Elsevier

Journal of Electrocardiology

Volume 58, January–February 2020, Pages 171-175
Journal of Electrocardiology

Substantial prevalence of subclinical cardiovascular diseases in patients with hemophilia A evaluated by advanced electrocardiography

https://doi.org/10.1016/j.jelectrocard.2019.12.008Get rights and content

Highlights

  • Advanced-ECG can be used to detect CVD in patients with hemophilia A (PWHA).

  • PWHA have a higher probability for CVD than controls as evaluated by A-ECG.

  • Patients with severe HA have a lower probability for CVD than non-severe patients.

  • CVD risk factors should be monitored in adult PWHA.

Abstract

Background

Patients with hemophilia A (PWHA) have reportedly lower mortality due to cardiovascular disease (CVD) compared to the general population.

Aim

To evaluate signs of CVD in asymptomatic PWHA using advanced electrocardiography (A-ECG).

Methods

PWHA (n = 29, median [interquartile range] age 57 [47–70] years) and age-matched male controls (n = 29, 59 [48–68] years) were evaluated. Digital resting 12‑lead ECGs were retrospectively analysed using both conventional and A-ECG techniques including derived vectorcardiography and waveform complexity. Previously validated multivariate A-ECG scores designed to detect: 1) cardiac disease in general, 2) left ventricular systolic dysfunction (LVSD), 3) coronary artery disease or coronary microvascular disease (CAD/CMVD), or 4) left ventricular hypertrophy defined as left ventricular electrical remodelling (LVH/LVER), were quantified and compared between PWHA and controls.

Results

Compared to controls, PWHA had a higher probability of having cardiac disease (median [interquartile range] 84.6 [32.5–99.5] vs. 0.6 [0.2–8.2]%), LVSD (4.1 [1.3–12.9] vs. 0.9 [0.5–3.2]%), CAD/CMVD (84.3 [35.6–96.6] vs. 6.7 [0.8–24.4]%), and LVH/LVER (17 [5/29] vs. 0 [0/29]%). Compared to patients with non-severe HA (n = 20), patients with severe HA (n = 9) showed a non-significant trend towards lower probability of cardiac disease, CAD/CMVD, LVSD and LVH/LVER.

Conclusion

In PWHA, A-ECG exhibits changes more indicative of overt or subclinical CVD compared to controls, and there is a tendency for lower scores for CVD in patients with severe compared to non-severe HA. These results suggest that PWHA ≥ 40 years could be at higher risk for CVD than age-matched controls and that A-ECG could potentially be used for early detection.

Introduction

Hemophilia A (HA) is an inherited bleeding disorder caused by the deficiency of coagulation factor VIII (FVIII) [1]. Based on residual FVIII levels, patients with HA (PWHA) are classified as having severe (<1%), moderate (1–5%) or mild (5–40%) HA. Most patients with severe HA may require regular prophylaxis with coagulation FVIII concentrate, while the rest are treated on-demand. Improved healthcare and early initiation of prophylaxis have significantly increased the life expectancy of PWHA, from 30 years in the mid-1960s to almost normal lifespan [2]. According to the latest annual global survey of world federation of hemophilia, 19% and 31% of PWHA are over 45 years old in Americas and in Europe, respectively [3]. Since the risk of developing cardiovascular disease (CVD) increases with age [4], older PWHA are also considered to be at risk for CVD events.

Several cohort studies have shown that PWHA have lower mortality from CVD than the general population [2,5,6]. This observation cannot be explained by a difference in prevalence of CVD risk factors and atherosclerosis. For example, the prevalence of hypertension in PWHA has been reported to be either similar [7], or even higher than the general population [8,9]. Furthermore, PWHA are not protected from developing atherosclerosis compared to a matched control group [10,11]. It is suggested that due to decreased FVIII, PWHA experience less atherosclerotic plaque rupture, or their plaque ruptures are “silent” and do not lead to arterial occlusion [12].

