Major adverse cardiac events and mortality in chronic obstructive pulmonary disease following percutaneous coronary intervention: a systematic review and meta-analysis

Background We aimed to systematically compare Major Adverse Cardiac Events (MACEs) and mortality following Percutaneous Coronary Intervention (PCI) in patients with and without Chronic Obstructive Pulmonary Diseases (COPD) through a meta-analysis. Methods Electronic databases (Cochrane library, EMBASE and Medline/PubMed) were searched for English publications comparing in-hospital and long-term MACEs and mortality following PCI in patients with a past medical history of COPD. Statistical analysis was carried out by Revman 5.3 whereby Odds Ratio (OR) and 95% Confidence Intervals (CI) were considered the relevant parameters. Results A total number of 72,969 patients were included (7518 patients with COPD and 65,451 patients without COPD). Results of this analysis showed that in-hospital MACEs were significantly higher in the COPD group with OR: 1.40, 95% CI: 1.19–1.65; P = 0.0001, I2 = 0%. Long-term MACEs were still significantly higher in the COPD group with OR: 1.58, 95% CI: 1.38–1.81; P = 0.00001, I2 = 29%. Similarly, in-hospital and long-term mortality were significantly higher in patients with COPD, with OR: 2.25, 95% CI: 1.78–2.85; P = 0.00001, I2 = 0% and OR: 2.22, 95% CI: 1.33–3.71; P = 0.002, I2 = 97% respectively. However, the result for the long-term death was highly heterogeneous. Conclusion Since in-hospital and long-term MACEs and mortality were significantly higher following PCI in patients with versus without COPD, COPD should be considered a risk factor for the development of adverse clinical outcomes following PCI. However, the result for the long-term mortality was highly heterogeneous warranting further analysis.


Background
In this new era of 2016-2017, where the total number of smokers has increased drastically among the youngsters and the older population of males and females (smoking cigarettes began at a very young age, female smokers are on the rise, and passive smokers are becoming more and more common), Chronic Obstructive Pulmonary Diseases (COPD) might soon overcome other major worldwide causes of death [1,2]. Even though this severe chronic respiratory disease is known from decades, limited data are available on patients who suffer coronary comorbidities and who are candidates for Percutaneous Coronary Intervention (PCI).
Among the few researches which were undertaken, controversies have already been observed among COPD patients who underwent PCI. Insights from the National Heart, Lung and Blood Institute Dynamic Registry showed COPD to be associated with higher Major Adverse Cardiac Events (MACEs) and mortality rates following PCI [3]. In contrast, in a research which was carried out in a hospital in Taiwan, the authors concluded that COPD was not an independent predictor of major adverse clinical outcomes in patients with STEMI following PCI [4]. This same study also unexpectedly showed no difference in hospital mortality between COPD and non-COPD patients following PCI. Therefore, we aimed to systematically compare MACEs and mortality following PCI, in patients with and without COPD, through a meta-analysis.

Data sources and search strategies
Electronic databases (Cochrane library, EMBASE and Medline/PubMed) were searched for English publications related to COPD and PCI by typing the following words or phrases each at a time: (a) Chronic obstructive pulmonary disease and percutaneous coronary intervention (b) COPD and PCI (c) COPD and coronary angioplasty (d) COPD and myocardial infarction If an article appeared to be fully relevant, its reference list was also checked for any suitable study.

Inclusion criteria
Studies which satisfied the inclusion criteria were those studies that: (a) Were randomized trials or observational studies which compared PCI in patients with versus without COPD.
(b) Reported either MACEs or death among their clinical outcomes. (c) Involved data which were relevant to this current analysis.

Exclusion criteria
Studies were excluded if: (a) They were other types of studies apart from randomized trials or observational studies. (b) They did not involve patients with COPD. (c) They did not report either MACEs or mortality among their clinical endpoints. (d) They were either duplicates or involved the same cohort or trial.
Definitions, outcomes and follow ups COPD was defined differently in different studies. COPD, as defined in each of the study has been listed in Table 1.
The main outcomes which were analyzed included: The definitions of the outcomes were listed in Table 2 and the reported outcomes and follow up periods were summarized in Table 3.

