O7.2 Cardiac disease

COPD patients possess an increased burden of cardiovascular disease (CVD), cardiac arrhythmia and heart failure when compared to the normal population . Chen’s systematic review and meta-analysis pooled the results from 29 datasets and reported that COPD patients were more likely to be diagnosed with cardiovascular disease (ischaemic heart disease, dysrhythmia, heart failure, pulmonary circulatory disorders and arterial diseases) than controls (OR 2.46, CI 2.02-3.00, p<0.0001). This result was mainly driven by angina (OR 8.16) (Chen 2015) [evidence level III-2].  In addition, Feary’s study of 1,204,100 patients who were followed for a median of 895 days in the primary care setting, also demonstrated an association of COPD with increased rates of first myocardial infarction (MI) (HR 10.34, CI 3.28 to 32.6), and stroke (HR 3.44, CI 0.85 to 13.84), stratified by age and adjusted for gender and smoking status (Feary 2010) [evidence level III-2].

CVD is an important cause of mortality and hospital presentations in COPD, even affecting those with mild disease. In addition to the high individual prevalences of COPD and CVD, these conditions share conventional risk factors of advanced age, smoking, low socioeconomic status (SES) and sedentary lifestyle. Systemic inflammation, autonomic dysregulation, hypoxia, acidosis and haemodynamic derangements are likely to also contribute (Fuschillo 2012). Independent of smoking and other risk factors,  impaired lung function per se is a major risk factor for CVD and arrhythmia (on par with hypercholesterolaemia), with the relationship being strongest for fatal events (Hole 1996), (Agarwal 2012) [evidence level III-2]. Arterial stiffness has been proposed as one potential mechanism for this excess of CVD as it strongly predicts CVD events and mortality. In COPD, arterial stiffness increases during exacerbation and is associated with COPD severity (measured as airflow obstruction or degree of emphysema), inflammation, oxidative stress and sympathetic nervous system (SNS) tone.  COPD also predicted lipid core (OR 2, CI 1.25-3.69, p=0.0058),  plaque component vulnerable to rupture (Lahousse 2013) [evidence level III-2], which increases risk of CVD events.

One review (Vivodtzev 2014)[evidence III-2] demonstrates results across multiple studies showing increased arterial stiffness (n=18), endothelial dysfunction (n=4) and carotid intima-media thickness (n=3) in COPD. Several trials showed a gradated effect, with an increase in COPD subjects compared with non-COPD smokers, and in smokers compared with healthy non-smokers. This group also summarised preliminary data suggesting that current therapeutic interventions may impact on increased arterial stiffness; included studies reported a statistically significant improvement in arterial stiffness after standard pulmonary rehabilitation, after treatment with combination ICS/ LABA or LAMA, and possible improvement with supplemental oxygen.

Konecny’s group sought to explore cardiac arrhythmia as a potential source of the excess CVD mortality in COPD in a retrospective record review of 7441 subjects who underwent 24 hour Holter monitoring and spirometry during the course of clinical assessment. The 3121 (49%) COPD patients demonstrated more arrhythmias than those without COPD; atrial fibrillation/flutter were identified in 23.3% versus 11% (p<0.0001), and non-sustained ventricular tachycardia in 13% versus 5.9% (p<0.0001). Both results remained statistically significant after adjustment for multiple confounders (Konecny 2014) [evidence level III-2]. The study population was a highly select group, which potentially limits the broad application of the results. However, the study reports a “COPD dose effect”, based on spirometry criteria, which adds weight to its conclusions

Medications used in the treatment of COPD also have potential to impact on cardiac morbidity and mortality, due to intrinsic effects on chronotropy and muscle action potentials or due to side effects such as hypokalaemia. Medications implicated include beta-agonist and antimuscarinic bronchodilators and methylxanthines. More recently, macrolide antibiotics, which in chronic dosing have been shown to reduce respiratory exacerbations, have been added to the list, due to an association with QT prolongation and bradycardia. Randomised controlled trials (RCT) of chronically dosed azithromycin have not demonstrated adverse cardiac effects in the clinical setting, particularly when known drug interactions are avoided. Likewise, for most inhaled bronchodilators, when used at therapeutic dose in stable COPD, there are no proven adverse effects on safety. Despite being common clinical practice, there is less evidence about the safety of high dose, combined bronchodilator therapy in the setting of an exacerbation of COPD.

Markers of cardiac involvement during an exacerbation of COPD may be an important determinant of short-term prognosis. In a study of 250 consecutive admissions with an exacerbation of COPD and no evidence of acute cardiac disease over 12 months, elevated NT-pro BNP >220 pmol/L and troponin T >0.03 were present in 27% and 16.7% subjects and predicted 30 day mortality (OR 9, CI 3.1-26.2) and (OR 6.3, CI 2.4 – 16.5), respectively, after adjustment for other mortality predictors. Elevated troponin T level lost significance with both cardiac biomarkers included in the model, although the mortality association was additive for subjects in whom both biomarker levels were elevated (Chang 2011) [evidence level III-2].

