Comparing a new device to measure lung volumes with three traditional methods in individuals with different lung conditions

Introduction

The measurement of lung volumes, essential to identify abnormalities such as lung restriction, can be performed through different techniques including lung function tests and thoracic imaging.¹ Body plethysmography (BP), the best studied method, measures ventilated and non-ventilated intrathoracic gas volume (ITGV) based on Boyle’s law, and requires that the subject be inside a closed cabin during the maneuver.^2,3. However, the device is expensive and large in size making it hardly accessible, especially in the office setting and, in addition, the quality control is complex. For patients, it can be intimidating, uncomfortable and the maneuvers a challenge to perform, especially in the presence of certain conditions such as claustrophobia or advanced lung disease.³

Measuring ventilated lung volume with a tracer gas, such as that estimated during single breath diffusing lung capacity for carbon monoxide single breath (DLCO_sb), is simple and widely used method for evaluation of lung function status, but it tends to underestimate intrathoracic gas volume in the presence of airflow obstruction, maldistribution of ventilation, or lung bullae.⁴

Currently it is also possible to measure pulmonary volumes with imaging methods such as computed tomography (CT). In recent years, various tools have been developed that have improved the quantitative analysis of CT and have been studied for pre-operative assessment, to evaluate therapeutic interventions,^5–9 and to define repeatability.^10,11 However, this method is expensive and involves exposure to radiation; thus, CT has not proven its cost-effectiveness compared with the conventional pulmonary function test (PFT).³

The MiniBox+™ is a new cabin-less compact device, claiming to measure rapidly and accurately absolute lung volumes in subjects with either normal, obstructive or restrictive lung status with minimal cooperation from subjects.^12–14

Our objective is to compare lung volume measurements (LVM) by different methods, such as BP, DLCO_sb, CT and the new cabin-less plethysmography in patients with varied lung volumes and conditions, as well as in healthy subjects.

Methods

The study, approved by the Institutional Ethics Committee (Code C63-18), was conducted at the Department of Research on Smoking and COPD at the Instituto Nacional de Enfermedades Respiratorias (INER) Ismael Cosío Villegas, a referral center for respiratory diseases mainly for uninsured patients in Mexico City, at 2,240 m above sea level. All participants signed an informed consent document.

Between October 2018 and May 2019 we prospectively identified and recruited adult patients who attended the clinic who met study inclusion criteria and healthy subjects who wanted to participate in the research. We considered a total of four groups with conditions often requiring lung volume analysis, and with a wide range of lung volumes: patients with airflow obstruction, patients with lung bullae both with known difficulty for lung volume measurements by inert gas dilution methods, also a group with lung restriction and subjects healthy from the respiratory point of view.

The group with airflow obstruction, defined as post-bronchodilator (Albuterol 400 µg) spirometry with FEV₁/FVC <lower limit of normality (LLN), had all a clinical diagnosis of chronic obstructive pulmonary disease (COPD).

Another group of patients had lung bullae by chest CT scanning, with or without airflow obstruction, and had a separate outpatient clinic and follow up, aimed to identify candidates for bullectomy or invasive procedures.

The group with restrictive lung disease had a post-bronchodilator spirometry with normal FEV₁/FVC, FVC <80% predicted value, and a TLC <80% of predicted, and several clinical diagnosis including interstitial pulmonary disease due to varied causes such as hypersensitivity pneumonitis, connective tissue disease, or with obesity-hypoventilation syndrome.

In the group of healthy subjects all were volunteers with normal spirometry, never smokers (<100 smoked cigarettes in a lifetime), lacking a personal history of pulmonary, cardiac, renal, hepatic, or metabolic diseases, or respiratory symptoms in the last 4 weeks or persistent respiratory symptoms (>3 days/week) in the previous 12 months. All the volunteers were employees of the institute or relatives of the patients who accepted participate.

We excluded pregnant women, patients with acute respiratory disease or exacerbation, and those admitted to hospital within the last 4 weeks.

After assisting at an explanation of the study in detail, the subjects read and signed informed consent, answered the modified Medical Research Council dyspnea scale (mMRC), and the Short Form 12 (SF-12) health-related quality of life instrument. We also documented exacerbations in the past 12 months, use of supplemental oxygen, and the presence of comorbidities. Subsequently, weight and height were measured using a SECA^® digital scale and stadiometer (models mBCA 514 and 274, respectively).

