Prognostic implication of monocytes in atrial fibrillation: The West Birmingham Atrial Fibrillation Project

Background and objectives High monocyte counts are related to adverse outcomes in cardiovascular disease. Their role in prognostication in patients with atrial fibrillation (AF) is unknown. We investigated whether monocyte counts are useful as a marker of prognosis in patients with AF. Methods Monocyte counts were obtained from blood samples in 881 AF patients. Study outcomes were (i) all-cause death; (ii) major adverse cardiovascular events; (iii) stroke, TIA or other systemic embolism (SSE); and (iv) major bleeding. Results Median follow up was 7.2 years; 44% of patients died, 48% developed MACE; 9% had SSE and 5% had major bleeding. On Cox regression, after adjustment for CHA2DS2-VASc score, the highest quartile of monocyte counts (i.e., ≥580 μL vs. other quartiles) was associated with increased risk of death (hazard ratio [HR] 1.64, 95% confidence interval [CI] 1.31–2.05, p<0.001) and MACE (HR 1.58, 95% CI 1.28–1.96, p<0.001). Persistent monocyte levels ≥580 per μL during follow up were associated with further increase in risk of death (HR 1.52, 95% CI 1.10–2.11, p = 0.01) and MACE (HR 1.54, 95% CI 1.13–2.09, p = 0.006). Persistent monocyte levels ≥580 per μL during were associated with a significant increase in major bleeding events (HR 2.77, 95% CI 1.36–5.67, p = 0.005, after adjustment for HAS-BLED score). Conclusion High monocyte counts independently predict the occurrence of MACE, major bleeding and mortality, but not SSE. Understanding the pathophysiological mechanisms involved would help understand the relationships between monocytes, and adverse thrombotic and bleeding outcomes in AF patients.


Introduction
Circulating monocytes have been closely linked to outcomes in patients with cardiovascular disease [1]. The primary role of monocytes is to detect and replenish the stores of macrophages and dendritic cells, and to provide phagocytosis of pathogens [2]. Monocytes make up to 8% of the peripheral blood white cells and play a central role in the host response to infective agents, such as bacteria and viruses. Additionally, monocytes modulate the inflammatory processes, producing both pro-and anti-inflammatory cytokines and developing macrophages with proand anti-inflammatory phenotype [3].
Research into the role of inflammation in cardiovascular disease has found increased monocyte counts in patients with a myocardial infarction and other forms of acute cardiovascular pathology [1,4,5]. Monocyte-derived "foam cell" macrophages are a substrate for atherosclerosis and thus facilitate the progress to myocardial infarction. Overall, monocytes have been used as indicators of prognosis in humans with their high numbers being associated with increased risk of recurrent myocardial infarction, hospitalization and cardiac death [1]. Available data indicate that monocyte mobilization in acute cardiac disease does not simply reflect a response to cardiac damage, as they are actively involved in the pathological processes themselves [6,7].
Introduction of oral anticoagulation has dramatically reduced the risk of stroke. However, the contemporary outcomes in Atrial Fibrillation (AF) are increasingly driven by non-embolic events and complication of oral anticoagulation (bleeding). The role of monocytes in determining outcomes amongst AF patients is unknown. Such data could help identify patients at high risk of adverse outcomes and subsequently highlight those in need of targeted therapy to control cardiovascular risk factors as well as novel therapeutic strategies aimed at modulating the inflammatory response in AF patients.
Our aim was to investigate the prognostic roles of monocyte counts in AF for the occurrence of death, major adverse cardiovascular events (MACE), stroke and systemic embolism (SSE), as well as significant bleeding events in a longer term observational study cohort of AF patients. We tested the hypothesis that high monocyte counts confer an increased risk of these adverse outcome.

Methods
Patients with documented AF were recruited from outpatient Atrial fibrillation clinics in Sandwell and West Birmingham Hospitals Trust and Oral anticoagulation clinics in the West Birmingham area between August 2008 and August 2010 (Table 1). There has been no patient selection based on co-morbidities. All recruited patients were included into the analysis if they had data on monocyte counts after the diagnosis of AF (36 [4%] of the patients were excluded for this reason). A total of 881 patients with data on monocyte counts were included in this analysis. Data on blood monocyte counts during routine appointments after a diagnosis of AF were obtained from clinical records. Monocyte data from acute admissions were not included. Follow up monocyte data were collected from routine appointments at one-year time or nearest later date and were available for 670 patients.
Patients were prospectively followed up with clinical outcomes recorded based on clinical records. The study outcomes were (i) any-cause death; (ii) major adverse cardiovascular events (MACE: first event of death, myocardial infarction, ischemic stroke, transient ischemic attacks (TIA) or other systemic embolism); (iii) stroke, TIA or other systemic embolism (SSE); and (iv) major bleeding (including hemorrhagic stroke). The study was approved by the Black Country Research and Ethics Committee, UK and the study complied with the Declaration of Helsinki. All patients provided written informed consent for participation in the study.

