Defining the interaction between exercise and arrhythmogenic right ventricular cardiomyopathy

The opinions expressed in this article are not necessarily those of the Editors of the European Journal of Heart Failure or of the European Society of Cardiology.

This article refers to ‘Vigorous physical activity impairs myocardial function in patients with arrhythmogenic right ventricular cardiomyopathy and in mutation-positive family members’ by J. Saberniak et al., published in the December 2014 issue on pages1337–1344.

An association between exercise and arrhythmogenic right ventricular cardiomyopathy (ARVC) was appreciated soon after the first descriptions of the condition by Marcus et al. in 1982.1 The observations that deaths associated with ARVC were disproportionately represented amongst competitive athletes2 and frequently occurred during exercise3 gave rise to the popular hypothesis that exercise acted as a trigger for arrhythmic events. However, recent pre-clinical and epidemiological investigations have associated intense endurance with earlier onset and more severe expression of the ARVC phenotype, thereby implying that exercise not only acts as a trigger for arrhythmias but may also impact directly on the disease substrate.

Arrhythmogenic right ventricular cardiomyopathy is a familial cardiomyopathy associated with abnormalities in the structure, function, and electrophysiological properties of the myocardium. A familial history or mutation in one of five desmosomal genes can be identified in a majority of ARVC patients, but there is considerable phenotypic variability such that the same mutation may cause severe heart failure and arrhythmias in one patient and no symptoms in another. The reasons for this variability are incompletely understood, but hypotheses include additional genetic factors such as modifier polymorphisms or interaction with environmental factors. Similarly, mutations in novel genes or complex interactions between polymorphisms and environmental factors could also potentially explain the 30–40% of patients in whom a familial predisposition cannot be currently identified. In regard to potential environmental factors determining phenotypic expression, exercise is a logical candidate. Intense exercise increases ventricular wall stress and disproportionately affects the right ventricle,4 hence providing a haemodynamic explanation for why a defect in the proteins which bind all myocytes result in a pathological process which predominantly, though not exclusively, affects the right ventricle.

Kirchhof et al. provided the first support for the hypothesis that exercise could modify the expression of ARVC by exercising mice with a heterozygous deficiency in plakoglobin.5 As compared with the sedentary mice, exercise resulted in earlier right ventricular (RV) dysfunction and arrhythmias. The first observation that the same process may also affect humans came from Sen-Chowdhry et al. who documented larger RV volumes and lower RV function in 11 endurance athletes as compared with a much larger non-athletic ARVC cohort.6 This was further advanced by the Baltimore group of James et al. who conducted an exercise interview in 87 patients with a known desmosomal gene mutation in whom two-thirds were categorized as endurance athletes using a fairly liberal definition of >50 h per year of intense aerobic exercise.7 As compared with the non-athletic subjects, endurance athletes developed symptoms at a younger age, and were more likely to meet Task Force criteria for ARVC, to develop ventricular arrhythmias, and to suffer heart failure.

In the December issue of the European Journal of Heart Failure, Saberniak et al.8 further qualified the interaction between exercise and ARVC. Using a slightly more stringent definition of athletic status (≥4 h per week of strenuous exercise for at least 6 years), they identified 37 athletes amongst 110 subjects who either met ARVC Task Force criteria or were a family member in whom an ARVC gene mutation was identified. In this larger European cohort, they validated the findings of James et al.'s US experience that athletes more frequently met ARVC diagnostic criteria, were more often symptomatic, more frequently experienced ventricular arrhythmias, and were more likely to progress to cardiac transplant.8 In addition, comprehensive multimodality imaging was employed to demonstrate greater impairment of both LV and RV function in athletes as compared with non-athletes. Perhaps of greatest interest is the apparent dose–response relationship in which a composite measure of exercise intensity and duration correlated modestly with echocardiographic and magnetic resonance imaging measures of LV and RV dysfunction. Thus, one may draw the logical conclusion that the effect of exercise should be considered as a continuum of both intensity and duration rather than by an arbitrary categorization of athletic status.

