Arrhythmogenesis in Fabry Disease

Tachyarrhythmia is prevalent in FD and may be supraventricular or ventricular in origin. Atrial fibrillation (AF) is the most frequently reported supraventricular arrhythmia, and regular narrow complex tachycardias are less common, despite shortening of the PR interval [6]. Ventricular tachycardia (sustained and non-sustained) and ventricular fibrillation are the most frequent ventricular arrhythmias (VA), with overall prevalence between 13 and 18% [7‐11]. Figure 1 illustrates the non-sustained ventricular tachycardia (NSVT) detected on an implanted loop recorder (ILR) in a patient with FD. The incidence of SCD varies between 0.5 and 3% (follow-up range 1.2–8 years) [7, 12, 13, 14•, 15••, 16], which is much higher than the rate in both the general population (0.03–0.1%) and patients with sarcomeric hypertrophic cardiomyopathy (0.5–1%) [17, 18]. Details of studies describing incidence and prevalence of VA and SCD are summarised in Table 1. Predictors of VA and SCD include advancing age, an annual increase in myocardial fibrosis, severity of FD measured by the Mainz Severity Score Index (MSSI), and prolongation of QRS duration on electrocardiogram (ECG) [8, 12]. Single-centre studies have demonstrated a higher prevalence of VA in males than females [7, 19]. Risk factors for VA are illustrated in Fig. 2. No change in frequency of tachyarrhythmia has been demonstrated solely according to mutation-type in FD [20].

Bradyarrhythmia is more common than tachyarrhythmia in FD [21••]. In a recent systematic review, weighted estimates of event rate bradyarrhythmia were 10% (pooled median follow-up time 4.5 years) [21••]. Bradyarrhythmia that have been reported in FD patients include sinus node disease (specifically, symptomatic bradycardia, sinus pauses, and chronotropic incompetence) and atrioventricular (AV) nodal disease (including first-, second-, and third-degree AV block) [22]. In a study of 29 adults with FD (82% male; mean age 43 ± 11 years), the prevalence of resting bradycardia was reported as 72% (of which 7% had shortened PR interval, 7% prolonged PR interval, and 86% normal) [23], which may be attributed to early autonomic dysfunction and chronotropic incompetence. There is even data suggesting that sinus bradycardia may ensue in childhood with a study of 26 children with FD (46% male; mean age 9.7 ± 3.8 years) demonstrating that 23% had a resting bradycardia on 12-lead ECG [24]. Bradyarrhythmia relating to clinically significant sinus node disease (severe sinus bradycardia and sinus pauses) has been reported in up to 12% [25]. Bradyarrhythmia requiring permanent pacemaker implantation (PPM) varies between 2.5 and 11% (follow-up range 1.2–8 years) [7, 9, 12, 13, 16, 26‐29]. Figure 3 illustrates an asymptomatic daytime pause on ILR in a patient with FD. Details of studies describing incidence and prevalence of bradyarrhythmia are summarised in Table 2. Risk factors associated with bradyarrhythmia include increasing age, left atrial (LA) dysfunction, lower resting heart rate, prolonged PR and QRS duration, use of beta-blocker therapy, and impaired LV global longitudinal strain (GLS) [25, 26, 28]. Risk factors for bradycardia are illustrated in Fig. 4. No changes in frequency of bradycardia have been demonstrated according to gender or mutation-type in FD [20].

AF is a supraventricular arrhythmia that involves irregular and disorganised atrial conduction arising due to electrophysiological disruption usually originating from the pulmonary veins. AF is prevalent in FD with reported figures between 3 and 21% [9, 25, 26]. Incidence of AF in FD from longitudinal data is reported between 3 and 31% (follow-up range 1.2–8 years) [7, 9, 12, 13, 16, 26‐29]. The incidence is likely to be higher with numerous mechanisms in FD contributing to the structural, functional, and electrical substrate that is required for AF to initiate and persist. LVH raises left ventricular end-diastolic pressure which transmits to the LA, resulting in LA dilatation and impaired LA strain (see Fig. 5) [30]. LA strain impairment predicts new onset and recurrence of AF in the general population [31] and has also been found in patients with FD, with initiation of enzyme replacement therapy (ERT) improving function [32]. The same study demonstrated a higher incidence of AF and stroke in the cohort with impaired LA strain. Most of the patients had diastolic function without LVH suggesting that direct atrial sphingolipid accumulation may contribute to LA dilatation and impaired strain rather than passive effects of the LV [28].