Data extraction and review
The following data were extracted by two independent reviewers (PKB and CG): (j) Number of events in the study (COPD) as well as the control (non-COPD) groups.
Any disagreement which followed were discussed carefully with each other. However, any unsolved issue was further discussed by the third author (GMX) and a final decision was made by him. In this meta-analysis, the PRISMA guideline was followed [5].
The Newcastle Ottawa Scale (NOS) was used to assess the methodological quality of the studies (non-randomized studies) and NOS has been refined based on expertise and experience whereby it was used in several projects [6].
This NOS consists of eight items, which have been categorized in three different groups: selection, comparability, and outcome or exposure. These three groups involved several sub-items whereby 'stars' were given if Almagro 2015 COPD was defined as a post-bronchodilator forced expiratory volume in the 1st second (FEV1)/forced vital capacity (FVC) ratio < 0.70.

Berger 2004
COPD was defined by the requirement of chronic bronchodilator therapy or a forced expiratory volume in 1 s < 75% of the predicted value or a room air pO2 < 60 or a pCO2 > 50.

Campo 2013
A patient was considered to have COPD combining different sources of data: i) documented history of hospital admission for COPD; ii) treatment with pharmacologic therapies specific for COPD (e.g., inhaled steroids, inhaled anticholinergics, inhaled β-agonists or theophylline).
Enriquez 2011 COPD was defined as a history or presence of physician-diagnosed COPD. Additionally, the patients were required to be on chronic pharmacologic therapy and/or have an FEV 1 < 75% of predicted value.

Jatene 2016
The presence of COPD was determined clinically by local investigators, based on history, clinical presentation, previous examinations, and medications, recorded as COPD in the case report form at enrollment.

Nishiyama 2009
A patient was considered to have COPD if it was listed as a comorbid condition in our database and its diagnosis was confirmed by a simple test called spirometry. Such a diagnosis should be considered in any patient who has symptoms of cough, sputum production, or dyspnea (difficult or labored breathing), and/or a history of exposure to risk factors for the disease. In cases where spirometry is unavailable, the diagnosis of COPD should be made using all available tools. Clinical symptoms and signs such as abnormal shortness of breath and increased forced expiratory time can be used to arrive at the diagnosis.

Selvaraj 2005
The diagnosis of COPD was based on the clinical history or obtained from chart review and recorded as a co-morbidity in the database.
Sung 2013 COPD was defined according to one of the following criteria: (1) Information on COPD status was obtained by reviewing chart record of the need for pharmacologic therapy using bronchodilator agent; (2) Past history of a 1-s forced expiratory volume < 70% of the predicted value (by pulmonary function test); (3) Physical examination (by auscultation) showed expiratory wheezing and further confirmed by blood gas and chest radiograph (i.e., emphysematous change); or (4) Current use of bronchodilators prior to acute myocardial infarction.

Zhang 2012
A diagnosis of COPD should be considered in any patient who has symptoms of cough, sputum production, or dyspnea, and/or a history of exposure to risk factors for the disease. The diagnosis is confirmed by spirometry. The presence of a postbronchodilator FEV1 < 80% of the predicted value in combination with an FEV1/FVC < 70% confirms the presence of airflow limitation that is not fully reversible. Where spirometry is unavailable, the diagnosis of COPD should be made using all available tools.
Abbreviations: COPD chronic obstructive pulmonary disease, FEV forced expiratory volume, FVC forced vital capacity these items were present. Each item deserved one star, however, a maximum of 2 stars could be given for comparability. Selection included (maximum 4 stars): representativeness of the exposed cohort, selection of the non-exposed cohort, ascertainment of exposure, demonstrating the fact that any outcome of interest was not present at the beginning of the study.
Comparability included (maximum 2 stars): comparability of the cohorts on the basis of the design or analysis.
Outcome included (maximum 3 stars): assessment of the outcome, longer duration of the follow up, adequacy of follow-up of cohorts.
NOS assessment involved a minimum number of zero star to a maximum number of nine stars depending on the quality of the study being assessed. The total number of scores allotted were listed in Table 4.

