· web view£ corresponding author: tina manon-jensen, nordic bioscience a/s, herlev hovedgade...
TRANSCRIPT
End-product of fibrinogen is elevated in emphysematous chronic
obstructive pulmonary disease and is predictive of mortality in the
ECLIPSE cohort
Tina Manon-Jensen1,£,#, Lasse L. Langholm1,2,#, Sarah Rank Rønnow1,3, Morten Asser Karsdal1, Ruth
Tal-Singer4*, Jørgen Vestbo5*, Diana Julie Leeming1, Bruce E. Miller4 and Jannie Marie Bülow
Sand1
*The Evaluation of COPD Longitudinally to Identify Surrogate Endpoints (ECLIPSE) study,
investigators. #Contributed equally.
1Nordic Bioscience A/S, Herlev, Denmark, 22University of Copenhagen, Faculty of Health and
Medical Sciences, Department of Biomedical Sciences, Copenhagen, Denmark, 3University of
Southern Denmark, The Faculty of Health Science, Odense, Denmark, 4GSK R&D, Collegeville, PA,
USA, 5Division of Infection Immunity and Respiratory Medicine, The University of Manchester and
Manchester University NHS Foundation Trust, Manchester Academic Health Sciences Centre,
Manchester, England.
£ Corresponding author: Tina Manon-Jensen, Nordic Bioscience A/S, Herlev Hovedgade 205-207
2730 Herlev, Denmark. Tel. +45 4452 5252, Fax: +45 4452 5251, E-mail: [email protected]
RUNNING TITLE: Fibrinogen processing in patients with COPD
KEYWORDS: fibrinogen, fibrin, ECLIPSE, COPD, fibrosis, D-dimer, fibrinopeptide A, exacerbations,
mortality
Word count: Abstract 310, Text 3561, Figure 4, tables 4
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
ABSTRACT
Background:
Chronic obstructive pulmonary disease (COPD) is characterized by an abnormal epithelial repair
process that may result in intra-airway accumulation of fibrin. Fibrinogen deposition and fibrin-
fibrin cross-links help strengthen the tether formation of the fibrinogen platelets, which are
essential for wound repair processes. Given that plasma fibrinogen is the only FDA qualified
biomarker that predicts mortality and COPD exacerbations, we hypothesized that changes in the
processing of fibrinogen may provide additional characterization of disease endotype and risk of
COPD progression. This was investigated by targeting neo-epitopes of fibrinogen that are released
during tissue repair.
Methods: A subpopulation of participants with COPD, (n=953) smoker (n=205) and non-smoker
controls (n=98) were included from The Evaluation of COPD Longitudinally to Identify Predictive
Surrogate End-points (ECLIPSE) cohort. Two blood biomarkers that specifically target the
thrombin-mediated conversion of fibrinogen into fibrin (FPA), and plasmin-mediated degradation
of cross-linked fibrin (D-dimer) were measured and compared with fibrinogen measurements.
Results: Plasma D-dimer had a predictive value for two-year mortality, with an adjusted hazard
ratio of 1.48 per SD (n=980; 95% Cl 1.18-1.84; p<0.0001). This was comparable to the fibrinogen
hazard ratio of 1.59 per SD (n=983; 95% Cl 1.29-1.96; p=0.0003), whereas FPA was not significantly
associated with mortality. D-dimer (p<0.001), fibrinogen (p<0.0001) and FPA (p<0.05) were
significantly elevated in participants with dyspnea (mMRC ≥ 2) as compared to participants
without dyspnea. By contrast, D-dimer was the only biomarker that was associated with
emphysema (p<0.001), and only plasma fibrinogen (p<0.05) was weakly associated with future
exacerbations.
Conclusion: There is a need for biochemical markers to characterize the heterogeneity of COPD, to
continuously improve clinical trial design and to identify individuals at risk for disease progression
to optimize treatment and early intervention. In the current study, we demonstrate that three
applied fibrinogen biomarkers provide information which reflect the distinct aspects of COPD,
rendering them potential attractive endotype biomarkers in COPD.
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
INTRODUCTION
Chronic obstructive pulmonary disease (COPD) is a heterogeneous and progressive lung disease
with different pathological processes leading to recognition of patient subgroups that may have
their own disease characteristics. Therefore, there is an immediate need for accurate and precise
biomarkers enabling identification of patients in most need of treatment, for optimal use of health
care resources, and to allow for better clinical trial design1. COPD is characterized by an
heightened airway inflammation and an abnormal epithelial repair process2 that may result in
intra-airway accumulation of fibrin causing long-term breathlessness and predisposes patients to
exacerbations, hospitalizations and premature death3,4. Further understanding of the wound
healing cascade which is central to COPD and the processing of fibrinogen may add valuable
information to the heterogeneous nature of COPD and provide better endotype-driven differential
diagnostics tools for patients, drug developers and payers of healthcare.