Simple helical CT (Siemens Somatom Definition A 128) was the first study performed. The mediastinal window images were acquired with 110 mA, 120 kV, Pitch 1.4, FOV 362, and reconstructed with 3mm thickness, utilizing the algorithm for mediastinum I26 homogeneous. For the lung window, acquisition was with 110 mA, 120 kV, Pitch 1.4, FOV 362, and reconstructed with 1 mm thickness, with a lung algorithm I70f very defined. We explained in detail to the patient how to perform a TLC maneuver and how to practice it before conducting the study; then the patient was instructed by means of a simple video. Subsequently, Osirix-Lite software (V10.10) was used to estimate lung volume, through semi-automatic segmentation of pulmonary fields, with density thresholds of −1,024 to −200 Hownsfield units. No correction was made for any study, and volume includes anatomic dead space. CT was performed in all patients except for healthy subjects.

The PFT were performed in the morning within a period of up to 3 h, at the same laboratory and by the same two technicians. None of the subjects used bronchodilators 24 h prior to performing the tests. Before testing, the maneuvers were explained and demonstrated. All tests were performed according to ATS/ERS standards.^3,4,15,16 Flow calibration was verified daily with a certified 3-L syringe, flow linearity was tested weekly at three different flow rates, considering as acceptable a variation of ±2.5%. Ambient conditions (barometric pressure, temperature and humidity) were updated automatically and cabin’s pressure transducer calibration was performed daily. The DLCO_sb was performed with the following concentration of gases: Helium 10.4 cmol/mol; CO 0.313 cmol/mol, and O₂ 21.69 cmol/mol, balanced with nitrogen.

The spirometry (Easy One Pro Lab, Ndd®, Zurich, Switzerland, and PulmoOne Advance Medical Devices, Ltd., Ra’anana, Israel) was carried out with a nasal clip and with the subject seated. Individuals took a quick full inspiration followed by an explosive forced expiration. At least three technically acceptable maneuvers were performed, with two best FVC and FEV₁ within 150 mL to fulfill repeatability criteria. BP (Masterscreen Body, Lab Manager, CareFusion, Germany) was performed with the subject seated inside the closed cabin, using a nasal clip and holding hands against cheeks. The subject then began to take tidal breaths until stabilization. When told, the subject performed a panting maneuver against a closed shutter, to measure ITGV (at functional residual capacity) with a frequency of >0.5-<1.5 Hz, followed by an inspiratory capacity maneuver, and finally, slow vital capacity (SVC). At least three technically acceptable maneuvers were performed, with an ITGV variance of <5% and vital capacity (VC) difference of <0.15L to ensure test repeatability.

Measurement of TLC by cabin-less plethismography (PulmoOne Advance Medical Devices, Ltd. Ra’anana, Israel) was carried out with a nasal clip and hands held against cheeks; then, after reaching stable tidal breathing, the patient performed an inspiratory capacity maneuver and finally, a slow vital capacity maneuver. During tidal breathing, in the inspiratory phase, several transient airway occlusions occur in which airflow to and from the reservoir and pressure, are measured. These measurements are utilized for estimation of total lung capacity by regression equations obtained from a group of healthy individuals as well from patients with several diseases.¹⁷ A maneuver was considered acceptable when tidal breathing was stable and at least six occlusions took place. Good repeatability criteria was considered as at least three technically satisfactory maneuvers with an ITGV of <5% variance, a VC difference of <0.15 L between the two highest, and a TLC of <0.20L between the two highest.

The DLCO_sb (Easy One Pro Lab (Ndd®, Zurich, Switzerland) was performed with the subject seated with a nasal clip. The participant performed 4-6 tidal breaths until stabilization; then the participant was asked to expire maximally, followed by maximal quick inspiration (of the test gas) with 10-s breath-holding and then a full exhalation. The time between procedures was at least 4 min. At least two technically acceptable maneuvers of A or B grade quality were completed, in order to reach two DLCO_sb within 2 mL/min/mmHg of each other for repeatability criteria. Reference values used for this study were NHANES III for spirometry¹⁸, Quanjer’s for plethysmography¹⁹ and Vazquez’s for DLCO ²⁰

Results

A total of 69 patients were recruited, but seven were eliminated, five because of poor quality criteria in pulmonary function tests and two because airflow obstruction disappeared in the post-bronchodilator test. Sixty-two subjects were analyzed, including 21 healthy subjects, 15 with lung bullae, 14 with obstructive lung disease, and 12 with restrictive lung disease.

