Antiplatelet Treatment Reduces All-Cause Mortality in COPD Patients: A Systematic Review and Meta-Analysis

ABSTRACT

Previous studies clearly showed that patients with chronic obstructive pulmonary disease (COPD) are at high risk for cardiovascular events. Platelet activation is significantly heightened in these patients, probably because of a chronic inflammatory status. Nevertheless, it is unclear whether antiplatelet treatment may contribute to reduce all-cause mortality in COPD patients. To clarify this issue, we performed a systematic review and meta-analysis including patients with COPD (outpatients or admitted to hospital for acute exacerbation). The primary endpoint was all-cause mortality. We considered studies stratifying the study population according the administration or not of antiplatelet therapy and reporting its relationship with the primary endpoint. Overall, 5 studies including 11117 COPD patients were considered (of those 3069 patients were with acute exacerbation of COPD). IHD was present in 33% of COPD patients [95%CI 31%–35%). Antiplatelet therapy administration was common (47%, 95%CI 46%–48%), ranging from 26% to 61%. Of note, IHD was considered as confounding factor at multivariable analysis in all studies. All-cause mortality was significantly lower in COPD patients receiving antiplatelet treatment (OR 0.81; 95%CI 0.75–0.88). The data was consistent both in outpatients and in those with acute exacerbation of COPD. The pooled studies analysis showed a very low heterogeneity (I² : 8%). Additional analyses (meta-regression) showed that antiplatelet therapy administration was effective independently (to potential confounding factors as IHD, cardiovascular drugs and cardiovascular risk factors. In conclusion, our meta-analysis suggested that antiplatelet therapy might significantly contribute to reduce all-cause mortality in COPD patients.

KEYWORDS:

• aspirin
• antiplatelet agents
• chronic obstructive pulmonary disease
• mortality

Introduction

Ischemic heart disease (IHD) and chronic obstructive pulmonary disease (COPD) are the first and the fourth cause of death in Western countries, respectively (1, 2). COPD patients are at high risk for cardiovascular (CV) events and IHD represents their cause of death in at least one-third of cases (1–3). During acute exacerbation of COPD (AECOPD), cardiac troponin elevation, even without symptoms and signs of concomitant acute coronary syndrome (ACS), is significantly related to increased risk of all-cause mortality (4). IHD and COPD share common pathological correlates as chronic inflammation, endothelial dysfunction and increased platelet reactivity (1). In COPD patients, increased levels of circulating platelet-monocyte aggregates, heightened on-treatment platelet reactivity, and abnormalities in the number and mean volume of platelets have been reported (5–8). Accordingly, some Authors suggested a potential benefit from antiplatelet therapy administration in COPD patients, independently from the presence of IHD (8). Nevertheless, available data are conflicting and no randomized clinical trial has been performed to address this issue (1,9). Thus, we performed a systematic review to select all studies investigating this issue and a meta-analysis to determine whether antiplatelet therapy administration could contribute to reduce all-cause mortality in COPD patients.

Methods

We performed a systematic review and meta-analysis following Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) amendment to the Quality of Reporting of Meta-analyses (QUOROM) statement and recommendations from The Cochrane Collaboration and Meta-analysis of Observational Studies in Epidemiology (MOOSE) (10–13).

Search strategy

Appropriate articles were found following MESH strategy and searching in MEDLINE, Cochrane Library, Google Scholar and BioMed Central database (Figure 1). Only those published in English were considered. The terms searched were: (“pulmonary disease, chronic obstructive”(MeSH Terms) OR (“pulmonary”AND “disease”AND “chronic”AND “obstructive”) OR “chronic obstructive pulmonary disease” OR “copd”) AND (“cardiovascular”AND “therapy”) OR “cardiovascular therapy” OR “clopidogrel” OR “aspirin”(MeSH Terms) OR “ticagrelor” OR “prasugrel” OR “ticlopidine”(MeSH Terms) OR “antiplatelet”).

Figure 1. Outline of search strategy. n = number.

The research was carried out between January 2014 and March 2015. Two independent reviewers (GC, SB) analyzed the articles in terms of title and abstract, valuing which articles have to be analyzed as full paper. All reviewers agreed to the final number of studies to be included in the analysis (Figure 1). The following inclusion criteria were necessary: i) observational studies regarding patients with diagnosis of COPD; ii) including more than 50 patients; iii) stratification of patients according the administration or not of antiplatelet therapy (aspirin and/or clopidogrel and/or ticagrelor and/or prasugrel) iv) reporting all-cause mortality, expressed as hazard ratio (HR) or odds ratio (OR) at multivariate analysis. Duplicate, interventional or on animals studies were excluded. Similarly, studies dated before 1990 were excluded.