Wound healing plays a central role in the development of COPD. Cigarette smoke and other
injurious agents disrupt the protective epithelial barrier of the airways and induce injuries that
activate the coagulation cascade and initiate a pro-inflammatory response5,6. Patients with COPD
are repetitively exposed to injurious agents that renew these responses, causing a continued
release of molecules associated with wound healing and repair which results in lasting tissue
alterations such as emphysema and small airways fibrosis7,8. Plasma fibrinogen has been qualified
as a drug development tool by the U.S Food and Drug Administration (FDA)3 and the European
Medicines Agency (EMA)9 for use in clinical trials as a prognostic enrichment biomarker of patients
at risk for COPD exacerbation and mortality3,10. Fibrinogen processing (i.e. fibrin deposition and
fibrin-fibrin cross-links) is on the critical path for early stages of wound healing and determines the
outcome of tissue repair11 (Fig. 1). During early wound healing, platelets adhere to and are
activated by collagens at the wound site11 and in parallel, the coagulation cascade is initiated and
results in thrombin-mediated conversion of soluble fibrinogen into insoluble fibrin fibers, leading
to the release of the pro-peptides fibrinopeptides A (FPA) and B (FPB)12. These two processes
combined, form the platelet-fibrin-rich clot which is further strengthened when the fibrin-
stabilizing transglutaminase factor 13 (FXIII) generates fibrin-fibrin cross-links. Once hemostasis
has been achieved, the abundant pro-coagulatory signals that drive coagulation are promptly
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
removed to facilitate fibrinolysis which is required for the wound healing to proceed. Fibrinolysis,
the process that degrades cross-linked fibrin, is initiated when the zymogen plasminogen is
converted to plasmin, resulting in the release of D-dimer, a small fibrinolysis-specific degradation
product of fibrin that can be cleared by the liver13.
In this study, we evaluated plasma fibrinogen in addition to two novel biochemical markers
reflecting neo-epitopes of fibrinogen that are released during tissue repair. We hypothesized that
assessing fibrin deposition and fibrin-fibrin cross-links and not just plasma fibrinogen may provide
additional characterization of disease endotypes and risk of progression in COPD. We examined
serine protease-generated neo-epitopes of fibrin(ogen) with the biomarkers FPA and D-dimer, in a
subpopulation of the Evaluation of COPD Longitudinally to Identify Predictive Surrogate End-points
(ECLIPSE) cohort. These biomarkers reflect a thrombin-generated neo-epitope of fibrinopeptide A
(FPA), that allows quantification of the amount of active wound healing, and the plasmin-
generated neo-epitope of D-dimer that quantifies the absolute level of FXIII-cross-linked fibrin
which is considered accurate wound healing.
MATERIALS AND METHODS
Study population
The analysis was based on the three-year observational longitudinal study ECLIPSE
(ClinicalTrials.gov. number, NCT00292552), described previously14. The full ECLIPSE study included
2163 participants with COPD. The enrollment criteria included a forced expiratory volume in one
second (FEV1) of less than 80% of the predicted value and FEV1/forced vital capacity (FVC) ratio of
0.7 or less, assessed after the use of bronchodilators. Participants with COPD were clinically
evaluated at baseline, month three, six and subsequently every six months in three years. The
current biomarker analysis was performed on heparinized plasma samples from 1286 subjects
consisting of 983 COPD subjects, 205 smoking control participants and 98 non-smoking control
participants. Importantly, fibrinogen was measured in ethylenediaminetetraacetic acid (EDTA)
plasma15. Control participants had to be free of significant comorbidities, determined from
screening investigation, physical examination and medical history and only 95 COPD participants
were age, BMI and gender matched with the control groups. For the analyses, we used clinical
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
data obtained at baseline and available biomarker data obtained at year one. All-cause mortality
was assessed at year three. The study was conducted according to the Declaration of Helsinki and
Good Clinical Practice guidelines and was approved by relevant ethics and review boards.
Participants provided informed consent.
Biomarker measurements
Fibrin(ogen) formation and degradation products were assessed by enzyme-linked
immunosorbent assays (ELISAs) (Nordic Bioscience A/S, Herlev, Denmark). In both ELISAs, mouse
monoclonal antibodies were used to target a neo-epitope of thrombin-mediated degradation of
fibrinogen α-chain (competitive ELISA, FPA, unpublished) and a neo-epitope of plasmin-mediated
degradation of cross-linked fibrin β-chain (Sandwich ELISA, D-dimer, unpublished) (Table 1). The
neo-epitope is defined as a specific amino acid sequence that has been generated by enzymatic
cleavage of fibrin(ogen). The antibodies react only with enzymatically processed, and not with
intact fibrin(ogen).
For FPA competitive ELISA, 96-well streptavidin plate was coated with biotinylated synthetic
peptide dissolved in assay buffer (50mM Tris, 137 mM NaCl, 1% BSA, 0.05% Tween-20, 0.36%
Bronidox L5, pH 8.0) and incubated 30 minutes at 20°C. 20 µL of standard peptide or samples
diluted in assay buffer were added to appropriate wells, followed by 100 µL of monoclonal
antibody 84-23, and incubated 20 hours at 4°C. After plate wash (20 mM Tris, 50 mM NaCl, pH
7.2), 100 µL of Horseradish Peroxidase (HRP) labeled rabbit anti-mouse secondary antibody was
added and incubated at 20°C for one hour. For D-dimer sandwich ELISA, streptavidin plate was
coated with biotinylated antibody dissolved in coating buffer (50mM Tris, 137 mM NaCl, 1% BSA,
0.05% Tween-20, 0.36% Bronidox L5, pH 8.0) and incubated 30 minutes at 20°C. 20µL of standard
dilution or samples followed by 80µL incubation buffer (50mM Tris, 137 mM NaCl, 1% BSA, 0.05%
Tween-20, 0.36% Bronidox L5, 5% liquid II, pH 8.0) were added to appropriate wells and incubated
20 hours at 4°C. After plate wash, 100µL of HRP labeled antibody was added and incubated at 20°C
for one hour. For both assays, the plate was washed and added 100 µl of tetramethylbenzidine
(TMB) and incubated for 15 min at 20oC in the dark, before stopping the HRP reaction by adding
100 µl of stopping solution (1% H2SO4). Plates were read in a SpectraMax M5 (Molecular Devices,
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
CA, USA) at 450nm with 650nm as reference.
