- Email: [email protected]
Small airways disease in mild and moderate chronic obstructive pulmonary disease: a cross-sectional study
Articles
Small airways disease in mild and moderate chronic obstructive pulmonary disease: a cross-sectional study Hyun-Kyoung Koo*, Dragoş M Vasilescu*, Steven Booth*, Aileen Hsieh, Orestis L Katsamenis, Nick Fishbane, W Mark Elliott, Miranda Kirby, Peter Lackie, Ian Sinclair, Jane A Warner, Joel D Cooper, Harvey O Coxson, Peter D Paré, James C Hogg, Tillie-Louise Hackett
Summary
Background The concept that small conducting airways less than 2 mm in diameter become the major site of airflow obstruction in chronic obstructive pulmonary disease (COPD) is well established in the scientific literature, and the last generation of small conducting airways, terminal bronchioles, are known to be destroyed in patients with very severe COPD. We aimed to determine whether destruction of the terminal and transitional bronchioles (the first generation of respiratory airways) occurs before, or in parallel with, emphysematous tissue destruction. Methods In this cross-sectional analysis, we applied a novel multiresolution CT imaging protocol to tissue samples obtained using a systematic uniform sampling method to obtain representative unbiased samples of the whole lung or lobe of smokers with normal lung function (controls) and patients with mild COPD (Global Initiative for Chronic Obstructive Lung Disease [GOLD] stage 1), moderate COPD (GOLD 2), or very severe COPD (GOLD 4). Patients with GOLD 1 or GOLD 2 COPD and smokers with normal lung function had undergone lobectomy and pneumonectomy, and patients with GOLD 4 COPD had undergone lung transplantation. Lung tissue samples were used for stereological assessment of the number and morphology of terminal and transitional bronchioles, airspace size (mean linear intercept), and alveolar surface area. Findings Of the 34 patients included in this study, ten were controls (smokers with normal lung function), ten patients had GOLD 1 COPD, eight had GOLD 2 COPD, and six had GOLD 4 COPD with centrilobular emphysema. The 34 lung specimens provided 262 lung samples. Compared with control smokers, the number of terminal bronchioles decreased by 40% in patients with GOLD 1 COPD (p=0·014) and 43% in patients with GOLD 2 COPD (p=0·036), the number of transitional bronchioles decreased by 56% in patients with GOLD 1 COPD (p=0·0001) and 59% in patients with GOLD 2 COPD (p=0·0001), and alveolar surface area decreased by 33% in patients with GOLD 1 COPD (p=0·019) and 45% in patients with GOLD 2 COPD (p=0·0021). These pathological changes were found to correlate with lung function decline. We also showed significant loss of terminal and transitional bronchioles in lung samples from patients with GOLD 1 or GOLD 2 COPD that had a normal alveolar surface area. Remaining small airways were found to have thickened walls and narrowed lumens, which become more obstructed with increasing COPD GOLD stage. Interpretation These data show that small airways disease is a pathological feature in mild and moderate COPD. Importantly, this study emphasises that early intervention for disease modification might be required by patients with mild or moderate COPD. Funding Canadian Institutes of Health Research. Copyright © 2018 Elsevier Ltd. All rights reserved.
Introduction Small conducting airways less than 2 mm in internal diameter, which offer less than 10% of the total resistance to airflow in the normal lung,1–3 become the major site of airflow obstruction in chronic obstructive pulmonary disease (COPD).4–6 To develop effective treatments for COPD, an understanding of disease pathogenesis within these small airways is vital. Current CT imaging protocols applied in longitudinal COPD studies, with 1–2 mm inplane spatial resolution and low-dose radiation exposure (1·5 mSv), can only resolve airways larger than 2·5 mm in internal diameter.7,8 A study9 that used ultra-highresolution CT and high-dose radiation exposure (11·2 mSv) has shown that it is possible to visualise small airways that are 0·8 mm in diameter. However, to date,
only micro-CT has the spatial resolution required to resolve the alveolar structure and reliably identify and characterise the smallest generation of conducting airways, the terminal bronchioles, which have an average lumen diameter of 424 µm.10 By use of micro-CT on fixed and dried lung samples, McDonough and colleagues10 showed that compared with controls (n=4) the number of terminal bronchioles was reduced by 72% in seven patients with very severe COPD (Global Initiative for Chronic Obstructive Lung Disease [GOLD] stage 4)11 with panlobular emphysema and by 89% in four patients with GOLD 4 COPD with centrilobular emphysema. Additionally, they showed that the number of terminal bronchioles was reduced even in tissue samples that had no detectable emphysema as measured by mean linear
www.thelancet.com/respiratory Published online July 4, 2018 http://dx.doi.org/10.1016/S2213-2600(18)30196-6
Lancet Respir Med 2018 Published Online July 4, 2018 http://dx.doi.org/10.1016/ S2213-2600(18)30196-6 See Online/Comment http://dx.doi.org/10.1016/ S2213-2600(18)30290-X *Joint first authors Centre for Heart Lung Innovation, University of British Columbia, St Paul’s Hospital, Vancouver, BC, Canada (H-K Koo MD, D M Vasilescu PhD, S Booth MSc, A Hsieh, N Fishbane MSc, W M Elliott PhD, M Kirby PhD, Prof H O Coxson PhD, Prof P D Paré MD, Prof J C Hogg MD, T-L Hackett PhD); Department of Anesthesiology, Pharmacology and Therapeutics (H-K Koo, S Booth, T-L Hackett), Department of Medicine (Prof P D Paré), and Department of Pathology (Prof J C Hogg), University of British Columbia, Vancouver, BC, Canada; µ-VIS X-ray Imaging Centre, Faculty of Engineering and the Environment (O L Katsamenis PhD, Prof I Sinclair PhD); Clinical and Experimental Sciences, Faculty of Medicine (P Lackie PhD, J A Warner PhD), University of Southampton, Southampton, UK; and Division of Thoracic Surgery, University of Pennsylvania, Philadelphia, PA, USA (Prof J D Cooper MD) Correspondence to: Dr Tillie-Louise Hackett, Centre for Heart Lung Innovation, University of British Columbia, St Paul’s Hospital, Vancouver, BC V6Z 1Y6, Canada [email protected]
1
Articles
Research in context Evidence before this study Results from several physiological studies have shown that small conducting airways less than 2 mm in internal diameter are the major site of airflow obstruction in patients who have very severe chronic obstructive pulmonary disease (COPD). On Jan 18, 2018, we searched the scientific literature in PubMed (with no date or language restrictions) for the following search terms “small airways”, “emphysema”, and “COPD”. We found no publications using multiresolution imaging to investigate small airway pathology and its association with emphysema in mild and moderate COPD. Added value of this study This study uses a novel approach that combines stereology, multiresolution CT imaging, and histology that, to our knowledge, for the first time provides direct evidence that substantial loss of small airways (terminal and transitional bronchioles) occurs in patients with mild and moderate COPD, before the detection of emphysema by clinical CT. Furthermore, these new data show that destruction of terminal and transitional bronchioles can occur in regions of the lung with normal alveolar surface area, a robust measure of internal lung structure measured with micro-CT. We also
intercept. These data suggested that terminal bronchiole obliteration might precede emphysematous tissue destruction in COPD.10 The hypothesis that terminal bronchioles are lost before emphysematous destruction cannot be tested in a longitudinal study of disease progression within the same patients. We aimed to determine whether destruction of the terminal and, to our knowledge, for the first time transitional bronchioles (the first generation of respiratory airways)12,13 occurs before, or in parallel with, emphy sematous tissue destruction. Furthermore, we assessed emphysematous tissue destruction using the measure ment of alveolar surface area. This approach is superior to mean linear intercept because, in addition to airspace enlargement, it measures the destruction of alveolar septae, which is an important pathological change in emphysema,14,15 and it is a robust measure of internal lung structure, which is calculated from airspace size and volume fractions of alveolar tissue, and is therefore not affected by inflation status and elastic properties of the lung.16–18 Since lungs resected for lung cancer are invariably fixed before study, we also used a new micro-CT imaging protocol for formalin-fixed, paraffin-embedded lung tissue samples,19 which enabled a matched histological investigation of the remaining small airways.
Methods
Participants We did a cross-sectional analysis of smokers with normal lung function, patients with mild COPD, patients with 2
used histology to show that the remaining small airways have thickened walls and narrowed lumens that become more obstructed with increased COPD severity. These data indicate that small airways disease is a feature of mild and moderate COPD. Implications of all the available evidence Although these results must be considered preliminary because of the small number of cases that have been studied to date, they strongly support the concept that small airways disease is well established by the time the diagnosis of mild or moderate COPD is made. Furthermore, the data suggest that the reason most clinical trials investigating COPD treatments in severe COPD did not show beneficial effects is because they were initiated after a substantial number of terminal and transitional bronchioles were already lost, and that early intervention in patients with mild and moderate COPD might be required for disease modification. We expect that the findings from this research will prompt discussion of guidelines and policies to improve early diagnosis and effective management of patients with mild and moderate COPD.
moderate COPD, and patients with very severe COPD, who donated lungs or lung samples after pneumonectomy or lobectomy surgery. The severity of COPD was graded with the 2011 GOLD staging system, on the basis of postbronchodilator FEV1, in which airflow limitation is indicated as mild for GOLD 1 (FEV1 ≥80% predicted), moderate for GOLD 2 (50% ≤ FEV1 <80% predicted), and very severe for GOLD 4 (FEV1 <30% predicted). Informed consent was obtained from patients treated for small peripheral lung cancers (primary tumours with no metastasis) by lobectomy and pneumonectomy surgery at St Paul’s Hospital (Vancouver, BC, Canada), with approval by the Providence Health Care Research Ethics Board (PHCREB; Vancouver, BC, Canada) and from patients who underwent lung transplantation for GOLD 4 COPD at the University of Pennsylvania Hospital (Philadelphia, PA, USA), with the approval of the University of Pennsylvania Hospital Institutional Review Board. All specimens were stored in a dedicated archival Lung Registry between 1991 and 2006, with approval of the PHCREB (number H13-02173).
CT analysis CT scans were acquired before surgery at suspended full inspiration by coaching the patient to inhale to total lung capacity in the supine position, as previously described.20 The CT image acquisition parameters were 120–140 kV, 80–345 mA, 5–10 mm slice thickness, and 0–10 mm space between slices. Semi-automatic lung segmentation was done on the CT scans with a custom software package,
www.thelancet.com/respiratory Published online July 4, 2018 http://dx.doi.org/10.1016/S2213-2600(18)30196-6
Articles
EmphylxJ21 (University of British Columbia, Vancouver, BC, Canada), to isolate the specific resected lung or lobe and calculate lung volumes. To quantify emphysema on the 10 mm thick slice CT scans, we used a density threshold of –910 Hounsfield units (HU; percentage low-attenuation area [LAA] <–910 HU), since this pathology-validated threshold defined by Muller and colleagues22 has been shown to yield the best correlation with emphysema by use of 10 mm thick slice CT datasets.22
Lung tissue preparation Briefly, resected lobes and whole lungs were inflated, frozen, and sliced into contiguous 2 cm thick slices in the transaxial plane, as previously described.10,23 We excluded the tumour and normal margin and extracted tissue samples measuring 1·5 cm in diameter from the lung slices using a systematic uniform sampling method to obtain representative unbiased samples of the lobe or lung. For this study, we obtained one tissue sample per slice. All samples were fixed in pre-cooled alcohol-based formalin overnight and then infiltrated with low-meltingpoint paraffin by use of a tissue processor (model ASP6025; Leica, Ontario, CA, USA). We measured the diameter, cross-sectional area, and volume of each tissue sample before and after processing to assess tissue shrinkage, which was applied as a correction factor to the quantitative measurements.
Micro-CT imaging The paraffin-embedded tissue samples were scanned with a micro-CT scanner (HMX225; Nikon Metrology, Brighton, MI, USA), with a previously developed imaging protocol19 (50 kV, 180 µA, and reconstructed with an isotropic voxel size of 6·7 µm). Images were processed with ImageJ software, version 1.47q (National Institutes of Health, Bethesda, MD, USA), and data were collected and analysed by three observers masked to patient characteristics. To ensure measurements obtained by micro-CT imaging of formalin-fixed, paraffin-embedded tissue were similar to the method of glutaraldehyde fixation and critical point drying previously used by McDonough and colleagues,10 we used both methods to process adjacent lung tissue samples from the GOLD 4 donor lungs, and we found no difference between the two methods when measuring alveolar surface area (appendix p 4).