Over the past two decades, several advanced electrocardiography (A-ECG) measures, including those from derived vectorcardiography [13] and waveform complexity [14], have been used to detect cardiac disease with greater sensitivity using the standard resting 12‑lead ECG. Abnormal A-ECG results have been found in patients with coronary artery disease or coronary microvascular disease (CAD/CMVD), various cardiomyopathies, left ventricular hypertrophy (LVH), or left ventricular systolic dysfunction (LVSD) [[15], [16], [17]]. A-ECG results can also be statistically integrated to maximize diagnostic/prognostic accuracy for particular disease conditions, with notably higher accuracy than conventional ECG [[15], [16], [17], [18], [19], [20]]. Another advantage of A-ECG is that it can be performed rapidly and inexpensively in clinical practice, including in patients unable to participate in cardiac exercise or stress testing [15].

The present study aimed to determine whether PWHA (≥40 years) are protected from CVD compared to age-matched controls, as determined by A-ECG. The secondary objective was to use A-ECG to compare the A-ECG results between severe versus non-severe (moderate and mild) PWHA. To our knowledge, this study is the first to use A-ECG to investigate subclinical cardiac disease in PWHA.

Section snippets

Study design

The study was performed retrospectively as a clinical audit of a cohort of PWHA followed up at the Coagulation Unit, Department of Hematology, Karolinska University Hospital from 2005 to 2018. The study was approved by the Regional Ethical Review Board in Stockholm, and was conducted in accordance with the Declaration of Helsinki.

Patient inclusion criteria: 1) male patients diagnosed with HA; 2) ≥40 years old at the time of inclusion, since they are more likely to develop CVD than younger

Cardiovascular risk factors

Data on CVD risk factors were collected from historical medical documentation of all study participants. In both the PWHA and control groups, hypertension, diabetes and dyslipidaemia were defined per physician annotations, and obesity as body mass index (BMI) ≥30 kg/m2. Family history of CVD and smoking were not reliably recorded for all the participants, and were therefore not included as variables in the study. Since only one measurement of blood pressure (BP) was obtained for control

Advanced ECG acquisition and analysis

Digital 10-second resting 12‑lead ECGs were extracted from the local ECG storage system (MUSE® Cardiology Information System, Version 9.0 SP6, GE Healthcare, Chicago, IL, USA) and exported into de-identified .xml files. The .xml files were analysed using dedicated semi-automatic A-ECG software [15]. If a patient had multiple ECG files in the hospital system, the latest was analysed.

Measures of A-ECG included conventional ECG intervals, axes and voltages; multiple vectorcardiographic measures

Results

Initially, medical records of 190 PWHA (≥40 years) were reviewed, and 44 were found to have electronic ECG files and were thusly eligible for inclusion in the study (Fig. 1). Eleven of the 44 PWHA were then excluded, due to previous myocardial infarction (n = 5), angina (n = 2), transient ischemic attack (n = 2), retinal embolism and carotid stenosis (n = 1) or heart failure (n = 1), and another 4 due to paced rhythm (n = 2), atrial fibrillation with wide QRS complex (n = 1) and noisy ECG

Characteristics and CVD risk factors of the cohorts

Twenty-nine PWHA (median [interquartile range] 57 [47–70] years, all male) and 29 age-matched male controls (59 [48–68] years) were included in the study. Nine (31%) PWHA had severe HA, while 20 (69%) had non-severe HA, including 16 (55%) with mild HA and 4 (14%) with moderate HA. Nine (31%) PWHA were on prophylaxis, including 7 with severe and 2 with non-severe HA. Twelve (41%) patients had hepatitis C virus (HCV) infection, and 7 of them had anti-virus therapy; another 2 (7%) patients had

