Usefulness of the Exhaled Breath Temperature Plateau in Asthma Patients Free

In recent years, there has been increasing interest in developing noninvasive methods for measuring inflammation in asthma, such as analysis of induced sputum cells and measurement of inflammatory markers in the exhaled breath condensate and in the fraction of exhaled nitric oxide (FeNO). Inflammation is a protective response characterized by several classic signs, one of which is heat. However, temperature has been little used in the exploration of inflammatory respiratory diseases.

Exhaled breath temperature (EBT) has been proposed as a noninvasive marker of airway inflammation in asthma [1]. The basis for this proposal is that inflammation of the bronchial mucosa increases local vascularization, consequently increasing bronchial temperature. This increase could be measured in exhaled air [2,3,4,5]. EBT has also been related to markers of inflammation such as FeNO levels, eosinophil counts, forced expiratory volume in 1 s (FEV₁), and bronchial challenge tests [1,6,7,8,9], but results to date have been contradictory. Furthermore, most studies have been performed in small, pediatric populations with varying equipment and measurements.

The use of EBT as a marker of inflammation in asthma has been explored in different ways. Paredi et al. [1] calculated the exhaled air temperature increase (Δe°T) by measuring the time it took the temperature to reach 63% of the final reading in a slow and controlled expiratory maneuver. Piacentini et al. [4] measured the EBT with a high-performance temperature indicator connected to a thermocouple during a controlled expiratory maneuver. They reported the peak expiratory temperature and the plateaus valued at the end of the expiration. Popov et al. [6] designed and marketed a simple, potentially individual device to measure the EBT plateau (EBTp). They took measurements at 1-min intervals until no further increments were noted.

The present work was designed to analyze the usefulness of the EBTp [6] for asthma patients. Our aim was to correlate EBTp measurements with asthma control, asthma severity, bronchial obstruction, and bronchial inflammation.

We performed an observational cross-sectional study of data collected from January 2012 to July 2013.

The study design complied with the principles of the Declaration of Helsinki (18th World Medical Assembly, 1964) and was approved by the Clinical Research Ethics Committee at the Hospital de la Santa Creu i Sant Pau in Barcelona (COD 26/2012). The patients gave written informed consent prior to participation in the study. No data were recorded that could identify individual patients; their identities were known only by their physicians and were not disclosed to any third parties. The ClinicalTrials.gov identifier is NCT02064686.

The study group consisted of 69 patients on maintenance treatment for asthma. All were nonactive smokers aged ≥18 years. According to the Global Initiative for Asthma (GINA) criteria [10], we defined asthma as a history of variable respiratory symptoms and evidence of variable expiratory airflow limitation. All patients had a positive bronchodilator test or daily peak expiratory flow variability >20%, or a positive methacholine challenge test documented in their case history. High-dose inhaled corticosteroids (ICS) were defined according to the GINA 2014 criteria [10], i.e. when the patients had treatment with budesonide or equivalent ICS at ≥800 µg/day.

Patients were enrolled from an outpatient asthma unit at a tertiary referral hospital. Exclusion criteria were respiratory tract infection or asthma exacerbation in the 30 days prior to inclusion, treatment with oral steroids or immunomodulator drugs, programmed hospitalization during the study, and a cognitive disorder affecting comprehension of the study.

After signing the informed consent form, the patients answered the Asthma Control Test (ACT) questionnaire [11] and underwent the following examinations on the same day: FeNO, spirometry, induced sputum, extraction of 10 ml of peripheral venous blood for later analysis, and the skin prick test with standardized allergenic extracts (based on the modified test of Pepys) [12]. Total IgE was measured by EliA (ImmunoCAP, Phadia 250). Induced sputum was analyzed within 2 h. The body mass index (BMI) was recorded as well.

The EBTp was measured using the X-halo Breath Thermometer (Delmedica, Singapore) according to the method validated by Popov et al. [6]. The patients were instructed to breathe tidally, inhaling freely through the nose and exhaling into the device. The maneuver was continued until the software of the instrument indicated that the measured value was stable. The EBTp was defined as 20 s with changes <0.01°C. For all patients, the EBT was measured in the morning (9.00-10.00 h) after a minimum of 2 h of fasting. The environmental temperature was controlled to maintain a stable condition of 24 ± 1°C and a relative humidity of 40-50% [7].

