Silence of the lungs: comparing measures of slow and noncommunicating lung units from pulmonary function tests with computed tomography

Abstract

Multiple breath washout (MBW) has successfully assessed the silent lung zone particularly in cystic fibrosis lung disease, however, it is limited to the communicating lung only. There are a number of different pulmonary function methods that can assess what is commonly referred to as trapped air, with varying approaches and sensitivity. Twenty-five people with cystic fibrosis (pwCF) underwent MBW, spirometry, body plethysmography, and spirometry-controlled computed tomography (spiro-CT) on the same day. PwCF also performed extensions to MBW that evaluate air trapping, including our novel extension (MBW_ShX), which reveals the extent of underventilated lung units (UVLU). In addition, we used two previously established 5-breath methods that provide a volume of trapped gas (VTG). We used trapped air % from spiro-CT as the gold standard for comparison. UVLU derived from MBW_ShX showed the best agreement with trapped air %, both in terms of correlation (R_S 0.89, P < 0.0001) and sensitivity (79%). Bland-Altman analysis demonstrated a significant underestimation of the VTG by both 5-breath methods (-249 mL [95% CI -10,796; 580 mL] and -203 mL [95% CI -997; 591 mL], respectively). Parameters from both spirometry and body plethysmography were suboptimal at assessing this pathophysiology. The parameters from MBW_ShX demonstrated the best relationship with spiro-CT and had the best sensitivity compared with the other pulmonary function methods assessed in this study. MBW_ShX shows promise to assess and monitor this critical pathophysiological feature, which has been shown to be a driver of lung disease progression in pwCF.NEW & NOTEWORTHY We consider the term "trapped air" either in the use of imaging or pulmonary function testing, something of a misnomer that can lead to an inaccurate assessment of an important physiological feature. Instead, we propose the term underventilated lung units (UVLU). Of the many pulmonary function methods we used in this study, we found that the use of multiple breath washout with short extension (MBW_ShX) to be the best nonimaging method.

Keywords: air trapping; collateral ventilation; cystic fibrosis (CF); pulmonary function testing (PFT); pulmonary physiology.

Graphical abstract

Figure 1.

Correlations between trapped air % from spirometry-controlled computed tomography and pulmonary function parameters including percent predicted (pp) forced expiratory volume in 1 second (FEV₁), forced expiratory flow at 25–75% of the forced vital capacity (FVC) (FEF_25–75%), flow when 75% of FVC has been exhaled (FEF_75%) from spirometry and derivatives of Short extension multiple breath washout (MBW_ShX). Spearman’s correlation coefficient was used. Individual dots are color coded for a patient’s total lung capacity (TLC) which is measured in liters and is consistent on all scatter plots. UVLU, underventilated lung units.

Figure 2.

Spearman’s correlation coefficient of the volume of trapped gas (VTG) over residual volume % from the 5 inspiratory capacity breaths (5IC) method (left) and 5 vital capacity (5VC) method (right) with trapped air % from spirometry-controlled computed tomography. There is a clear under-read of the of VTG/residual volume (RV)% from both 5-breath methods compared with the trapped air %. Straight black line is a linear regression line, with dashed lines either side the 95% confidence intervals.

Figure 3.

We used the linear regression slope values from Fig. 2, to update the fundamental assumption of 78% N₂ (16) concentration to find a better agreement between the 5-breath methods and trapped air %. For the 5 inspiratory capacity breaths (5IC) method, the 1/slope value was 3.823 whereas for the 5 vital capacity (5VC) method, the 1/slope value was 2.979. Therefore, we recalculated the volume of trapped gas (VTG)/residual volume (RV)% for the 5IC method using a N₂ of 20.40% (78/3.823) and 5VC breaths using a N₂ of 26.183% (78/2.979) (Fig. 4). As would be expected, the correlation coefficient for both parameters stayed the same, but now the agreement on the extent of trapped air % was almost perfect with a 1/slope 1.00 for both parameters.

Figure 4.

As seen in Fig. 3, a fixed correction factor applied to the assumption of N₂ concentration of “trapped air” spaces still did not improve the correlation coefficient or remove large outliers, when comparing volume of trapped gas with trapped air%. Therefore, we used a molarity calculator to model the N₂% for an individual patient that was likely present in the “trapped air” spaces which would generate a perfect agreement with trapped air % from spirometry-controlled computed tomography. Each dot is color coded for the N₂% modeled to generate that perfect agreement. Note the patients in red had a larger volume of trapped gas (VTG) over residual volume % (RV%) than trapped air %, therefore N₂ concentration could not be calculated. The range for N₂% for the 5 inspiratory capacity breaths (5IC) method was 3.5% to 34.8% and for the 5 vital capacity (5VC) was 8.6% to 55.3%. Without having the accompanying trapped air% values calculating an accurate VTG/RV% is not possible.

Figure 5.

Axial computed tomography (CT) with inspiratory image left (A) and expiratory image right (B), subsegmental posterior-basal left lower lobe. The airway pointed out with the red arrow is patent, but ectatic on the inspiratory image, but collapses on the forced expiratory image. The adjacent low attenuation (inside red dashed line) is generally interpreted as “small airways disease” or “trapped air,” but the nonphysiologic larger airways collapse could contribute to this appearance. This participant had a percent predicted forced expiratory volume in 1 second (ppFEV₁) of 83% and Z-score of −1.36. The location of the collapsed airway is at the 6th generation, which is still within the proximal airways.