The general characteristics of the population are presented in Table 1. Women predominated in the healthy and restrictive lung groups. The healthy group was the youngest, and the obstructive lung group, the oldest subjects. Body mass index (BMI) in the restrictive lung group was highest among the groups (27 kg/m²).

Restrictive lung group had a TLC expressed as percentage of predicted (TLC%p) of 61(57-66) measured by BP; patients with obstructive lung disease had a post bronchodilator FEV₁%p of 54 (36-78). The obstructive group had the highest number of exacerbations (71%) and highest smoking intensity (40 pack-years). Home oxygen therapy was more common in the restrictive lung group.

TLC measured by BP ranged between 3.8 L and 6.6 L in the participants, and the ITGV, between 2.2 L and 4.3 L. We analyzed the agreement of TLC measured by BP (our gold standard) and the other methods in all groups as presented in Table 2, including the CCC, the intercept (alpha α) and slope (beta β) of a linear regression. In a perfect concordance, CCC would be 1, with α=0 and β=1. Best agreement was observed between BP and CT (CCC=0.95, α −0.18, β 0.97, and mean difference, −0.35L). Agreement between BP and DLCO_sb was significantly lower with CCC=0.65 and a large average difference of −1.05L, while BP vs. Minibox+™ was moderate (CCC=0.91, with a mean difference of −0.19 L).

In Table 3, we observe the agreement of TLC measured by BP and by the other methods studied separated by groups. Concordance in restrictive subjects between BP and cabin-less plethysmography was highest comparing with the other groups (CCC=0.90), following by healthy and bullae lung groups (CCC=0.87 and CCC=0.85, respectively). Concordance with DLCO_sb in the obstructive group and in patients with bullae were the lowest (CCC=0.18 and 0.47, respectively); also for healthy and restrictive groups the concordance was < 0.90. Volumes measured by CT scan had highest agreement with BP for lung bullae and obstructive groups (CCC 0.95 and CCC 0.76, respectively).

Agreement is shown graphically in Bland-Altman plots (Figure 1). The method with highest agreement compared with BP was chest CT, then with cabin-less plethysmography following close behind, and DLCO_sb with higher discrepancy, especially in patients with obstructive and lung bullae groups.

• Download figure
• Open in new tab

Figure 1.

Bland-Altman plots shows Panel A) TLCbody vs. TLCmini; Panel B) TLCbody vs. TLCdlco; Panel C) TLCbody vs. TLCct. Subjects groups: · healthy, X lung bullae, ▪ obstructive disease, Δ restrictive disease.

The agreement of VC, ITGV, and RV in all subjects between cabin-less plethismography and BP was 0.85, 0.81, and 0.76, respectively. We calculated the intratest-CV and compared this among devices (except by CT scanning, performed only once). For TLC, the CV of cabin-less plethismography was the lowest with 2.7%, followed by BP (3.3%) and DLCO_sb (4.6%).

Time spent in performing each method had a mean of 12.5 min (SD 3.4) for cabin-less plethismography, 19.2 (SD 9.6) for BP, and 12.6 (SD 6.3) for DLCO_sb. Participants were asked about the difficulty they encountered in to performing each test, with five levels (1-very hard to 5-very easy); 76% considered cabin-less plethismography a test that can be carried out between easily and very easily; measuring lung volumes by BP was considered easy and very easy by 35% of the participants.

Discussion

We compared several methods of lung volume measurement (LVM) in different groups of participants with a wide variety of lung volumes and health conditions, including a new method: Minibox™. ATS/ERS guidelines describe several methods for LVM but for the purposes of this work, we considered BP the gold standard.³ Our results demonstrated that the best agreement observed of TLC was between BP and CT scan with a substantial concordance (CCC=0.95) including among individuals in the lung bullae group, as found in previous studies^21–25, whereas agreement was moderate between BP and cabin-less plethysmography (CCC=0.91) in the whole group of individuals tested. We also identified variations in the concordance for the different participant groups and volumes measured.