Another prospective cohort study (Hoiseth 2012) [evidence level III-2] reported results for 99 COPD patients with 217 exacerbations and a median follow up duration of 1.9 years and found NT-pro BNP to be an independent risk factor for mortality after an exacerbation of COPD. Dividing NT-pro BNP levels into tertiles, mortality rates were 8.6, 35 and 62 per 100 patient years (age-adjusted log-rank p<0.0001) and, compared to the lowest tertile, adjusted HR for death were 2.4(CI 0.95-6.0) and 3.2 (CI 1.3-8.1) in the intermediate and highest tertiles, respectively

Preliminary research suggests that cardiac pathology contributes to a proportion of exacerbations of COPD. A small study (Bhatt 2012) [evidence level III-2] investigated a potential role for arrhythmia in an exacerbation; comparing ECG indices during an exacerbation with stable state. They reported that P wave duration was more variable during exacerbation. Moreover, “frequent exacerbator patients” (defined as two or more exacerbations of COPD within 12 months) had increases in ECG PR interval during stable state compared with “infrequent exacerbators”. Although methodology was not robust, the results probably justify further research into this issue. In addition, Abusaid’s group (Abusaid 2009)[evidence level III-2] proposed a contributory role for diastolic dysfunction (DD) in an exacerbation of COPD. Their retrospective single centre cohort study reported that DD was associated with prolonged length of hospital stay (4.02 versus 3.24 days, p= 0.005) and increased frequency of hospitalisation for an exacerbation (1.28 versus 0.67 per patient year, p=0.0067) in the absence of traditional precipitating factors.

Donaldson et al (Donaldson 2010) sought to quantify the increased risk of cardiac adverse events (CAE) associated with an exacerbation of COPD. Using self-controlled case series methodology, they identified 25,857 COPD patients and their CAEs (524 myocardial infarctions (MI) in 426 patients and 633 ischaemic strokes in 482 patients) using health care database diagnostic codes and defining an exacerbation by receipt of systemic corticosteroid course (at minimum daily dose) and/or specified antibiotics. Comparing CAE incidence during the period immediately after an exacerbation with that in stable state and adjusting for seasonality, they demonstrated increased risk for MI (RR 2.27, CI 1.1 to 4.7) in the five days following exacerbation onset, if combined antibiotics and corticosteroids were required and increased risk for stroke (RR 1.26, CI 1.0 to 1.6) for 49 days, for an exacerbation requiring antibiotics only [evidence level III-2].

Two studies attempted to evaluate the extra morbidity burden conferred by heart disease in COPD [evidence level III-2]. De Miguel Diez (de Miguel-Diez 2010) recruited patients meeting diagnostic criteria for stable COPD from the Spanish primary health care setting and assessed chronic morbidity and health resource utilisation according to the presence of ICD-9 codes for heart disease. Of 9,390 COPD patients, 18.8% had documented heart disease. Compared to patients without heart disease this group had worse lung function, worse quality of life (QoL), required more respiratory medications, consumed more health resources and generated greater expenses differences which were all statistically significant. The authors identified admission duration as a major contributor to increased costs in these patients [evidence level III-2]. In the study by Patel’s group (Patel 2012), data from the London Cohort (1995 – 2009), comprising prospectively collected exacerbation data via symptom diaries from 386 subjects with COPD (as defined by spirometry) and at least 12 months’ diary data. Health status assessment occurred whilst in stable phase and comparison was made regarding frequency and duration of an exacerbation of COPD between patients with and without ischaemic heart disease (IHD). The 16% of the cohort with IHD scored worse on QOL assessment (St George Respiratory questionnaire), MRC dyspnoea scale and six minute walk distance. There was no difference in frequency of respiratory exacerbations or the need for antibiotics and systemic corticosteroid therapy. However, subjects with IHD recovered more slowly and so endured more days with increased levels of symptoms. The subjects did not differ in COPD treatments received, but the authors provided no information on treatments received for IHD [evidence III–2].

Conversely, two studies have looked at the impact of COPD on outcomes after first MI (Bursi 2010, Andell 2014) [evidence level III–2]. Prevalence of clinically diagnosed COPD in these studies was 12% and 6%, respectively. In Bursi’s American cohort, COPD prevalence increased significantly over time, and was associated with increased mortality (adjusted HR 1.3, CI 1.1 to 1.54), independent of age, traditional indicators of poor prognosis and comorbidities. Likewise, Andell’s group reported worse outcomes for COPD patients in their Swedish cohort: one year mortality 1.14 (1.07 – 1.21), and development of heart failure 1.35 (1.24 – 1.47). Bursi’s group found that the association of COPD with survival remained unchanged over time, despite an overall decline in mortality after MI (seen with improvements in medical care). The difference in clinical presentation and therapeutic interventions received reported by Andell’s group, may partially explain the discrepant outcomes seen in COPD patients (COPD patients were more likely to present with atypical symptoms, less likely to undergo percutaneous revascularisation procedures or to receive secondary prevention medications).