Clinical Roles of Lung Volumes Detected by Body Plethysmography and Helium Dilution in Asthmatic Patients: A Correlation and Diagnosis Analysis

Roles of lung volumes in asthma remain controversial. We aimed to evaluate the efficacy of lung volumes in differentiating asthma severity levels. Consecutive outpatients with chronic persistent asthma were enrolled, and body plethysmography (BP) and helium dilution (HD) were performed simultaneously to extract RV%pred, TLC%pred, and RV/TLC. Significant negative correlations were found between FEV1%pred and RV%pred (r = −0.557, P < 0.001), TLC%pred (r = −0.387, P < 0.001), and RV/TLC (r = −0.485, P < 0.001) measured by BP, as well as difference in volumes between these two techniques (ΔRV%pred, ΔTLC%pred and ΔRV/TLC). In mild and moderate asthma, AUC of RV%pred detected by BP and ΔTLC%pred was 0.723 (95%CI 0.571–0.874, P = 0.005) and 0.739 (95%CI 0.607–0.872, P = 0.002) with sensitivity and specificity being 79.41% and 88.24%, and 65.22% and 56.52% at cut-off of 145.40% and 14.23%, respectively. In moderate and severe asthma, AUC of RV%pred detected by BP and ΔTLC%pred was 0.782 (95%CI 0.671–0.893, P < 0.001) and 0.788 (95%CI 0.681–0.894, P < 0.002) with sensitivity and specificity being 77.78% and 97.22%, and 73.53% and 52.94% at cut-off of 179.85% and 20.22%, respectively. In conclusion, lung volumes are reliable complement of FEV1 in identifying asthma severity levels.

of the lung and airways to undergo unequal and asynchronous compression or decompression during panting maneuvers [12][13][14] . Helium dilution (HD) is an alternative method for measuring alveolar volume but may lead to an underestimate because gas contained within the poorly ventilated regions is not included in the helium estimate of lung volume [13][14][15] . In a small study by Woolcock and colleagues, functional residual capacity (FRC) and TLC were found to be significantly higher by plethysmography than those by dilution method, and the differences between these methods were the greatest when the FEV 1 was lowest, and these differences decreased during clinical recovery 16 . Therefore, we hypothesized that the difference in volume between these two methods may provide additional clinical value in identifying individuals with differing asthma severity.
To test this hypothesis, we conducted a prospective correlation and diagnosis analysis in an attempt to further assess the correlation between lung volumes and FEV 1 , and the values of individual lung volumes as well as the corresponding differences between the two methods in distinguishing asthma severity.

Demographics.
A total of 93 patients (48 male and 45 female) were enrolled in our final analysis, of which 23 (24.73%) had mild asthma, 34 (36.6%) had moderate asthma, and 36 (38.7%) had severe asthma. The mean age of patients with mild, moderate, and severe asthma was 52.0, 53.4, and 54.4 years old, respectively; but there was no significant difference (P = 0.777). No difference was observed in the duration of asthma (4.5 ± 4.0 vs. 4.7 ± 2.5 vs. 6.1 ± 2.9 years, P = 0.083) or smoking history (6.7 ± 8.7 vs. 10.5 ± 12.4 vs. 10.9 ± 10.8 pack*year, P = 0.321) between groups (Table 1). Despite a significant difference of gender between different asthmatic severity groups (6 vs. 18 vs. 24, P = 0.010), the between group difference was only significant between patients with mild and severe asthma (P = 0.003) ( Table 1).  Table 1. Baseline characteristics in patients with mild, moderate and severe asthma. ANOVA, analysis of variance; FEV 1 , forced expiratory volume in one second; FEV 1 %pred, predicted percentage of forced expiratory volume in one second; FEV 1 /FVC, ratio of forced expiratory volume in one second to forced vital capacity; MMEF%pred, predicted percentage of maximal mid-expiratory flow; PEF%pred, predicted percentage of peak expiratory flow; RVhe%pred, predicted percentage of residual volume measured by helium dilution method; RVpleth%pred, predicted percentage of residual volume measured by body plethysmography; RV/TLChe, ratio of residual volume to total lung capacity measured by helium dilution method; RV/TLCpleth, ratio of residual volume to total lung capacity measured by body plethysmography; TLChe%pred, predicted percentage of total lung capacity measured by helium dilution method; TLCpleth%pred, predicted percentage of total lung capacity measured by body plethysmography; Δ RV%pred, predicted percentage of difference of residual volume between body plethysmography and helium dilution method; Δ RV/TLC, difference of ration of residual volume to total lung capacity between body plethysmography and helium dilution method; Δ TLC%pred, predicted percentage of difference of total lung capacity between body plethysmography and helium dilution method.  (Fig. 1C). We also found a significant increasing trend of RV%pred (139.0 ± 38.1% vs. 164.2 ± 24.5% vs. 198.1 ± 37.8%, P < 0.001), TLC%pred (108.8 ± 15.6% vs. 120.2 ± 12.2% vs. 127.0 ± 13.3%, P < 0.001), and RV/TLC (45.9 ± 7.5% vs. 47.2 ± 7.2% vs. 56.4 ± 9.1%, P < 0.001) measured by BP, rather than HD, as asthma severity levels ( Table 1). In addition, a similar elevation of differences between BP and HD in predicted percentage of RV (Δ RV%pred), TLC (Δ TLC%pred) and RV/TLC (Δ RV/TLC) was also observed.
A cut-off of 145.4% and 179.9% in RV%pred exhibited a sensitivity and specificity of 79.41% and 77.78%, and 65.22% and 73.53%, respectively, in identifying different asthma severity levels; while a cut-off of 14.2% and 20.2% in Δ TLC%pred was estimated to have a sensitivity and specificity of 88.24% and 97.22%, and 56.52% and 52.94%, respectively, to retrieve moderate and severe asthma from mild and moderate asthma. (Table 3).

Discussion
In our study, we found that lung volumes including RV%pred, TLC%pred and RV/TLC measured by BP were significantly higher than those measured by HD, and that the volumes measured by BP as well as the difference in volume between these techniques were positively correlated with increasing severity of asthma but negatively correlated with FEV 1 %pred. Furthermore, we also identified that RV%pred measured by BP and Δ TLC%pred could reliably distinguish both mild and moderate asthma and moderate and severe asthma with a high AUC and sensitivity. , and RV/TLC (C) measured by BP (black) and HD (white) showed that RV%pred, TLC%pred, and RV/TLC were significantly higher in BP than in HD in all asthmatic severity levels. BP, body plethysmography; HD, helium dilution method; RV%pred, predicted percentage of residual volume; RV/TLC, ratio of residual volume to total lung capacity; TLC%pred, predicted percentage of total lung capacity. † P < 0.001; ‡ P = 0.001.
Scientific RepoRts | 7:40870 | DOI: 10.1038/srep40870 BP measures thoracic gas volumes (TGV) on the basis of Boyle's Law, which states that, under isothermal conditions, the product of gas volume and pressure is constant at any given moment, and results in an equation of TGV = − (Δ V/Δ P) × P A2 × (P A1 /P B ), of which Δ V is the change in volume of the thorax before and after compression or rarefraction of the gas in thorax, Δ P is the change in the alveolar pressure measured at the airway opening under conditions of no flow during the panting maneuver, P A1 and P A2 are the alveolar pressure before and after compression or rarefraction of the gas in thorax under the assumption that pressure measured at the airway opening is representative of alveolar pressure, and P B indicates the barometric pressure 17 . In comparison, HD measures lung volume from communicating regions of the lung only, and the FRC at the time the subject is connected to the spirometry apparatus of a known volume (Vapp) and helium fraction (F He1 ) is calculated from the helium fraction at the time of equilibration (F He2 ) as the following equation: FRC = Vapp × (F He1 − F He2 )/F He2 12,18 . Therefore, lung volumes measured by HD rely on the gas volume exhaled by the patient; and consequently gas contained within the poorly ventilated regions is not incorporated in the helium estimate of lung volume, leading to higher volumes measured by BP compared with HD in patients with gas trapping 14,17,19 . Previous studies have also demonstrated that lung volumes measured by BP may be more sensitive than HD in distinguishing different levels of current asthma status 16,20 . In our study, we also found such a difference in lung volumes between BP and HD, and the difference between the two methods further increased with increasing severity, which we speculate to be a result of an increasing gas trapping as demonstrated by the decreasing FEV 1 %pred in our study.