Results
We included 881 patients with a median follow up of 7.2 (4.8-10) years. The index monocyte assessment was done 0.7 (0.1-3.0) years after the initial diagnosis of AF. Patient demographic and clinical characteristics are presented in Table 1. At the time of the recruitment 525 (59%) patients received OAC. Follow up monocyte data were available for 670 patients with 23 (17-33) months duration from the first sample. Persistent monocyte count !580 per μL was recorded in 84 (13%) patients.

All-cause death
Three hundred and ninety (44%) patients died during follow up (  (Fig 3).

Interactions of monocyte status with age, ethnicity and AF characteristics
On additional regression models, there were no significant interactions between the monocyte status and use of oral anticoagulants, non-White ethnicity, advanced age and presence of long standing persistent or permanent AF for all tested outcomes ( Table 2).

Translational perspectives
The precise factors driving risk of death and MACE in AF patients with higher (but mostly still within normal limit) monocyte counts are unclear. The likely mechanisms include the inflammatory and profibrotic properties of monocytes. Admittedly the mechanistic links between monocytes and unfavorable outcomes could not be established in this analysis and monocytes could be bystanders of ongoing pathological processes rather than direct cause of the outcome. It is likely contributed by the fact that monocytes are a major source of inflammatory cytokines and reactive oxygen species in the circulation. [8] With regards to bleeding, fibrinolysis is tightly controlled by a series of cofactors, inhibitors, and receptors. [9] Plasmin is the primary fibrinolytic enzyme, and is activated from plasminogen by either of two primary serine proteases, tissue-type plasminogen activator and urokinase-type plasminogen activator. Urokinase-type plasminogen activator is primarily produced by monocytes and macrophages. [10,11] Excessive elevation in monocyte counts or their abnormal functional state may shift the tightly controlled local hemostasis state towards the bleeding state. Monocyte capacity to produce matrix metalloproteases may further amplify the process and thus overcome procoagulant monocyte properties. [12][13][14] Physicians in clinical practice should be aware that higher monocyte counts signal high risk of major bleeding in the AF population. This can aid in the decision-making process when considering oral anticoagulation and measures to reduce bleeding risk. Future areas of research into this field should aim to delineate possible roles of monocyte subsets in bleeding risk and possible therapeutic measures to attenuate this bleeding risk.

Conclusion
High monocyte counts independently predict the occurrence of MACE, major bleeding and mortality, but not SSE. This study demonstrates for the first time a strong association between monocyte counts and the subsequent adverse outcomes of all cause death, MACE and major bleeding in patients with AF. The similar rates of stroke/SSE in patients on or without oral anticoagulation may be reflective of the younger average age of the population not receiving the treatment. Furthermore, the lower CHA 2 DS 2 -VASc score in this group reflects a lower overall baseline stroke risk. The powerful associations were present despite adjustment for well-established scores for prognostication of thromboembolic and bleeding complications in AF. Indeed, the top quartile of monocyte counts were associated with a 64% rise in all-cause death and 58% increase in risk of MACE even after adjustment for the CHA 2 DS 2 -VASc score, which accounts for major prognosticators of unfavorable cardiovascular outcomes and age. Specifically, persistent monocyte levels at 580 per μL or above were associated with over 50% increase in risk of death or MACE irrespectively of the baseline monocyte status.
The study also demonstrates for the first time that persistently high monocyte levels (i.e., !580 per μL) are associated with almost 3-fold increase in major bleeding events. These data support clinical relevance of previous, mostly experimental data demonstrating pro-fibrinolytic activity of monocytes. [12] Of note, risk of stroke was not related to monocyte counts, despite the increased rates of other cardiovascular events. This may reflect the multidirectional effects of monocytes, including excessive inflammation, driving risk of myocardial infarction and death, and exaggerated fibrinolysis and connective tissue degradation predisposing to bleeding. Accumulating data point toward the potential implication monocyte in bleeding complications. This can be largely mediated by their three functions: production of matrix metalloproteases, that can break the barrier between the vascular lumen and the surrounding space, (ii) clot resolution via phagocytosis and (iii) providing surface for fibrinolysis. Further research is needed to get further insights into these processes. [12] Understanding the pathophysiological mechanisms involved would help understand the relationships between monocytes, and adverse thrombotic and bleeding outcomes in AF patients.

Limitations
The study has several limitations. Although confounding factors were considered with regards to AF-related risk stratification, possible effects of various co-morbidities (overt or subclinical) affecting monocyte counts have not been analysed. This is an observational analysis, and mechanistic insights into molecular mechanisms of the findings and monocyte functional characteristics have not been studied. [15] As the study included patients with monocytes measured as part of routine clinical management, this could lead to some selection bias due to not inclusion of patients with no full blood count tested at all during the routine assessment. The medications used by the participants are likely undergone modifications during the years of followed and all changes may not have been captured by this study.