To objectively assess the level of asthma control, we used the ACT. This is a simple 5-question questionnaire that measures the degree of control that asthmatic patients assign to their illness over the previous 4 weeks. The answers to each question are assigned 1-5 points, resulting in a total score ranging from 5 (worst possible control) to 25 (best possible control). A score >20 corresponds to ‘controlled asthma' and <19 points to ‘not well-controlled asthma' [11].

We classified asthma severity retrospectively, based on the level of treatment required to control symptoms and exacerbation using the GINA classification. Mild asthma was asthma that could be controlled with step 1 or 2 treatment, while severe asthma required step 4 or 5 treatment [10].

Spirometry was performed using a Datospir 500 (Sibelmed SA, Barcelona, Spain) following the recommendations of the European Respiratory Society [13] and the Spanish Respiratory Society (SEPAR) [14]. Patients were classified into two groups: those with an FEV₁ level ≥80% and those with an FEV₁ level <80%.

Fraction of Exhaled Nitric Oxide

FeNO measurements were performed using an NO Vario Analyzer device (FILT GmbH, Berlin, Germany); the patients performed a maximal inspiration followed by a controlled expiratory maneuver at a flow of 50 ml/s [15]. FeNO concentration was expressed in parts per billion (ppb), and the reference values were taken from the literature [16].

Inflammatory Asthma Phenotypes

Cell counts in induced sputum were determined using the method of Pizzichini et al. [17]. Briefly, mucus plugs were manually selected and weighed, incubated (15 min at room temperature) in 4 times the weight (in ml) of the selected plug (in mg) in 0.1% dithiothreitol (Calbiochem, San Diego, Calif., USA), washed with 4 times the plug weight (in ml) in Dulbecco's PBS and gravity filtered through a 41-μm-pore nylon net filter (Millipore, Billerica, Mass., USA). After dithiothreitol homogenization, each specimen was aliquoted into two parts of equal volume. Total cell counts were performed using a Neubauer hemocytometer. Visually identifiable squamous epithelial cells were not counted or included in the total cell count. Samples that did not produce adequate sputum cell numbers (<1,000 × 10⁶ cells/g) were excluded. Cell viability was determined by light microscopic assessment using trypan blue exclusion staining. After centrifuging the cell preparation, we obtained a cell pellet and a supernatant. The cell pellet was used for differential cell counts (macrophages, eosinophils, neutrophils, lymphocytes, and bronchial epithelial cells) performed on May-Grünwald-Giemsa-stained preparations. A differential leukocyte analysis of nonsquamous cells (Diff-Quik stained) was performed on a minimum of 400 cells. Differential cell counts are expressed as the percentage of total nonsquamous nucleated cells. Reference values of the cell counts were taken as previously described in the literature [18].

Patients with a neutrophil count >61% were classified as having neutrophilic asthma, those with eosinophil counts >3% were classified as having eosinophilic asthma, and those with neutrophil counts ≤61% and eosinophil counts ≤3% were classified as having paucigranulocytic asthma [17,18,19].

Descriptive baseline values are presented as percentages and frequencies for qualitative data, and means and SD for quantitative data. The comparison between the two groups was performed by analysis of the t test for quantitative or ordinal variables. The comparison between the groups according to the induced sputum inflammatory phenotypes was conducted using analysis of variance for the primary and the other quantitative variables. Results are presented as mean values and SD.

Categorical variables were compared with contingency tables using the χ² test. Results are given as numbers of cases and percentages. Correlation analysis was performed by Pearson's r. In all cases, the level of statistical significance was 5% (α = 0.05). Statistical analysis was performed using SPSS software (version 18.0) for Windows (SPSS Inc., Chicago, Ill., USA).

Table 1 shows the demographic and asthma characteristics of the 69 patients included in the study.

