Introduction
Acute exacerbations of chronic obstructive pulmonary disease (AECOPD) cause the majority
of the morbidity, mortality, and health-care expenditure associated with COPD. The
pathogenesis of AECOPD is complex, but oxidative stress has been implicated in AECOPD.
Club cell secretory protein (CCSP), a molecule secreted by the non-ciliated bronchiolar
Club cells, is thought to protect airways from oxidative stress and may be a therapeutic
target in AECOPD.1 Previous studies have linked CCSP with COPD and lung function decline.2
However, no studies have investigated the association between CCSP levels and AECOPD.
Our objective was to determine if CCSP levels correlate with AECOPD by measuring plasma
CCSP levels at the time of AECOPD hospitalization, and recovery compared to stable
outpatients. We used data from two multicenter clinical trials of patients at increased
risk of AECOPD to test the hypothesis that plasma CCSP levels are reduced in AECOPD
compared to stable COPD.
Methods
We performed a case–control study (ratio 1:2, matched on age, sex, tobacco pack-years,
and FEV1% predicted) where cases were hospitalized AECOPD patients from the Zileuton
to Treat Adults with Chronic Obstructive Pulmonary Disease (LEUKO) study. All participants
in the LEUKO study with plasma samples obtained at both hospital day 0 and 30 days
post-hospitalization were included in the analyses. Matched stable controls were outpatients
at high risk for AECOPD, but without exacerbations in the 4 weeks prior to study entry
and blood draw, in the Macrolide Azithromycin to Prevent Rapid Worsening of Symptoms
Associated with Chronic Obstructive Pulmonary Disease (MACRO) study. Plasma CCSP levels
at hospital day 0 and day 30 post-hospitalization for cases and at baseline for controls
were quantified using commercially available ELISA kits (BioVendor, Candler, NC, USA).
CCSP levels were transformed to a natural logarithmic scale to achieve normal distribution.
To account for case–control matching, conditional logistic regression was used to
determine the association between CCSP and case–control status. Serum creatinine was
a covariate in adjusted models since CCSP levels are affected by kidney function and
creatinine clearance predicts plasma CCSP levels.3
T-tests were used to compare CCSP means between cases at hospital day 0, day 30 post-hospitalization,
and controls.
Results
We measured plasma CCSP levels in 30 cases and 60 controls. After exclusion of one
case with extreme plasma CCSP levels (day 0: 125.56 ng/mL; day 30: 136.73 ng/mL) and
the two matched controls for this case, we have 29 cases and 58 matched controls for
analysis (Table 1). Median [interquartile range (IQR)] age of the cases was 61 [55–68]
years, 45% were females, 35% were current smokers, median [IQR] smoking history of
40 [28–57] pack-years, mean FEV1% predicted was 28.8% and median [IQR] serum creatinine
was 1.0 [0.8–1.1]. The controls had a median [IQR] age of 60 [56–68] years, 41% were
female, 26% were current smokers, median [IQR] smoking history of 40 [30–50] pack-years,
mean FEV1% predicted was 34.6%, and median [IQR] serum creatinine was 0.9 [0.8–1.1]
mg/dL.
Table 1
Baseline Characteristics
Cases (n=29)
Controls (n=58)
Age, years (median [IQR])
61 [55-68]
60 [56-68]
Female, no. (%)
13 (45%)
24 (41%)
Race or ethnic group, no. (%)
White
15 (52%)
42 (72%)
African-American
13 (45%)
10 (17%)
Other
1 (3%)
6 (10%)
Current smoker, no. (%)
10 (35%)
15 (26%)
Smoking history, pack-years (median [IQR])
40 [28-57]
40 [30-50]
Spirometry (mean±SD)
FEV1, L
0.8 ± 0.4
1.0 ± 0.5
FEV1% predicted
28.8 ± 13.1
34.6 ± 15.9
FVC, L
2.0 ± 0.6
2.5 ± 0.9
FEV1/FVC ratio
0.4 ± 0.1
0.4 ± 0.1
Inhaled therapies, no (%)
ICS
24 (83%)
43 (74%)
LABA
22 (76%)
43 (74%)
LAMA
17 (59%)
39 (67%)
Serum creatinine, mg/dL (median [IQR])
1.0 [0.8-1.1]
0.9 [0.8-1.1]
Notes: Definitions: Cases, hospitalized acute exacerbation in COPD; controls, stable
COPD.
Abbreviations: FEV1, forced expiratory volume in one second; FVC, forced vital capacity;
ICS, inhaled corticosteroid; LABA, long-acting beta2-agonist; LAMA, long-acting muscarinic
antagonist; and SD, standard deviation.