The incidence and prevalence of both tachyarrhythmia, bradyarrhythmia, and AF are variably reported because the detection rate depends in part on the duration of monitoring. Data from small single-centre cohorts using continuous monitoring with advanced disease have detected a much higher rate of arrhythmia requiring intervention than studies using intermittent 24-h Holter monitoring [13, 33]. This is important because the annual Holter monitoring recommended by current guidelines is likely to under-estimate the frequency of arrhythmia [34]. The variation in reporting of arrhythmia is also likely to be attributable to differences in the cohorts studied, in terms of age, sex, genotype, and classical or non-classical phenotype. Studies suggest a greater number of cardiac events in males with classical FD compared with males with non-classical disease or females. However, these studies do not specify whether these relate specifically to arrhythmia [35]. Within these limitations, however, it is interesting to reflect on the possibility that FD patients seem to have a much higher risk of VA and SCD than sarcomeric hypertrophic cardiomyopathy, which is the paradigm for arrhythmogenesis and SCD [36]. Future studies using longer-term monitoring with implanted loop recorders have completed recruitment and outcomes are awaited [37].

Whilst intracellular sphingolipid deposition within conduction tissue may have direct pro-arrhythmic effects, accumulation can trigger other deleterious processes either within the cell or because of ‘over-spill’ into the extracellular environment. Firstly, lysosomes are directly involved in cellular autophagy (Fig. 6). A key regulator of autophagy-lysosomal fusion and mitochondrial function is the mTOR-dependent signalling pathway [44]. mTOR is located on lysosomal surfaces and upon activation through amino acid accumulation, triggers autophagy [45]. Sphingolipid accumulation has been found to disrupt mTOR activation in FD cardiac fibroblasts and thereby inhibits autophagy-lysosome fusion [46]. mTOR is also involved in mitochondrial metabolism by promoting translation of mitochondrial-related proteins, which is adversely affected by lysosomal accumulation [47]. Reduced autophagy and impaired mitochondrial metabolism may play a vital role in inducing structural and functional changes within cardiac cells that could directly trigger arrhythmia [48]. Secondly, endomyocardial biopsies of FD patients demonstrate perinuclear vacuoles of glycosphingolipid. On electron microscopy, these vacuoles had the appearance of single-membrane-bound vesicles which displaced cardiac myofibrils to the periphery of the cell and induced myofibrillolysis. Cardiomyocyte area and percentage occupied by these vacuoles were similar, indicating that increased cardiomyocyte size may be explained by these vacuoles. The same study demonstrated that in FD patients with abnormal tissue Doppler indices on transthoracic echocardiography (TTE) indicative of diastolic dysfunction, in-vitro experiments on the cardiomyocytes of these patients also demonstrated stiffening [49]. Myofibrillolysis, vacuolation, and mitochondrial dysfunction may each contribute as trophic stimuli for hypertrophy [50], which forms a substrate for arrhythmia. Thirdly, data emerging from studies modelling FD in induced pluripotent stem cells (iPSC) have identified pro-arrhythmic intracellular mechanisms. A recent study characterised the structural and electrophysiological properties of FD patient-derived IPSC-derived ventricular cardiomyocytes with iPSCs obtained from patients with FD [51]. This study demonstrated that Fabry iPSC-derived cardiomyocytes had a higher spontaneous action potential (AP) frequency, shorter AP duration with an increased peak sodium current density, and upstroke velocity, suggesting increased excitability. Calcium transients had a greater amplitude and reduction in peak width duration, which indicate increased excitability and disruption to intracellular calcium handling. It has previously been shown that ion channel expression is affected by Gb3 accumulation in neuronal and endothelial cells [52, 53]. Therefore, α-GAL A deficiency and the accumulation of glycosphingolipid may alter the expression of calcium and sodium ion channels, causing severe effects on the electrophysiological properties of the cell. These early cellular changes likely cause an increased propensity for VA during disease development [51]. There is further scope using in-vivo and ex-vivo models for evaluation of arrhythmia used in combination in FD [54].

Prevalence of conventional risk factors for atherosclerosis in FD is higher than in the general population suggesting accelerated atherosclerosis [82]. Interestingly, novel mechanisms of atherosclerotic disease have also been shown specifically in FD suggesting accelerated atherosclerosis via non-conventional mechanisms [82]. Although there is no randomised evidence of benefit specifically in FD, these data highlight the importance of aggressive risk factor modification including lipid-lowering therapy, antihypertensive therapy, and strict glycaemic control. Lifestyle modification including smoking cessation and regular physical exercise is also recommended [83]. Elevated systolic blood pressure has an incremental impact on progression of cardiomyopathy and subsequent arrhythmic risk due to relation with myocardial hypertrophy and geometry [84]. The benefits of angiotensin-converting enzyme (ACE) inhibitors and angiotensin receptor blocker (ARB) therapy in reducing cardiovascular risk have been demonstrated in sarcomeric HCM [85].