Data extraction, definition and end-points

The independent reviewers (GC, SB) completed the database, related to authors, journal, year of publication, baseline characteristics of the population and COPD phase (stable vs. acute exacerbation), as well as all variants considered at univariate and multivariate analysis (Tables 1 and 2). The primary endpoint is incidence of all-cause mortality. We focused on studies reporting the HR or OR values for all-cause mortality in patients receiving antiplatelet therapy as compared to those without. As complementary analysis, we stratified patients according to COPD state and repeated the analysis.

Table 1. Baseline data and cardiovascular risk factors of study population.

				Age					Beta-	ACE-inhibitors		Mean
	Study	Pts	Female	(years)	Diabetes	Hypertension	CVD	CVD	blockers	or ARB	Statins	follow-up
References	design	(n)	(%)	mean	(n, %)	(n, %)	(n, %)	definition	(n, %)	(n, %)	(n, %)	length
Soyseth et al. (2007)	R	854	52%	71	–	179 (21)	243 (28)	IHD	111 (13)	179 (21)	736 (86)	1.9 years
Short et al. (2011)	R	5977	49%	69	855 (14)	–	2878 (48)	Hospital admission for IHD, HF, PVD	796 (13)	–	–	4 years
Ekstrom et al. (2012)	P	2249	59%	75	291(13)	–	1494 (66)	At least one of hypertension, HF, IHD, PVD, CVA	829 (37)	763 (34)	431 (19)	1 years
Harrison et al. (2014)	P	1343	51%	72 *	203(15)	–	377 (28) 70 (5) 188 (14)	Prior MI Prior surgical CR Stable angina	228 (17)	666 (50)	566 (42)	1 year
Campo et al. (2015)	R	694	47%	77	172(25)	376 (54)	235 (34)	IHD	203 (29)	350 (50)	441 (64)	1.9 years

Pts: patients. n: number. R: retrospective. P: prospective. CVD: cardiovascular disease. ACE: angiotensin converting enzyme. ARB: angiotensing II receptor blocker. IHD: ischemic heart disease. HF: heart failure. PVD: peripheral vascular disease. CVA: cerebrovascular accident. MI: myocardial infarction. CR: coronary revascularization. *: median.

Table 2. Main features regarding design and COPD diagnosis criteria.

References	Pts receiving antiplatelets (n, %)	Sources of data	COPD diagnosis criteria	Pts with AECOPD at the time of the enrolment	AECOPD diagnosis criteria	FEV1	Adjusting factors in univariate and multivariate analysis
Soyseth et al. (2007)	224 (26)	Medical reports	Previous clinical diagnosis, spirometry available in 72% of cases.	Y	ICD codes for AECOPD, verified by a physician	Y	age, sex, DM, FEV1, smoking, IHD, CV drugs, RD, hypertension, HF, AF, TE, cancer.
Short et al. (2011)	—	Scottish morbidity records and General Register Office for Scotland using ICD codes.	Prior COPD diagnosis confirmed by spirometry	N	—	Y	age, sex, DM, FEV1, smoking, CV drugs, CV and respiratory hospital readmission, resting arterial O2 sat, deprivation index.
Ekstrom et al. (2012)	887 (39)	National Swedish patient register using ICD codes.	Prior COPD diagnosis confirmed by spirometry at the baseline	N	—	Y	age, sex, DM, FEV1, smoking, hospitalization, CV disease, CV drugs, RD, PaO2 and CO2 air, BMI, WHO performance status, anemia, renal failure.
Harrison et al. (2014)	713 (53)	Medical records and hospital records and morbidity database using ICD codes.	Prior COPD confirmed by spirometry in stable phase	Y	As reported in the EXODUS registry, a multicentre audit database	Y	age, sex, FEV1, smoking, IHD, LTOT, CV disease, platelet count, CV drugs, cancer, BMI, MRC dyspnoea score, oxygen therapy, acidosis, PaO2 and CO2, BAP65, long-term care.
Campo et al. (2015)	424 (61)	Programmazione e controllo di gestione, Statistica Sanitaria using ICD codes.	Prior COPD confirmed by spirometry	Y	Acute change of baseline dyspnea, caught, sputum production	N	age, sex, DM, hypertension, dyslipidemia, smoking, IHD history, WBC, Hb, PLT, fibrinogen, CRP, CV drugs, arterial blood gases, creatinine, troponin elevation.

