Introduction

Chronic obstructive pulmonary disease (COPD) is a progressive disease primarily of the lungs, and is characterised by persistent respiratory symptoms and airflow limitation due to airway and alveolar abnormalities. COPD is generally attributed to exposure to tobacco and/or harmful air particles [1]. Changes to the small airways in COPD, such as obstructive bronchiolitis and parenchymal destruction (emphysema), can occur alone or in combination, and the relative contributions and severity of each vary between individuals [1].

In the era of precision medicine, phenotyping patients with COPD based on one or more disease characteristics could improve our understanding of pathogenesis and disease progression, as patients do not generally all follow the same trajectory in their disease. As a result, this may allow for more targeted, individualised, and hence efficacious pharmacological and non-pharmacological management. Prospective longitudinal studies can help to assess whether soluble biomarkers (e.g., from whole blood, serum, plasma or sputum), imaging biomarkers and clinical parameters, or combinations therein, may serve as predictors of disease progression in COPD. Several studies in COPD have either completed or are underway to help address this question, including COPDGene [2], ECLIPSE [3], SPIROMICS [4, 5], COSYCONET [6], CanCOLD [7] and the British Lung Foundation early COPD study [8].

FOOTPRINTS® (ClinicalTrials.gov: NCT02719184) is a 3-year study aiming to assess the association between soluble biomarkers of inflammation/lung tissue destruction, and imaging and physiological biomarkers, with COPD progression in ex-smokers with mild-to-severe COPD (as per the Global Initiative for Chronic Obstructive Lung Disease [GOLD] classification) [9]. In addition, this study includes a subgroup of 19 subjects with COPD and alpha-1-antitrypsin deficiency (A1ATD), a genetic condition associated with a higher risk of developing emphysema [10]. In A1ATD, emphysema occurs at a younger age and progresses faster as compared with COPD patients without A1ATD [11, 12].

The data from FOOTPRINTS® may allow identification of different COPD phenotypes, the biomarkers that help differentiate those phenotypes, and their associated risks for emphysema progression. More detailed information on the methodology of the FOOTPRINTS® study has been published previously [13]. Here, baseline demographic and clinical characteristics (lung function, soluble biomarkers, functional imaging findings and respiratory symptoms) for the study subjects are reported, as well as a descriptive statistical comparison of soluble biomarker levels in the individual GOLD groups and A1ATD group versus ex-smokers without airflow limitation (EwAL).

Methods

Study design

Details of the methodology for the FOOTPRINTS® study (NCT02719184) have been published previously [13]. In brief, the FOOTPRINTS® study is a multinational, prospective, longitudinal, observational biomarker study.

The study population included subjects with a range of COPD severities (GOLD stage 1: forced expiratory volume in 1 s [FEV1] ≥ 80% predicted; GOLD stage 2: FEV1 ≥ 50– < 80% predicted; and GOLD stage 3: FEV1 ≥ 30– < 50% predicted), subjects with COPD and A1ATD, and an EwAL control group. All subjects in this study should have stopped smoking at least 9 months prior to inclusion [9]. The EwAL control group was included to provide a comparison for biomarkers between subjects with and without airflow limitation while controlling for the potential effects of previous smoking.

Inclusion and exclusion criteria

Full details on the inclusion and exclusion criteria for participation in the FOOTPRINTS® study have been published previously [13]. Post-bronchodilator FEV1/forced vital capacity [FVC] < 70% was required for inclusion in the COPD and A1ATD groups, and post-bronchodilator FEV1/FVC ≥ lower limit of normal (as defined by the GOLD 2015 strategy report) was required for inclusion in the EwAL group. Subjects that were heterozygous or homozygous for the A1AT Z or S alleles were not included in the COPD GOLD groups 1–3 and EwAL group. Subjects with A1ATD were required to have a documented (prior to Visit 2) A1ATD of ZZ genotype. Patients in the COPD and A1ATD groups who were currently receiving or planned to receive A1AT augmentation therapy were excluded.

Baseline parameters of interest

At study entry, data were collected for demographic and disease characteristics, history of comorbidities and COPD exacerbations, and soluble biomarkers in various biofluids (whole blood, serum, plasma and induced sputum). History of comorbidities or COPD exacerbations included an event/disease that had occurred at least once during the lifetime of a subject.