Risk of cardiac diseases predicted by A-ECG scores

The probability of having cardiac disease in general by A-ECG was higher in PWHA than in controls (84.6 [32.5–99.5] vs 0.6 [0.2–8.2] %, p < .001) (Fig. 2). Although patients with severe HA had a numerically lower probability for disease than those with non-severe HA, the difference was not statistically significant (70.8 [6.3–99.8] vs 85.3 [36.7–99.2] %, p = .46). While the probability for cardiac disease never exceeded 50% in any control, it did in 66% (19/29) of PWHA. Among the nineteen PWHA,

Discussion

The results from this study suggest that PWHA are not protected from developing CVD. Specifically, PWHA in this study had more subclinical cardiac disease by A-ECG than age-matched male controls, including higher likelihoods of heart disease in general, CAD/CMVD, and LVSD, as well as a higher prevalence of LVER.

The development of CVD among PWHA is known to be associated with hypertension, dyslipidemia and other traditional CVD risk factors [25]. Several studies have also shown that CVD risk

Conclusions

Overall results from this study suggest that PWHA are not protected from developing subclinical CVD. PWHA showed higher probability of developing LVSD, CAD/CMVD and LVH/LVER than age-matched controls as evaluated by A-ECG. These results indicate that it is important to monitor the risk of developing CVD in PWHA older than 40 years, especially considering that life expectancy in PWHA now approaches that of the non-HA population. In addition, patients with severe HA in this study tended to have

Funding

This research was supported grants from Svenska Sällskapet för Trombos och Hemostas (SSTH), and Stiftelsen för Koagulationsforskning vid Karolinska Institutet (grant number: stiftelse245).

CRediT authorship contribution statement

Yanan Zong: Formal analysis, Writing - original draft. Maren Maanja: Methodology, Investigation. Roza Chaireti: Investigation, Resources, Writing - review & editing. Todd T. Schlegel: Methodology, Writing - review & editing. Martin Ugander: Conceptualization, Writing - review & editing. Jovan P. Antovic: Conceptualization, Writing - review & editing, Supervision, Project administration.

Declaration of competing interest

TTS is a principal of Nicollier-Schlegel SARL (Switzerland), a company that performs ECG research consultancy with the software used in the present study. The other authors declare there is no conflict of interest with the contents of this article.

Acknowledgements

Thanks to Ann-Michele Francoeur for editing the first draft of the manuscript.

References (35)

  • N Bhatnagar et al.

    Major bleeding disorders: diagnosis, classification, management and recent developments in haemophilia

    Arch Dis Child

    (2018)
  • Hemophilia WFo

    Annual global survey mini report 2: age

  • AS Go et al.

    Heart disease and stroke statistics-2013 update a report from the american heart association

    Circulation

    (2013)
  • S Lovdahl et al.

    Incidence, mortality rates and causes of deaths in haemophilia patients in Sweden

    Haemophilia

    (2013)
  • DE Fransen van de Putte et al.

    Unfavourable cardiovascular disease risk profiles in a cohort of Dutch and British haemophilia patients

    Thromb Haemost

    (2013)
  • AS Sait et al.

    Risk assessment for coronary heart disease in patients with haemophilia: A single centre study in the United States

    Haemophilia

    (2014)
  • A Tuinenburg et al.

    Coronary artery calcification in hemophilia A: no evidence for a protective effect of factor VIII deficiency on atherosclerosis

    Arterioscler Thromb Vasc Biol

    (2012)
  • Cited by (5)

    • Predicting peri-operative troponin elevation by advanced electrocardiography

      2021, Journal of Electrocardiology
      Citation Excerpt :

      Existing digital 10-s, 12‑lead ECGs were collected for each patient from the local ECG database (MUSE Cardiology Information System, GE Healthcare, IL, USA) and analysed digitally using semi-automatic software developed in-house. The following parameters were analysed: conventional ECG parameters including scalar amplitudes, durations and axes of the ECG wave forms in each lead; derived vectorcardiography parameters including spatial axes, duration and magnitudes; and complexity measures of the QRS wave and T waveforms derived through singular value decomposition [14]. A composite of these detailed analyses then provided diagnostic scores.

    View full text