Stereology Following the American Thoracic Society and European Respiratory Society guidelines for stereology,24 we applied systematic uniform random (SUR) sampling to obtain ten images per sample for analysis. The total number of images in the micro-CT scan was first divided by ten to generate ten substacks, and a random number generator was then used to pick the random image used in each substack. SUR-sampled images were used for measure ment of mean linear intercept with a line grid24,25 and
alveolar surface area with a point counting grid16,25 in Image-Pro Plus, version 5.1 (Media Cybernetics, Silver Spring, MD, USA). Measurements of mean linear intercept and alveolar surface area obtained by micro-CT were validated by comparison to the gold standard of histology (appendix p 5). Terminal bronchioles were identified within the microCT scan by following consecutive airway branches until transitional bronchioles13 were identified by the first occurrence of individual alveoli along the airway wall (appendix p 3). Identification of transitional bronchioles subsequently enabled the parent airway to be counted as a terminal bronchiole. The numbers of terminal bronchioles per mL of lung (TB/mL) and transitional bronchioles per mL of lung (TrB/mL) were calculated by dividing the number of respective bronchioles by the sample volume corrected for shrinkage. Terminal bronchioles were further classified as nondiseased, thickened, or obstructed. Thickened terminal Control group (smokers with normal lung function; n=10) Lung tissue samples
75
GOLD 1 group (mild COPD; n=10) 81
See Online for appendix
GOLD 2 group (moderate COPD; n=8) 62
GOLD 4 group (very severe COPD; n=6) 44
Sex* Female
6 (60%)
4 (40%)
3 (38%)
1 (17%)
Male
4 (40%)
6 (60%)
5 (63%)
5 (83%)
Age (years)*
62·0 (7·9)
67·4 (7·3)
Height (cm)*
167·9 (8·7)
168·6 (9·9)
62·9 (11·3) 167·4 (6·7)
59·2 (2·1) 170·3 (6·3)
Weight (kg)*
68·4 (14·7)
77·1 (18·1)
73·1 (15·6)
71·5 (10·8)
Smoking history (pack-years)*
34·5 (10·5)
45·5 (25·3)
33·6 (12·7)
37·5 (15·1)
FEV1 (% predicted)†
91·8 (6·4)
88·3 (6·2)
62·1 (9·5)
22·3 (6·7)
FVC after bronchodilator (% predicted)‡
96·7 (5·4)
108·3 (8·5)
88·5 (6·4)
60·3 (19·7)
FEV1/FVC (%)§
74·9 (4·4)
63·5 (4·7)
60·1 (7·6)
29·5 (7·8)
DLCO:VA (mL/min per mm Hg per L)¶
3·85 (0·9)
2·83 (0·7)
2·64 (0·9)
1·71 (0·8)
Total lung volume (L)
4·86 (1·4)
5·37 (1·5)
4·92 (0·9)
7·91 (1·0)||
% LAA <–910 (HU)
1·05 (3·7)
1·84 (7·9)
8·72 (12·4)
67·68 (16·8)**
Right upper lobe
2
3
2
Left upper lobe
2
3
1
Right middle lobe and right lower lobe
1
Surgical resection
··
Left lower lobe
1
1
Right whole lung
2
2
Left whole lung
2
1
·· ·· ··
1
·· ··
·· 3
2 4
Data are n (%) or mean (SD). GOLD=Global Initiative for Chronic Obstructive Lung Disease. COPD=chronic obstructive pulmonary disease. FVC=forced vital capacity. DLCO:VA=diffusing capacity of the lung for carbon monoxide adjusted for alveolar volume. LAA=low-attenuation area. HU=Hounsfield units. *One-way ANOVA with a Tukey’s pairwise comparison showed no significant differences. †Differences in FEV1 were significant (p<0·001) between all groups except between the control and GOLD 1 groups. ‡Differences were significant (p<0·01) between all groups except between the control and GOLD 1 groups, and control and GOLD 2 groups. §Differences were significant (p<0·01) between all groups except between the GOLD 1 and GOLD 2 groups. ¶Differences were significant between the control and GOLD 2 groups (p<0·05), and control and GOLD 4 (p<0·05) groups. ||Increased (p<0·01) compared with all other groups. **Increased compared with the control (p=0·0002) and GOLD 1 (p=0·024) groups.
Table 1: Preoperative patient characteristics
www.thelancet.com/respiratory Published online July 4, 2018 http://dx.doi.org/10.1016/S2213-2600(18)30196-6
3
Articles
using Movat’s pentachrome stain (see appendix p 2 for more details).
A 10
Mean number of terminal bronchioles per mL of lung per person
p=0·0001
Statistical analysis
p=0·036
8
We averaged all samples per patient and presented per case. We assessed outcomes for Gaussian distribution and tested for normality using the Kolmogorov-Smirnov test, D’Agostino-Pearson omnibus normality test, and Shapiro-Wilk normality test. Measurements of airways per mL of lung, diseased airways per mL of lung, and alveolar surface area per mL of tissue were normally distributed, and assessed with a one-way ANOVA with a Tukey’s pairwise comparison. The Shapiro-Wilk nor mality test showed that measurements of mean linear intercept and CT density data were not normally distributed, and therefore a non-parametric Kruskal Wallis with Dunn’s post-hoc test was used. We used GraphPad Software, version 5, for this analysis. We generated kernel density plots with multiple comparisons using the Bergmann Hommel method (scmamp R package). We assessed the number of terminal and transitional bronchioles in cores with normal alveolar surface area using a linear mixed-effect model (nlme package), and we calculated Pearson’s correlations with false discovery rate corrections using the statistical software R3.3.1. Data are expressed as the mean (SD) or median (IQR) for non-parametric data. p values of less than 0·05 were considered significant.
p=0·014
6
4
2
0
B
Mean number of transitional bronchioles per mL of lung per person
25
p=0·0001 20
p=0·0001 p=0·0001
p=0·029
15
10
Role of the funding source 5
0
Control (n=10)
GOLD 1 (n=10)
GOLD 2 (n=8)
GOLD 4 (n=6)
Figure 1: Quantification of terminal and transitional bronchioles in mild to moderate chronic obstructive pulmonary disease Comparison of the mean number of terminal (A) and transitional (B) bronchioles per mL of lung per person grouped by Global Initiative for Chronic Obstructive Lung Disease (GOLD) stage. Means and SD per group are shown.
bronchioles were defined as a wall area percentage greater than the 95th percentile of non-diseased airways in controls (appendix p 2). Obstructed terminal bronchioles were defined as those with 100% luminal obstruction (see appendix p 2 for more details).
Histological analysis After micro-CT scanning, the lung samples were used for histological assessment of terminal bronchioles. The first three 5 µm histological sections from the tissue sample were used to reorientate the micro-CT volumetric dataset in the same plane as the sample with image registration. The exact coordinates of the terminal bronchioles within the volumetric micro-CT scan were used to locate and section the exact regions of interest within the sample. Sections were stained 4
The study funder had no role in study design, data collection, data analysis, data interpretation, or writing of the report. The corresponding author had full access to all the data in the study and had final responsibility for the decision to submit for publication.
Results Of the 34 patients included in this study, ten were controls (smokers with normal lung function), ten patients had GOLD 1 COPD, eight had moderate GOLD 2 COPD, and six had GOLD 4 COPD with centrilobular emphysema. 28 of the patients underwent lobectomy and pneumo nectomy surgery, and six underwent lung transplantation. Age, height, smoking history, sex, and weight did not differ between groups (table 1). Preoperative inspiratory thoracic CT scans were available for 30 of the 34 patients because four patients with GOLD 1 or GOLD 2 COPD did not have preoperative scans. As we expected, patients with GOLD 2 and GOLD 4 COPD had a lower diffusing capacity (diffusing capacity of the lung for carbon monoxide adjusted for alveolar volume [DLCO:VA]) than controls (p<0·05; table 1). Total lung volume computed by CT was larger in patients with GOLD 4 COPD than in patients in all other groups (p<0·01; table 1). We obtained between six and 11 tissue samples per patient, given the varying sizes of lobes and lungs, which yielded 262 lung samples. Compared with control smokers with a mean of 4·7 TB/mL (SD 1·3), we found a decrease of 40% in
www.thelancet.com/respiratory Published online July 4, 2018 http://dx.doi.org/10.1016/S2213-2600(18)30196-6
Articles
Lung or lobe location
Sex
Whole lung volume* (L)
Terminal bronchioles per lung
Transitional bronchioles per lung
Alveolar surface area per lung (m²)
32 824
64·73 66·87
Control 1
Right middle lobe and right Female lower lobe†
2·39
13 481
2
Right upper lobe†
Female
2·54
16 163
33 711
3
Left upper lobe†
Male
3·14
12 848
32 121
94·20
4
Left whole lung
Male
2·20
9204
26 529
68·71
5
Left upper lobe†
Female
1·34
7898
15 232
37·45
6
Right whole lung
Female
2·87
10 566
23 114
58·10
7
Right whole lung
Female
1·56
6628
13 698
37·59
8
Right upper lobe†
Male
4·02
12 682
32 761
97·81
9
Left whole lung
Male
2·11
14 061
28 607
44·11
10
Left lower lobe†
Female
2·18
7382
14 763
53·84
Mean (SD)
··
··
2·43 (0·77)
25 336 (8122)
62·34 (21·07)
11
Right upper lobe†
Male
3·54
9744
15 428
86·32
12
Left lower lobe†
Male
2·13
12 424
20 992
51·98
13
Left upper lobe†
Female
2·19
5707
9218
30·16
14
Right whole lung
Male
4·11
4682
4682
82·92
16
Left upper lobe†
Male
2·71
10 710
21 420
69·35
18
Left whole lung
Female
2·27
6364
11 932
51·16
19
Right upper lobe†
Female
2·68
4551
8452
42·58
20
Right upper lobe†
Male
2·45
3462
11 541
58·59
Mean (SD)
··
··
2·76 (0·71)
7205 (3302)
12 958 (5954)
59·13 (19·41)
21
Left lower lobe†
Male
2·31
9753
14 630
49·09
24
Left upper lobe†
Male
2·78
5588
9500
54·05
25
Left upper lobe†
Female
2·19
7109
9297
37·29
26
Left whole lung
Female
1·86
3260
9781
44·43
27
Left whole lung
Male
2·06
3141
4398
53·26
28
Right upper lung
Female
2·29
4762
4082
27·63
Mean (SD)
··
··
2·25 (0·31)
5602 (2520)
8615 (3927)
44·29 (10·25)
29
Left whole lung
Male
4·30
4536
3402
41·25
30
Left whole lung
Female
3·10
3797
2531
27·82
31
Right whole lung
Male
4·64
12 897
11 725
36·13
32
Left whole lung
Male
3·30
1509
1509
33·62
33
Left whole lung
Male
3·95
3054
3054
29·52
34
Right whole lung
Male
4·68
10 436
Mean (SD)
··
··
3·99 (0·67)
11 091 (3225)
GOLD 1
GOLD 2
GOLD 4
6038 (4541)
1491
37·02
3952 (3888)
34·23 (4·98)
Mean (SD) are for 30 of the 34 patients who had a preoperative thoracic CT scan. GOLD=Global Initiative for Chronic Obstructive Lung Disease. *Whole lung volume denotes the right or left lung volume obtained from the clinical CT scans. †Samples were obtained within a lobe and not the whole lung, and therefore values calculated from these samples should only be taken as estimates of the whole lung.
Table 2: Estimates of small airway counts and alveolar surface area in whole lungs for each patient
patients with GOLD 1 COPD (2·8 TB/mL [1·4], p=0·014), 43% in patients with GOLD 2 COPD (2·7 TB/mL [1·1], p=0·036), and 68% in patients with GOLD 4 COPD (1·5 TB/mL [1·0], p=0·0001; figure 1A). The number of transitional bronchioles per mL of lung was reduced from a mean of 10·8 TrB/mL (SD 2·9) in the control smokers by 56% in patients with GOLD 1 COPD (4·8 TrB/mL [2·6], p=0·0001), 59% in patients with GOLD 2 COPD (4·4 TrB/mL [2·2], p=0·0001), and 90%
in patients with GOLD 4 COPD (1·1 TrB/mL [0·9], p=0·0001; figure 1B). Using these stereological counts, we estimated the total numbers of terminal and transitional bronchioles per whole lung volume determined from the thoracic CT scan (table 2). Controls had a mean of 11 091 terminal bronchioles per lung (SD 3225), and this number was reduced to 7205 (3302) in patients with GOLD 1 COPD, to 5602 (2520) in patients with GOLD 2 COPD, and to
www.thelancet.com/respiratory Published online July 4, 2018 http://dx.doi.org/10.1016/S2213-2600(18)30196-6
5
Articles
B
A
p=0·0002
p=0·0002 p=0·032
60 40 95th percentile of controls (20·3)
20 0 –20
Control (n=10)
GOLD 2 (n=6)
GOLD 1 (n=8)
p=0·024
80 Resected lung emphysema on clinical CT (% LAA < –910 HU)
Whole-lung emphysema on clinical CT (% LAA < –910 HU)
80
60 40 20 0 –20
GOLD 4 (n=6)
95th percentile of controls (7·4)
Control (n=10)
GOLD 1 (n=8)
p<0·0001
p<0·0001 p=0·027
1000
95th percentile of controls (449·1)
500
Control (n=10)
GOLD 1 (n=10)
GOLD 2 (n=8)
p=0·0021
150
GOLD 4 (n=6)
Alveolar surface area per mL of lung per person (cm2/mL)
1500
Mean linear intercept (µm)
GOLD 4 (n=6)
D
C
0
GOLD 2 (n=6)
p=0·019 p=0·0043
100
5th percentile of controls (41·3)
50
0
Control (n=10)
GOLD 1 (n=10)
GOLD 2 (n=8)
GOLD 4 (n=6)
Figure 2: Quantification of emphysema in mild to moderate chronic obstructive pulmonary disease Comparison of percentage low-attenuation area less than –910 HU (% LAA <–910 HU) on thoracic CT per patient grouped by Global Initiative for Chronic Obstructive Lung Disease (GOLD) stage in both the whole lung (A) and region of surgically resected lung lobe (B). The dashed line indicates the 95th percentile of controls for each graph. (C) Comparison of mean linear intercept measured by micro-CT per person grouped by GOLD stage. The dashed line denotes the upper 95th percentile of the mean linear intercept (449·1 µm) in the control group. (D) Comparison of alveolar surface area per mL of tissue measured by micro-CT per person grouped by GOLD stage. The dashed line denotes the lower 5th (41·3 cm²/mL) percentile of alveolar surface area per mL of tissue in the control group. Median and IQR per group are shown for non-parametric data (A, B, and C), and means and SD per group are shown for parametric data (D). LAA=low-attenuation area. HU=Hounsfield units.