Data on the 6-min walk test (6MWT) were collected for all subjects and symptom questionnaires were conducted for subjects in the COPD and A1ATD groups. Pulmonary symptoms were assessed using the following questionnaires: the COPD Assessment Test (CAT), the modified Medical Research Council (mMRC) dyspnoea score, and St. George’s Respiratory Questionnaire (SGRQ). Exercise capacity was assessed via the 6MWT. Baseline quantitative and qualitative X-ray computed tomography (CT) data were collected for all subjects.

Biomarker characterisation

Biomarkers were analysed in several biofluids at study entry and are being assessed up to Week 156. The serum/plasma biomarkers collected include, but are not limited to, neutrophil elastase-specific elastin fragment, other protease-generated neoepitopes such as cathepsin S-cleaved decorin, soluble receptor for advanced glycation end product (sRAGE), surfactant protein D (SP-D), lysyl oxidase-like 2, interleukin-6 (IL-6), high-sensitivity C-reactive protein (hs-CRP), white blood cell (WBC) count and fibrinogen.

Soluble biomarkers from sputum included biomarkers for the neutrophil-derived serine proteases neutrophil elastase, cathepsin G and proteinase 3 activity. These are not reported in this publication.

Quantitative and qualitative CT

Chest CT scans were performed to assess lung disease, including emphysema, air trapping, small airway disease and airway wall thickness. At four time points (Week 0 [i.e., Day 1], Week 52, Week 104 and Week 156), two different scans were performed: (1) a low-dose inspiratory chest CT scan with dose modulation to assess airway walls and emphysema, and (2) an expiratory scan to assess air trapping. The inspiratory CT technique was similar to the reduced dose technique developed for Phase 3 of COPDGene [14]. To optimise assessment of non-reversible air trapping, the chest CT scan was conducted between 1 and 4 h after salbutamol (albuterol) administration.

The Fleischner criteria for CT imaging in COPD were used for visual characterisation of centrilobular and paraseptal emphysema and of airway disease [15]. Quantitative analysis of lung images was performed centrally using LungQ (Thirona B.v., Nijmegen, the Netherlands). Assessments included inspiratory and expiratory volume, emphysema, air trapping and the square root of the wall area of bronchus with an internal perimeter of 10 mm. For emphysema, analyses included lung density at the 15th percentile point of the CT histogram (PD15), adjusted lung density (ALD) and percentage of lung voxels with attenuation ≤ –950 Hounsfield Units in inspiration (LAA-950) [13]. For air trapping, the analyses included percentage of lung voxels with attenuation ≤ –856 Hounsfield Units (LAA-856) not including parametric response mapping, mean lung density during expiration and the ratio of expiratory and inspiratory volume (function residual capacity [FRC]/total lung capacity [TLC] ratio) [13].

Statistical methods

All baseline data are presented descriptively for each of the subject subgroups, either as mean and standard deviation (SD) for normally distributed variables, median and interquartile range (IQR) for variables deviating from normality, or as number and percentage for categorical variables. To evaluate whether the continuous baseline variables showed a trend across the subject subgroups ordered by disease severity, linear regression (for normally distributed variables) or the Jonckheere-Terpstra test (for non-parametric comparisons) were conducted. For categorical variables with ≥ 2 categories, a Cochran–Armitage test for trend or a Cochran–Mantel–Haenszel test was applied, respectively. If a trend was not established, a Chi-squared test was performed to investigate associations. All soluble biomarkers were log2-transformed. Due to a high number of values below the limit of quantification, IL-6 was analysed as a categorical variable, with a lower limit of quantification equal to 3.12 ng/L used as a cut-off for categorisation. All analyses were hypothesis-generating and no correction for multiple testing was performed. Analyses were conducted using SAS v9.4.

Results

Subject disposition

In total, 463 subjects were enrolled into the FOOTPRINTS® study (Fig. 1). Of these, 123 subjects were categorised as GOLD stage 1, 130 as GOLD stage 2, and 129 as GOLD stage 3. Nineteen subjects had COPD and homozygous ZZ A1ATD and 62 were EwAL. Of the 19 subjects in the A1ATD group, 1 subject (5.3%) met the criteria for GOLD stage 1, 5 (26.3%) for GOLD stage 2, 12 (63.2%) for GOLD stage 3, and 1 (5.3%) for GOLD stage 4.