Statistics
All biomarker data were not normally distributed (D’agostino Pearson test), thus for comparison,
all data were analyzed as non-parametric. Demographics data was analyzed using Kruskal-Wallis
one-way ANOVA and chi-squared test as appropriate. The Mann-Whitney t-test and Kruskal-Wallis
testing were used for between-group biomarker comparison. Biomarker cutoffs (Youden Index
criterion) estimated using receiver operating curve (ROC) analysis was based on mortality data and
plotted using Kaplan-Meier survival curves. Plasma D-dimer data were dichotomized based on ROC
statistics. Mortality risk for groups below and above the cutoff was compared using Cox
proportional hazards analysis with or without relevant confounders as determined by univariate
Cox proportional hazards analysis. Cox proportional hazard regression analysis used to estimate
hazard ratios (HR) per 1 standard deviation (SD) change for a better comparison between the
biomarkers. HRs were adjusted for confounding factors for mortality: age, smoking history, 6-
minute walking distance (6MWD), prior hospitalizations due to COPD exacerbations and modified
Medical Research Council (mMRC) dyspnea score. Software used was GraphPad Prism version 7.00
for Windows (GraphPad Software, La Jolla California USA) and MedCalc Statistical Software version
14.8.1 (MedCalc Software bvba, Ostend, Belgium). A p-value < 0.05 was considered statistically
significant.
RESULTS
Assessment of fibrinogen turnover in participants with COPD
The basic demographic description of the population and standard clinical parameters of lung
function and disease stage are presented in Table 2. Mean age of the participants was 63.1 years
for the COPD cohort and 55.0 years and 58.9 years for the smoker and non-smoker control group,
respectively.
To investigate the change in fibrinogen turnover in COPD, we compared COPD participants with
smoker and non-smoker controls. Degradation of cross-linked fibrin, represented by the plasmin-
generated fragment D-dimer, increased in participants with COPD compared to smokers and non-
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
smokers (Fig. 2B, p<0.0001, p=0.035), which was similar to the change in fibrinogen (Fig. 2C,
p<0.0001). Notably, in age-, gender- and BMI-matched samples, D-dimer did not show any
significant differences between groups, whereas fibrinogen shows significant difference between
COPD and smoker controls (p<0.05) and non-smoker controls (p<0.0001) (data not shown).
Notably, the matched analysis included only 95 COPD participants and represented mostly less
diseased patients, as compared to the complete cohort (data not shown). The concentration of the
thrombin-generated fibrin fragment FPA, representing wound healing activation, followed the
pattern of fibrinogen, although it did not reach statistical significance (Fig. 2A).
Fibrinogen turnover related to disease phenotype (dyspnea, exacerbation, and emphysema)
To examine the underlying symptoms or endotypes related to COPD, we assessed the relationship
between fibrinogen processing and dyspnea, emphysema and exacerbations. To investigate the
change in fibrinogen turnover related to breathlessness, participants from all groups were divided
into an asymptomatic/mild (n=765) and a symptomatic group (n=475), based on a modified mMRC
dyspnea score cutoff of ≥2, as previously described16. Plasma levels of the three biomarkers of
fibrinogen turnover show significant difference between asymptomatic and symptomatic
participants. Formation of fibrin, represented by the thrombin-generated fragment FPA, increased
significantly in symptomatic participants (p=0.0282; Fig.3A). D-dimer concentration, represented
by the neo-epitope of plasmin-mediated degradation of the cross-linked fibrin beta-chain,
increased substantially corresponding to a 62 % increase in the D-dimer level of symptomatic
participants (p=0.0002; Fig.3B). Among all three biomarkers, the level of fibrinogen shows the
strongest marked difference between asymptomatic and symptomatic participants (p<0.0001,
Fig.3C). None of the biomarkers show significant difference between asymptomatic and
symptomatic participants when assessing the overall health based on a reference cutoff point of
St. George´s Respiratory Questionnaire (SGRQ) ≥25, as suggested by GOLD17 (data not shown).
To investigate the change in fibrinogen turnover related to emphysematous COPD, the COPD
subgroup was separated based on whether or not they had significant emphysema, defined as low
attenuation area at -950 Hounsfield Units (% LAA) ≥ 10% on chest computed tomography (CT)
scans. D-Dimer was the only biomarker that differed significantly between emphysematous
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
(n=613) and non-emphysematous (n=321) COPD with concentrations of 74.00 (95 % CI 46.67–
101.3) ng/mL and 48.48 (95 % CI 38.01-58.94) ng/ mL, respectively (p = 0.0002), corresponding to
a 53 % increase in the levels (Fig. 3H). Biomarkers FPA (Fig.3G) and fibrinogen (Fig.3I) did not show
any significant difference between emphysematous and non-emphysematous COPD.
To investigate change in fibrinogen turnover related to COPD exacerbations, the COPD subgroup
was divided into two groups based on if they had experienced one or more exacerbations (n=547)
or none (n=437) in the two years follow-up period after the blood-sampling. Exacerbations were
characterized as sustained worsening of respiratory symptoms, such as breathlessness or
increased sputum volume, but were self-reported by the patients. Fibrinogen levels were
significantly elevated in patients with future exacerbations (p=0.0164, Fig.3F), while FPA and D-
dimer concentrations were unchanged (FPA, p=0.1897, Fig.3D; D-dimer, p=0.5983, Fig.3E).
Fibrinogen turnover related to mortality
Next, we wanted to investigate how the fibrinogen turnover markers related to mortality, using
data from plasma fibrinogen as a reference to further explore the full fibrinogen turnover profile.
