INTRODUCTION
Patients with heart failure (HF) are at elevated risk of cerebrovascular events, including stroke [1]. They are often burdened with many cardiovascular risk factors, in particular, atrial fibrillation (AF), a known risk factor for both stroke and HF [2–4].
It has been demonstrated that in AF patients, the risk of stroke is markedly elevated on the first days after successful electrical cardioversion of AF [5–7]. Regardless of their appropriateness, high-voltage shock therapies delivered by implantable cardioverter-defibrillators (ICDs) often result in termination of supraventricular arrhythmias, including AF cardioversion, potentially increasing the risk of stroke in those patients. However, the association between ICD shocks and the risk of stroke in AF patients is unknown. Therefore, the present analysis aimed to define whether any association exists between ICD shocks and strokes in remotely monitored HF patients.
METHODS
Data of patients enrolled in the remote monitoring (RM) program after implantation of an ICD were obtained from the COMMIT-HF registry. The rationale and the major results of the registry were published before, and in brief, among other findings, the registry encompasses detailed demographic, clinical, echocardiographic, and angiographic data of patients with heart failure with left ventricular ejection fraction ≤35% not caused by an acute coronary syndrome at index hospitalization, who have been admitted and treated in our facility (NCT02536443) [2, 8]. During the entire follow-up, patients were treated according to the respective guidelines [9, 10]. Routinely, apart from RM based on the continuous connectivity and alert transmissions, patients with ICDs visited the outpatient clinic every 12 months for an in-person evaluation. The baseline status of AF was based on the patient’s medical history records, while any newly diagnosed AF during the follow-up period, was defined as de novo AF. A decision regarding the introduction and/or continuation of oral anticoagulation was based on the thromboembolic risk assessment using the CHA2DS2-VASc score [11].
Data regarding the occurrence of the first stroke after inclusion in the registry were obtained (based on codes I63–I64 of the International Classification of Diseases) from the National Health Fund — the sole Polish healthcare provider. The RM data of all patients were obtained from the investigator-initiated and maintained registry of RM, which summarizes all major findings in remotely monitored patients in our department. The registry data are particularly focused on the occurrence of alert transmissions, especially if they are caused by arrhythmic episodes or device therapies. In the registry, each patient’s data are updated annually [12]. Each patient with a stroke had their RM history evaluated for the period of the prior 30 days, with emphasis on the last 10 days preceding stroke. The period of 10 days was selected on the basis of prior, mechanistic and clinical studies, indicating that the risk of stroke after cardioversion of AF persists for 10 days [6]. The study protocol was approved by an appropriate institutional review board and ethics committee.
Statistical analysis
The categorical variables were presented as absolute numbers and percentages. After assessment for normality using the Shapiro–Wilk test, due to non-normal distribution of numerical variables, they were presented as medians with quartiles 1 and 3. Pearson’s χ2 test was used to compare the categorical variables, and the Mann-Whitney U test was used to compare non-normally distributed numerical variables. A two-tailed P-value <0.05 was considered statistically significant. STATISTICA 10 (StatSoft Inc., Tulsa, OK, US) was used for all calculations.
RESULTS AND DISCUSSION
During the period between 2011 and 2020, 1299 patients with detailed clinical characteristics were enrolled in the RM program, and their clinical data are summarized in Table 1. In the studied population, 523 (40.3%) patients had cardiac resynchronization therapy defibrillator devices, 438 (33.7%) had dual-chamber ICDs, and 338 (26.0%) had single-chamber ICDs, of whom 38 had an atrial sensing dipole. There were 434 (33.4%) patients with AF at baseline, and AF de novo was identified during follow-up in further 182 patients (14.0%). Median (Q1–Q3) CHA2DS2-VASc scores in patients without and with AF at baseline were, respectively, 3 (2–4) and 3 (2–5) points (P <0.001). Among patients without and with AF at baseline, 267 (30.9%) and 148 (34.1%), respectively, received any (either appropriate or inappropriate) ICD shocks during follow-up (P = 0.24). In 25.3% of all patients, at least one appropriate shock occurred, while inappropriate shocks were reported in 11.5% of patients. In 38.2% of patients without AF at baseline, who developed de novo AF, at least one shock occurred.