Pts: patients. COPD: chronic obstructive pulmonary disease. AECOPD: acute exacerbation of COPD. FEV₁: forced expiratory volume 1 second. Y: yes. N: no. ICD: international classification of disease. DM: diabetes mellitus. IHD: ischemic heart disease. CV cardiovascular. RD: respiratory drugs. HF: heart failure. AF: atrial fibrillation. TE: thrombo-embolism. BMI: body mass index. WHO: world health organization. BNP: brain natriuretic peptide. LTOT: long-term oxygen therapy. WBC: white blood cells. Hb: haemoglobin. PLT: platelets. CRP: C-reactive protein.

Internal validity and quality appraisal

Two unblinded reviewers (RP, FDA) evaluated the quality of included studies using pre-specified electronic forms that were piloted over the first 3 cases. The divergences were resolved by consensus. Modifying the MOOSE item list to include the specific features of included studies, we separately abstracted and appraised study design, setting, data source, and statistical methods for multivariable analysis, as well as, in keeping with The Cochrane Collaboration approach, the risk of analytical, selection, adjudication, detection, and attrition bias (expressed as low, moderate, or high risk of bias, as well as incomplete reporting leading to inability to ascertain the underlying risk of bias).

Data analysis and synthesis

Continuous variables are reported as mean (± standard deviation) or median (interquartile range). Categorical variables are expressed as number and percentage (%). Point estimates and standard errors were extracted from individual studies and were combined by the generic inverse variance method (14), computing risk estimates with 95% confidence intervals according to logarithmic transformation of the hazard measures, using RevMan 5 (The Cochrane Collaboration, The Nordic Cochrane Centre, Copenhagen, Denmark).

Considering the high likelihood of between-study variance, we used a random effect model. Statistical heterogeneity was assessed using the Cochran's Q-test. This statistic is complemented with the I² statistic, which quantifies the proportion of total variation across studies that is due to heterogeneity rather than chance. A value of I² of 0 to 25% represents insignificant heterogeneity, 26 to 50% low heterogeneity, 51 to 75% moderate heterogeneity, and >75% high heterogeneity (15). Finally, random effect meta-regression analysis was performed to assess the effect of some potential confounding factors (beta-blocker, statin, ACE-inhibitor, previous history of cardiovascular disease, smoking habits, diabetes mellitus, female sex) on results. Meta-regression analysis was done by an independent reviewer (FG), using Comprehensive Meta-analysis 2.0 (Biostat Inc., Englewood, NJ, USA).

Results

Search results, study selection and patient characteristics

After removal of duplicates and first evaluation of the title and abstract, 10 full-text articles were assessed. After further evaluation of inclusion and exclusion criteria, we selected 5 studies (8, 9, 16–18). In all studies, COPD diagnosis was present before the enrolment and it was based on spirometry data. Of note, in the study of Soyseth et al., ICD codes for AECOPD were used to screen the population and spirometry was available in the 72% of patients (18). A total of 11,117 COPD patients were included (Table 1) (8, 9, 16–18). The 51% were female, with a mean age of 71 years. Overall, 8048 (72%) patients were outpatients with COPD, the remaining 3069 (28%) with acute exacerbation of COPD (AECOPD) (Table 2).

Baseline data and the occurrence of CV risk factors were reported in Table 1. Of note, the presence of IHD is openly stated only in three studies (Table 1) (8, 9, 18). IHD was present in the 33% of COPD patients (95%CI 31%-35%). In the remaining two studies, the Authors reported the presence of CV disease including both IHD and other cardiac disorders (Table 1) (16, 17). IHD and/or CV disease were in all cases included as confounding factors in multivariable analysis (Table 2). All studies openly reported the number of patients receiving antiplatelet therapy, with the exception of the one by Short et al. (Table 2) (17). Antiplatelet therapy administration was common (47%, 95%CI 46%-48%), ranging from 26% to 61% (Table 2). Aspirin was the only agent administered in the studies of Short et al. (17) and Soyseth et al. (18). Aspirin was the most frequent antiplatelet agent also in the other studies (89%, 88% and 92% in Campo et al. (9), Harrison et al. (8) and Ekstrom et al. (16).

Quantitative analysis

All-cause mortality was significantly lower in COPD patients receiving antiplatelet treatment (OR 0.81; 95%CI 0.75–0.88) (Figure 2). Importantly, the pooled studies analysis showed a very low heterogeneity (I² : 8%) (Figure 2). Similar findings were obtained limiting the analysis to patients admitted to hospital for AECOPD (OR 0.78; 95%CI 0.62–0.99; I² = 40%) or to COPD outpatients (OR 0.82; 95% CI 0.76-0.89; I² = 0%). Finally, random effect meta-regression disclosed no significant interaction between the reduced risk of all-cause mortality associated with antiplatelet agents and the following variables: smoking habit (β = 0.05, p = 0.5), female sex (β = −0.002, p = 0.9), diabetes mellitus (β = −0.009, p = 0.8), history of cardiovascular disease (β = −0.03, p = 0.3), beta-blocker (β = 0.06, p = 0.6), statin (β = −0.18, p = 0.7) and ACE-inhibitor (β = −0.08, p = 0.5).