6038 (4541) in patients with GOLD 4 COPD (table 2). For transitional bronchioles, this loss was even greater with an estimated number of 25 336 (8122) transitional bronchioles per lung in controls that was reduced to 958 (5954) in patients with GOLD 1 COPD, to 12 8615 (3927) in patients with GOLD 2 COPD, and to 3952 (3888) in patients with GOLD 4 COPD (table 2). We quantitatively analysed 30 preoperative CT scans on the segmented right or left lung (figure 2A) and the segmented resected lobes (figure 2B) for patients with a lobectomy sample. In both analyses, emphysema (%LAA <–910 HU, >95th percentile of controls) was only detectable in patients with GOLD 4 COPD compared with controls (p=0·0002) and patients with GOLD 1 COPD (p=0·032, figure 2A; p=0·024, figure 2B). Additionally, we found that the CT density measure ments on the 15 segmented lobes overlapped well with 6
the density measurements of the whole right lung CT scans (appendix p 6). When micro-CT was used to assess the degree of emphysema with mean linear intercept (figure 2C), we found patients with GOLD 4 COPD had an increase in mean linear intercept compared with controls (p<0·0001) and patients with GOLD 1 COPD (p=0·027). Importantly, when we assessed alveolar surface area per mL of lung available for gas exchange (figure 2D), we found a reduction in alveolar surface area per mL of lung in all COPD groups compared with controls, who had a mean alveolar surface area of 73·6 cm²/mL (SD 23·6) of lung. A 33% loss of alveolar surface area was seen in patients with GOLD 1 COPD (49·0 cm²/mL [17·9], p=0·019), a 45% loss was seen in patients with GOLD 2 COPD (40·3 cm²/mL [13·6], p=0·0021), and a 79% loss was seen in patients with GOLD 4 COPD (14·8 cm²/mL [4·3], p<0·0001; figure 2D).
www.thelancet.com/respiratory Published online July 4, 2018 http://dx.doi.org/10.1016/S2213-2600(18)30196-6
Articles
A
B Control GOLD 1 GOLD 2 GOLD 4
0·006
Control GOLD 1 GOLD 2 GOLD 4
0·06
Density
Density
0·04 0·004
0·02
0·002
0
0
500
1000
1500
0
2000
50
0
p=0·016 8
p=0·0006 20
p=0·039 Mean number of transitional bronchioles per mL of lung
Mean number of terminal bronchioles per mL of lung
150
D
C
6
4
2
0
100
Alveolar surface area per mL of lung (cm²/mL)
Mean linear intercept (μm)
Control (n=61)
GOLD 1 (n=37)
GOLD 2 (n=23)
All samples within 5th and 95th percentiles of control alveolar surface area per mL of lung
p=0·0016
15
10
5
0
Control (n=61)
GOLD 1 (n=37)
GOLD 2 (n=23)
All samples within 5th and 95th percentiles of control alveolar surface area per mL of lung
Figure 3: Association of emphysema with terminal and transitional bronchiole number in mild to moderate chronic obstructive pulmonary disease (A) Kernel density plot of mean linear intercept with micro-CT for all 262 samples, segregated by Global Initiative for Chronic Obstructive Lung Disease (GOLD) stage. The dashed lines denote the lower 5th (267·7 µm) and upper 95th (449·1 µm) percentiles of the mean linear intercept in the control group. A darker colour on the shaded bar represents increased mean linear intercept, indicating more emphysema. (B) Kernel density plot of alveolar surface area per mL of tissue measured by micro-CT for all 262 samples, segregated by GOLD stage. The dashed lines denote the lower 5th (41·3 cm²/mL) and upper 95th (126·2 cm²/mL) percentiles of alveolar surface area per mL of tissue in the control group. A darker colour on the shaded bar represents decreased alveolar surface area per mL of tissue, indicating more emphysema. Comparison of terminal (C) and transitional (D) bronchiolar number per mL of lung in all cores (n=121) with an alveolar surface area per mL of tissue within the 5th and 95th percentiles of controls, segregated by GOLD stage. We assessed data using a linear-mixed effect model, and the mean and SD per group are shown.
When total alveolar surface area was estimated for the whole lung (table 2), our data show that compared with controls (mean 62·34 m² [SD 21·07] ) the total alveolar surface area available for gas exchange was reduced to 59·13 m² (19·41) in patients with GOLD 1 COPD, 44·29 m² (10·25) in patients with GOLD 2 COPD, and 34·23 m² (4·98) in patients with GOLD 4 COPD. The alveolar surface area did not differ significantly between upper, lower, and whole lung samples obtained from micro-CT (appendix p 7). The kernel density plot (figure 3A) of all 262 lung samples shows that by use of the mean linear intercept, it was possible to distinguish emphysematous changes in GOLD 4, but not GOLD 1 and GOLD 2 samples,
compared with controls. When the same samples were plotted with alveolar surface area per mL of lung (figure 3B), GOLD 1, GOLD 2, and GOLD 4 samples segregated from control samples. To determine whether terminal and transitional bronchioles were lost in the absence of emphysematous destruction, we plotted the number of terminal and transitional bronchioles in the 121 of 262 cores that had an alveolar surface area per mL of lung within the normal range (41·3 cm²/mL for the 5th percentile to 126·2 cm²/mL for the 95th percentile) of control cases. In lung samples that had a normal alveolar surface area, a 29% reduction in the number of terminal bronchioles occurred in patients with GOLD 1 COPD (mean 3·3 TB/mL [SD 2·6], p=0·039),
www.thelancet.com/respiratory Published online July 4, 2018 http://dx.doi.org/10.1016/S2213-2600(18)30196-6
7
Articles
and a 40% reduction occurred in patients with GOLD 2 COPD (2·8 TB/mL [2·3], p=0·016) compared with controls (4·6 TB/mL [2·4]; figure 3C). A 41% reduction in the number of transitional bronchioles was observed
in patients with GOLD 1 COPD (mean 6·2 TrB/mL [SD 4·8], p=0·0016), and a 53% reduction in patients with GOLD 2 COPD (4·9 TrB/mL [3·7], p=0·0006) compared with controls (10·4 TrB/mL [5·4]; figure 3D). Two tissue
B
A p=0·0012 p=0·039 p=0·041
p=0·0023 p=0·083
80
60
40
20
4 3 2 1
C
GOLD 4 (n=6)
GOLD 2 (n=8)
D
Micro-CT as a scouting tool to identify airways of interest
Obstructed
5
0
GOLD 1 (n=10)
Thickened
6
0 Control (n=10)
Non-diseased
7
Terminal bronchioles per mL of lung
Diseased terminal bronchioles (%)
100
Control
GOLD 1
GOLD 2
GOLD 4
Non-disease terminal bronchiole
1
2
3
2
3
2
3
2
3
250 µm
1 2 3
1 250 µm
1 2 3 250 µm
E
F
Thickened terminal bronchiole
1
2
3
250 µm
1
250 µm
8
Obstructed terminal bronchiole
1
250 µm
2
3
1
250 µm
www.thelancet.com/respiratory Published online July 4, 2018 http://dx.doi.org/10.1016/S2213-2600(18)30196-6
Articles
samples from all GOLD 4 lungs had an alveolar surface area within the normal range and therefore no statistical analysis was done. We next assessed the morphology of the remaining terminal bronchioles. In control lungs, 12% of the terminal bronchioles (mean 1·0 TB/mL [SD 1·4]) were found to be diseased (thickened or obstructed) and that number increased to 41% in GOLD 1 lungs (1·5 TB/mL [1·5], p=0·0023), 37% in GOLD 2 lungs (1·1 TB/mL [1·5], p=0·083), and 77% in GOLD 4 lungs (1·1 TB/mL [0·7], p=0·0012; figure 4A). Moreover, the majority of diseased terminal bronchioles were thickened in patients with GOLD 1 or GOLD 2 COPD, and the majority of diseased terminal bronchioles were 100% obstructed in patients with GOLD 4 COPD (figure 4B; statistical comparisons are shown in appendix p 8). Figure 4C shows how the spatial coordinates from the volumetric micro-CT scans were used to accurately cut histological sections at regular intervals (dashed lines) along the airway branch length for patients with GOLD 1 or GOLD 2 COPD. Representative images are shown for a non-diseased (figure 4D), thickened (figure 4E), and obstructed (figure 4F) terminal bronchiole observed in patients with GOLD 1 and GOLD 2 COPD. Figure 4F further shows how the morphology of the terminal bronchioles can change substantially along its branch length and highlights the importance of volumetric micro-CT imaging to understand the disease pathology. Movat’s pentachrome staining showed that thickened airways quantified by micro-CT had airway wall fibrosis with collagen deposition and inflammatory cell infiltration. Furthermore, obstructed airways had thickened walls, and the obstructions were not simply mucus plugs but composed of collagen infiltrated with structural and inflammatory cells. Comparisons of the micro-CT measurements of terminal and transitional bronchiole number and alveolar surface area with the patients’ FEV1, FEV1/forced vital capacity (FVC), and DLCO:VA show that the lung function measurements used to classify the GOLD stages of COPD are significantly correlated with both the loss of terminal and transitional bronchioles and alveolar surface area in mild to moderate COPD (table 3; see also appendix p 9 for a detailed comparison). Figure 4: Terminal bronchiole pathology (A) Comparison of the percentage of diseased terminal bronchioles per total number of terminal bronchioles per person grouped by Global Initiative for Chronic Obstructive Lung Disease (GOLD) stage. Mean (SD) per group are shown. (B) Terminal bronchiolar number per mL of lung segregated by non-diseased, thickened, and complete luminal obstruction. Each bar indicates one lung donor. (C) Micro-CT was used as a scouting tool to precisely locate terminal bronchioles and enable efficient histological sectioning. The numbered dashed lines indicate the corresponding histological sections taken in a cross-section at locations 1, 2, and 3 for a representative terminal bronchiole (D–F). Matched micro-CT and histological sections of a representative non-diseased (D), thickened (E), and obstructed terminal bronchiole (F). Histological sections were stained with Movat’s pentachrome to highlight various components of connective tissue; elastic fibres (black), collagen and reticular fibres (yellow), and fibrin and muscle structures (red).
Discussion This study, to our knowledge, provides the first direct evidence that the smallest airways within the lung, conducting terminal bronchioles and respiratory transitional bronchioles, are significantly lost in the lungs of patients with mild (GOLD 1) and moderate (GOLD 2) COPD compared with age-matched smokers with normal lung function. Using a robust measurement of emphysema, alveolar surface area, which translates to the functional tissue involved in gas exchange, we also report that terminal and transitional bronchioles are lost in lung tissue in which no emphysematous destruction is present, indicating that small airways disease is an early pathological feature of mild and moderate COPD. Although several histological studies have documented airway remodelling and inflammation in the small airways of patients with COPD, to our knowledge no study to date has been able to quantify the numbers of terminal and transitional bronchioles and show that loss of these small airways correlates with decreases in lung function in mild and moderate COPD. Furthermore, using the combination of micro-CT and histology, we show that the remaining small airways that have been extensively studied in COPD23,26,27 do indeed have thickened walls and, when examined along their branch lengths, are often completely obstructed by fibrotic tissue. This study highlights the small airways disease that occurs in patients with mild and moderate COPD, which could not be measured previously, and highlights the potential importance of early intervention for disease modification. Our results support and substantially extend an earlier report that used micro-CT10 in patients with very severe (GOLD 4) COPD by providing new evidence that the destruction of terminal, and now also transitional bronchioles, is well established in people with mild or moderate COPD. Furthermore, using a stereological sampling design, we show that small airways disease is an early pathological feature of mild and moderate COPD and is present in regions of the lung that have no emphysematous disease. These data strongly support the concept that the small airways are the major and earliest site of airflow limitation in COPD.4–6 In 1970, Mead28 proposed that the small airways might represent a socalled quiet zone within the lung, where disease could accumulate over many years without being noticed. In support of this hypothesis, our data show that even in mild (GOLD 1) COPD, classified by an FEV1/FVC of less than 0·7 and an FEV1 of 80% or more predicted, a patient would potentially have lost, on average, 40% of their terminal bronchioles and 56% of their transitional bronchioles. The reason that this substantial loss of small airways can occur with a small change in lung function can be explained by the extensive parallel arrangement of the small airways within the lung. Direct measurements of small airway resistance in both post-mortem human lungs4 and living humans6
www.thelancet.com/respiratory Published online July 4, 2018 http://dx.doi.org/10.1016/S2213-2600(18)30196-6
9
Articles
Alveolar surface area per mL of lung
Terminal bronchioles per mL of lung
Transitional bronchioles per mL of lung
All patients
All patients
All patients
Mild-to-moderate COPD
Mild-to-moderate COPD
Mild-to-moderate COPD
FEV1 (% predicted)
0·715 (p<0·0001)
0·502 (p=0·0078)
0·523 (p=0·0018)
0·326 (p=0·0988)
0·640 (p=0·0001)
0·423 (p=0·0301)
FEV1/FVC (%)
0·722 (p<0·0001)
0·547 (p=0·0045)
0·581 (p=0·0008)
0·465 (p=0·0189)
0·717 (p<0·0001)
0·627 (p=0·0009)
DLCO:VA (mL/min per mm Hg per L)
0·610 (p<0·001)
0·549 (p=0·0045)
0·591 (p=0·0011)
0·567 (p=0·0050)
0·715 (p<0·0001)
0·676 (p=0·0003)
Pearson’s correlations were calculated to investigate the association between micro-CT measures and pulmonary function. Data are r value (false discovery rate p value). COPD=chronic obstructive pulmonary disease. FVC=forced vital capacity. DLCO:VA=diffusing capacity of the lung for carbon monoxide adjusted for alveolar volume.