A total of 31 COPD subjects died within the three year study period (952 survivors). The biomarker
D-dimer was significantly elevated 7.7 fold (mean levels 415.7 ng/ml vs. 54.0 ng/ml) in subjects
who died compared to survivors (p=0.0018) (Fig. 4B). Plasma fibrinogen (Fig.4C), but not FPA
(Fig.4A), was also significantly elevated 1.2 fold (mean levels 457.9 ng/ml vs. 390.1 ng/ml) in non-
survivors (p=0.0012).
Fibrinogen has a published defined cutoff of 350 mg/dl for classifying patients as having high
fibrinogen levels and increased mortality and exacerbation events10. Notably, fibrinogen
concentrations in the ECLIPSE study were adjusted with -13.6% to account for the use of EDTA
plasma (the K-assay) instead of citrate plasma (Clauss method) for measuring fibrinogen based on
data provided by the manufacturer10. We estimated similar cutoff for the D-dimer using ROC curve
analysis, which resulted in an AUC of 0.66 (95%CI 0.63:0.69) for D-dimer with an associated
Youden index criterion of 32 ng/ml for predicting mortality. The corresponding AUC of fibrinogen
was 0.67 (95%CI 0.64:0.70). The biomarker FPA was excluded from these analyses as it did not
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
show any significant difference between survivors and non-survivors (Fig.4A). Kaplan-Meier plots
show a difference in survival times when subjects are separated by the dichotomized biomarker
groups for fibrinogen (Fig.4E) and D-dimer (Fig.4D). Subjects with high plasma levels of D-dimer
(≥32 ng/ml) had a significantly increased mortality risk within the two years of follow up, which is
similar to what we see with high fibrinogen (≥350 mg/dl) (Fig.4A, B). We used cox proportional
hazard regression to investigate if D-dimer was an independent predictor of mortality as indicated
by the Kaplan-Meier curve. Univariate cox regression analysis was performed to evaluate
confounders of mortality risk and here age, 6MWD, mMRC, prior hospitalizations due to COPD
exacerbations and current smoking status were identified as confounders (Table 3). Indeed, with
an adjusted hazard ratio of 1.48 per SD (n=980; 95% Cl 1.18-1.84; p<0.0001) (Table 4), D-dimer had
a predictive value for mortality which compares to plasma levels greater than or equal to 350
mg/dl for fibrinogen with an adjusted hazard ratio of 1.59 per SD (n=983; 95% Cl 1.29-1.96;
p=0.0003). We used cox regression analysis to assess the discriminatory power of a D-dimer model
for predicting death. We evaluated D-dimer alone and in an adjusted model using the
dichotomized biomarker cutoff 32 ng/ml and significant confounders (Table 3). The adjusted
model (including D-dimer, age, and current smoking status) showed D-dimer to be independently
associated with mortality, with the high biomarker group having a hazard ratio of 4.31 (95% CI
1.80-10.3) compared to the low biomarker group. No model could be obtained including
fibrinogen (using a cutoff of 350 mg/dL).
DISCUSSION
Given that fibrinogen is the only FDA and EMA qualified as a drug development tool biomarker for
predicting mortality and exacerbations in COPD3, we hypothesized that more refined assays of
fibrinogen, such as neo-epitopes of fibrinogen that are released during wound repair as activity
measures of healing may provide more specificity additional disease characteristics and enable
personalized health care in COPD. Here, we measured plasma levels of fibrinogen processing by
detecting the neo-epitope of thrombin-mediated degradation of fibrinogen α-chain, FPA, and the
neo-epitope of plasmin-mediated degradation of cross-linked fibrin β-chain, D-dimer. FPA and D-
dimer measure a specific processing of fibrinogen and may add additional biological value
compared to measuring fibrinogen alone. We found that D-dimer, a measure of ongoing fibrin
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
formation and degradation, had a predictive value for mortality, similar to fibrinogen, whereas
FPA, a measure of fibrin formation, was not significantly associated with mortality. Plasma D-
dimer, fibrinogen and FPA were significantly elevated in symptomatic participants as compared to
asymptomatic participants. In contrast, D-dimer was the only marker that was associated with
emphysema and only plasma fibrinogen was associated with exacerbations, albeit weakly.
A variety of fibrinogen assays exist18, in particular for D-dimer which utilizes antibodies to detect
different epitopes of degraded fibrinogen18. D-dimer levels in COPD subjects have previously
shown high variations across studies19–24. Our findings suggest an impaired regulation of fibrinogen
processing to be a part of COPD pathogenesis. In line with other studies25, D-dimer and fibrinogen
were increased in COPD compared to smoking and non-smoking controls. Current smoking, as
opposed to never smoking, is accompanied by increased levels of most circulating inflammatory
markers including plasma fibrinogen26–28. Unexpectedly, we found no significant difference
between smoking and non-smoking control subjects regardless of whether the subject groups
were matched for age, gender and BMI. Also, no significant difference between former and
current COPD smokers was found for any of the fibrinogen biomarkers (data not shown). Smoking
cessation may not fully attenuate the inflammatory process once COPD is developed and past
smoking may result in long-term changes in fibrinogen levels26. Notably, not all COPD participants
had a high degree of systemic inflammation defined by C-reactive protein levels despite high levels
of fibrinogen, fully supporting the notion that impaired deposition and possible fibrinolysis are
pathological events in COPD3.