Fig. 1
figure 1

CONSORT flow diagram. A1ATD alpha-1-antitrypsin deficiency, GOLD Global Initiative for Chronic Obstructive Lung Disease

Full size image

All demographics and analyses are based on the per-protocol set, which comprised 456 subjects (Table 1). Seven subjects were excluded from the per-protocol set due to newly added anti-inflammatory treatment or change to therapy prior to Visit 1, presenting with an immunocompromising condition, significant pulmonary disease other than COPD, or A1ATD subjects with a genotype other than ZZ.

Table 1 Subject demographics
Full size table

Baseline characteristics

Selected baseline characteristics are shown in Table 1. The majority of subjects were male (n = 294, 64.5%) and the mean (± SD) age was 60.7 (6.9) years. Mean (± SD) body mass index was 26.6 kg/m2 (3.8) and mean (± SD) smoking pack history was 43.1 pack-years (19.5). Of note, a higher proportion of subjects in the A1ATD group were male (84.2%) compared with subjects in the GOLD 1–3 groups (59.8–69.7%) and EwAL group (62.3%). The mean age was lower in the A1ATD group (52.8 ± 9.0) and EwAL group (53.9 ± 6.8) than in the GOLD 1–3 groups (61.8 ± 5.4–62.9 ± 5.6). Smoking pack history was also lower in the A1ATD group (25.2 ± 12.3) and EwAL (32.9 ± 14.6) than in the GOLD 1─3 groups (41.5 ± 17.6–49.2 ± 21.6).

As disease severity increased, the proportion of subjects with multiple comorbidities increased (Table 1); 27.0% of GOLD 1 subjects had ≥ 3 comorbidities compared with 39.4% of GOLD 2 and 40.2% of GOLD 3 subjects. The most common comorbidities (> 10% prevalence) in GOLD 1, 2 and 3 subjects were, respectively, systemic arterial hypertension (36.1%, 40.9%, 49.6%), hyper-/dyslipidaemia (20.5%, 31.5%, 29.1%), pneumonia (4.9%, 19.7%, 22.8%), gastroesophageal reflux disease (14.8%, 15.7%, 15.7%), depression (9.0%, 17.3%, 15.7%), osteoarticular disorders (10.7%, 18.1%, 14.2%), diabetes (13.1%, 13.4%, 11.0%), anxiety (6.6%, 14.2%, 14.2%) and obesity (4.9%, 12.6%, 12.6%). Further details are given in the supplementary material (Additional file 1). Of the subjects in the A1ATD group, 21.1% had ≥ 3 comorbidities, but, as noted above, this was a younger population, with subjects being on average 10 years younger.

The percentage of prior COPD exacerbations (defined as an exacerbation that had occurred at least once during the lifetime of a subject) increased with increasing disease severity (as per GOLD classification): 24.6%, 37.8% and 55.9% of GOLD 1, GOLD 2 and GOLD 3 subjects, respectively, had experienced a previous COPD exacerbation. Overall, 40.8% of COPD subjects had a documented COPD exacerbation in their medical history.

Further information on baseline characteristics are available in Additional file 1.

Symptom questionnaires and 6MWT

Symptom questionnaires were conducted for subjects in the GOLD 1–3 and A1ATD groups, but not for the EwAL group. As shown in Table 2, increasing disease severity was associated with an increase in symptoms. Symptom scores in the A1ATD group were intermediate between those of the GOLD 1 and 3 groups. Increasing disease severity was associated with increases in the mean CAT score (GOLD 1: 11.2 [SD 5.7]; GOLD 2: 14.7 [SD 6.2]; GOLD 3: 16.6 [SD 7.2]; A1ATD: 13.4 [SD 7.4], P < 0.0001), the proportion of subjects with a CAT score ≥ 10 (GOLD 1: 62.3%; GOLD 2: 78.0%; GOLD 3: 79.5%; A1ATD: 63.2%, P = 0.0335), the proportion of subjects with an mMRC dyspnoea score ≥ 2 (GOLD 1: 11.5%; GOLD 2: 32.3%; GOLD 3: 60.6%; A1ATD: 47.4%, P < 0.0001) and the proportion of subjects with an SGRQ score ≥ 25 (GOLD 1: 50.8%; GOLD 2: 69.3%; GOLD 3: 83.5%; A1ATD: 73.7%, P < 0.0001).