Statistical analysis
Statistical analysis was carried out by the latest version of Revman software (5.3) whereby odds ratio (OR) and 95% confidence intervals (CI) were considered relevant.
During the subgroup analysis, heterogeneity [7] was assessed by the Q statistic test focusing on the P value with a cut-off point of 0.05. A P value less or equal to 0.05 was considered statistically significant or else, the result was considered insignificant.
Heterogeneity was also dependent on the I 2 test. A low heterogeneity was denoted by a low percentage of I 2 whereas an increasing percentage denoted an increasing heterogeneity.
The decision to use a fixed effects model (I 2 < 50%) or a random effects model (I 2 > 50%) was also dependent upon the I 2 value.
Publication bias was visually estimated through funnel plots.
Sensitivity analyses were also carried out by the exclusion method (each study was excluded one by one and a new analysis was carried out each time).
Ethical Board Review approval was not required.

Results
Flow of study selection Figure 1 represents the process of the study selection. In all, a total number of 138 publications was obtained through the electronic search. After a careful assessment of the titles and a close check of the abstracts, 112 articles were eliminated (not related to the idea of this research). Twenty-six full text articles were assessed for eligibility. Further articles were deselected since they were either case studies (2), letter to editors (1), they did not report the relevant endpoints (3), they were duplicates (8) or they were associated with the same trial or cohort (2). Finally, 10 studies [3,4,[8][9][10][11][12][13][14][15] were selected for this analysis.

General features of the studies which were included
A total number of 72,969 patients were included (7518 patients with COPD and 65,451 patients without COPD) in this analysis. Most of the studies were observational studies and the enrollment period of the patients ranged from 1997 to 2011 (Table 5).

Baseline features of the studies which were included
Baseline features have been summarized in Table 6. A mean age ranging from 66.1 to 70.0 in the COPD group and 60.9 to 66.0 in the non-COPD group were observed. Most of the patients were males with a percentage above 50% in each study. The percentage of patients with comorbidities such as hypertension, dyslipidemia, and diabetes mellitus has been listed in Table 6. Current smokers were slightly higher in the COPD group compared to the non-COPD group. Overall, almost no significant difference was observed in the baseline features of the patients in both of the groups.     Fig. 2. When MACEs were analyzed during a longer follow up period, MACEs were significantly higher in the COPD group with OR: 1.58, 95% CI: 1.38-1.81; P = 0.00001, I 2 = 29%. This time, it came to our attention that the result might have been influenced by study Enriquez 2011. Therefore, when study Enriquez 2011 was excluded and another analysis was conducted,   Fig. 3.

Mortality following PCI in COPD versus non-COPD
When mortality was analyzed, in-hospital death was significantly higher in patients with COPD, with OR: 2.25, 95% CI: 1.78-2.85; P = 0.00001, I 2 = 0% (Fig. 4). During a longer follow up period, mortality was still significantly higher in the COPD group with OR: 2.22, 95% CI: 1.33-3.71; P = 0.002, I 2 = 97%. However, even if it was fully relevant to the literature, this long-term result was highly heterogeneous (Fig. 5).
When a longer follow up was considered, MI and CR were still not significantly different with OR: 1.37, 95%   (Fig. 7).

Analysis including patients with COPD which was confirmed by a spirometry test
Another subgroup analysis was carried out including patients with COPD defined based on a spirometry test.
In addition, long-term death in this particular subgroup of patients was not significantly different with OR: 2.05, 95% CI: 0.90-4.68; P = 0.09, I 2 = 98% as shown in Fig. 9.
Sensitivity analyses obtained consistent results. Table 7 summarized the overall results of this analysis.