Table 3: Correlation of small airways disease and emphysema with lung function
have shown that the small airways account for a small proportion of the total resistance to airflow in the normal human lung but become the major site of increased resistance in COPD. Because the small airways are arranged in parallel, a 50% reduction in their number is expected to double their resistance. Previous direct measurements of peripheral airway resistance in living humans show that doubling the mean value of 0·70 cm H2O/L per s [SD 0·26] measured in individuals with normal lung function to 1·4 cm H2O/L per s falls well below the 2·78–4·59 cm H2O/L per s measured in patients with severe COPD.6 Although these findings are consistent with the much greater reduction in terminal and transitional airways reported here and previously for patients with end-stage GOLD 4 COPD,10 we do not mean to imply that these are the only airways narrowed and lost in COPD. By use of thoracic CT scans, several previous studies have shown that airways ranging from 0·8 mm to 3 mm in diameter are also destroyed in all GOLD categories of COPD.7,9,10 The relative importance of small airways disease and emphysema for lung function decline has been debated greatly. In this study, we show direct evidence that small airways disease is a well established pathological feature of patients with mild and moderate COPD. Furthermore, these data also show that the reduction in both terminal and transitional bronchioles is well established before emphysematous destruction of the alveolar surface area can be visualised with the spatial resolution (1000 μm) provided by thoracic CT scans. This finding strongly suggests that the appearance of emphysema on thoracic CT scans seems to be a good predictor of a rapid decline in FEV121,29,30 because a substantial number of terminal and transitional bronchioles has already been lost before the emphysematous disease can be detected. Therefore, approaches to identify susceptible smokers before they reach the GOLD 1 classification7,31,32 warrant further attention. Additionally, these new data substantially extend earlier reports on the pathology present in the remaining small airways by showing that the remaining airways are thickened and become more obstructed with disease progression. This finding is consistent with the concept that airways develop thickened walls and narrowed lumens with COPD progression.23 What 10
remains unclear is whether thickening, obstruction, and obliteration occur sequentially in individual airways. In several reviews, we have postulated that terminal and transitional bronchioles might be vulnerable to deposition of fine particulate matter (especially from tobacco smoke) because they are located in the region of the lung in which the transition from bulk flow to diffusion of gas occurs.33 The answer to these questions will require much further, but necessary, work. These data do, however, suggest that small airways disease is related to the pathology of the airways themselves, and that loss of elastic recoil due to emphysematous destruction is an independent phenomenon that potentially causes further airway obstruction. It has been postulated that destruction of the axial elastic network might allow the airways to shorten and obliterate the airway lumen.34 Our current working hypothesis is that the disappearance of small airways is preceded by a localised constrictive bronchiolitis that first obliterates the airway lumen and then separates the affected airway into two parts that leaves closed end buds coming off both the parent airway and the distal daughter branch. Thus, future studies designed to answer this question with the combination of micro-CT and histology are still required to definitively understand the nature of this destructive disease process. This study has limitations that should be mentioned. We note that this study used archived lungs and lobes from surgeries done in the early 2000s for lung cancer resection of small peripheral tumours with no metastasis or lung transplantation, which excluded patients with severe GOLD 3. Therefore, because the study is crosssectional, it does not allow us to observe the progress of the disease pathology over time, or to determine whether the patients with mild or moderate COPD might have been born with a lower number of airways. However, only lobectomy and pneumonectomy surgeries provide the possibility to collect intact lung tissue for inflation, preoperative pulmonary function, and CT data from smokers with and without mild or moderate COPD. Furthermore, patients had an inspiratory CT scan for surgical planning but not for COPD assessment, which now use inspiratory and expiratory CT scans. Thus, we were unable to assess the association between small
www.thelancet.com/respiratory Published online July 4, 2018 http://dx.doi.org/10.1016/S2213-2600(18)30196-6
Articles
airways disease and gas trapping in this study. We also acknowledge that for 15 (44%) of the 34 individuals, samples were obtained only from a single lobe, which did not allow us to capture disease heterogeneity within the whole lung. However, when we compared the clinical CT-based density histograms of the whole lung with the resected lung region or alveolar surface area obtained from micro-CT measures, we found no significant variation between upper and lower lung samples. Despite these limitations, the numbers of terminal and transitional bronchioles and alveolar surface area reported for the control smokers with normal lung function in this study are consistent with those previously reported in stereological studies of the normal human lung.35,36 In conclusion, these findings suggest that, like the kidney in which a substantial proportion of nephrons are lost before renal dysfunction appears,37 the development of airflow limitation in COPD involves progressive destruction and loss of the terminal and transitional bronchioles before a decline in lung function is observed. Additionally, these data also indicate that destruction of terminal and transitional bronchioles can occur in the absence of emphysematous destruction in mild and moderate COPD and that the remaining airways have narrowed lumens and thickened walls. Most importantly, these data suggest that several large clinical trials investigating COPD treatments in severe COPD did not show beneficial treatment effects because they were initiated after a substantial number of terminal and transitional bronchioles were already lost. Therefore, future clinical trials need to be done in cohorts of patients with mild and moderate COPD if we are to have any chance to modify disease outcomes. Contributors H-KK, DMV, SB, AH, OLK, NF, WME, MK, PL, IS, JAW, JDC, HOC, and T-LH contributed to data collection. H-KK, DMV, SB, HOC, PDP, JCH, and T-LH contributed to study design, data interpretation, figures, and writing. Declaration of interests DMV is supported by Canadian Thoracic Society and Alpha-1 Foundation fellowships. MK is supported by the Canadian Institutes of Health Research (CIHR) Banting fellowship. T-LH is supported by the CIHR, Michael Smith Health Research Foundation, Parker B Francis, and Providence Health Care Research Institute New Investigator awards. HOC is supported by Spiration Inc and has personal fees from Samsung. All other authors declare no competing interests. Acknowledgments This study was funded by a Canadian Institutes of Health Research operating grant MOP 130504. The authors thank the James Hogg Lung Registry (Darren Sutherland), Histology Core (Amrit Samra), and the Bio-Imaging Core (Aaron M Barlow) from the Centre for Heart Lung Innovation, St Paul’s Hospital, Vancouver, BC, Canada, and the µ-VIS X-Ray Imaging Centre at the University of Southampton, Southampton, UK. IS is funded by Nikon and EPSRC grant EPH01506X. References 1 Weibel ER. Morphometry of the human lung. New York, NY: Academic, 1963. 2 Green M. How big are the bronchioles? St Thomas Hosp Gaz 1965; 63: 136–39.
3 4 5
6 7
8 9 10 11
12 13 14 15 16
17 18 19 20 21 22 23 24
25
Macklem PT, Mead J. Resistance of central and peripheral airways measured by a retrograde catheter. J Appl Physiol 1967; 22: 395–401. Hogg JC, Macklem PT, Thurlbeck WM. Site and nature of airway obstruction in chronic obstructive lung disease. N Engl J Med 1968; 278: 1355–60. Van Brabandt H, Cauberghs M, Verbeken E, Moerman P, Lauweryns JM, Van de Woestijne KP. Partitioning of pulmonary impedance in excised human and canine lungs. J Appl Physiol Respir Environ Exerc Physiol 1983; 55: 1733–42. Yanai M, Sekizawa K, Ohrui T, Sasaki H, Takishima T. Site of airway obstruction in pulmonary disease: direct measurement of intrabronchial pressure. J Appl Physiol 1992; 72: 1016–23. Kirby M, Tanabe N, Tan WC, et al. Total airway count on computed tomography and the risk of chronic obstructive pulmonary disease progression. Findings from a population-based study. Am J Respir Crit Care Med 2018; 197: 56–65. Hasegawa M, Nasuhara Y, Onodera Y, et al. Airflow limitation and airway dimensions in chronic obstructive pulmonary disease. Am J Respir Crit Care Med 2006; 173: 1309–15. Kurashima K, Hoshi T, Takaku Y, et al. Changes in the airway lumen and surrounding parenchyma in chronic obstructive pulmonary disease. Int J Chron Obstruct Pulmon Dis 2013; 8: 523–32. McDonough JE, Yuan R, Suzuki M, et al. Small-airway obstruction and emphysema in chronic obstructive pulmonary disease. N Engl J Med 2011; 365: 1567–75. Pauwels RA, Buist AS, Calverley PM, Jenkins CR, Hurd SS, GOLD Scientific Committee. Global strategy for the diagnosis, management, and prevention of chronic obstructive pulmonary disease. NHLBI/WHO Global Initiative for Chronic Obstructive Lung Disease (GOLD) Workshop summary. Am J Respir Crit Care Med 2001; 163: 1256–76. Rodriguez M, Bur S, Favre A, Weibel ER. Pulmonary acinus: geometry and morphometry of the peripheral airway system in rat and rabbit. Am J Anat 1987; 180: 143–55. Haefeli-Bleuer B, Weibel ER. Morphometry of the human pulmonary acinus. Anat Rec 1988; 220: 401–14. Muhlfeld C, Hegermann J, Wrede C, Ochs M. A review of recent developments and applications of morphometry/stereology in lung research. Am J Physiol Lung Cell Mol Physiol 2015; 309: L526–36. Fehrenbach H. Animal models of pulmonary emphysema: a stereologist’s perspective. Eur Respir Rev 2006; 15: 136–47. Knudsen L, Weibel ER, Gundersen HJ, Weinstein FV, Ochs M. Assessment of air space size characteristics by intercept (chord) measurement: an accurate and efficient stereological approach. J Appl Physiol 2010; 108: 412–21. Ochs M. Estimating structural alterations in animal models of lung emphysema. Is there a gold standard? Ann Anat 2014; 196: 26–33. Knust J, Ochs M, Gundersen HJ, Nyengaard JR. Stereological estimates of alveolar number and size and capillary length and surface area in mice lungs. Anat Rec 2009; 292: 113–22. Scott AE, Vasilescu DM, Seal KA, et al. Three dimensional imaging of paraffin embedded human lung tissue samples by micro-computed tomography. PLoS One 2015; 10: e0126230. Sieren JP, Newell JD Jr, Barr RG, et al. SPIROMICS protocol for multicenter quantitative computed tomography to phenotype the lungs. Am J Respir Crit Care Med 2016; 194: 794–806. Yuan R, Hogg JC, Pare PD, et al. Prediction of the rate of decline in FEV1 in smokers using quantitative computed tomography. Thorax 2009; 64: 944–49. Muller NL, Staples CA, Miller RR, Abboud RT. “Density mask”. An objective method to quantitate emphysema using computed tomography. Chest 1988; 94: 782–87. Hogg JC, Chu F, Utokaparch S, et al. The nature of small-airway obstruction in chronic obstructive pulmonary disease. N Engl J Med 2004; 350: 2645–53. Hsia CC, Hyde DM, Ochs M, et al. An official research policy statement of the American Thoracic Society/European Respiratory Society: standards for quantitative assessment of lung structure. Am J Respir Crit Care Med 2010; 181: 394–418. Vasilescu DM, Klinge C, Knudsen L, et al. Stereological assessment of mouse lung parenchyma via nondestructive, multiscale micro-CT imaging validated by light microscopic histology. J Appl Physiol 2013; 114: 716–24.
www.thelancet.com/respiratory Published online July 4, 2018 http://dx.doi.org/10.1016/S2213-2600(18)30196-6
11
Articles
26 Di Stefano A, Turato G, Maestrelli P, et al. Airflow limitation in chronic bronchitis is associated with T-lymphocyte and macrophage infiltration of the bronchial mucosa. Am J Respir Crit Care Med 1996; 153: 629–32. 27 O’Shaughnessy TC, Ansari TW, Barnes NC, Jeffery PK. Inflammation in bronchial biopsies of subjects with chronic bronchitis: inverse relationship of CD8+ T lymphocytes with FEV1. Am J Respir Crit Care Med 1997; 155: 852–57. 28 Mead J. The lung’s “quiet zone”. N Engl J Med 1970; 282: 1318–19. 29 Nishimura M, Makita H, Nagai K, et al. Annual change in pulmonary function and clinical phenotype in chronic obstructive pulmonary disease. Am J Respir Crit Care Med 2012; 185: 44–52. 30 Mohamed Hoesein FA, de Hoop B, Zanen P, et al. CT-quantified emphysema in male heavy smokers: association with lung function decline. Thorax 2011; 66: 782–87. 31 Allinson JP, Hardy R, Donaldson GC, Shaheen SO, Kuh D, Wedzicha JA. The presence of chronic mucus hypersecretion across adult life in relation to chronic obstructive pulmonary disease development. Am J Respir Crit Care Med 2016; 193: 662–72.