Figure 2. Meta-analysis of studies testing the association between antiplatelet treatment and all-cause mortality in COPD patients. Data are displayed for each available study. Error bars represent 95% confidence intervals. SE: standard error. CI: confidence interval.

Discussion

At the best of our knowledge, this is the first study-level meta-analysis trying to assess the relationship between antiplatelet therapy and all-cause mortality in patients with COPD, independently to the presence of IHD.

COPD is characterized by a chronic inflammatory status (1), both in patients with stable phase of the disease and during/after acute exacerbation (2). Patients with concomitant IHD and COPD showed a worse outcome as compared to patients without COPD (1, 19). Similarly, patients with COPD are at higher risk of MI, especially after AECOPD (1, 20). Chronic inflammation significantly contributes to platelet activation (1,5). Recently, we reported significantly higher platelet reactivity in patients with COPD as compared to those without (6). Similarly, other Authors remarked the presence of abnormal platelet function in COPD patients (1, 2, 5, 7, 8).

Overall, these findings represent the pathophysiological support for a potential benefit of antiplatelet therapy in COPD patients (1, 2, 5). No prospective and/or randomized clinical trials tested this hypothesis. We have data only from observational studies. To overcome these biases and to obtain preliminary results that may help in the planning of future studies, we performed a study-level meta-analysis. We analyzed data from five studies. The mean age of our study population was about 70 years. This is not surprising considering that both COPD and IHD are typically age-related conditions. We found that the treatment with antiplatelet agents may be beneficial in improving outcome of COPD patients. Of note, aspirin was the most used antiplatelet agents I all studies (91%, 95%CI 89%–92%).

We are not able to discriminate which mechanisms are involved in the observed benefit. One possible explanation could be the modulation, by antiplatelet agents, of the inflammatory pathways secondary to the heightened activation state of platelet in COPD patients. Otherwise, the direct reduction by antiplatelet agents of the hemostatic process may justify the better outcome. Interestingly, the beneficial effect of antiplatelet agents seems to be consistent across different phases of COPD disease. We observed it in both outpatients and in those admitted to hospital for AECOPD. Although our results are preliminary and should be considered “hypothesis-generating,” we found that antiplatelet treatment may be beneficial in COPD patients. These results should not be considered conclusive and exhaustive, but they strongly support further investigations and randomized clinical trials to confirm and validate our findings.^{Figure 3}

Study limitations

Several limitations should be considered in the interpretation of our data. First, our data were extrapolated from observational studies designed to test different hypothesis. Secondly, we observed significant differences between study populations and it could be a limitation in the comparison and interpretation of data. Third, we are not able to discriminate if only patients with IHD drive the observed benefit. As clearly stated, the definition of IHD was not reported in all studies and several discrepancies in the definition of CV disease were present (Table 1). Nevertheless, the percentage of antiplatelet therapy administration was superior to that of patients with IHD. It is plausible that all IHD patients received antiplatelet agents and so the observed benefit could not be related to differences in their treatment.

In addition, in all studies IHD and/or CV disease and/or hospital admission for CV disorders were included in the multivariable analysis (Table 2). Our HR and/or OR values describing the relationship between antiplatelet treatment and all-cause mortality were corrected for these important confounding factors. Similar considerations should be made for other cardiac specific medications such as beta-blockers, ace-inhibitors or statins. Nevertheless, their prescriptions were quite low (Table 1), and in all studies were included in the multivariable analysis (Table 2). In addition, we performed meta-regression analysis to further support our hypothesis. The overall result is consistent.

The benefit of antiplatelet agents on all-cause mortality seems to be independent to these potential confounding factors. Data regarding the cardiovascular death could further support our findings. Unfortunately, this information is not available in all studies. Similarly, we are aware that other potential variables (COPD severity, social status, overall quality of care) should be considered. Our study-level meta-analysis should be regarded as hypothesis-generating. Further studies are clearly on demand to confirm our preliminary data. Finally, it is important to highlight that the overall publication bias in our meta-analysis was low (Figure 3).

Figure 3. Funnel plot of standard error.

Conclusions

Our meta-analysis suggested that antiplatelet therapy might significantly contribute to reduce all-cause mortality in COPD patients, independently to presence or not of IHD.

Declaration of Interest Statement

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.