Patients with FD tolerate arrhythmia supraventricular tachycardia and AF poorly due to impaired diastolic filling. Therefore, rhythm control is the favoured strategy with either pharmacological intervention, direct current cardioversion (DCCV), or ablation, although long-term success is compromised by progressive atrial dilatation, impaired atrial function, and change in atrial structure. Although there is now general support for early ablation in all patients presenting with AF, there remain concerns regarding long-term efficacy in FD [86]. In those with refractory AF not amenable to cardioversion, careful rate control is important, with the additional concern regarding co-existing risk of bradyarrhythmia and conduction disease. Beta blockers or calcium channel antagonists are generally favoured as first line for rate-limiting strategies, although amiodarone is generally avoided long term in FD. Amiodarone may alter lysosomal pH, subsequent enzyme activity, and trigger lysosomal dysfunction. It has also been shown to induce phospholipidosis via inhibition of lysosomal phospholipase activity, triggering phospholipid accumulation and development of lamellar bodies [87]. Amiodarone may induce acute heart failure decompensation with features of amiodarone toxicity confirmed on endomyocardial biopsy in a case study of amiodarone initiation in FD [88]. Frequent Holter or wearable ECG monitoring is recommended.

Although anticoagulation is recommended for life in all FD patients with a history of AF due to the high incidence of stroke, none of the current scoring systems for assessing risk of stroke, such as the CHA₂DS₂VAS_c scoring system [89], have been validated for use in FD. There is little systematic data on choice of anticoagulation (direct oral anticoagulant versus warfarin). However, direct oral anticoagulation may be advantageous due to lower risks of intracranial bleeding, as FD may be associated with cerebral micro-bleeding [90].

As described, cardiac device implantation with PPM or ICD is high due to increased risk of bradyarrhythmia and VA, particularly in advanced disease [21, 76]. However, with no specific guide for device implantation in FD currently available and as FD is specifically excluded from conventional risk calculators for HCM for primary prevention, device implantation is generally for secondary prevention after a significant arrhythmic event or aborted SCD [83].

The mainstay of systemic FD-related therapy is ERT (recombinant α-GAL A) to replace deficient α-GAL A. Current recommendations are that in adult males with classical variants and enzyme activity < 5%, FD therapy is to be considered at diagnosis. In adult females and males with non-classical variants, FD therapy is recommended in those with an LV wall thickness > 13 mm in males and > 12 mm in females, indexed LV mass on TTE/CMR above normal for age and sex or the presence of LGE on CMR [34]. As with all FD therapy, better outcomes are reported with early initiation. There is limited efficacy if commenced > 50 years of age in terms of stabilising LV mass compared to if started < 30 years of age [72]. Those without LGE have seen improvements in LV mass, strain, and exercise capacity with ERT, which is not observed in those with LGE [91]. Efficacy in FD cardiomyopathy is mixed. Once many of the cardinal features of FD cardiomyopathy have developed, which increase arrhythmic risk, the benefits of ERT initiation are significantly limited. There is currently no definitive evidence that ERT reduces the burden of arrhythmia in FD [72, 92].

Oral chaperone therapy (OCT) is a pharmacological chaperone licensed for use in adults with FD with residual enzyme and an amenable GLA mutation. The mechanism is to stabilise the mutant enzyme, increase bioavailability, and traffic this to the lysosome whereby metabolism of sphingolipid can take place [93]. As with ERT, efficacy of OCT in established cardiomyopathy is limited and there is no evidence demonstrating the benefit of OCT on reducing arrhythmia burden. OCT may slow organ damage with studies showing a stabilisation and, in some cases, mild reduction in LVMi [94, 95]. However, in those with evidence of LGE, cardiomyopathy still progresses even with OCT [96].

In patients with significant symptomatic heart failure, intractable arrhythmia despite all optimal therapy (medical and cardiac devices including synchronisation), transplantation may be a viable option. The major barrier to cardiac transplantation in FD is the progressive and multi-organ nature of the disease. Patients often have co-existing end-organ renal disease and so combined cardiac and renal transplantation may need to be considered. The involvement of multiple organs and increased disease burden has a significant impact on exercise capacity and mental health which affects transplant candidacy [97]. Finally, in patients with non-classical disease including cardiac variants, these are often late-onset in age which presents another barrier to transplantation [98].