FEV 1 is commonly used as a gold standard to diagnose chronic airway diseases, including asthma, and to evaluate their severity in accordance with the measurements of bronchodilator reversibility test or bronchial provocation test 3,21,22 . Nevertheless, these classifications were derived from expert opinion and were not validated in clinical studies. An increasing number of clinical observations and studies have demonstrated that FEV 1 is poorly related to symptoms 23 . Bacharier and his colleagues found in a prospective cohort study on 219 asthmatic children that FEV 1 %pred was not significantly different between severity levels of asthma (99.6% vs. 97.2% vs. 101.0% vs. 93.7% in mild intermittent and persistent, and moderate and severe persistent asthma, respectively, P = 0.3) 7 . Mahut and his colleagues divided 180 asthmatic children with documented airflow reversibility into three groups of severe exacerbation, asthma symptoms without severe exacerbation, and asymptomatic asthma, and found that FEV 1 %pred tested before or after bronchodilator did not differ significantly between groups (pre-bronchodilator: 94 ± 15 vs. 96 ± 12 vs. 98 ± 15, P = 0.53; post-bronchodilator: 105 ± 13 vs. 108 ± 10 vs. 107 ± 12, P = 0.59) 24 . In addition, Stănescu and his colleagues found a lung functional pattern with decreased VC and FEV 1 but increased RV and RV/TLC 10 , therefore, the central phenomenon of this pattern is the increase in RV and RV/TLC, which resulted in the concomitant decrease of VC and FEV 1 and the lack of correlation between FEV 1 and asthma severity due to the following potential mechanisms: 1) loss of elastic recoil (extrinsic obstruction), airways closed at the same transpulmonary pressure as in healthy people but at a higher lung volume, which limited further emptying of the lungs 25 ; 2) small airways obliteration (intrinsic obstruction), small airways narrowed at a higher transpulmonary pressure that in healthy people, which limited further expiration 26 .
Meanwhile, studies have also demonstrated that in many patients with asthma, lung volume is increased. Sorkness and his colleagues analyzed the plethysmographic lung function in patients with no asthma, nonsevere asthma, and severe asthma; and they found that RV, TLC, and ratio of RV to TLC (RV/TLC) were significantly higher in patients with severe asthma than in patients without asthma, and that air trapping was a characteristic feature of severe asthma population 27 . Furthermore, Hartley et al. have elucidated that air trapping was significantly increased in patients with asthma, and multiple regression analyses showed that in asthma subgroups with postbronchodilator predicted percentage of forced expiratory volume (FEV 1 %pred) less than 80%, air trapping was a strong predictor of lung function impairment 28 . In addition, a recent study found that higher RV/TLC was a significant determinant of decreased deep inspiration-induced bronchodilation, which was strongly correlated  Table 2. Correlation analysis between lung volumes and FEV 1 %pred. FEV 1 %pred, predicted percentage of forced expiratory volume in one second; RVhe%pred, predicted percentage of residual volume measured by helium dilution method; RVpleth%pred, predicted percentage of residual volume measured by body plethysmography; RV/TLChe, ratio of RV to TLC measured by helium dilution method; RV/TLCpleth, ratio of RV to TLC measured by body plethysmography; TLChe%pred, predicted percentage of total lung capacity measured by helium dilution method; TLCpleth%pred, predicted percentage of total lung capacity measured by body plethysmography; Δ RV%pred, predicted percentage of difference of residual volume between body plethysmography and helium dilution method; Δ RV/TLC, difference of ration of residual volume to total lung capacity between body plethysmography and helium dilution method; Δ TLC%pred, predicted percentage of difference of total lung capacity