12
32 de Marco R, Accordini S, Cerveri I, et al. Incidence of chronic obstructive pulmonary disease in a cohort of young adults according to the presence of chronic cough and phlegm. Am J Respir Crit Care Med 2007; 175: 32–39. 33 Hogg JC, Pare PD, Hackett TL. The contribution of small airway obstruction to the pathogenesis of chronic obstructive pulmonary disease. Physiol Rev 2017; 97: 529–52. 34 Mitzner W. Emphysema—a disease of small airways or lung parenchyma? N Engl J Med 2011; 365: 1637–39. 35 Weibel ER. Morphometric analysis of the number, volume and surface of the alveoli and capillaries of the human lung. Z Zellforsch Mikrosk Anat 1962; 57: 648–66 (in German). 36 Wiebe BM, Laursen H. Lung morphometry by unbiased methods in emphysema: bronchial and blood vessel volume, alveolar surface area and capillary length. APMIS 1998; 106: 651–56. 37 Remuzzi G, Bertani T. Pathophysiology of progressive nephropathies. N Engl J Med 1998; 339: 1448–56.
www.thelancet.com/respiratory Published online July 4, 2018 http://dx.doi.org/10.1016/S2213-2600(18)30196-6
Small airways disease in mild and moderate chronic obstructive pulmonary disease: a cross-sectional study Hyun-Kyoung Koo*, Dragoş M Vasilescu*, Steven Booth*, Aileen Hsieh, Orestis L Katsamenis, Nick Fishbane, W Mark Elliott, Miranda Kirby, Peter Lackie, Ian Sinclair, Jane A Warner, Joel D Cooper, Harvey O Coxson, Peter D Paré, James C Hogg, Tillie-Louise Hackett
Summary
Background The concept that small conducting airways less than 2 mm in diameter become the major site of airflow obstruction in chronic obstructive pulmonary disease (COPD) is well established in the scientific literature, and the last generation of small conducting airways, terminal bronchioles, are known to be destroyed in patients with very severe COPD. We aimed to determine whether destruction of the terminal and transitional bronchioles (the first generation of respiratory airways) occurs before, or in parallel with, emphysematous tissue destruction. Methods In this cross-sectional analysis, we applied a novel multiresolution CT imaging protocol to tissue samples obtained using a systematic uniform sampling method to obtain representative unbiased samples of the whole lung or lobe of smokers with normal lung function (controls) and patients with mild COPD (Global Initiative for Chronic Obstructive Lung Disease [GOLD] stage 1), moderate COPD (GOLD 2), or very severe COPD (GOLD 4). Patients with GOLD 1 or GOLD 2 COPD and smokers with normal lung function had undergone lobectomy and pneumonectomy, and patients with GOLD 4 COPD had undergone lung transplantation. Lung tissue samples were used for stereological assessment of the number and morphology of terminal and transitional bronchioles, airspace size (mean linear intercept), and alveolar surface area. Findings Of the 34 patients included in this study, ten were controls (smokers with normal lung function), ten patients had GOLD 1 COPD, eight had GOLD 2 COPD, and six had GOLD 4 COPD with centrilobular emphysema. The 34 lung specimens provided 262 lung samples. Compared with control smokers, the number of terminal bronchioles decreased by 40% in patients with GOLD 1 COPD (p=0·014) and 43% in patients with GOLD 2 COPD (p=0·036), the number of transitional bronchioles decreased by 56% in patients with GOLD 1 COPD (p=0·0001) and 59% in patients with GOLD 2 COPD (p=0·0001), and alveolar surface area decreased by 33% in patients with GOLD 1 COPD (p=0·019) and 45% in patients with GOLD 2 COPD (p=0·0021). These pathological changes were found to correlate with lung function decline. We also showed significant loss of terminal and transitional bronchioles in lung samples from patients with GOLD 1 or GOLD 2 COPD that had a normal alveolar surface area. Remaining small airways were found to have thickened walls and narrowed lumens, which become more obstructed with increasing COPD GOLD stage. Interpretation These data show that small airways disease is a pathological feature in mild and moderate COPD. Importantly, this study emphasises that early intervention for disease modification might be required by patients with mild or moderate COPD. Funding Canadian Institutes of Health Research. Copyright © 2018 Elsevier Ltd. All rights reserved.
Introduction Small conducting airways less than 2 mm in internal diameter, which offer less than 10% of the total resistance to airflow in the normal lung,1–3 become the major site of airflow obstruction in chronic obstructive pulmonary disease (COPD).4–6 To develop effective treatments for COPD, an understanding of disease pathogenesis within these small airways is vital. Current CT imaging protocols applied in longitudinal COPD studies, with 1–2 mm inplane spatial resolution and low-dose radiation exposure (1·5 mSv), can only resolve airways larger than 2·5 mm in internal diameter.7,8 A study9 that used ultra-highresolution CT and high-dose radiation exposure (11·2 mSv) has shown that it is possible to visualise small airways that are 0·8 mm in diameter. However, to date,
only micro-CT has the spatial resolution required to resolve the alveolar structure and reliably identify and characterise the smallest generation of conducting airways, the terminal bronchioles, which have an average lumen diameter of 424 µm.10 By use of micro-CT on fixed and dried lung samples, McDonough and colleagues10 showed that compared with controls (n=4) the number of terminal bronchioles was reduced by 72% in seven patients with very severe COPD (Global Initiative for Chronic Obstructive Lung Disease [GOLD] stage 4)11 with panlobular emphysema and by 89% in four patients with GOLD 4 COPD with centrilobular emphysema. Additionally, they showed that the number of terminal bronchioles was reduced even in tissue samples that had no detectable emphysema as measured by mean linear
www.thelancet.com/respiratory Published online July 4, 2018 http://dx.doi.org/10.1016/S2213-2600(18)30196-6
Lancet Respir Med 2018 Published Online July 4, 2018 http://dx.doi.org/10.1016/ S2213-2600(18)30196-6 See Online/Comment http://dx.doi.org/10.1016/ S2213-2600(18)30290-X *Joint first authors Centre for Heart Lung Innovation, University of British Columbia, St Paul’s Hospital, Vancouver, BC, Canada (H-K Koo MD, D M Vasilescu PhD, S Booth MSc, A Hsieh, N Fishbane MSc, W M Elliott PhD, M Kirby PhD, Prof H O Coxson PhD, Prof P D Paré MD, Prof J C Hogg MD, T-L Hackett PhD); Department of Anesthesiology, Pharmacology and Therapeutics (H-K Koo, S Booth, T-L Hackett), Department of Medicine (Prof P D Paré), and Department of Pathology (Prof J C Hogg), University of British Columbia, Vancouver, BC, Canada; µ-VIS X-ray Imaging Centre, Faculty of Engineering and the Environment (O L Katsamenis PhD, Prof I Sinclair PhD); Clinical and Experimental Sciences, Faculty of Medicine (P Lackie PhD, J A Warner PhD), University of Southampton, Southampton, UK; and Division of Thoracic Surgery, University of Pennsylvania, Philadelphia, PA, USA (Prof J D Cooper MD) Correspondence to: Dr Tillie-Louise Hackett, Centre for Heart Lung Innovation, University of British Columbia, St Paul’s Hospital, Vancouver, BC V6Z 1Y6, Canada [email protected]
1
Articles
Research in context Evidence before this study Results from several physiological studies have shown that small conducting airways less than 2 mm in internal diameter are the major site of airflow obstruction in patients who have very severe chronic obstructive pulmonary disease (COPD). On Jan 18, 2018, we searched the scientific literature in PubMed (with no date or language restrictions) for the following search terms “small airways”, “emphysema”, and “COPD”. We found no publications using multiresolution imaging to investigate small airway pathology and its association with emphysema in mild and moderate COPD. Added value of this study This study uses a novel approach that combines stereology, multiresolution CT imaging, and histology that, to our knowledge, for the first time provides direct evidence that substantial loss of small airways (terminal and transitional bronchioles) occurs in patients with mild and moderate COPD, before the detection of emphysema by clinical CT. Furthermore, these new data show that destruction of terminal and transitional bronchioles can occur in regions of the lung with normal alveolar surface area, a robust measure of internal lung structure measured with micro-CT. We also
intercept. These data suggested that terminal bronchiole obliteration might precede emphysematous tissue destruction in COPD.10 The hypothesis that terminal bronchioles are lost before emphysematous destruction cannot be tested in a longitudinal study of disease progression within the same patients. We aimed to determine whether destruction of the terminal and, to our knowledge, for the first time transitional bronchioles (the first generation of respiratory airways)12,13 occurs before, or in parallel with, emphy sematous tissue destruction. Furthermore, we assessed emphysematous tissue destruction using the measure ment of alveolar surface area. This approach is superior to mean linear intercept because, in addition to airspace enlargement, it measures the destruction of alveolar septae, which is an important pathological change in emphysema,14,15 and it is a robust measure of internal lung structure, which is calculated from airspace size and volume fractions of alveolar tissue, and is therefore not affected by inflation status and elastic properties of the lung.16–18 Since lungs resected for lung cancer are invariably fixed before study, we also used a new micro-CT imaging protocol for formalin-fixed, paraffin-embedded lung tissue samples,19 which enabled a matched histological investigation of the remaining small airways.
Methods
Participants We did a cross-sectional analysis of smokers with normal lung function, patients with mild COPD, patients with 2
used histology to show that the remaining small airways have thickened walls and narrowed lumens that become more obstructed with increased COPD severity. These data indicate that small airways disease is a feature of mild and moderate COPD. Implications of all the available evidence Although these results must be considered preliminary because of the small number of cases that have been studied to date, they strongly support the concept that small airways disease is well established by the time the diagnosis of mild or moderate COPD is made. Furthermore, the data suggest that the reason most clinical trials investigating COPD treatments in severe COPD did not show beneficial effects is because they were initiated after a substantial number of terminal and transitional bronchioles were already lost, and that early intervention in patients with mild and moderate COPD might be required for disease modification. We expect that the findings from this research will prompt discussion of guidelines and policies to improve early diagnosis and effective management of patients with mild and moderate COPD.
moderate COPD, and patients with very severe COPD, who donated lungs or lung samples after pneumonectomy or lobectomy surgery. The severity of COPD was graded with the 2011 GOLD staging system, on the basis of postbronchodilator FEV1, in which airflow limitation is indicated as mild for GOLD 1 (FEV1 ≥80% predicted), moderate for GOLD 2 (50% ≤ FEV1 <80% predicted), and very severe for GOLD 4 (FEV1 <30% predicted). Informed consent was obtained from patients treated for small peripheral lung cancers (primary tumours with no metastasis) by lobectomy and pneumonectomy surgery at St Paul’s Hospital (Vancouver, BC, Canada), with approval by the Providence Health Care Research Ethics Board (PHCREB; Vancouver, BC, Canada) and from patients who underwent lung transplantation for GOLD 4 COPD at the University of Pennsylvania Hospital (Philadelphia, PA, USA), with the approval of the University of Pennsylvania Hospital Institutional Review Board. All specimens were stored in a dedicated archival Lung Registry between 1991 and 2006, with approval of the PHCREB (number H13-02173).
CT analysis CT scans were acquired before surgery at suspended full inspiration by coaching the patient to inhale to total lung capacity in the supine position, as previously described.20 The CT image acquisition parameters were 120–140 kV, 80–345 mA, 5–10 mm slice thickness, and 0–10 mm space between slices. Semi-automatic lung segmentation was done on the CT scans with a custom software package,
www.thelancet.com/respiratory Published online July 4, 2018 http://dx.doi.org/10.1016/S2213-2600(18)30196-6
Articles
EmphylxJ21 (University of British Columbia, Vancouver, BC, Canada), to isolate the specific resected lung or lobe and calculate lung volumes. To quantify emphysema on the 10 mm thick slice CT scans, we used a density threshold of –910 Hounsfield units (HU; percentage low-attenuation area [LAA] <–910 HU), since this pathology-validated threshold defined by Muller and colleagues22 has been shown to yield the best correlation with emphysema by use of 10 mm thick slice CT datasets.22
Lung tissue preparation Briefly, resected lobes and whole lungs were inflated, frozen, and sliced into contiguous 2 cm thick slices in the transaxial plane, as previously described.10,23 We excluded the tumour and normal margin and extracted tissue samples measuring 1·5 cm in diameter from the lung slices using a systematic uniform sampling method to obtain representative unbiased samples of the lobe or lung. For this study, we obtained one tissue sample per slice. All samples were fixed in pre-cooled alcohol-based formalin overnight and then infiltrated with low-meltingpoint paraffin by use of a tissue processor (model ASP6025; Leica, Ontario, CA, USA). We measured the diameter, cross-sectional area, and volume of each tissue sample before and after processing to assess tissue shrinkage, which was applied as a correction factor to the quantitative measurements.
Micro-CT imaging The paraffin-embedded tissue samples were scanned with a micro-CT scanner (HMX225; Nikon Metrology, Brighton, MI, USA), with a previously developed imaging protocol19 (50 kV, 180 µA, and reconstructed with an isotropic voxel size of 6·7 µm). Images were processed with ImageJ software, version 1.47q (National Institutes of Health, Bethesda, MD, USA), and data were collected and analysed by three observers masked to patient characteristics. To ensure measurements obtained by micro-CT imaging of formalin-fixed, paraffin-embedded tissue were similar to the method of glutaraldehyde fixation and critical point drying previously used by McDonough and colleagues,10 we used both methods to process adjacent lung tissue samples from the GOLD 4 donor lungs, and we found no difference between the two methods when measuring alveolar surface area (appendix p 4).