It is intriguing to extrapolate beyond the range of exercise practised by the cohort of Saberniak et al. The most practised athletes trained for the equivalent of 10 h at an intensity of 10 METS (the equivalent of 35 mL/min/kg which is moderate intensity exercise for well-trained athletes). This is less than a quarter of the dose of exercise practised by professional endurance athletes such as cyclists, rowers, and triathletes. Could massive doses of exercise promote an ARVC phenotype in athletes with a mild genetic predisposition or, taking the concept a step further, could extreme exercise cause ARVC in its own right? Figure 1 provides a graphical summary of this concept using the ‘threshold theory’ of phenotypic expression. This hypothesis that extreme exercise could cause an ARVC-like phenotype in its own right was first proposed by Heidbuchel et al.9 in 2003, thus pre-dating any clear evidence of an interaction between exercise and ARVC substrate. After observing a recurring clinical pattern of RV arrhythmias and mild RV dysfunction amongst professional cyclists presenting with palpitations, Heidbuchel collated thorough diagnostic evaluations under the diagnostic framework of the ARVC criteria in spite of the fact that evidence of familial disease seemed to be absent in the vast majority of cases.10 Nearly two-thirds of the professional cyclists met the diagnostic criteria for ARVC, and the high rate of malignant arrhythmias during follow-up suggested that this condition portended a prognosis similar to that of familial disease. A subsequent comprehensive genetic analysis revealed desmosomal mutations in only 12.8% of the athletic cohort and in none of the athletes performing >14 h of intense exercise per week despite all of these athletes manifesting features sufficient for a diagnosis of ARVC (60%) or probable ARVC (40%). We have subsequently conducted a range of experiments seeking to elucidate the pathophysiological mechanism underpinning the susceptibility of the right ventricle to pro-arrhythmic remodelling11 including an increase in RV wall stress during exercise,4 transient RV injury if intense exercise is sustained for many hours,12, 13 and greater chronic remodelling of the right ventricle in highly trained endurance athletes, as compared with the left ventricle.4 Figure 2 summarizes two potential models of the ARVC phenotype with genetically determined desmosomal frailty contrasting with disruption of desmosomal integrity as a result of the haemodynamic stress of exercise. The ‘Heidbuchel syndrome’ of exercise-induced ARVC has remained controversial, but is supported by the ‘marathon rat’ studies from the Barcelona group of Lluis Mont. In a comparison of untrained rats without any genetic predisposition for ARVC, Benito et al.14 demonstrated that rats undergoing an intensive endurance training regime developed RV-specific fibrosis and more readily inducible ventricular arrhythmias. Thus, with a combination of pre-clinical and human data we are starting to understand the full spectrum of the exercise–genetic interaction.

The clinical implications of the work of Saberniak et al. are immediate. There is now a consistent signal that intense exercise can accelerate disease progression and should thus be moderated in all patients with established ARVC, and probably also in asymptomatic geno-positive family members in whom avoidance of more intense exercise may delay or prevent phenotypic expression. But what constitutes intense exercise? One of the major limitations in these studies is that there is no standard definition of exercise intensity, and accurate quantification of exercise dose has proved difficult. Given the importance of trying to establish a threshold of activity which may be safe, future studies should attempt to quantify exercise intensity and duration directly with any of the plethora of devices which can accurately track activity duration and intensity. It is important to note that in the context of general population standards, the exercise range described by Saberniak et al. is substantial but, relative to competitive athlete training, it is modest. For example, a rider in a professional cycling tour would typically perform more exercise in a single day than the maximum weekly exercise described in the study. It is quite remarkable that a dose–response relationship could be delineated within this reasonably modest range of exercise and suggests that exercise restriction should be recommended for leisure time activities in addition to current recommendations for competitive sport.15 A recent analysis of a registry of athletes who continued competitive sport after implantation of a defibrillator concluded that continued sport participation was relatively safe, with no deaths observed and a relatively modest incidence of device therapies during sports practice.16 This provides some reassurance for many athletes, but the fact that that ARVC patients constituted only 5% of the athletes involved in regular competition and the non-randomized nature of the trial makes it impossible to conclude that these athletes may compete safely post-implantable cardioverter defibrillator (ICD) implantation. Athletes should be informed that the current weight of evidence would suggest that exercise may lead to measurable decrements in cardiac function, more frequent arrhythmias, and a greater likelihood of heart failure requiring transplantation.

These studies of the interaction between exercise and ARVC serve as an important model for the wider investigation of the interaction between genetic and environmental risk. It is likely that other co-morbid, dietary, and lifestyle factors may also influence ARVC disease expression, and it is also possible that the extent of the interaction may differ according to the specific genetic mutation. Perhaps of even more immediate importance is whether similar interactions may be present in other inherited cardiomyopathies such as hypertrophic cardiomyopathy (HCM) which is the leading cause of sudden cardiac death amongst athletes. Exercise serves as a potential trigger for arrhythmias in HCM, but it would be important to establish whether a carefully tailored exercise programme may improve cardiac function as a result of exercise-associated improvements in the diastolic properties of the myocardium or whether, on the other hand, exercise exacerbates the degree of hypertrophy. Hypotheses of benefit or harm can also be considered for other causes of SCD such as the channelopathies and, now that the use of ICDs afford some protection against the risk of exercise-triggered arrhythmic events, we enter the brave new world in which we need to establish the effect of exercise on disease substrate for each of these conditions. The work of Saberniak et al. in ARVC patients will long serve as the prototype for such investigations.

Conflict of interest: none declared.