The presented data not only confirm that D-dimer is elevated in COPD, but also that D-dimer could
be a marker of emphysema. Emphysema is a condition in which the alveoli are destroyed as a
result of an aberrant inflammatory reaction to irritating gases like cigarette smoke. Interestingly,
neither FPA nor intact fibrinogen were associated with emphysema indicating that fibrinogen is
not just fibrinogen and that assessing neo-epitope protein fragments of fibrinogen provide
different information to the pathogenesis of COPD, which is not provided by assessing intact
fibrinogen. As expected, we found the D-dimer concentration significantly highest in symptomatic
participants compared to asymptomatic participants, supporting D-dimer concentrations to be
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
highest during active disease, where regenerative capacity and activity of tissue turnover are
expected to be at their highest. Data derived from multiple large clinical trials indicate that the
average lung function decline appears to be higher in an earlier state of COPD compared to later
stage where large parts of the lung tissue have already been damaged29. However, FEV1 is not the
optimal tool for staging patients with COPD, as a low FEV1 can be obtained with both normal and
accelerated decline, depending the patient’s starting point30. Assessment of active wound healing
as well as tissue damage and repair may provide a more accurate picture of the current
pathological state31–33.
Plasma fibrinogen is a useful biomarker to stratify individuals with COPD into those with a high or
low risk of future exacerbations, and may be used to identify those with a higher risk of
mortality34. Exacerbations occur when COPD subjects experience that their symptoms are much
worse than usual causing a sudden accelerated decline in respiratory functions35, often leading to
hospitalizations. COPD associated with frequent exacerbations is regarded as a distinct and
persistent phenotype of COPD that is largely independent of lung function36. Contrary to intact
fibrinogen, neither FPA nor D-dimer concentrations were increased in patients experiencing
exacerbations. Our results may reflect exhaustion of COPD or a feedback mechanism
downregulating activation of fibrin formation to counteract a sudden burst in inflammation.
Contrary to our findings, elevated concentration of D-dimer has been associated with the risk of
developing acute exacerbations37. Other papers have established D-dimer as an independent risk
factor for in-hospital mortality38 and for poor long-term prognosis in patients admitted for acute
exacerbations39. Also, D-dimer has been suggested to be used as an evaluation index for the
severity of COPD with acute exacerbations40. Notably, the diagnostic efficiency of D-dimer tests is
continuously debated as D-dimer levels are conflicting. Multiple D-dimer assays exist using
antibodies which detect different epitopes of degraded fibrinogen. Contrary to the D-dimer
assessed in this study, some of the commercial kits have obvious cross-reaction to fibrinogen18,
which will affect the accuracy of the D-dimer test18.
In the present study, D-dimer was independently associated with increased risk of death, with a
hazard ratio similar to fibrinogen, while FPA was not. Furthermore, a model including
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
dichotomized D-dimer showed that high D-dimer (≥32 ng/ml) was associated with a significantly
increased risk of mortality within the two years of follow-up. Surprisingly, we did not find any
association between fibrinogen and increased mortality using the established 350 mg/dl cutoff.
This may be explained by the smaller sample size used in our study. Furthermore, for comparison,
we looked at a clinical model consisting of BODE, age and previous hospitalizations as previously
described41. The AUC of predicting mortality for this model were 0.76, which could be increased to
0.83 by including fibrinogen. This was comparable to the D-Dimer cutoff model with an AUC of
0.81 (data not shown). This indicates that D-Dimer show comparably predictive power to
fibrinogen.
A major limitation of our study is the lack of information about when blood samples were taken in
relation to time since last exacerbation. Polosa et al. showed elevated levels of D-dimer in COPD
subjects during acute exacerbation, which declined substantially when the patients reached
clinical stability42. Furthermore, the relatively small mortality rate would need a larger validation
cohort to properly evaluate the prognostic value. Likewise, the current D-dimer cutoff used for
assessing mortality needs further validation in large independent cohorts before it can be
compared to the well-defined cutoff of fibrinogen, but it does indicate that biomarkers of
fibrinogen turnover could add prognostic value.
CONCLUSION
There is an immediate need for accurate and precise biochemical markers to characterize the
heterogeneity of COPD, to identify progressors of disease, for efficient use of health care
resources, and to allow for better clinical trial design. We demonstrate that the three applied
fibrinogen markers each provide information that mark distinct aspects of COPD, making them
potential attractive endotype biomarkers. D-dimer, neo-epitope detection of plasmin-degraded
cross-linked fibrin, may serve as a biomarker for predicting disease phenotype by identifying
patients with emphysematous COPD and may show value as a prognostic marker for all-cause
mortality in COPD. Given the heterogeneity of COPD, and to enable personalized health care, it is
most likely that multiple prognostic or enrichment tools alongside plasma fibrinogen are needed
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
to facilitate drug development and to provide add-on information to the pathological wound
healing process central to COPD.
CONFLICT-OF-INTEREST DISCLOSURE: TMJ, LLL, SRR, JMBS, DJL, TMJ and MAK are employed by
Nordic Bioscience, Biomarkers and Research. MAK, DJL and TMJ hold stocks in Nordic Bioscience.
Nordic Bioscience is a privately owned small-medium sized enterprise, partly focused on the
development of biomarkers for connective tissue disorders and rheumatic diseases. None of the
authors from Nordic Bioscience received any kind of financial benefits or other bonuses for the
work described in this manuscript. RTS and BEM are employees and shareholders of GSK. JV has
received honoraria for presenting and advising from Astra Zeneca, Boehringer-Ingelheim, Chiesi,
GlaxoSmithKline and Novartis, outside the submitted work.
AUTHOR CONTRIBUTIONS.
Two representatives of GlaxoSmithKline (RTS, BEM) and one academic (JV), together representing
the ECLIPSE study investigators, developed the current study design and concept in collaboration
with representatives of Nordic Bioscience (DJL, MAK, TMJ, JMBS). They approved the plan for the
current analyses, had full access to the data, and were responsible for the decision to publish. The
study sponsor did not place any restrictions with regard to statements made in the final paper.
The biomarkers were measured by LLL and SRR. The data analysis and interpretation of data was
done by LLL, SRR and TMJ. Drafting the manuscript was done by TMJ and LLL. All authors read and
approved the final version of the manuscript, and provided interpretation of data, and revising
critically for important intellectual content.