Table 2 Symptom questionnaires and 6-min walk test
Full size table

Increasing disease severity was associated with a reduction in distance covered in the 6MWT (GOLD 1: 489.3 m [SD 91.7 m]; GOLD 2: 458.0 m [SD 101.3 m]; GOLD 3: 403.8 m [SD 121.7 m], P < 0.0001). The 6MWT distance for subjects with A1ATD (508.6 m [SD 114.4 m]) was numerically higher than in any of the GOLD groups, but lower than in the EwAL group (573.1 m [SD 100.3 m]).

Lung function and lung volume

Airflow limitation

At baseline, patients with the greatest disease severity (as defined by the GOLD 1–4 classification system) had the poorest lung function (Table 3; Fig. 2). Patients with COPD and A1ATD also had worse lung function than EwAL. Lung function parameters for the A1ATD group were similar to the GOLD stage 2/3 groups, with the exception of FVC.

Table 3 Post-bronchodilator lung function and lung volume measurements
Full size table
Fig. 2
figure 2

Lung function parameters. A1ATD alpha-1-antitrypsin deficiency, FEV1 forced expiratory volume in 1 s, FVC forced vital capacity, GOLD Global Initiative for Chronic Obstructive Lung Disease

Full size image

Across GOLD groups 1–3, mean FEV1 was highest in GOLD 1 (2.66 L [SD 0.54 L]) and lowest in GOLD 3 subjects (1.17 L [SD 0.29 L]). Mean FEV1 was lower in GOLD groups 1–3 and subjects with A1ATD (1.85 L [SD 0.85 L]) than in EwAL (3.48 L [SD 0.77 L]). Mean percent predicted FEV1 was 87.4% [SD 9.4%] in GOLD 1 subjects and decreased with increasing disease severity to 39.6% [SD 7.6%] in GOLD 3 subjects (P < 0.0001). The mean percent predicted FEV1 was 50.4% (SD 20.3%) in subjects with A1ATD and 102.1% (SD 12.2%) in the EwAL group. Mean FVC also decreased with increasing disease severity (GOLD 1: 4.22 L [SD 0.85 L]; GOLD 2: 3.66 L [SD 0.90 L]; GOLD 3: 3.16 L [SD 0.89 L], P < 0.0001), and, overall, was highest in subjects with A1ATD (4.54 L [SD 1.38 L]. The EwAL control group had a mean FVC of 4.42 L (SD 0.97 L). The mean FEV1/FVC ratio was 63.4% (SD 5.4%) in GOLD 1 subjects and reduced with increasing disease severity to 38.2% (SD 9.0%) in GOLD 3 subjects (P < 0.0001). Subjects with A1ATD had a mean FEV1/FVC ratio of 40.6% (SD 13.4%) and the EwAL control group FEV1/FVC ratio was 78.8% (SD 5.7%).

Plethysmography

Table 3 shows post-bronchodilator lung function and lung volume measurements at baseline. Increasing disease severity was associated with increasing mean residual volume (GOLD 1: 2.51 L [SD 0.66 L]; GOLD 2: 2.91 L [SD 0.96 L]; GOLD 3: 3.64 L [SD 1.02 L], P < 0.0001). Residual volume was high in subjects with A1ATD (3.63 L [SD 0.81 L]) and comparable with GOLD 3 COPD subjects, whereas it was lowest in EwAL (2.27 L [SD 0.54 L]).

TLC values generally increased with increasing disease severity (GOLD 1: 6.70 [SD 1.29]; GOLD 2: 6.60 [SD 1.44]; GOLD 3: 7.02 [1.58], P = 0.0217). The highest mean TLC was observed in the A1ATD group (8.16 L [SD 1.53 L]), whereas TLC for EwAL was 6.81 L (SD 1.27 L).

Additional lung function tests

Increasing disease severity and A1ATD were associated with reduced diffusing capacity of the lung for carbon monoxide (DLCO), (P < 0.0001) (Table 3). The mean percent predicted DLCO was 79.8% (SD 21.3%) in GOLD 1 subjects, 63.4% (SD 18.1%) in GOLD 2 subjects, 51.5% (SD 17.5%) in GOLD 3, subjects and 51.1% (SD 12.7%) in subjects with A1ATD.

