What's new?

In this study, the number of nonsignificant findings on coronary angiography was 1.6‑fold higher in patients with atrial fibrillation (AF) than in patients without AF. Atrial fibrillation was associated with absence of significant coronary lesions and patients with AF less frequently underwent coronary angioplasty as compared with those with sinus rhythm. This raises the possibility that it may be difficult to select patients with AF for invasive procedures, given the increased findings of nonsignificant chronic coronary syndromes (CCS) on coronary angiography. The reason for these findings might be multifactorial: AF symptoms mimic CCS symptoms, ST‑segment depression during AF poorly predicts obstructive CCS, stress tests are rarely proceeded in patients with AF, and rapid rhythm makes computed tomography scan difficult to interpret. A more efficacious noninvasive diagnostic approach is needed in patients with AF and suspected CCS.

Introduction

Atrial fibrillation (AF) is the most common cardiac arrhythmia associated with excessive risk of ischemic stroke and heart failure as well as reduced life expectancy.^1,2 On the other hand, chronic coronary syndrome (CCS) remains the main cause of morbidity and mortality in an aging population. Both disease entities share common risk factors such as hypertension, diabetes, and obesity.^3-5

The prevalence of CCS in patients with AF varies from 17% to 47%,^3,6,7 whereas AF in patients with CCS is much less common, ranging from 0.2% to 5%.^8-12 Patients with AF more often have concomitant CCS as compared with sinus rhythm controls.^13,14 Additionally, AF is more frequently present in patients with either peripheral or cerebrovascular artery disease than in nonatherosclerotic individuals.¹⁵

Patients with AF in the course of arrhythmia may present with chest pain, which can be accompanied by transient ischemic‑type ST‑segment changes, sometimes with marginally elevated cardiac necrosis markers, thus mimicking symptoms of CCS.¹⁶ There are contradictory data on angiographic findings among patients with AF.¹⁴ It is important to look at the prevalence of significant CCS lesions in patients with AF compared with those with sinus rhythm to prevent unnecessary coronary angiography in patients with AF.

On the other hand, there are several studies including the ARIC (the Atherosclerosis Risk in Communities) study, the MESA (Multi‑Ethnic Study of Atherosclerosis), and the Rotterdam Study which found an association between subclinical atherosclerosis and incident AF.¹⁷

The aim of the present study was to investigate the hypothesis indicating the absence of significant coronary lesions in patients with AF as compared with those with sinus rhythm referred for elective coronary angiography. This indicates that some patients with AF may have undergone coronary angiography needlessly.

Patients and methods

We reviewed medical records of 26 985 patients hospitalized in the Department of Invasive Cardiology of the Medical University of Bialystok, Poland, who had coronary angiography performed because of exacerbated angina (recurrent chest pain, classical stable angina, long history of chest pain / angina, or other symptoms such as dyspnea) between 2007 and 2016. The Bialystok Coronary Project is an observational research project that includes hospitalized patients with AF and CCS. It is focused on diagnostics and therapy of this population. We excluded patients with acute coronary syndromes (ACS), Takotsubo cardiomyopathy, and history of ischemic heart disease, as well as those referred for coronary angiography before heart valve surgery. Prior cardiosurgical valve replacement was also the exclusion criterion. The set of extracted variables included demographic data, medical history, physical examination, resting electrocardiogram, routine transthoracic echocardiogram, coronary angiogram, and percutaneous coronary intervention (PCI). Eventually, our final study cohort included 8288 patients. Concomitant diseases were defined via the International Statistical Classification of Diseases, Tenth Revision (ICD‑10) codes, not prescription claims.

We analyzed whether history of AF was associated with the absence of significant coronary lesions on coronary angiography. We also investigated factors connected with the absence of significant coronary lesions on angiography. The study protocol conformed to the ethical guidelines of the 1975 Declaration of Helsinki, and was approved by the local ethics committee.

Results

Out of 26 985 patients admitted for elective coronary angiography, a total of 8288 patients were included in the final analysis (Figure 1, Table 1). Over 80% of study participants had hypertension, 66.9% had hyperlipidemia, and 25.7% had diabetes. Chronic kidney disease (CKD) and chronic heart failure (CHF) were diagnosed in 32.5% and 20.6% of the study population, respectively. Detailed characteristics of the study population is provided in Table 2.

A total of 1674 (20.2%) patients had AF. Paroxysmal AF was the most frequent type of the analyzed arrhythmia. However, we found only a marginally lower prevalence of permanent AF (Figure 1).