ACKNOWLEDGEMENTS: The study was sponsored by GlaxoSmithKline; the Danish Agency for
Science, Technology and Innovation; and the Danish Research Foundation. JV is supported by the
National Institute of Health Research Manchester Biomedical Research Centre (NIHR Manchester
BRC). The authors acknowledge all participants, medical, nursing, and technical staff involved in
the ECLIPSE study.
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
REFERENCE
1. Agusti, A. et al. Treatable traits: toward precision medicine of chronic airway diseases. Eur.
Respir. J. 47, 410–9 (2016).
2. Vestbo, J. COPD: definition and phenotypes. Clin. Chest Med. 35, 1–6 (2014).
3. Miller, B. E. et al. Plasma fibrinogen qualification as a drug development tool in chronic
obstructive pulmonary disease: Perspective of the chronic obstructive pulmonary disease
biomarker qualification consortium. American Journal of Respiratory and Critical Care
Medicine (2016). doi:10.1164/rccm.201509-1722PP
4. Patel, I. S., Roberts, N. J., Lloyd-Owen, S. J., Sapsford, R. J. & Wedzicha, J. A. Airway epithelial
inflammatory responses and clinical parameters in COPD. Eur. Respir. J. 22, 94–9 (2003).
5. Petecchia, L. et al. Bronchial airway epithelial cell damage following exposure to cigarette
smoke includes disassembly of tight junction components mediated by the extracellular
signal-regulated kinase 1/2 pathway. Chest 135, 1502–1512 (2009).
6. Patel, I. S. et al. Airway epithelial inflammatory responses and clinical parameters in COPD.
(2003). doi:10.1183/09031936.03.00093703
7. Hogg, J. C. et al. The nature of small-airway obstruction in chronic obstructive pulmonary
disease. N. Engl. J. Med. 350, 2645–53 (2004).
8. Cosio, M. et al. The relations between structural changes in small airways and pulmonary-
function tests. N. Engl. J. Med. 298, 1277–81 (1978).
9. Medicines Agency, E. Product Development Scientific Support Department Qualification
opinion on Plasma Fibrinogen as a Prognostic Biomarker (Drug Development Tool) for All-
Cause Mortality and COPD Exacerbations in COPD subjects. (2018).
10. Mannino, D. M. et al. Plasma Fibrinogen as a Biomarker for Mortality and Hospitalized
Exacerbations in People with COPD. Chronic Obstr. Pulm. Dis. (Miami, Fla.) 2, 23–34 (2015).
11. Manon-Jensen, T., Kjeld, N. G. & Karsdal, M. A. Collagen-mediated hemostasis. J. Thromb.
Haemost. n/a-n/a (2016). doi:10.1111/jth.13249
12. Stalker, T. J. et al. Hierarchical organization in the hemostatic response and its relationship
to the platelet-signaling network. Blood 121, 1875–85 (2013).
13. Danese, S. et al. Inflammation and coagulation in inflammatory bowel disease: The clot
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
thickens - CME. American Journal of Gastroenterology 102, 174–186 (2007).
14. Vestbo, J. et al. Evaluation of COPD Longitudinally to Identify Predictive Surrogate End-
points (ECLIPSE). Eur. Respir. J. 31, 869–73 (2008).
15. Miller, J. et al. Comorbidity, systemic inflammation and outcomes in the ECLIPSE cohort.
Respir. Med. (2013). doi:10.1016/j.rmed.2013.05.001
16. Vogelmeier, C. F. et al. Global Strategy for the Diagnosis, Management, and Prevention of
Chronic Obstructive Lung Disease 2017 Report. GOLD Executive Summary. Am. J. Respir.
Crit. Care Med. 195, 557–582 (2017).
17. Rabe, K. F. et al. Global strategy for the diagnosis, management, and prevention of chronic
obstructive pulmonary disease: GOLD executive summary. Am. J. Respir. Crit. Care Med.
176, 532–55 (2007).
18. Dempfle, C. E., Zips, S., Ergül, H., Heene, D. L. & Fibrin Assay Comparative Trial study group.
The Fibrin Assay Comparison Trial (FACT): evaluation of 23 quantitative D-dimer assays as
basis for the development of D-dimer calibrators. FACT study group. Thromb. Haemost. 85,
671–8 (2001).
19. Alessandri, C. et al. Hypercoagulability state in patients with chronic obstructive pulmonary
disease. Chronic Obstructive Bronchitis and Haemostasis Group. Thromb. Haemost. 72,
343–6 (1994).
20. Arregui, M. A. A., Ezquerra, K. L., López, F. C. & Lacasa, R. C. [Hypercoagulability state and
endotelial injury in stable chronic obstructive pulmonary disease patients]. An. Sist. Sanit.
Navar. 33, 43–50
21. Hartmann, I. J., Hagen, P. J., Melissant, C. F., Postmus, P. E. & Prins, M. H. Diagnosing acute
pulmonary embolism: effect of chronic obstructive pulmonary disease on the performance
of D-dimer testing, ventilation/perfusion scintigraphy, spiral computed tomographic
angiography, and conventional angiography. ANTELOPE Study Group. Ad. Am. J. Respir. Crit.
Care Med. 162, 2232–7 (2000).
22. Silva, D. R., Coelho, A. C., Gazzana, M. B., Menna Barreto, S. S. & Knorst, M. M. D-dimer
levels in stable COPD patients: a case-control study. COPD 9, 426–31 (2012).
23. Akgun, M. et al. Comparison of clinical characteristics and outcomes of patients with COPD
exacerbation with or without venous thromboembolism. Respiration. 73, 428–33 (2006).