The mean DLCO/alveolar volume ratio was highest in GOLD 1 subjects (2.25 mmol/min/kPa [SD 1.36 mmol/min/kPa]) and decreased with disease severity (GOLD 3: 1.61 mmol/min/kPa [SD 1.20 mmol/min/kPa]). Overall, mean DLCO/alveolar volume was lowest in the A1ATD group (0.83 mmol/min/kPa [SD 0.19 mmol/min/kPa] vs 1.88 mmol/min/kPa [SD 1.20 mmol/min/kPa] in EwAL). Further DLCO parameters are available in Additional file 1.

Soluble biomarkers

For the majority of biomarkers presented in Fig. 3, the mean expression levels or concentration increased with disease severity. A linear increase in the level of WBC (P < 0.0001), fibrinogen (P = 0.0249), hsCRP (P = 0.0002) and plasma SP-D (P = 0.0194) was observed whereas sRAGE levels linearly decreased with disease severity (P < 0.0001), indicative of more severe emphysema. In all GOLD groups (GOLD 1–3), levels of WBC, fibrinogen, hs-CRP and plasma SP-D were numerically higher compared with the EwAL group. Further soluble biomarker parameters are available in Additional file 1.

Fig. 3
figure 3

Distribution of inflammatory biomarkers at baseline by diagnosis group: a WBC, b fibrinogen, c hs-CRP, d IL-6, e sRAGE, f plasma SP-D. All biomarker values (except for IL-6) are log2 transformed. WBC, sRAGE: P ≤ 0.0001; hs-CRP: P = 0.0002; plasma SP-D: P = 0.0194; fibrinogen: P = 0.0249, linear regression for subject groups ordered by disease severity (Kruskal–Wallis test used for fibrinogen). No association was observed between positive expression of IL-6 and subject groups (chi-square test, P = 0.0701). A1ATD alpha-1-antitrypsin deficiency, Abs absolute, EwAL ex-smokers without airflow limitation, GOLD Global Initiative for Chronic Obstructive Lung Disease, hs-CRP high-sensitivity C-reactive protein, IL-6 interleukin 6, SP-D surfactant protein D, sRAGE soluble receptor for advanced glycation end product, WBC white blood count

Full size image

No association was detected between IL-6 positive expression and disease severity (P = 0.0701). IL-6 was expressed less often in the EwAL subjects compared with GOLD 1–3 and A1ATD subjects, of whom 13%–26% expressed IL-6 (Fig. 3d).

Qualitative CT data

Increasing disease severity was associated with a greater prevalence of centrilobular emphysema (Cochran–Mantel–Haenszel test, P < 0.0001) (Fig. 4). Emphysema was present in all subjects in the A1ATD group, and in 41.0% of the EwAL group. Among GOLD subjects, the prevalence of advanced destructive emphysema was higher in GOLD 3 (GOLD 1: 9.0%; GOLD 2: 13.4%; GOLD 3: 31.5%), as well as the prevalence of confluent emphysema (GOLD 1: 13.1%; GOLD 2: 24.4%; GOLD 3: 26.8%). The Fleischner classification does not provide a separate category for panlobular emphysema [15], which would be graded as either confluent or advanced destructive; these findings were present in 14 (82%) of 17 subjects with A1ATD.

Fig. 4
figure 4

Centrilobular emphysema on CT (A) and CT images showing typical emphysema findings for each category i) absent, ii) mild, iii) moderate, iv) confluent, v) advanced destructive (B). Percentages do not total 100% due to missing data (missing: EwAL, 3.3%; GOLD 1, 2.5%; GOLD 2, 1.6%; GOLD 3, 3.9%; A1ATD, 10.5%)