Stereology Following the American Thoracic Society and European Respiratory Society guidelines for stereology,24 we applied systematic uniform random (SUR) sampling to obtain ten images per sample for analysis. The total number of images in the micro-CT scan was first divided by ten to generate ten substacks, and a random number generator was then used to pick the random image used in each substack. SUR-sampled images were used for measure ment of mean linear intercept with a line grid24,25 and
alveolar surface area with a point counting grid16,25 in Image-Pro Plus, version 5.1 (Media Cybernetics, Silver Spring, MD, USA). Measurements of mean linear intercept and alveolar surface area obtained by micro-CT were validated by comparison to the gold standard of histology (appendix p 5). Terminal bronchioles were identified within the microCT scan by following consecutive airway branches until transitional bronchioles13 were identified by the first occurrence of individual alveoli along the airway wall (appendix p 3). Identification of transitional bronchioles subsequently enabled the parent airway to be counted as a terminal bronchiole. The numbers of terminal bronchioles per mL of lung (TB/mL) and transitional bronchioles per mL of lung (TrB/mL) were calculated by dividing the number of respective bronchioles by the sample volume corrected for shrinkage. Terminal bronchioles were further classified as nondiseased, thickened, or obstructed. Thickened terminal Control group (smokers with normal lung function; n=10) Lung tissue samples
75
GOLD 1 group (mild COPD; n=10) 81
See Online for appendix
GOLD 2 group (moderate COPD; n=8) 62
GOLD 4 group (very severe COPD; n=6) 44
Sex* Female
6 (60%)
4 (40%)
3 (38%)
1 (17%)
Male
4 (40%)
6 (60%)
5 (63%)
5 (83%)
Age (years)*
62·0 (7·9)
67·4 (7·3)
Height (cm)*
167·9 (8·7)
168·6 (9·9)
62·9 (11·3) 167·4 (6·7)
59·2 (2·1) 170·3 (6·3)
Weight (kg)*
68·4 (14·7)
77·1 (18·1)
73·1 (15·6)
71·5 (10·8)
Smoking history (pack-years)*
34·5 (10·5)
45·5 (25·3)
33·6 (12·7)
37·5 (15·1)
FEV1 (% predicted)†
91·8 (6·4)
88·3 (6·2)
62·1 (9·5)
22·3 (6·7)
FVC after bronchodilator (% predicted)‡
96·7 (5·4)
108·3 (8·5)
88·5 (6·4)
60·3 (19·7)
FEV1/FVC (%)§
74·9 (4·4)
63·5 (4·7)
60·1 (7·6)
29·5 (7·8)
DLCO:VA (mL/min per mm Hg per L)¶
3·85 (0·9)
2·83 (0·7)
2·64 (0·9)
1·71 (0·8)
Total lung volume (L)
4·86 (1·4)
5·37 (1·5)
4·92 (0·9)
7·91 (1·0)||
% LAA <–910 (HU)
1·05 (3·7)
1·84 (7·9)
8·72 (12·4)
67·68 (16·8)**
Right upper lobe
2
3
2
Left upper lobe
2
3
1
Right middle lobe and right lower lobe
1
Surgical resection
··
Left lower lobe
1
1
Right whole lung
2
2
Left whole lung
2
1
·· ·· ··
1
·· ··
·· 3
2 4
Data are n (%) or mean (SD). GOLD=Global Initiative for Chronic Obstructive Lung Disease. COPD=chronic obstructive pulmonary disease. FVC=forced vital capacity. DLCO:VA=diffusing capacity of the lung for carbon monoxide adjusted for alveolar volume. LAA=low-attenuation area. HU=Hounsfield units. *One-way ANOVA with a Tukey’s pairwise comparison showed no significant differences. †Differences in FEV1 were significant (p<0·001) between all groups except between the control and GOLD 1 groups. ‡Differences were significant (p<0·01) between all groups except between the control and GOLD 1 groups, and control and GOLD 2 groups. §Differences were significant (p<0·01) between all groups except between the GOLD 1 and GOLD 2 groups. ¶Differences were significant between the control and GOLD 2 groups (p<0·05), and control and GOLD 4 (p<0·05) groups. ||Increased (p<0·01) compared with all other groups. **Increased compared with the control (p=0·0002) and GOLD 1 (p=0·024) groups.
Table 1: Preoperative patient characteristics
www.thelancet.com/respiratory Published online July 4, 2018 http://dx.doi.org/10.1016/S2213-2600(18)30196-6
3
Articles
using Movat’s pentachrome stain (see appendix p 2 for more details).
A 10
Mean number of terminal bronchioles per mL of lung per person
p=0·0001
Statistical analysis
p=0·036
8
We averaged all samples per patient and presented per case. We assessed outcomes for Gaussian distribution and tested for normality using the Kolmogorov-Smirnov test, D’Agostino-Pearson omnibus normality test, and Shapiro-Wilk normality test. Measurements of airways per mL of lung, diseased airways per mL of lung, and alveolar surface area per mL of tissue were normally distributed, and assessed with a one-way ANOVA with a Tukey’s pairwise comparison. The Shapiro-Wilk nor mality test showed that measurements of mean linear intercept and CT density data were not normally distributed, and therefore a non-parametric Kruskal Wallis with Dunn’s post-hoc test was used. We used GraphPad Software, version 5, for this analysis. We generated kernel density plots with multiple comparisons using the Bergmann Hommel method (scmamp R package). We assessed the number of terminal and transitional bronchioles in cores with normal alveolar surface area using a linear mixed-effect model (nlme package), and we calculated Pearson’s correlations with false discovery rate corrections using the statistical software R3.3.1. Data are expressed as the mean (SD) or median (IQR) for non-parametric data. p values of less than 0·05 were considered significant.
p=0·014
6
4
2
0
B
Mean number of transitional bronchioles per mL of lung per person
25
p=0·0001 20
p=0·0001 p=0·0001
p=0·029
15
10
Role of the funding source 5
0
Control (n=10)
GOLD 1 (n=10)
GOLD 2 (n=8)
GOLD 4 (n=6)
Figure 1: Quantification of terminal and transitional bronchioles in mild to moderate chronic obstructive pulmonary disease Comparison of the mean number of terminal (A) and transitional (B) bronchioles per mL of lung per person grouped by Global Initiative for Chronic Obstructive Lung Disease (GOLD) stage. Means and SD per group are shown.
bronchioles were defined as a wall area percentage greater than the 95th percentile of non-diseased airways in controls (appendix p 2). Obstructed terminal bronchioles were defined as those with 100% luminal obstruction (see appendix p 2 for more details).
Histological analysis After micro-CT scanning, the lung samples were used for histological assessment of terminal bronchioles. The first three 5 µm histological sections from the tissue sample were used to reorientate the micro-CT volumetric dataset in the same plane as the sample with image registration. The exact coordinates of the terminal bronchioles within the volumetric micro-CT scan were used to locate and section the exact regions of interest within the sample. Sections were stained 4
The study funder had no role in study design, data collection, data analysis, data interpretation, or writing of the report. The corresponding author had full access to all the data in the study and had final responsibility for the decision to submit for publication.
Results Of the 34 patients included in this study, ten were controls (smokers with normal lung function), ten patients had GOLD 1 COPD, eight had moderate GOLD 2 COPD, and six had GOLD 4 COPD with centrilobular emphysema. 28 of the patients underwent lobectomy and pneumo nectomy surgery, and six underwent lung transplantation. Age, height, smoking history, sex, and weight did not differ between groups (table 1). Preoperative inspiratory thoracic CT scans were available for 30 of the 34 patients because four patients with GOLD 1 or GOLD 2 COPD did not have preoperative scans. As we expected, patients with GOLD 2 and GOLD 4 COPD had a lower diffusing capacity (diffusing capacity of the lung for carbon monoxide adjusted for alveolar volume [DLCO:VA]) than controls (p<0·05; table 1). Total lung volume computed by CT was larger in patients with GOLD 4 COPD than in patients in all other groups (p<0·01; table 1). We obtained between six and 11 tissue samples per patient, given the varying sizes of lobes and lungs, which yielded 262 lung samples. Compared with control smokers with a mean of 4·7 TB/mL (SD 1·3), we found a decrease of 40% in
www.thelancet.com/respiratory Published online July 4, 2018 http://dx.doi.org/10.1016/S2213-2600(18)30196-6
Articles
Lung or lobe location
Sex
Whole lung volume* (L)
Terminal bronchioles per lung
Transitional bronchioles per lung
Alveolar surface area per lung (m²)
32 824
64·73 66·87
Control 1
Right middle lobe and right Female lower lobe†
2·39
13 481
2
Right upper lobe†
Female
2·54
16 163
33 711
3
Left upper lobe†
Male
3·14
12 848
32 121
94·20
4
Left whole lung
Male
2·20
9204
26 529
68·71
5
Left upper lobe†
Female
1·34
7898
15 232
37·45
6
Right whole lung
Female
2·87
10 566
23 114
58·10
7
Right whole lung
Female
1·56
6628
13 698
37·59
8
Right upper lobe†
Male
4·02
12 682
32 761
97·81
9
Left whole lung
Male
2·11
14 061
28 607
44·11
10
Left lower lobe†
Female
2·18
7382
14 763
53·84
Mean (SD)
··
··
2·43 (0·77)
25 336 (8122)
62·34 (21·07)
11
Right upper lobe†
Male
3·54
9744
15 428
86·32
12
Left lower lobe†
Male
2·13
12 424
20 992
51·98
13
Left upper lobe†
Female
2·19
5707
9218
30·16
14
Right whole lung
Male
4·11
4682
4682
82·92
16
Left upper lobe†
Male
2·71
10 710
21 420
69·35
18
Left whole lung
Female
2·27
6364
11 932
51·16
19
Right upper lobe†
Female
2·68
4551
8452
42·58
20
Right upper lobe†
Male
2·45
3462
11 541
58·59
Mean (SD)
··
··
2·76 (0·71)
7205 (3302)
12 958 (5954)
59·13 (19·41)
21
Left lower lobe†
Male
2·31
9753
14 630
49·09
24
Left upper lobe†
Male
2·78
5588
9500
54·05
25
Left upper lobe†
Female
2·19
7109
9297
37·29
26
Left whole lung
Female
1·86
3260
9781
44·43
27
Left whole lung
Male
2·06
3141
4398
53·26
28
Right upper lung
Female
2·29
4762
4082
27·63
Mean (SD)
··
··
2·25 (0·31)
5602 (2520)
8615 (3927)
44·29 (10·25)
29
Left whole lung
Male
4·30
4536
3402
41·25
30
Left whole lung
Female
3·10
3797
2531
27·82
31
Right whole lung
Male
4·64
12 897
11 725
36·13
32
Left whole lung
Male
3·30
1509
1509
33·62
33
Left whole lung
Male
3·95
3054
3054
29·52
34
Right whole lung
Male
4·68
10 436
Mean (SD)
··
··
3·99 (0·67)
11 091 (3225)
GOLD 1
GOLD 2
GOLD 4
6038 (4541)
1491
37·02
3952 (3888)
34·23 (4·98)
Mean (SD) are for 30 of the 34 patients who had a preoperative thoracic CT scan. GOLD=Global Initiative for Chronic Obstructive Lung Disease. *Whole lung volume denotes the right or left lung volume obtained from the clinical CT scans. †Samples were obtained within a lobe and not the whole lung, and therefore values calculated from these samples should only be taken as estimates of the whole lung.
Table 2: Estimates of small airway counts and alveolar surface area in whole lungs for each patient
patients with GOLD 1 COPD (2·8 TB/mL [1·4], p=0·014), 43% in patients with GOLD 2 COPD (2·7 TB/mL [1·1], p=0·036), and 68% in patients with GOLD 4 COPD (1·5 TB/mL [1·0], p=0·0001; figure 1A). The number of transitional bronchioles per mL of lung was reduced from a mean of 10·8 TrB/mL (SD 2·9) in the control smokers by 56% in patients with GOLD 1 COPD (4·8 TrB/mL [2·6], p=0·0001), 59% in patients with GOLD 2 COPD (4·4 TrB/mL [2·2], p=0·0001), and 90%
in patients with GOLD 4 COPD (1·1 TrB/mL [0·9], p=0·0001; figure 1B). Using these stereological counts, we estimated the total numbers of terminal and transitional bronchioles per whole lung volume determined from the thoracic CT scan (table 2). Controls had a mean of 11 091 terminal bronchioles per lung (SD 3225), and this number was reduced to 7205 (3302) in patients with GOLD 1 COPD, to 5602 (2520) in patients with GOLD 2 COPD, and to
www.thelancet.com/respiratory Published online July 4, 2018 http://dx.doi.org/10.1016/S2213-2600(18)30196-6
5
Articles
B
A
p=0·0002
p=0·0002 p=0·032
60 40 95th percentile of controls (20·3)
20 0 –20
Control (n=10)
GOLD 2 (n=6)
GOLD 1 (n=8)
p=0·024
80 Resected lung emphysema on clinical CT (% LAA < –910 HU)
Whole-lung emphysema on clinical CT (% LAA < –910 HU)
80
60 40 20 0 –20
GOLD 4 (n=6)
95th percentile of controls (7·4)
Control (n=10)
GOLD 1 (n=8)
p<0·0001
p<0·0001 p=0·027
1000
95th percentile of controls (449·1)
500
Control (n=10)
GOLD 1 (n=10)
GOLD 2 (n=8)
p=0·0021
150
GOLD 4 (n=6)
Alveolar surface area per mL of lung per person (cm2/mL)
1500
Mean linear intercept (µm)
GOLD 4 (n=6)
D
C
0
GOLD 2 (n=6)
p=0·019 p=0·0043
100
5th percentile of controls (41·3)
50
0
Control (n=10)
GOLD 1 (n=10)
GOLD 2 (n=8)
GOLD 4 (n=6)
Figure 2: Quantification of emphysema in mild to moderate chronic obstructive pulmonary disease Comparison of percentage low-attenuation area less than –910 HU (% LAA <–910 HU) on thoracic CT per patient grouped by Global Initiative for Chronic Obstructive Lung Disease (GOLD) stage in both the whole lung (A) and region of surgically resected lung lobe (B). The dashed line indicates the 95th percentile of controls for each graph. (C) Comparison of mean linear intercept measured by micro-CT per person grouped by GOLD stage. The dashed line denotes the upper 95th percentile of the mean linear intercept (449·1 µm) in the control group. (D) Comparison of alveolar surface area per mL of tissue measured by micro-CT per person grouped by GOLD stage. The dashed line denotes the lower 5th (41·3 cm²/mL) percentile of alveolar surface area per mL of tissue in the control group. Median and IQR per group are shown for non-parametric data (A, B, and C), and means and SD per group are shown for parametric data (D). LAA=low-attenuation area. HU=Hounsfield units.