Publication bias
Based on the funnel plots obtained (Figs. 10 and 11), there was only little evidence of publication bias among the studies which assessed most of the clinical endpoints.

Discussion
Since the total number of patients with COPD is on the rise, this analysis aimed to compare MACEs and mortality following PCI in patients with and without COPD. Results of this analysis showed in-hospital as well as long-term MACEs to be higher in COPD patients following coronary angioplasty. Mortality was also significantly higher during the in-hospital and longer follow-up periods. However, other clinical outcomes which were analyzed (MI and CR) were not significantly different between these 2 groups of patients.
Several studies have shown an association of cardiovascular diseases with COPD. Cardiovascular diseases accounts for a high portion of mortality in such chronic pulmonary patients. A recent systematic review summarizing the existing data regarding subclinical cardiovascular events in patients with COPD on the base of identifying screening strategies in such patients showed a high subclinical burden of coronary artery disease in these chronic pulmonary patients [16]. Other studies have shown carotid-intima media thickness to significantly increase in patients with COPD suffering from coronary artery disease [17]. Recent research has also shown COPD to also be very prevalent in European patients with atrial fibrillation, and these patients were at a higher risk of several cardiovascular complications and death [18].
Well, to support the results of this current analysis, a study involving 1 of 3 tertiary medical centers in New York City showed COPD to be independently associated with long-term mortality following PCI [9]. Another study published by Selvaraj et al. and including 10,994 patients also showed a higher in-hospital and longterm mortality to be associated with COPD [14]. In addition, Insights from the National Heart, Lung and Blood Institute Dynamic Registry also showed worse prognosis in patients with COPD following PCI [3]. The authors even concluded that a lower rate of guidelines recommended class I medications which were prescribed at discharge might be hugely responsible for such higher death rates. Our results were further supported by the REAL registry [10].
Nevertheless, a few studies did not report significantly high post-angioplasty MACEs or mortality associated with COPD. The study published by Sung et al. which aimed to report the incidence and prognostic outcome in COPD patients with acute coronary syndrome (NSTEMI) showed COPD not to be an independent predictor of short and medium-term major adverse clinical outcomes in such patients following PCI [4].
Finally, according to our observations, it is recommended that special care and strict medical adherence have to be considered when managing COPD patients following PCI to avoid or reduce re-admission to the hospital, exacerbation of the obstructive disease, or any increase in mortality or MACEs following PCI. An overview of the pharmacological challenges facing physicians in the management of patients with concomitant cardiovascular disease and COPD strongly suggests that evidence-based treatment in such cases should not be changed [19]. In daily practice in clinics, obtaining the optimal titration of cardiovascular and respiratory drugs is a vital element. Early identification of co-morbidities and counselling about the harm of cigarette smoking might help to improve prognosis in such patients.

Novelty
This research contributes to a novel aspect in clinical medicine due to the fact that it is the first meta-analysis of COPD versus non-COPD and PCI. The larger population size might also contribute to its novelty. In addition, a low level of heterogeneity which was obtained among several subgroups which were analyzed could represent another new feature of this analysis.

Limitations
Limitations of this research were: -Even though a large number of patients was used, this number might still be small compared to other studies outside this scope. -Most of the studies which were included were observational studies with heterogeneous data. Therefore, the subgroup analyzing long-term mortality involved a very high level of heterogeneity.
-In addition, different studies had different follow-up periods further contributing to this high level of heterogeneity when analyzing long-term mortality. -When other clinical outcomes (MI and CR) were analyzed, in some cases, the number of patients were adjusted to avoid the influence of studies with larger number of patients. This might also have affect the results for other clinical outcomes.

Conclusion
Since in-hospital and long-term MACEs and mortality were significantly higher following PCI in patients with versus without COPD, COPD should be considered a risk factor for the development of adverse clinical outcomes following PCI. However, the result for the long-term mortality was highly heterogeneous warranting further analysis.
Abbreviations COPD: Chronic obstructive pulmonary disease; MACEs: Major adverse cardiac events; PCI: Percutaneous coronary intervention