The patients were classified into two groups according to the level of asthma control. Asthma was controlled in 43 patients (62.31%; ACT ≥20) and not well controlled in 26 (37.69%). EBTp levels showed no significant differences between groups: 34.29°C (SD 0.61) in the controlled versus 34.02°C (0.72) in the not-well-controlled asthma group (p = 0.104). No significant correlations were observed between EBTp and ACT levels (R = 0.179, p = 0.142).

The patients were classified into two groups according to disease severity: 24 (34.78%) had intermittent or mild and 45 (65.22%) moderate or severe persistent asthma. The EBTp was 34.33°C (0.28) in the intermittent-mild and 34.11°C (0.78) in the moderate-severe group. No differences were observed between groups (p = 0.200).

According to the FEV₁ values (if FEV₁ was <80% of the predicted normal value), 28 patients (40.58%) had a bronchial obstruction and 41 (59.42%) did not. The EBTp was 34.30°C (0.43) in the first and 34.02°C (0.88) in the second group. These differences were not statistically significant (p = 0.128). We found no significant correlation between EBTp measurements and FEV₁% (R = 0.165, p = 0.176) in any of the samples analyzed.

Fraction of Exhaled Nitric Oxide

Eighteen patients (26.08%) had a high FeNO (≥50 ppb), and 51 (73.92%) had normal FeNO values (<50 ppb). The EBTp was 34.26°C (0.57) in the first and 34.17°C (0.69) in the second group. The difference was not significant (p = 0.604). No significant correlation was found between EBTp measurements and FeNO values (R = -0.07, p = 0.568).

Inflammatory Asthma Phenotypes

Twenty-three patients (33.33%) had a paucigranulocytic phenotype, 35 (50.72%) had an eosinophilic phenotype, and 11 (15.95%) had a neutrophilic asthma phenotype. The EBTp observed per group was 34.22°C (0.78) for the paucigranulocytic asthma patients, 34.12°C (0.53) for the eosinophilic asthma patients, and 34.34°C (0.78) for the neutrophilic asthma patients. No significant differences in EBTp measurements were found between the inflammatory asthma phenotype groups (p = 0.607). However, a significant negative correlation was shown between eosinophil counts in the induced sputum and EBTp measurements (R = -0.348, p = 0.003) (fig. 1).

There was no significant correlation between the EBTp and age (R = -0.052, p = 0.670), BMI (R = 0.091, p = 0.459), ambient temperature (R = -0.075, p = 0.586), or body temperature (R = 0.156, p = 0.262). There was a significant difference in EBTp according to gender: the EBTp was 34.07°C (0.74) in women and 34.38°C (0.46) in men (p = 0.038).

Figure 2 compares EBTp values between subgroups: ACT <20 versus ≥20; FEV₁ below versus above the median value; FeNO ≥50 versus <50 ppb; ICS high versus middle/low dose; age below versus above the median value, and men versus women.

In this study in adult patients with asthma, we did not find any relationship between EBTp and asthma control, disease severity, airway obstruction, or bronchial inflammation. The rationale for EBT as a new, simple, and noninvasive biomarker of airway inflammation is based on the inflammatory nature of asthma, characterized by increased vascularization that may lead to increased heat loss in exhaled air [1,4]. EBT measurements have been performed in various studies using different nonstandardized methods, but their clinical usefulness is still uncertain. In one of the first published studies that compared a group of asthma patients (n = 18) with healthy volunteers (n = 16), Paredi et al. [1] found a faster Δe°T in patients with asthma, but no differences in EBTp. They also failed to find a correlation between EBTp and FeNO. In another study, using a thermometer connected to a high-performance thermocouple, Piacentini et al. [4] found a significant correlation between plateau temperature measured at the end of expiration and FeNO, but no correlation between peak expiratory temperature and FeNO.

Results regarding the use of an X-halo Breath Thermometer to measure EBTp to control asthma vary among studies. Popov et al. [6] found that the EBT was increased in 14 not-well-controlled asthmatics, but that it decreased after anti-inflammatory treatment had been started. In a group of 100 patients, García et al. [20] observed higher levels of EBTp in not-well-controlled asthmatics than in well-controlled asthma patients and healthy subjects. However, using the same equipment, we could not confirm these findings, observing no correlation between EBTp and levels of asthma control. The discrepancies between these three studies could be explained by the lower proportion of not-well-controlled asthma patients in our study (37.69%).