6038 (4541) in patients with GOLD 4 COPD (table 2). For transitional bronchioles, this loss was even greater with an estimated number of 25 336 (8122) transitional bronchioles per lung in controls that was reduced to 958 (5954) in patients with GOLD 1 COPD, to 12 8615 (3927) in patients with GOLD 2 COPD, and to 3952 (3888) in patients with GOLD 4 COPD (table 2). We quantitatively analysed 30 preoperative CT scans on the segmented right or left lung (figure 2A) and the segmented resected lobes (figure 2B) for patients with a lobectomy sample. In both analyses, emphysema (%LAA <–910 HU, >95th percentile of controls) was only detectable in patients with GOLD 4 COPD compared with controls (p=0·0002) and patients with GOLD 1 COPD (p=0·032, figure 2A; p=0·024, figure 2B). Additionally, we found that the CT density measure ments on the 15 segmented lobes overlapped well with 6
the density measurements of the whole right lung CT scans (appendix p 6). When micro-CT was used to assess the degree of emphysema with mean linear intercept (figure 2C), we found patients with GOLD 4 COPD had an increase in mean linear intercept compared with controls (p<0·0001) and patients with GOLD 1 COPD (p=0·027). Importantly, when we assessed alveolar surface area per mL of lung available for gas exchange (figure 2D), we found a reduction in alveolar surface area per mL of lung in all COPD groups compared with controls, who had a mean alveolar surface area of 73·6 cm²/mL (SD 23·6) of lung. A 33% loss of alveolar surface area was seen in patients with GOLD 1 COPD (49·0 cm²/mL [17·9], p=0·019), a 45% loss was seen in patients with GOLD 2 COPD (40·3 cm²/mL [13·6], p=0·0021), and a 79% loss was seen in patients with GOLD 4 COPD (14·8 cm²/mL [4·3], p<0·0001; figure 2D).
www.thelancet.com/respiratory Published online July 4, 2018 http://dx.doi.org/10.1016/S2213-2600(18)30196-6
Articles
A
B Control GOLD 1 GOLD 2 GOLD 4
0·006
Control GOLD 1 GOLD 2 GOLD 4
0·06
Density
Density
0·04 0·004
0·02
0·002
0
0
500
1000
1500
0
2000
50
0
p=0·016 8
p=0·0006 20
p=0·039 Mean number of transitional bronchioles per mL of lung
Mean number of terminal bronchioles per mL of lung
150
D
C
6
4
2
0
100
Alveolar surface area per mL of lung (cm²/mL)
Mean linear intercept (μm)
Control (n=61)
GOLD 1 (n=37)
GOLD 2 (n=23)
All samples within 5th and 95th percentiles of control alveolar surface area per mL of lung
p=0·0016
15
10
5
0
Control (n=61)
GOLD 1 (n=37)
GOLD 2 (n=23)
All samples within 5th and 95th percentiles of control alveolar surface area per mL of lung
Figure 3: Association of emphysema with terminal and transitional bronchiole number in mild to moderate chronic obstructive pulmonary disease (A) Kernel density plot of mean linear intercept with micro-CT for all 262 samples, segregated by Global Initiative for Chronic Obstructive Lung Disease (GOLD) stage. The dashed lines denote the lower 5th (267·7 µm) and upper 95th (449·1 µm) percentiles of the mean linear intercept in the control group. A darker colour on the shaded bar represents increased mean linear intercept, indicating more emphysema. (B) Kernel density plot of alveolar surface area per mL of tissue measured by micro-CT for all 262 samples, segregated by GOLD stage. The dashed lines denote the lower 5th (41·3 cm²/mL) and upper 95th (126·2 cm²/mL) percentiles of alveolar surface area per mL of tissue in the control group. A darker colour on the shaded bar represents decreased alveolar surface area per mL of tissue, indicating more emphysema. Comparison of terminal (C) and transitional (D) bronchiolar number per mL of lung in all cores (n=121) with an alveolar surface area per mL of tissue within the 5th and 95th percentiles of controls, segregated by GOLD stage. We assessed data using a linear-mixed effect model, and the mean and SD per group are shown.
When total alveolar surface area was estimated for the whole lung (table 2), our data show that compared with controls (mean 62·34 m² [SD 21·07] ) the total alveolar surface area available for gas exchange was reduced to 59·13 m² (19·41) in patients with GOLD 1 COPD, 44·29 m² (10·25) in patients with GOLD 2 COPD, and 34·23 m² (4·98) in patients with GOLD 4 COPD. The alveolar surface area did not differ significantly between upper, lower, and whole lung samples obtained from micro-CT (appendix p 7). The kernel density plot (figure 3A) of all 262 lung samples shows that by use of the mean linear intercept, it was possible to distinguish emphysematous changes in GOLD 4, but not GOLD 1 and GOLD 2 samples,
compared with controls. When the same samples were plotted with alveolar surface area per mL of lung (figure 3B), GOLD 1, GOLD 2, and GOLD 4 samples segregated from control samples. To determine whether terminal and transitional bronchioles were lost in the absence of emphysematous destruction, we plotted the number of terminal and transitional bronchioles in the 121 of 262 cores that had an alveolar surface area per mL of lung within the normal range (41·3 cm²/mL for the 5th percentile to 126·2 cm²/mL for the 95th percentile) of control cases. In lung samples that had a normal alveolar surface area, a 29% reduction in the number of terminal bronchioles occurred in patients with GOLD 1 COPD (mean 3·3 TB/mL [SD 2·6], p=0·039),
www.thelancet.com/respiratory Published online July 4, 2018 http://dx.doi.org/10.1016/S2213-2600(18)30196-6
7
Articles
and a 40% reduction occurred in patients with GOLD 2 COPD (2·8 TB/mL [2·3], p=0·016) compared with controls (4·6 TB/mL [2·4]; figure 3C). A 41% reduction in the number of transitional bronchioles was observed
in patients with GOLD 1 COPD (mean 6·2 TrB/mL [SD 4·8], p=0·0016), and a 53% reduction in patients with GOLD 2 COPD (4·9 TrB/mL [3·7], p=0·0006) compared with controls (10·4 TrB/mL [5·4]; figure 3D). Two tissue
B
A p=0·0012 p=0·039 p=0·041
p=0·0023 p=0·083
80
60
40
20
4 3 2 1
C
GOLD 4 (n=6)
GOLD 2 (n=8)
D
Micro-CT as a scouting tool to identify airways of interest
Obstructed
5
0
GOLD 1 (n=10)
Thickened
6
0 Control (n=10)
Non-diseased
7
Terminal bronchioles per mL of lung
Diseased terminal bronchioles (%)
100
Control
GOLD 1
GOLD 2
GOLD 4
Non-disease terminal bronchiole
1
2
3
2
3
2
3
2
3
250 µm
1 2 3
1 250 µm
1 2 3 250 µm
E
F
Thickened terminal bronchiole
1
2
3
250 µm
1
250 µm
8
Obstructed terminal bronchiole
1
250 µm
2
3
1
250 µm
www.thelancet.com/respiratory Published online July 4, 2018 http://dx.doi.org/10.1016/S2213-2600(18)30196-6
Articles
samples from all GOLD 4 lungs had an alveolar surface area within the normal range and therefore no statistical analysis was done. We next assessed the morphology of the remaining terminal bronchioles. In control lungs, 12% of the terminal bronchioles (mean 1·0 TB/mL [SD 1·4]) were found to be diseased (thickened or obstructed) and that number increased to 41% in GOLD 1 lungs (1·5 TB/mL [1·5], p=0·0023), 37% in GOLD 2 lungs (1·1 TB/mL [1·5], p=0·083), and 77% in GOLD 4 lungs (1·1 TB/mL [0·7], p=0·0012; figure 4A). Moreover, the majority of diseased terminal bronchioles were thickened in patients with GOLD 1 or GOLD 2 COPD, and the majority of diseased terminal bronchioles were 100% obstructed in patients with GOLD 4 COPD (figure 4B; statistical comparisons are shown in appendix p 8). Figure 4C shows how the spatial coordinates from the volumetric micro-CT scans were used to accurately cut histological sections at regular intervals (dashed lines) along the airway branch length for patients with GOLD 1 or GOLD 2 COPD. Representative images are shown for a non-diseased (figure 4D), thickened (figure 4E), and obstructed (figure 4F) terminal bronchiole observed in patients with GOLD 1 and GOLD 2 COPD. Figure 4F further shows how the morphology of the terminal bronchioles can change substantially along its branch length and highlights the importance of volumetric micro-CT imaging to understand the disease pathology. Movat’s pentachrome staining showed that thickened airways quantified by micro-CT had airway wall fibrosis with collagen deposition and inflammatory cell infiltration. Furthermore, obstructed airways had thickened walls, and the obstructions were not simply mucus plugs but composed of collagen infiltrated with structural and inflammatory cells. Comparisons of the micro-CT measurements of terminal and transitional bronchiole number and alveolar surface area with the patients’ FEV1, FEV1/forced vital capacity (FVC), and DLCO:VA show that the lung function measurements used to classify the GOLD stages of COPD are significantly correlated with both the loss of terminal and transitional bronchioles and alveolar surface area in mild to moderate COPD (table 3; see also appendix p 9 for a detailed comparison). Figure 4: Terminal bronchiole pathology (A) Comparison of the percentage of diseased terminal bronchioles per total number of terminal bronchioles per person grouped by Global Initiative for Chronic Obstructive Lung Disease (GOLD) stage. Mean (SD) per group are shown. (B) Terminal bronchiolar number per mL of lung segregated by non-diseased, thickened, and complete luminal obstruction. Each bar indicates one lung donor. (C) Micro-CT was used as a scouting tool to precisely locate terminal bronchioles and enable efficient histological sectioning. The numbered dashed lines indicate the corresponding histological sections taken in a cross-section at locations 1, 2, and 3 for a representative terminal bronchiole (D–F). Matched micro-CT and histological sections of a representative non-diseased (D), thickened (E), and obstructed terminal bronchiole (F). Histological sections were stained with Movat’s pentachrome to highlight various components of connective tissue; elastic fibres (black), collagen and reticular fibres (yellow), and fibrin and muscle structures (red).
Discussion This study, to our knowledge, provides the first direct evidence that the smallest airways within the lung, conducting terminal bronchioles and respiratory transitional bronchioles, are significantly lost in the lungs of patients with mild (GOLD 1) and moderate (GOLD 2) COPD compared with age-matched smokers with normal lung function. Using a robust measurement of emphysema, alveolar surface area, which translates to the functional tissue involved in gas exchange, we also report that terminal and transitional bronchioles are lost in lung tissue in which no emphysematous destruction is present, indicating that small airways disease is an early pathological feature of mild and moderate COPD. Although several histological studies have documented airway remodelling and inflammation in the small airways of patients with COPD, to our knowledge no study to date has been able to quantify the numbers of terminal and transitional bronchioles and show that loss of these small airways correlates with decreases in lung function in mild and moderate COPD. Furthermore, using the combination of micro-CT and histology, we show that the remaining small airways that have been extensively studied in COPD23,26,27 do indeed have thickened walls and, when examined along their branch lengths, are often completely obstructed by fibrotic tissue. This study highlights the small airways disease that occurs in patients with mild and moderate COPD, which could not be measured previously, and highlights the potential importance of early intervention for disease modification. Our results support and substantially extend an earlier report that used micro-CT10 in patients with very severe (GOLD 4) COPD by providing new evidence that the destruction of terminal, and now also transitional bronchioles, is well established in people with mild or moderate COPD. Furthermore, using a stereological sampling design, we show that small airways disease is an early pathological feature of mild and moderate COPD and is present in regions of the lung that have no emphysematous disease. These data strongly support the concept that the small airways are the major and earliest site of airflow limitation in COPD.4–6 In 1970, Mead28 proposed that the small airways might represent a socalled quiet zone within the lung, where disease could accumulate over many years without being noticed. In support of this hypothesis, our data show that even in mild (GOLD 1) COPD, classified by an FEV1/FVC of less than 0·7 and an FEV1 of 80% or more predicted, a patient would potentially have lost, on average, 40% of their terminal bronchioles and 56% of their transitional bronchioles. The reason that this substantial loss of small airways can occur with a small change in lung function can be explained by the extensive parallel arrangement of the small airways within the lung. Direct measurements of small airway resistance in both post-mortem human lungs4 and living humans6
www.thelancet.com/respiratory Published online July 4, 2018 http://dx.doi.org/10.1016/S2213-2600(18)30196-6
9
Articles
Alveolar surface area per mL of lung
Terminal bronchioles per mL of lung
Transitional bronchioles per mL of lung
All patients
All patients
All patients
Mild-to-moderate COPD
Mild-to-moderate COPD
Mild-to-moderate COPD
FEV1 (% predicted)
0·715 (p<0·0001)
0·502 (p=0·0078)
0·523 (p=0·0018)
0·326 (p=0·0988)
0·640 (p=0·0001)
0·423 (p=0·0301)
FEV1/FVC (%)
0·722 (p<0·0001)
0·547 (p=0·0045)
0·581 (p=0·0008)
0·465 (p=0·0189)
0·717 (p<0·0001)
0·627 (p=0·0009)
DLCO:VA (mL/min per mm Hg per L)
0·610 (p<0·001)
0·549 (p=0·0045)
0·591 (p=0·0011)
0·567 (p=0·0050)
0·715 (p<0·0001)
0·676 (p=0·0003)
Pearson’s correlations were calculated to investigate the association between micro-CT measures and pulmonary function. Data are r value (false discovery rate p value). COPD=chronic obstructive pulmonary disease. FVC=forced vital capacity. DLCO:VA=diffusing capacity of the lung for carbon monoxide adjusted for alveolar volume.