Full size image

Quantitative CT data

When assessing lung volumes by quantitative CT, a linear increase in expiratory volume and a linear decrease in PD15 and ALD were observed with increasing disease severity (P < 0.0001 for all comparisons) (Table 4). Mean expiratory volume increased from 2.83 L (SD 0.67 L) in EwAL subjects to 4.78 L (SD 1.24 L) in GOLD 3 subjects. PD15 decreased from 91.2 g/L (SD 19.4 g/L) in EwAL subjects to 47.8 g/L (SD 24.9 g/L) in GOLD 3 subjects. PD15 was 44.4 g/L (SD 21.1 g/L) in A1ATD subjects. Similarly, ALD decreased from 88.50 g/L (SD 13.90 g/L) in EwAL subjects to 54.8 g/L (SD 23.1 g/L) in GOLD 3 subjects. ALD was 50.9 g/L (SD 18.2 g/L) in A1ATD subjects. A linear increase was also established for inspiratory volume with increasing disease severity (EwAL: 5.80 L [SD 1.30 L]; GOLD 1: 6.42 L [SD 1.28 L]; GOLD 2: 6.31 L [SD 1.35 L]; GOLD 3: 6.63 L [SD 1.45 L], P < 0.0001). Subjects in the A1ATD group had higher mean inspiratory (8.07 L [SD 1.73 L]) and expiratory (5.33 L [SD 1.51 L]) volumes than subjects in the GOLD 1–3 groups and the EwAL group.

Table 4 Qualitative and quantitative CT scans
Full size table

The mean FRC/TLC ratio, a measure of air trapping, was 0.570 (SD 0.118) in GOLD 1 subjects versus 0.722 (SD 0.097) in GOLD 3 subjects. Ex-smoker controls had the lowest FRC/TLC ratio (0.499 [SD 0.112]). Values for GOLD 2 subjects (0.629 [SD 0.102]) and A1ATD subjects (0.653 [SD 0.121]) were between the values for GOLD 1 and GOLD 3 subjects.

Discussion

The longitudinal observational FOOTPRINTS® study aims to identify biomarkers that are associated with COPD progression, in particular emphysema progression, over a 3-year period. Results from this study are expected to help address an unmet need to better understand different COPD phenotypes and their associated biomarkers. This could allow the use of biomarkers for diagnosis, to identify high-risk patients and to guide treatment options in a more targeted, precision-based manner. COPD diagnosis and treatment requires more than assessment of lung function. Biomarkers may help to better understand the biology, prognosis and treatment options suitable for different COPD phenotypes. In a study by Lowe et al. [2], for example, CT scans identified that 41% of the “healthy” EwAL subjects had centrilobular emphysema, indicating that emphysema progression had already occurred, although it was not possible to detect by traditional spirometry.

The baseline demographics and clinical characteristics of the FOOTPRINTS® study subjects, including lung function parameters, soluble biomarkers, qualitative and quantitative CT, and symptoms, are provided here. These data allowed us to compare characteristics between subjects grouped by COPD severity (GOLD stages 1–3), subjects with A1ATD and EwAL subjects.

As expected, at baseline, increasing COPD severity was associated with worsening lung function, increased expiratory lung volume and a reduction in DLCO. Compared with EwAL, A1ATD was also associated with worsening lung function, increased lung volumes and a reduction in DLCO. Of note, TLC was highest in the A1ATD group compared with the other groups; however, this could be a result of the higher proportion of male and younger subjects in the A1ATD group. Levels of soluble inflammatory biomarkers such as WBC count, fibrinogen, hs-CRP and plasma SP-D increased linearly by disease severity, with higher levels observed in the GOLD 3 group. Previous studies have demonstrated increases in the expression of these inflammatory biomarkers to be associated with the progression of COPD [16, 17]. Of note, levels of fibrinogen were numerically higher in the GOLD 1 group than in the GOLD 2 group. Compared with EwAL, mean biomarker values for the A1ATD group were generally higher, except for fibrinogen, which was slightly lower for the A1ATD group. IL-6 levels were numerically lower in the EwAL group compared with GOLD 1–3 and A1ATD subjects.

Levels of sRAGE were highest in EwAL subjects and decreased linearly with increasing disease severity, with lowest values observed in the GOLD 3 group, indicative of more severe emphysema. This is in keeping with previous studies, which have suggested lower blood sRAGE to be associated with more severe airflow obstruction and emphysema [18]. Similarly, in the A1ATD group, levels of sRAGE were numerically lower than in the EwAL group, again indicative of more severe emphysema.