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
24. Zhang, M. et al. D-dimer as a potential biomarker for the progression of COPD. Clin. Chim.
Acta. 455, 55–9 (2016).
25. Zhang, M. et al. D-dimer as a potential biomarker for the progression of COPD. Clin. Chim.
Acta. 455, 55–9 (2016).
26. Gan, W. Q., Man, S. F. P., Senthilselvan, A. & Sin, D. D. Association between chronic
obstructive pulmonary disease and systemic inflammation: a systematic review and a meta-
analysis. Thorax 59, 574–80 (2004).
27. Levitzky, Y. S. et al. Relation of smoking status to a panel of inflammatory markers: the
framingham offspring. Atherosclerosis 201, 217–24 (2008).
28. Gan, W. Q., Man, S. F. P. & Sin, D. D. The interactions between cigarette smoking and
reduced lung function on systemic inflammation. Chest 127, 558–64 (2005).
29. Bihlet, A. R. et al. Biomarkers of extracellular matrix turnover are associated with
emphysema and eosinophilic-bronchitis in COPD. Respir. Res. (2017). doi:10.1186/s12931-
017-0509-x
30. Lange, P. et al. Lung-Function Trajectories Leading to Chronic Obstructive Pulmonary
Disease. N. Engl. J. Med. 373, 111–122 (2015).
31. Stolz, D. et al. Systemic Biomarkers of Collagen and Elastin Turnover Are Associated With
Clinically Relevant Outcomes in COPD. Chest 151, 47–59 (2017).
32. Sand, J. M. B. et al. Accelerated extracellular matrix turnover during exacerbations of COPD.
Respir. Res. (2015). doi:10.1186/s12931-015-0225-3
33. Groenewegen, K. H. et al. Increased Systemic Inflammation Is a Risk Factor for COPD
Exacerbations. Chest 133, 350–357 (2008).
34. Duvoix, A. et al. Blood fibrinogen as a biomarker of chronic obstructive pulmonary disease.
Thorax 68, 670–6 (2013).
35. Patil, S. P., Krishnan, J. A., Lechtzin, N. & Diette, G. B. In-Hospital Mortality Following Acute
Exacerbations of Chronic Obstructive Pulmonary Disease. Arch. Intern. Med. 163, 1180
(2003).
36. Hurst, J. R. et al. Susceptibility to exacerbation in chronic obstructive pulmonary disease. N.
Engl. J. Med. 363, 1128–38 (2010).
37. Ishikawa, G., Acquah, S. O., Salvatore, M. & Padilla, M. L. Elevated serum D-dimer level is
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
associated with an increased risk of acute exacerbation in interstitial lung disease. Respir.
Med. 128, 78–84 (2017).
38. Hu, G. et al. Prognostic role of D-dimer for in-hospital and 1-year mortality in exacerbations
of COPD. Int. J. Chron. Obstruct. Pulmon. Dis. 11, 2729–2736 (2016).
39. Fruchter, O., Yigla, M. & Kramer, M. R. D-dimer as a prognostic biomarker for mortality in
chronic obstructive pulmonary disease exacerbation. Am. J. Med. Sci. 349, 29–35 (2015).
40. Liu, B. H. et al. Detection and study of plasma D-dimer change in patients with acute
exacerbation of chronic obstructive pulmonary disease. J. Biol. Regul. Homeost. Agents 30,
839–845
41. Celli, B. R. et al. Inflammatory biomarkers improve clinical prediction of mortality in chronic
obstructive pulmonary disease. Am. J. Respir. Crit. Care Med. 185, 1065–1072 (2012).
42. Polosa, R. et al. Effect of acute exacerbations on circulating endothelial, clotting and
fibrinolytic markers in COPD patients. Intern. Emerg. Med. 8, 567–74 (2013).
478
479
480
481
482
483
484
485
486
487
488
489
490
491492
493
494
TABLES
Table 1. Biomarker description
Structural biomarker
Specification Measure
FPA Neo-epitope of thrombin mediated degradation of fibrinogen (competitive ELISA)
Fibrin formation. Fibrinopeptide-A released from conversion of fibrinogen into fibrin
Quantify the amount of active wound healing
D-dimer Neo-epitope of plasmin-mediated degradation of cross-linked fibrin (sandwich ELISA)
Fibrinolysis - Fibrin cross-linking
Quantify the absolute level of factor 13 transglutaminase (FXIII) generated cross-links
Fibrinogen Previously measured in the ECLIPSE study by the Kamiya (K) assay
495
496
497
498
Table 2. Main clinical characteristics of the study population by subgroup at baseline
COPD(n=983)
Smoker Controls(n=205)
Non-smoker controls (n=98)
P-value
Male sex, n (%) 627 (63.8) 85 (41.5) 50 (51.0) P<0.0001#
Age, years 64 (63-65) 53 (52-56) 59 (57-61) P < 0.001
BMI, kg/m2 26.8 (5.9) 26.6 (4.3) 28.4 (4.6) P < 0.001
FEV1, post-bronchodilator, baseline, L
1.36 (1.31-1.39) 3.07 (2.99-3.17) 3.11 (2.96-3.46) P < 0.001
FEV1, post-bronchodilator, year 1, L
1.33 (1.28-1.38) 2.95 (2.88-3.04) 3.07 (2.91-3.38) P < 0.001
FEV1, % of predicted year 1 45.8 (44.2-46.9) 101.9 (100.4-
104.5) 111.3 (107.3-115.0) P < 0.001
%LAA, year 1 14.4 (13.5-15.7) 1.5 (1.3-2.0) 4.1 (-)* P < 0.001
Current smokers, n (%) 363 (36.9) 205 (100) 0 (0) P = 0.79#
Smoking history (pack years) 43 (42-45) 26 (23-28) 0 (0) P < 0.001
mMRC 1 (1-2) 0 (0) 0 (0) P < 0.001
6MWD, mean meters (SD) 385 (378-395) - -
BODE 3 (2-3) - -
GOLD grade n (%)
2 487 (49.5) - -
3 396 (40.3) - -
4 100 (10.2) - -
Data presented as median (95% Confidence interval), unless stated otherwise. BMI, Body Mass Index; FEV1, Forced Expiratory Volume in one second; FVC, Forced Vital Capacity; LAA%, Low attenuation area; 6MWD, 6 minute walking distance; BODE, Body mass index, airflow Obstruction, Dyspnea, and Exercise capacity; GOLD, the Global Initiative for Obstructive Lung Disease classification; mMRC, modified Medical Research Council dyspnea score. *, sample size too small (n=5). One-way ANOVA analysis used to determine P-values between groups, # Chi-squared test. The P-values were considered statistically significant if P < 0.05.