between body plethysmography and helium dilution method. with lower FEV 1 and airway distensibility 29 . In the present study, we also found that lung volumetric parameters (RV, TLC and RV/TLC) varied significantly in asthmatics with different severity levels and revealed a significant negative correlation between lung volumes and FEV 1 %pred. Hence, it seems reasonable to believe that lung volumetric parameters have potential diagnostic values in distinguishing asthmatic severity. For the first time, to our knowledge, our study specifically evaluated the roles of lung volumetric parameters in differentiating levels of asthma severity. In an attempt to determine whether lung function measures were consistent with levels of asthma severity, Bacharier and his colleagues classified asthma severity via symptom frequency and medication use in 219 asthmatic children 7 . Although FEV 1 %pred did not differ by severity, as we cited above, they found a significant negative correlation between asthma severity and ratio of FEV 1 to forced vital capacity (FEV 1 /FVC) (mild: 86.3 ± 8.5 vs. moderate: 83.0 ± 10.3 vs. severe: 79.8 ± 11.8, P < 0.001). They attributed such a negative correlation to the dysanapsis, an incongruence between the growth of the airways and lung parenchyma, which worsened in asthma with airways smaller than the lung parenchyma and could resulted in higher lung volumes; but they did not examine related parameters such as RV or RV/TLC. In our study, we found that RV%pred  TLCpleth%pred (B), and RV/TLCpleth (C) and FEV 1 %pred showed that RVpleth%pred, TLCpleth%pred, and RV/TLCpleth were negatively correlated with FEV 1 %pred with a R 2 of 0.310, 0.150, and 0.235, respectively. FEV 1 %pred, predicted percentage of forced expiratory volume in one second; RVpleth%pred, predicted percentage of residual volume measured by body plethysmography; RV/ TLCpleth, ratio of residual volume to total lung capacity measured by body plethysmography; TLCpleth%pred, predicted percentage of total lung capacity measured by body plethysmography. , and Δ RV/TLC (C) and FEV 1 %pred showed that Δ RV%pred, Δ TLC%pred, and Δ RV/TLC were negatively correlated with FEV 1 %pred with a R 2 of 0.209, 0.446, and 0.141, respectively. FEV 1 %pred, predicted percentage of forced expiratory volume in one second; Δ RV%pred, predicted percentage of difference of residual volume between body plethysmography and helium dilution method; Δ RV/TLC, difference of ration of residual volume to total lung capacity between body plethysmography and helium dilution method; Δ TLC%pred, predicted percentage of difference of total lung capacity between body plethysmography and helium dilution method. measured by BP and Δ TLC%pred, with a high sensitivity ranging from 77.78% to 97.22% and specificity between 52.94% and 73.53%, could effectively differentiate asthma severity. We also detected that TLC% measured by BP had a sensitivity as high as 82.35% in distinguishing mild and moderate asthma. In discriminating moderate and severe asthma, we found a relatively low sensitivity of 63.89% with the P value being on the borderline, which might be attributed to the limited patient samples. By contrast, RV/TLC measured by BP, Δ RV%pred and Δ RV/ TLC rendered high diagnostic values only in discriminating moderate and severe asthma, and the diagnostic specificity significantly surpassed the sensitivity (RV/TLC: 85.29% and 69.44%; Δ RV%pred: 85.29% and 58.33%), except for an equivalence in Δ RV/TLC (64.71% and 69.44%), which may result from the higher increase of gas trapping in severe asthma than that in mild asthma, as well as a similar increase of RV and TLC in mild asthma but a higher increase of RV than TLC in severe asthma. Nevertheless, a definitive conclusion of the influencing factors could not be drawn because we were unable to conduct spirometry in these patients before the onset of asthma to collect the baseline lung volumes.