When assessing lung density by quantitative CT, we noted a clear and expected decrease in PD15 with increasing COPD severity, as has been shown in previous studies [19]. There was also an increase in expiratory and inspiratory volume with increasing disease severity. As disease severity increased, the presence of centrilobular emphysema increased (as assessed by qualitative CT), as did the presence of advanced destructive emphysema. As expected, subjects in the A1ATD group had the highest prevalence of confluent or advanced destructive emphysema, which presumably corresponds to panlobular emphysema (panlobular emphysema is not separately classified in the Fleischner system) [15]. Although the EwAL group did not have airflow limitation (as defined by the 2015 GOLD strategy report [9]), centrilobular emphysema was present in 36% of subjects, indicating that emphysema progression had already occurred in many of these subjects, likely a result of their 33 pack-year smoking history. Several other studies have demonstrated a substantial proportion of smokers without airflow limitation to have CT evidence of structural changes such as emphysema and airway wall thickening [2, 20, 21]. For example, in a cohort of over 4000 current and former smokers from the COPDGene study who did not meet spirometric criteria for COPD, a significant number (24%; 72 of 300) had evidence of emphysema [20]. In a study by Tan and colleagues [7] comparing ever-smokers with and without COPD against never-smokers with and without COPD using the radiological methodology of Barr et al. [22], emphysema was found in 12.8% and 10.2% in never-smokers with and without a diagnosis of COPD, compared with 54.8% and 26.0% in ever-smokers with and without a diagnosis of COPD, respectively. Thus, emphysema is strongly related to smoking history. Further analyses of the COPDGene study data suggest that a substantial proportion of smokers who do not meet the criteria for COPD, but have respiratory symptoms and imaging abnormalities, are at significant risk of death and spirometric disease progression [2]. This result highlights the importance of finding appropriate markers to identify subjects early who are at risk of disease progression. Given the high proportion of EwAL subjects who had centrilobular emphysema in our study (41%), this raises the question of whether this group could be considered an early disease group rather than a true control group.

As expected, as disease severity increased, symptoms tended to worsen. Values for subjects in the A1ATD group were generally between those in the GOLD 2 and 3 groups. This may not be surprising, given that the majority of the A1ATD subjects met the criteria for COPD GOLD 3 (n = 12, 63.2%) or GOLD 2 (n = 5, 26.3%). Our study also found that mMRC dyspnoea score (< 2 and ≥ 2) and SGRQ total score (< 25 and ≥ 25) were better discriminators for the COPD GOLD 1–3 groups than CAT score (< 10 and ≥ 10), demonstrating a more consistent numerical trend across the groups.

In addition to the FOOTPRINTS® study, there are a number of longitudinal studies that are ongoing or completed in patients with COPD, including COPDGene [2], ECLIPSE [3], SPIROMICS [4, 5], COSYCONET [6], CanCOLD [7] and the British Lung Foundation early COPD study (Table 5) [8]. Comparatively, FOOTPRINTS® has a number of strengths. The inclusion of the A1ATD subpopulation is unique to the study and provides information on an important but rarely studied subpopulation of patients; these patients present with earlier onset and faster progression of emphysema [12]. Further, at the time FOOTPRINTS® was designed and began, no other study had presented an association between biomarkers of extracellular matrix degradation (in this case, sRAGE) and emphysema. Since then, this area has garnered the interest of several other studies [18, 23,24,25]. Another distinguishing factor is the method for visual assessment of emphysema used in our study; this has been conducted according to the statement by the Fleischner Society [15]. The Fleischner method has been shown to be complementary to quantitative CT assessment and to provide incremental information regarding the likelihood of disease progression, lung cancer [26] and mortality [27, 28].

Table 5 Summary of ongoing or completed longitudinal observational COPD studies
Full size table

Conclusions

The FOOTPRINTS® study is investigating biomarkers associated with emphysema over a 3-year period to increase the understanding of COPD patient phenotypes, pathogenesis and disease progression. The data reported here document the baseline disease characteristics and biomarkers of subjects in our study, highlighting the substantial burden of COPD and A1ATD. In addition to the baseline biomarker data presented, future analyses will be conducted on sputum biomarkers. Importantly, these findings provide important context to precede the upcoming results from the 3-year study, which aim to establish key correlations between disease progression and biomarkers over time in a representative COPD population.