|
Demographics at baseline |
Overall population (n = 1299) |
|
Female, n (%) |
230 (17.7) |
|
Age at implantation, years, median (Q1–Q3) |
62 (55–70) |
|
Indication for implantation |
Overall population (n = 1299) |
|
Ischemic cardiomyopathy, n (%) |
788 (60.7) |
|
Non-ischemic cardiomyopathy, n (%) |
511 (39.3) |
|
Arterial hypertension, n (%) |
695 (53.5) |
|
Atrial fibrillation, n (%) |
434 (33.4) |
|
Diabetes, n (%) |
475 (36.6) |
|
NYHA classification |
|
|
II, n (%) |
554 (42.6) |
|
III, n (%) |
532 (41.0) |
|
IV, n (%) |
91 (7.0) |
|
GFR ≤60 ml/min/1.73 m2, n (%) |
324 (24.9) |
|
NT-proBNP, pg/ml, median (Q1–Q3) [n/n] |
1885 (770–4195) [340/1299] |
|
LVEF, %, median (Q1–Q3) |
25 (20–30) |
|
Pharmacotherapy on discharge from baseline hospitalization |
Overall population (n = 1299) |
|
ACE-I/ARB, n (%) |
926/129 (71.3/9.9) |
|
Beta-blocker, n (%) |
1248 (96.1) |
|
Diuretics, n (%) |
1126 (86.7) |
|
Amiodarone, n (%) |
185 (14.2) |
|
Aldosterone antagonist, n (%) |
1107 (85.2) |
|
Ivabradine, n (%) |
70 (5.4) |
|
Digitalis, n (%) |
266 (20.5) |
|
Any oral anticoagulants, n (%) |
464 (35.7) |
|
Vitamin K antagonists, n (%) |
421 (32.4) |
|
Direct oral anticoagulants, n (%) |
45 (3.5) |
|
Low molecular-weight heparin, n (%) |
19 (1.5) |
|
Single antiplatelet therapy, n (%) |
508 (39.1) |
|
Dual antiplatelet therapy, n (%) |
331 (25.5) |
The median (Q1–Q3) follow-up was 4.4 (2.1–5.9) years, and the incidence of stroke was 45 (3.5%) in the studied population, with a rate of stroke of 25 (2.8%) in patients without AF at baseline and 20 (4.6%) in patients with AF (P = 0.11), as presented in Supplementary material, Figure S1. The median (Q1–Q3) time from ICD implantation to stroke was 1.5 (0.8–2.5) years.
In one patient without AF and one patient with AF, stroke occurred in less than 24 hours from ICD shock. In the first case, in a patient on acetylsalicylic acid monotherapy, the device appropriately terminated ventricular tachycardia which degenerated to ventricular fibrillation after anti-tachycardia pacing, and in the second case, shock therapy was delivered due to ventricular fibrillation in a 59-year-old male with HF and concomitant paroxysmal AF, who was on chronic treatment with warfarin.
Previous studies have identified AF cardioversion as a clinical event significantly elevating the risk of stroke. Successful cardioversion results in the return of rhythmic function of the atria, and thus potentially increases the risk of mobilization of the already-formed thrombus from the left atrial appendage, with increased risk of stroke persisting for approximately 10 days [5, 6]. To date, no study assessed the temporal association between ICD shocks and stroke.
In the studied population, the overall risk of stroke (3.5% during the median follow-up of 4.4 years) was comparable with other European populations of HF patients [13]. However, a temporal association between high-voltage ICD therapies and stroke was noted in only two patients. The first patient with AF had oral anticoagulation administered, and the second with VT had an appropriate shock. The post-shock stroke in the overall cohort was observed in approximately 0.5% of patients with any shock therapy.
A few limitations of our study are worth noting, as no exact information on the anticoagulation status of the studied patients was available — despite administration of oral anticoagulants in AF patients, according to individual assessment of thrombo-embolic risk, lack of adherence to therapy or being outside the therapeutic range cannot be excluded. Second, based on the International Classification of Diseases, no differentiation between types of stroke (e.g., hemorrhagic, ischemic, etc.) could be made for this analysis. Finally, although the study has been conducted on a meticulously remotely monitored cohort, the present analysis could be considered numerically too small to draw straightforward conclusions, and its retrospective nature does not allow causality determination. Nonetheless, our findings indicate an overall low risk of stroke associated with ICD shocks, both in patients with and without AF.
Supplementary material
Supplementary material is available at https://journals.viamedica.pl/polish_heart_journal.
Article information
Conflict of interest: None declared.
Funding: None.
Open access: This article is available in open access under Creative Common Attribution-Non-Commercial-No Derivatives 4.0 International (CC BY-NC-ND 4.0) license, which allows downloading and sharing articles with others as long as they credit the authors and the publisher, but without permission to change them in any way or use them commercially. For commercial use, please contact the journal office at polishheartjournal@ptkardio.pl