Despite the findings, the present study had three limitations, which may need cautious interpretation of our results. For one thing, the number of participants is small, especially those with mild asthma, which may result in an inaccuracy of our findings. Secondly, the severity levels of asthma used in this study are based on outdated guidelines (NAEPP 1997), which lack assessment of beta2-agonist reliever use. The severity classification scheme better reflects what is currently termed asthma control according to the most recent guidelines introduced by the Global Initiative for Asthma (GINA). Thirdly, in addition to a diagnosis of asthma, a large proportion of the patients in this study have significant smoking history across all severity groups and persistent airflow limitation could not be entirely excluded, such that chronic obstructive pulmonary disease (COPD) may be a confounder.  D) in differentiating mild and moderate asthma, and moderate and severe asthma. AUC, area under the curve; CI, confidence interval; ROC curve, receiver operating characteristic curve; RVpleth%pred, predicted percentage of residual volume measured by body plethysmography; RV/TLCpleth, ratio of residual volume to total lung capacity measured by body plethysmography; TLCpleth%pred, predicted percentage of total lung capacity measured by body plethysmography; Δ RV%pred, predicted percentage of difference of residual volume between body plethysmography and helium dilution method; Δ RV/TLC, difference of ration of residual volume to total lung capacity between body plethysmography and helium dilution method; Δ TLC%pred, predicted percentage of difference of total lung capacity between body plethysmography and helium dilution method. Finally, we used FEV 1 to identify asthma severity in our study, which might not entirely uncover the true relationship among FEV 1 , gas trapping and asthma severity.

Conclusions
In the diagnosis of asthma, body plethysmography is the optimal assessing method, and lung volumes assessed by body plethysmography as well as the difference between body plethysmography and helium dilution can differentiate between mild vs. severe, and moderate vs. severe asthma, but not mild vs. moderate asthma. Nevertheless, future investigations are still warranted to further reveal the correlation of lung volumes and asthma functional outcomes such as asthma controls, and to confirm the clinical applicability of lung volumes in discriminating asthma severity in clinical settings. Patients. From January 2014 to April 2015, consecutive outpatients with chronic persistent asthma were enrolled in West China Hospital, Sichuan University. The diagnostic criteria of asthma were recommended by the Global Initiative for Asthma (GINA) 2012 guideline including: 1) recurrent respiratory symptoms with episodic breathlessness, wheeze, cough, and chest tightness, which were often triggered by incidental allergen exposure, seasonal variability and upper respiratory tract infection, and might resolve spontaneously or in response to medication; 2) airflow reversibility or airway responsiveness documented by lung function test, which was an increase in FEV 1 of > 12% and > 200 ml from baseline or a diurnal variation if peak expiratory flow (PEF) of > 20% or a fall in FEV 1 from baseline of ≥ 20% with standard doses of methacholine 22 . We excluded patients with additional obstructive lung diseases, such as COPD, asthma-COPD overlap syndrome (ACOS), and cystic fibrosis.

Methods
Chronic persistence was defined as onset of asthmatic symptoms every week with various frequency or extent, and three severity levels were classified in accordance with the guideline released by National Asthma Education and Prevention Program (NAEPP) on the basis of daytime symptoms, night waking, activity limitation due to asthma, and FEV 1 %pred 21 : mild persistent (daytime symptoms ≥ 1 episode/week but not daily presenting, night waking > 2 episode/month but < 1 episode/week, probable activity limitation, and FEV 1 %pred > 80%), moderate persistent (daily daytime symptoms, night waking ≥ 1 episode/week, activity limitation, and 60% ≤ FEV 1 %pred ≤ 80%) and severe persistent (daily daytime symptoms, frequent night waking, activity limitation, and FEV 1 %pred < 60%).

Lung function test.
To detect lung volumes (RV, TLC), BP and HD were performed in all enrolled patients using a calibrated equipment (JAEGER, Jaeger Corp, Germany), consisting of a mixing fan, carbon dioxide (CO 2 )  Table 3. Cut-off points of RVpleth%pred and ΔRV%pred, TLCpleth%pred and ΔTLC%pred, RV/ TLCpleth and ΔRV/TLC, and the corresponding sensitivity, specificity, PPV, NPV and LR in distinguishing different asthma severity levels. LR, likelihood ratio; NPV, negative predictive value; PPV, positive predictive value; RVpleth%pred, predicted percentage of residual volume measured by body plethysmography; RV/ TLCpleth, ratio of RV to TLC measured by body plethysmography; TLCpleth%pred, predicted percentage of total lung capacity measured by body plethysmography; Δ RV%pred, predicted percentage of difference of residual volume between body plethysmography and helium dilution method; Δ RV/TLC, difference of ration of residual volume to total lung capacity between body plethysmography and helium dilution method; Δ TLC%pred, predicted percentage of difference of total lung capacity between body plethysmography and helium dilution method.