Hospitalized AECOPD (case) status was associated with reduced plasma CCSP levels compared
to stable COPD (control) status, matched on age, sex, tobacco pack years, and FEV1%
predicted. Each natural log unit increase in plasma CCSP was associated with a decreased
odds for hospitalized AECOPD status in both unadjusted (odds ratio [OR] 0.347, 95%
confidence interval [CI]: 0.121 to 0.998, p=0.0495) and adjusted models that included
serum creatinine as a covariate (adjusted OR 0.269, 95% CI: 0.089 to 0.811, p=0.0197)
(Table 2). The plasma CCSP (mean±standard deviation) at day 0 (1.53±0.62 ng/mL) vs
controls (1.79±0.51 ng/mL) (p=0.041) were significantly different, while day 0 vs
day 30 (1.64±0.60 ng/mL) (p=0.229) and day 30 vs controls were not significantly different
(p=0.495).
Table 2
Conditional Logistic Regression Analyses for Predicting Hospitalized Acute Exacerbation
in COPD (AECOPD) Case vs Matched Stable COPD Control Status
Hospitalized AECOPD Cases (n=29)
Stable COPD Controls (n=58)
Unadjusted OR (95% CI)
Adjusted OR (95% CI)
Plasma CCSP (mean±SD)
1.53±0.62
1.79±0.51
0.347 (0.121, 0.998)
p=0.0495
0.269 (0.089, 0.811)
p=0.0197
Notes: Cases and controls were matched on age, sex, tobacco pack-years, and FEV1%
predicted. Plasma CCSP levels in ng/mL are in natural logarithmic scale. Adjusted
model includes creatinine as a covariate.
Abbreviations: CCSP, club cell secretory protein; SD, standard deviation; CI, confidence
interval.
Discussion
To our knowledge, this study is the first to evaluate plasma CCSP during an AECOPD.
We found that lower plasma CCSP is associated with higher odds of hospitalized AECOPD
status relative to matched stable COPD and that plasma CCSP levels 30 days following
hospitalization are similar to stable controls.
CCSP is produced by the small airway club cell and is thought to protect airways from
oxidative stress.1 Therefore, a reduction in CCSP levels during an AECOPD would result
in increased susceptibility to oxidative stress, which could contribute to AECOPD
severity. The increasing trend in plasma CCSP 30 days later, though not significantly
different from hospitalization day 0, is suggestive of airway epithelial cells recovery
following an AECOPD.
We excluded one case with extreme plasma CCSP levels. This patient had a serum creatinine
of 12.1 mg/dL indicating renal failure. Because plasma CCSP levels are affected by
kidney function,3 it is likely that this extreme elevation in plasma CCSP levels is
due to renal failure. Therefore, we accounted for renal function in our adjusted model.
CCSP has been evaluated in several COPD cohorts. Bernard et al found that bronchoalveolar
lavage fluid CCSP levels were reduced in stable COPD compared to non-smoker controls.4
In the Evaluation of COPD Longitudinally to Identify Predictive Surrogate Endpoints
(ECLIPSE) study, a 3-year multicentre prospective observational cohort of 2747 participants
with COPD, Lomas et al found that median serum CCSP was lowest in COPD participants
(5.9 ng/mL) compared to smoker (6.3 ng/mL) and non-smoker controls (7.5 ng/mL) without
COPD.5 The median serum CCSP level among COPD participants in the ECLIPSE cohort is
similar to our stable COPD control group (5.50 ng/mL), consistent with CCSP deficiency
in patients with COPD. Low serum CCSP levels have been associated with accelerated
lung function decline,2 which brings into question the possible role of CCSP augmentation
therapy in COPD.
Pre-clinical models of CCSP augmentation have demonstrated protective effects in both
cellular and animal models.6 An ongoing clinical trial in premature infants with respiratory
distress syndrome has shown that recombinant human CCSP can be safely delivered intratracheally,
and a single dose reduces pulmonary inflammatory markers.7 Our findings support assessment
of therapies targeting restoration of CCSP levels in COPD, specifically during an
acute exacerbation or as a sustained, maintenance therapy in exacerbation-prone COPD
patients.
Our study has limitations including a modest sample size and being observational,
thus we are unable to establish causality. Whether reduction in CCSP levels is a risk
factor for versus a consequence of AECOPD will need to be addressed in larger longitudinal
and mechanistic studies. We did not have baseline plasma CCSP levels prior to hospitalization,
which could have helped determine if levels at 30 days return to baseline, but because
we were able to match on age, sex, tobacco pack-years, and FEV1% predicted, we theorize
that our matched controls reflect baseline, pre-hospitalization plasma CCSP levels
among exacerbation-prone COPD patients.
Conclusion
In conclusion, lower plasma CCSP is associated with hospitalized AECOPD. Clinical
studies are needed to evaluate the efficacy of CCSP as a novel treatment for COPD,
especially in AECOPD.