Predictors of nonsignificant coronary findings in the multivariable analysis

In the logistic regression model, the odds ratio of nonsignificant findings on coronary angiography were 1.6‑fold higher in patients with AF than in those without AF (OR, 1.57; 95% CI, 1.32–1.87; P <0.001). The indexed OR for a 5‑kg/m² increase in body mass index was 1.24 (95% CI, 1.15–1.35; P <0.001), and for a 5‑mg/ml increase in high‑density lipoprotein cholesterol concentration, 1.16 (95% CI, 1.1–1.24; P <0.001). The OR for CHF was 0.68 (95% CI, 0.56–0.83; P <0.001), other cardiovascular risk factors such as hypercholesterolemia, 0.64 (95% CI, 0.55–0.75; P <0.001), diabetes, 0.69 (95% CI, 0.59–0.81; P <0.001), hypertension, 0.68 (95% CI, 0.56–0.82; P <0.001), CKD, 0.79 (95% CI, 0.66–0.94; P <0.001) and women as compared with men, 0.39 (95% CI, 0.34–0.46; P <0.001). Additionally, for a 10‑year increase in age, the OR was 0.71 (95% CI, 0.65–0.76; P <0.001; Figure 2)

Discussion

In this study, nonsignificant findings on coronary angiography were 1.6‑fold higher in patients with AF than in patients without AF. Second, AF was associated with significantly less atherosclerotic narrowing on coronary angiography (627 [37.5%] vs 2718 [41.1%] in patients without AF; P <0.001) and patients with AF less frequently underwent coronary angioplasty as compared with the population with sinus rhythm. This raises the possibility that AF may result in difficulties in selecting patients with AF for invasive procedures, given the increased findings of nonsignificant CCS on coronary angiography.

The reason for these findings might be complex and multifactorial. Patients with AF often present with typical or atypical chest pain, transient ST‑segment depression, and elevated cardiac markers. These findings may mimic symptoms of CCS.²¹ Moreover, rapid tachycardia is often observed in patients with AF, and ST‑segment changes at these rates have been linked with myocardial ischemia. For example, Tsigkas et al²¹ reported ST‑segment depression in 38% of patients with AF and rapid rhythm and only about 50% of these patients had significant CCS on coronary angiography. ST‑segment depression may frequently occur during significant tachycardia, even without CCS, and it is not necessarily typical for cardiac ischemia if the ST segment depression is less than 2 mm.^21,22

Moreover, ST‑segment depression during rapid AF has less significant value than ST‑segment depression during exercise tests in patients with sinus rhythm.²³ Only 4% of patients without ST‑segment depression during rapid AF had positive noninvasive tests for myocardial ischemia and CCS on angiography.²³ Troponin release was shown in 15% of patients with AF and symptoms of ischemia, although CCS at angiography was absent.²⁵ In a database of patients referred for coronary angiography, history of AF correlated with no obstructive CCS.^24,26

On the other hand, Kralev et al²⁷ showed that left main stem and right coronary artery (RCA) disease was present more frequently in patients with AF compared with controls. They also suggested that significant atherosclerotic narrowing in the proximal RCA and the circumflex artery (Cx) prior to the takeoff of the atrial arteries raises the risk of AF. In our study, significant RCA and Cx disease were less common than in patients with sinus rhythm. Another analysis of 3220 patients referred for coronary angiography revealed that only 43% of patients with CCS and AF had a diseased RCA or Cx.²¹

Does AF contribute to CCS or vice versa? There is a debate on this topic and opinions are often contradictory. Motloch et al²⁸ found that the anatomical distribution of coronary artery stenoses did not contribute to AF in CCS patients; however, AF was linked to worse CCS severity, which might predispose individuals to AF by driving ischemic heart disease and changes in left ventricular function. In our study, patients with AF had less prevalent multivessel CCS. It is possible that in patients with AF, ischemia at the level of microcirculation of the myocardial muscle (rather than epicardial vessels) is responsible for the presence of typical angina. Unfortunately, our study did not explore the impact of cardiac ischemia induced by coronary small‑vessel disease.

Our findings reflect difficulties with selecting AF patients for invasive coronary investigations. In the general population, the sensitivity of exercise testing ranges from 60% to 70% and specificity is approximately 85%.²⁹ As mentioned before, ST‑segment depression is a common finding in AF patients,³⁰ and ST‑depression during rapid AF has relatively higher sensitivity and lower specificity, when compared with exercise testing. Thus, the absence of ST‑depression during rapid AF has satisfactory sensitivity for excluding significant CCS.²¹ Low positive predictive values for significant CCS of ST‑segment depression have been reported at approximately 32%.^23,24 Patients with AF are not optimal candidates for coronary CT angiography during a rapid cardiac rhythm. However, it is a good diagnostic modality for patients in sinus rhythm with low CCS probability.

Not only are patients with AF improperly referred for invasive coronary investigations, but they are at an increased risk of complications. In a study of 13 498 patients, it was shown that patients with a history of AF were more likely to have post procedural bleeding, more often needed blood transfusion, and more often had exacerbation of heart failure, cardiogenic shock, and in‑hospital mortality, but there were no differences in the incidence of postprocedural stroke or vascular complications in matched cohorts.^30,31 Similarly, several studies reported that AF patients were more likely to experience postprocedural heart failure and cardiogenic shock.^32-34 In the present study, there were 48 cases of significant vascular complications concerning femoral and radial artery related to the puncture of the site, and were more frequent in patients with AF than patients with sinus rhythm. Also, neurological disorders associated with the procedure were recorded more frequently in patients with AF than those with sinus rhythm. All CT examinations performed thereafter did not show signs of intracranial bleeding. Cardiac arrest during hospitalization was numerically higher in patients with AF. Apart from this, it was already demonstrated that AF patients with NSTEMI undergoing PCI can be predisposed to contrast induced nephropathy.³⁵