In addition, we did not find any association between EBTp level and severity of airflow obstruction. Previous studies have shown discrepant results regarding this association. Peroni et al. [21] found no correlation with EBTp in 50 children with asthma treated with ICS after exercise-induced bronchoconstriction, while in asthmatic children not treated with ICS they found an increase in EBTp and a decrease in FEV₁ after the same exercise. Svensson et al. [22] did not find changes in EBTp in asthma patients before or after two bronchoconstriction tests (eucapnic voluntary hyperventilation and methacholine challenge). These differences between studies could be explained by the anti-inflammatory effect of ICS. Most of the patients included in our study were on this medication.

Exhaled temperature has previously been related to airway inflammation and bronchial blood flow [3,5]. However, in a widely characterized sample of asthma patients, our study failed to demonstrate a correlation between EBTp and other markers of bronchial inflammation. This discrepancy has been observed in other studies. In a group of 19 asthmatic patients, for example, Paredi et al. [3] determined the Δe°T and found no correlation between FeNO levels and EBT - nor did Svensson et al. [23] observe any association between FeNO and EBTp in a group of 20 patients with mild asthma. However, unlike Paredi et al. [3], Piacentini et al. [4] found EBTp was associated with FeNO (41 children with allergic asthma) and with the percentage of eosinophils in induced sputum (16 children). Nevertheless, this association was lost when the Δe°T method was used. In the induced sputum from the 69 participants in our study, we observed a negative correlation between EBT and the proportion of eosinophils in sputum. One explanation for this discrepancy between studies could be that we studied three inflammatory phenotypes (paucigranulocytic, eosinophilic, and neutrophilic phenotypes), while the other authors investigated only eosinophilic phenotypes. Other explanations for the different findings could be the variability of the methods used to measure EBT, the lack of standardization of these methods, the differences in asthma populations, the effect of treatments, and the possibility that surrogate markers reflect different parts of the inflammatory cascade.

Higher EBT values have been associated with potential confounders such as male gender, younger age, physical activity, higher BMI, and higher room air temperature [7,23,24,25]. We found no correlation between EBTp and age, BMI, room temperature, or humidity, but men showed higher EBTp values than women. The lack of a correlation of EBTp with age and BMI was probably due to the homogeneity of the group. The lack of a correlation of EBTp with room temperature and humidity was most likely because these variables were always stable in our laboratory (24 ± 1°C, relative humidity 40-50%).

Our study has several limitations. First, we did not use other EBT methods such as Δe°T or peak EBT. Second, unlike some studies, we did not include children with asthma. Our study has several strengths, however. It is a comprehensive study in a large sample using multiple techniques to assess various aspects of the disease, while other studies to date have focused on specific or partial aspects. Furthermore, all clinical, functional, and inflammatory measurements were made on the same day for all patients.

We found no correlation between EBTp and asthma control, severity of disease, bronchial obstruction, or bronchial inflammation. According to our results, EBTp does not seem to provide useful clinical information as a single-point measurement. Further research regarding the potential use of EBTp measurements in asthma management should incorporate large series, longitudinal designs, and measurement standardization.

All authors made substantial contributions to the study and critically revised and approved the manuscript. Specifically, A.C.L. and V.P. generated the hypothesis, designed the study, enrolled patients, and wrote the first draft of the manuscript; A.C.L. collected clinical data on the patients, built the database, and performed the statistical analysis; J.G. performed the FeNO and EBT measurements; A.T. contributed to the data interpretation and writing of the manuscript; E.M. processed the sputum and performed the inflammatory cell counts, and M.T. and A.B. conducted sputum induction and spirometry.

We thank Carolyn Newey for linguistic and editorial assistance. This work was supported by: the Catalan Foundation of Pneumology (FUCAP), Barcelona; the Spanish Society of Pneumology and Thoracic Surgery (SEPAR), and the Social Research Private Foundation (MIA) of the Hospital de la Santa Creu i Sant Pau, Barcelona.

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.

This paper is part of the doctoral thesis of A.C.L.