Although CT could represent an accurate and reliable method to measure lung volumes, requires a breath-hold in maximal inspiration (TLC) during the study that could be challenging to perform in subjects with advanced lung disease. In addition, measurement is costly, depends on the respiratory cycle and the presence of pathologic conditions,²⁶, exposes the patients to radiation, and segmenting the lung fields is difficult. Nevertheless, it may be cost-effective if a CT ordered for other reasons, is utilized to measure lung volumes.

In healthy subjects the concordance between BP and cabin-less plethysmography was 0.87. Agreement of BP and DLCO_sb was 0.83. In this comparison we lacked the analysis with CT scan as was not performed in healthy subjects; in the case of restrictive group cabin-less plethysmography was the best method with a moderate concordance with BP (CCC=0.90), followed by CT scan with a CCC=0.87 (considered poor) and DLCO_sb with a low concordance (CCC=0.62). Other studies which included restrictive patients and healthy subjects had reported a concordance up to 0.81.^21,27

Among subjects with COPD, we found poor agreement (CCC=0.60) between BP and cabin-less plethysmography, with an underestimation of TLC by cabin-less plethysmography. Despite this, in the lung bullae group, which was challenging because of the presence of unventilated air spaces, agreement between BP and cabin-less plethysmography was better (CCC=0.85), demonstrating a reasonable measurement of intrathoracic gas, but one that is affected by airflow obstruction. We must also consider the reported systematic overestimation of lung volumes by plethysmography in the presence of severe airflow obstruction due to poor equilibration between alveolar and mouth pressure during the panting maneuver, with the rate of overestimation proportional to the FEV₁. However, in nearly all studies, this overestimation was corrected by performing this maneuver within a panting frequency in a range between >0.5 - <1Hz.^28–32

The agreement between BP and DLCO_sb was poor in the presence of lung bullae and airflow obstruction (CCC=0.47 and CCC=0.18, respectively), consistent with other studies, finding that gas dilution method by a single breath, only measures correctly well ventilated lung volume.^{19,27,33–37}

In a previous study including measurement of lung volumes with different methods³⁴, a systematic overestimation of TLC with BP was observed, when compared with DLCO rebreathing (DLCO_rb) (Δ0.63L) and CT (Δ1.08), despite adherence to proper technique. Likely the main reason for these discrepancies may be explained by the use of rebreathing DLCO instead of DLCO_sb done in a single breath, and the exclusion of the anatomical death space in CT^38,39. Also, in our study the measuring of TLC with CT includes the anatomical dead space.

We found that cabin-less plethysmography had the lowest intra-test CV in TLC (2.7%), the shortest time for the test and was perceived as easier to perform than BP and DLCO_sb regardless of the lung status. Both Cohen et al, and Fredberg, et al ^12–14 reported that BP and cabin-less method measured similar TLC in obstructive patients, but in our study, even though the cabin-less plethysmography performed well in healthy and restrictive subjects, in subjects with obstruction and pulmonary bullae both methods had a poor agreement. The difference found between the BP and cabin-less plethysmography (−0.19+0.75 L) or between BP and CT scan (−0.35+0.55) have to be taken into account for interpretation. Do not consider the different methods to measure lung volume as interchangeable.

Our study had several limitations. First, we only performed the LVM once (including several maneuvers except for CT scanning, only one study); therefore, we could not analyze the inter-test variability and evaluate its usefulness as a longitudinal assessment.^10,11 Second, the group of healthy subjects was younger compared with the other groups and lacked CT measurement. Third, we performed the CT measurement without a spirometric control, as did Tantucci and colleagues²¹, the best way to ensure that individuals were tested at TLC, but CT is performed reasonably well at TLC under routine conditions, and spirometric control of TLC is not available in the majority of centers. Last, our studied population was small, especially when the data is analyzed by group.

Conclusions

The cabin-less plethysmography proved to be an easier and faster method than any other presented here. Lung volumes measured by BP had a substantial concordance with those measured by CT scanning, and moderate with the Minibox. A poor concordance was seen when lung volumes were measured by DLCO_sb in individuals with bullae or airflow obstruction.