Table 3: Correlation of small airways disease and emphysema with lung function
have shown that the small airways account for a small proportion of the total resistance to airflow in the normal human lung but become the major site of increased resistance in COPD. Because the small airways are arranged in parallel, a 50% reduction in their number is expected to double their resistance. Previous direct measurements of peripheral airway resistance in living humans show that doubling the mean value of 0·70 cm H2O/L per s [SD 0·26] measured in individuals with normal lung function to 1·4 cm H2O/L per s falls well below the 2·78–4·59 cm H2O/L per s measured in patients with severe COPD.6 Although these findings are consistent with the much greater reduction in terminal and transitional airways reported here and previously for patients with end-stage GOLD 4 COPD,10 we do not mean to imply that these are the only airways narrowed and lost in COPD. By use of thoracic CT scans, several previous studies have shown that airways ranging from 0·8 mm to 3 mm in diameter are also destroyed in all GOLD categories of COPD.7,9,10 The relative importance of small airways disease and emphysema for lung function decline has been debated greatly. In this study, we show direct evidence that small airways disease is a well established pathological feature of patients with mild and moderate COPD. Furthermore, these data also show that the reduction in both terminal and transitional bronchioles is well established before emphysematous destruction of the alveolar surface area can be visualised with the spatial resolution (1000 μm) provided by thoracic CT scans. This finding strongly suggests that the appearance of emphysema on thoracic CT scans seems to be a good predictor of a rapid decline in FEV121,29,30 because a substantial number of terminal and transitional bronchioles has already been lost before the emphysematous disease can be detected. Therefore, approaches to identify susceptible smokers before they reach the GOLD 1 classification7,31,32 warrant further attention. Additionally, these new data substantially extend earlier reports on the pathology present in the remaining small airways by showing that the remaining airways are thickened and become more obstructed with disease progression. This finding is consistent with the concept that airways develop thickened walls and narrowed lumens with COPD progression.23 What 10
remains unclear is whether thickening, obstruction, and obliteration occur sequentially in individual airways. In several reviews, we have postulated that terminal and transitional bronchioles might be vulnerable to deposition of fine particulate matter (especially from tobacco smoke) because they are located in the region of the lung in which the transition from bulk flow to diffusion of gas occurs.33 The answer to these questions will require much further, but necessary, work. These data do, however, suggest that small airways disease is related to the pathology of the airways themselves, and that loss of elastic recoil due to emphysematous destruction is an independent phenomenon that potentially causes further airway obstruction. It has been postulated that destruction of the axial elastic network might allow the airways to shorten and obliterate the airway lumen.34 Our current working hypothesis is that the disappearance of small airways is preceded by a localised constrictive bronchiolitis that first obliterates the airway lumen and then separates the affected airway into two parts that leaves closed end buds coming off both the parent airway and the distal daughter branch. Thus, future studies designed to answer this question with the combination of micro-CT and histology are still required to definitively understand the nature of this destructive disease process. This study has limitations that should be mentioned. We note that this study used archived lungs and lobes from surgeries done in the early 2000s for lung cancer resection of small peripheral tumours with no metastasis or lung transplantation, which excluded patients with severe GOLD 3. Therefore, because the study is crosssectional, it does not allow us to observe the progress of the disease pathology over time, or to determine whether the patients with mild or moderate COPD might have been born with a lower number of airways. However, only lobectomy and pneumonectomy surgeries provide the possibility to collect intact lung tissue for inflation, preoperative pulmonary function, and CT data from smokers with and without mild or moderate COPD. Furthermore, patients had an inspiratory CT scan for surgical planning but not for COPD assessment, which now use inspiratory and expiratory CT scans. Thus, we were unable to assess the association between small
www.thelancet.com/respiratory Published online July 4, 2018 http://dx.doi.org/10.1016/S2213-2600(18)30196-6
Articles
airways disease and gas trapping in this study. We also acknowledge that for 15 (44%) of the 34 individuals, samples were obtained only from a single lobe, which did not allow us to capture disease heterogeneity within the whole lung. However, when we compared the clinical CT-based density histograms of the whole lung with the resected lung region or alveolar surface area obtained from micro-CT measures, we found no significant variation between upper and lower lung samples. Despite these limitations, the numbers of terminal and transitional bronchioles and alveolar surface area reported for the control smokers with normal lung function in this study are consistent with those previously reported in stereological studies of the normal human lung.35,36 In conclusion, these findings suggest that, like the kidney in which a substantial proportion of nephrons are lost before renal dysfunction appears,37 the development of airflow limitation in COPD involves progressive destruction and loss of the terminal and transitional bronchioles before a decline in lung function is observed. Additionally, these data also indicate that destruction of terminal and transitional bronchioles can occur in the absence of emphysematous destruction in mild and moderate COPD and that the remaining airways have narrowed lumens and thickened walls. Most importantly, these data suggest that several large clinical trials investigating COPD treatments in severe COPD did not show beneficial treatment effects because they were initiated after a substantial number of terminal and transitional bronchioles were already lost. Therefore, future clinical trials need to be done in cohorts of patients with mild and moderate COPD if we are to have any chance to modify disease outcomes. Contributors H-KK, DMV, SB, AH, OLK, NF, WME, MK, PL, IS, JAW, JDC, HOC, and T-LH contributed to data collection. H-KK, DMV, SB, HOC, PDP, JCH, and T-LH contributed to study design, data interpretation, figures, and writing. Declaration of interests DMV is supported by Canadian Thoracic Society and Alpha-1 Foundation fellowships. MK is supported by the Canadian Institutes of Health Research (CIHR) Banting fellowship. T-LH is supported by the CIHR, Michael Smith Health Research Foundation, Parker B Francis, and Providence Health Care Research Institute New Investigator awards. HOC is supported by Spiration Inc and has personal fees from Samsung. All other authors declare no competing interests. Acknowledgments This study was funded by a Canadian Institutes of Health Research operating grant MOP 130504. The authors thank the James Hogg Lung Registry (Darren Sutherland), Histology Core (Amrit Samra), and the Bio-Imaging Core (Aaron M Barlow) from the Centre for Heart Lung Innovation, St Paul’s Hospital, Vancouver, BC, Canada, and the µ-VIS X-Ray Imaging Centre at the University of Southampton, Southampton, UK. IS is funded by Nikon and EPSRC grant EPH01506X. References 1 Weibel ER. Morphometry of the human lung. New York, NY: Academic, 1963. 2 Green M. How big are the bronchioles? St Thomas Hosp Gaz 1965; 63: 136–39.
3 4 5
6 7
8 9 10 11
12 13 14 15 16
17 18 19 20 21 22 23 24
25
Macklem PT, Mead J. Resistance of central and peripheral airways measured by a retrograde catheter. J Appl Physiol 1967; 22: 395–401. Hogg JC, Macklem PT, Thurlbeck WM. Site and nature of airway obstruction in chronic obstructive lung disease. N Engl J Med 1968; 278: 1355–60. Van Brabandt H, Cauberghs M, Verbeken E, Moerman P, Lauweryns JM, Van de Woestijne KP. Partitioning of pulmonary impedance in excised human and canine lungs. J Appl Physiol Respir Environ Exerc Physiol 1983; 55: 1733–42. Yanai M, Sekizawa K, Ohrui T, Sasaki H, Takishima T. Site of airway obstruction in pulmonary disease: direct measurement of intrabronchial pressure. J Appl Physiol 1992; 72: 1016–23. Kirby M, Tanabe N, Tan WC, et al. Total airway count on computed tomography and the risk of chronic obstructive pulmonary disease progression. Findings from a population-based study. Am J Respir Crit Care Med 2018; 197: 56–65. Hasegawa M, Nasuhara Y, Onodera Y, et al. Airflow limitation and airway dimensions in chronic obstructive pulmonary disease. Am J Respir Crit Care Med 2006; 173: 1309–15. Kurashima K, Hoshi T, Takaku Y, et al. Changes in the airway lumen and surrounding parenchyma in chronic obstructive pulmonary disease. Int J Chron Obstruct Pulmon Dis 2013; 8: 523–32. McDonough JE, Yuan R, Suzuki M, et al. Small-airway obstruction and emphysema in chronic obstructive pulmonary disease. N Engl J Med 2011; 365: 1567–75. Pauwels RA, Buist AS, Calverley PM, Jenkins CR, Hurd SS, GOLD Scientific Committee. Global strategy for the diagnosis, management, and prevention of chronic obstructive pulmonary disease. NHLBI/WHO Global Initiative for Chronic Obstructive Lung Disease (GOLD) Workshop summary. Am J Respir Crit Care Med 2001; 163: 1256–76. Rodriguez M, Bur S, Favre A, Weibel ER. Pulmonary acinus: geometry and morphometry of the peripheral airway system in rat and rabbit. Am J Anat 1987; 180: 143–55. Haefeli-Bleuer B, Weibel ER. Morphometry of the human pulmonary acinus. Anat Rec 1988; 220: 401–14. Muhlfeld C, Hegermann J, Wrede C, Ochs M. A review of recent developments and applications of morphometry/stereology in lung research. Am J Physiol Lung Cell Mol Physiol 2015; 309: L526–36. Fehrenbach H. Animal models of pulmonary emphysema: a stereologist’s perspective. Eur Respir Rev 2006; 15: 136–47. Knudsen L, Weibel ER, Gundersen HJ, Weinstein FV, Ochs M. Assessment of air space size characteristics by intercept (chord) measurement: an accurate and efficient stereological approach. J Appl Physiol 2010; 108: 412–21. Ochs M. Estimating structural alterations in animal models of lung emphysema. Is there a gold standard? Ann Anat 2014; 196: 26–33. Knust J, Ochs M, Gundersen HJ, Nyengaard JR. Stereological estimates of alveolar number and size and capillary length and surface area in mice lungs. Anat Rec 2009; 292: 113–22. Scott AE, Vasilescu DM, Seal KA, et al. Three dimensional imaging of paraffin embedded human lung tissue samples by micro-computed tomography. PLoS One 2015; 10: e0126230. Sieren JP, Newell JD Jr, Barr RG, et al. SPIROMICS protocol for multicenter quantitative computed tomography to phenotype the lungs. Am J Respir Crit Care Med 2016; 194: 794–806. Yuan R, Hogg JC, Pare PD, et al. Prediction of the rate of decline in FEV1 in smokers using quantitative computed tomography. Thorax 2009; 64: 944–49. Muller NL, Staples CA, Miller RR, Abboud RT. “Density mask”. An objective method to quantitate emphysema using computed tomography. Chest 1988; 94: 782–87. Hogg JC, Chu F, Utokaparch S, et al. The nature of small-airway obstruction in chronic obstructive pulmonary disease. N Engl J Med 2004; 350: 2645–53. Hsia CC, Hyde DM, Ochs M, et al. An official research policy statement of the American Thoracic Society/European Respiratory Society: standards for quantitative assessment of lung structure. Am J Respir Crit Care Med 2010; 181: 394–418. Vasilescu DM, Klinge C, Knudsen L, et al. Stereological assessment of mouse lung parenchyma via nondestructive, multiscale micro-CT imaging validated by light microscopic histology. J Appl Physiol 2013; 114: 716–24.
www.thelancet.com/respiratory Published online July 4, 2018 http://dx.doi.org/10.1016/S2213-2600(18)30196-6
11
Articles
26 Di Stefano A, Turato G, Maestrelli P, et al. Airflow limitation in chronic bronchitis is associated with T-lymphocyte and macrophage infiltration of the bronchial mucosa. Am J Respir Crit Care Med 1996; 153: 629–32. 27 O’Shaughnessy TC, Ansari TW, Barnes NC, Jeffery PK. Inflammation in bronchial biopsies of subjects with chronic bronchitis: inverse relationship of CD8+ T lymphocytes with FEV1. Am J Respir Crit Care Med 1997; 155: 852–57. 28 Mead J. The lung’s “quiet zone”. N Engl J Med 1970; 282: 1318–19. 29 Nishimura M, Makita H, Nagai K, et al. Annual change in pulmonary function and clinical phenotype in chronic obstructive pulmonary disease. Am J Respir Crit Care Med 2012; 185: 44–52. 30 Mohamed Hoesein FA, de Hoop B, Zanen P, et al. CT-quantified emphysema in male heavy smokers: association with lung function decline. Thorax 2011; 66: 782–87. 31 Allinson JP, Hardy R, Donaldson GC, Shaheen SO, Kuh D, Wedzicha JA. The presence of chronic mucus hypersecretion across adult life in relation to chronic obstructive pulmonary disease development. Am J Respir Crit Care Med 2016; 193: 662–72.
12
32 de Marco R, Accordini S, Cerveri I, et al. Incidence of chronic obstructive pulmonary disease in a cohort of young adults according to the presence of chronic cough and phlegm. Am J Respir Crit Care Med 2007; 175: 32–39. 33 Hogg JC, Pare PD, Hackett TL. The contribution of small airway obstruction to the pathogenesis of chronic obstructive pulmonary disease. Physiol Rev 2017; 97: 529–52. 34 Mitzner W. Emphysema—a disease of small airways or lung parenchyma? N Engl J Med 2011; 365: 1637–39. 35 Weibel ER. Morphometric analysis of the number, volume and surface of the alveoli and capillaries of the human lung. Z Zellforsch Mikrosk Anat 1962; 57: 648–66 (in German). 36 Wiebe BM, Laursen H. Lung morphometry by unbiased methods in emphysema: bronchial and blood vessel volume, alveolar surface area and capillary length. APMIS 1998; 106: 651–56. 37 Remuzzi G, Bertani T. Pathophysiology of progressive nephropathies. N Engl J Med 1998; 339: 1448–56.
www.thelancet.com/respiratory Published online July 4, 2018 http://dx.doi.org/10.1016/S2213-2600(18)30196-6