0% with 96.4% sensitivity and 87.5% specificity. Similarly, the value of 4825 Ω at 5 cm defined AET ≥ 4.0% with 82.1% sensitivity and 71.9% specificity. The first study to evaluate MI measurements using a catheter during UGE included 69 patients in patients with GERD (presence of esophagitis or AET > 5.3%), the median value at 2 cm from the EGJ was 2096 Ω, whereas it was 3607 Ω in the absence of GERD (P = 0.008).18 The cutoff of 3200 Ω had a sensitivity of 88.0% and a specificity of 65.0% for GERD diagnosis.18
Subsequently, the same group of researchers validated this catheter in a larger sample (61 patients with erosive esophagitis, 81 without erosive esophagitis but with AET > 5.3%, 93 without GERD, 18 with achalasia, and 15 with eosinophilic esophagitis). The median MIs at 2 cm from the EGJ were 1427 Ω in erosive GERD, 1829 Ω in non-erosive GERD, 2956 Ω in healthy individuals, 5227 Ω in achalasia, and 1235 Ω in eosinophilic esophagitis. The cutoff of 1465 Ω had a sensitivity of 70.0% and a specificity of 91.0% for GERD diagnosis.31 However, previous studies used a catheter with sensors of different sizes and spacing and from another supplier (Sandhill Scientific, Highlands Ranch, CO, USA), in addition to an upper limit of AET (5.3%) for the definition of groups.18,31
The presence of liquid in the lumen and esophageal contractility can be a problem for the adequate contact of the catheter with the esophageal mucosa, therefore generating variability in the tests. Given this limitation, some authors have developed a new generation of catheters that consist of a balloon that can be inflated up to 2 cm in diameter and has 2-4 longitudinal columns of 10 impedance sensors (1 cm apart). Therefore, direct contact of the sensor with the mucosa is ensured, and simultaneous MI measurements at several points are possible.32,33
In conclusion, direct MI measurement during UGE is a simple and promising method for GERD diagnosis. Individuals with abnormal AET have lower MI measurements than those with normal AET. We believe that this method can replace prolonged reflux monitoring tests in patients who need diagnostic confirmation and, therefore, become a less uncomfortable option for those undergoing UGE. Future studies should continue to seek the development and validation of new devices in different populations to assess mucosal integrity, allowing for greater accuracy in GERD diagnosis.
None.
None.
Rafael B Lages, Ricardo C Barbuti, and Tomas Navarro-Rodriguez designed the study; Rafael B Lages and Luiz H de Souza Fontes collected the data; Rafael B Lages wrote the first draft of the manuscript; and Luiz H de Souza Fontes, Ricardo C Barbuti, and Tomas Navarro-Rodriguez revised the manuscript and supervised the study. All the listed authors actively participated in the study and reviewed and approved the final manuscript.
 |
 |