499
500
Table 3. Confounding factors of mortality risk assessed by Cox regression analysis
Parameter Hazard ratio
(95% confidence interval)
P-value
Age 1.13 (1.06;1.21) 0.002
Sex 0.96 (0.46;1.99) 0.910
BMI 1.00 (0.94;1.06) 0.960
BODE 1.41 (0.97;2.05) 0.070
6MWD 1.00 (0.99;1.00) 0.021
mMRC 1.51 (1.08;2.10) 0.015
Prior exacerbations 0.98 (0.72;1.33) 0.886
Hospitalizations 1.51 (1.01;2.24) 0.044
Current smoker 0.18 (0.06;0.59) 0.005
Smoking pack years 1.01 (1.00;1.02) 0.096
BMI, body mass index; BODE, Body mass index, airflow Obstruction, Dyspnea and Exercise capacity; 6MWD, 6 minute walking distance; mMRC, modified Medical Research Council dyspnea score; Prior exacerbations, exacerbations in year prior to screening; Hospitalizations, due to exacerbations.
Table 4. Hazard ratios for biomarkers
Covariate Hazard Ratio P-ValueD-Dimer (adjusted) 1.48 per 1 SD (95% Cl 1.18-1.84) <0.0001Fibrinogen (adjusted) 1.59 per 1 SD (95% Cl 1.29-1.96) 0.0003FPA (adjusted) 0.94 per 1 SD (95% Cl 0.65-1.35) 0.728D-Dimer model (≥ 32 ng/ml):
D-Dimer AgeFormer smoker
4.31 (95%CI 1.80:10.3)1.07 (95%CI 1.00:1.15)5.56 (95%CI 1.70:18.2)
0.0010.0490.005
Fibrinogen model (≥350 mg/dl):AgemMRCFormer smoker
1.09 (95%CI 1.02:1.17)1.41 (95%CI 1.01:1.98)5.15 (95%CI 1.21:21.8)
0.0120.0440.027
95% Cl, 95% confidence interval.
501
502
503504
505
FIGURES
Figure 1: Schematic processing of fibrinogen. During early wound healing, platelets are activated and recruited at the wound site where clot formation and fibrinolysis occurs. The coagulation cascade is initiated and results in thrombin-mediated conversion of soluble fibrinogen into insoluble fibrin fibers, leading to the release of the pro-peptides fibrinopeptides A (FPA) and B (FPB). Subsequently, the insoluble fibrin is crosslinked by transglutaminase factor 13 (FXIII) to strengthen the fibrin clot formation, and once hemostasis has been achieved, the fibrinolytic system, which includes a serial of enzymes and inhibitors, degrades fibrin in order to dissolve the clot, while D-dimer is plasmin-cleaved from fibrin and can be cleared by the liver.
Figure 2. Biomarker levels in COPD and control groups. A: FPA B: D-dimer C: Fibrinogen, assessed in plasma from year 1. All graphs present median ± 95% confidence interval. Differences were considered statistically significant if p<0.05 and significance levels are displayed as: *, p≤ 0.05. ****, p≤ 0.0001.
506
507
508
509510511512513514515516517518519520521522
523524525526527528529
530
Figure 3. Symptoms and endotypes related to COPD. Symptomatic (n=475) vs asymptomatic (n=765), future exacerbations (n=547) vs no future exacerbations (n=437), and emphysema (n=613) vs no emphysema (n=321) are represented for biomarkers FPA (A,D,G), D-dimer (B,E,H), and fibrinogen (C,F,I). Symptomatic participants defined as mMRC dyspnea score ≥ 2. Exacerbations defined as 1 ≥ exacerbations recorded in the year prior to biomarker measurement. Significant emphysema defined as LAA% ≥ 10%. All graphs present median ± 95% confidence interval. Differences between groups were considered statistically significant if p<0.05 and significance levels are displayed as: *, p≤ 0.05. **, p≤ 0.01. ***, p≤ 0.001, ****, p≤ 0.0001.
531532533534535536537538539
Figure 4. Mortality. Biomarker level comparison between survivors (n=953) and non-survivor (n=30) COPD participants. A: FPA; B: D-dimer; C: Fibrinogen. Kaplan-Meier Survival curve of D-dimer (D) and fibrinogen (E) with defined cutoffs. Log rank test for comparison of survival curves; D-dimer, 32 ng/ml cutoff, P<0.0001, E: Fibrinogen. 350 mg/dl cutoff, P=0.037. All graphs present median ± 95% confidence interval. Differences between survivors and non-survivors are considered statistically significant if p<0.05 and significance levels are displayed as: **, p≤ 0.01.
540
541542543544545546547548