Article Text

Subepicardial delayed gadolinium enhancement in asymptomatic athletes: let sleeping dogs lie?
  1. Frédéric Schnell1,2,
  2. Guido Claessen2,
  3. André La Gerche2,3,
  4. Jan Bogaert4,
  5. Pierre-Axel Lentz5,
  6. Piet Claus6,
  7. Philippe Mabo7,
  8. François Carré1,
  9. Hein Heidbuchel8
  1. 1Department of Sport Medicine, University Hospital Pontchaillou, Rennes, France
  2. 2Department of Cardiology, University Hospital Gasthuisberg, University of Leuven, Leuven, Belgium
  3. 3Baker IDI Heart and Diabetes Institute, Melbourne, Victoria, Australia
  4. 4Department of Radiology, University Hospital Gasthuisberg, University of Leuven, Leuven, Belgium
  5. 5Department of Radiology, University Hospital Pontchaillou, Rennes, France
  6. 6Department of Cardiovascular Imaging and Dynamics, University of Leuven, Leuven, Belgium
  7. 7Department of Cardiology, University Hospital Pontchaillou, Rennes, France
  8. 8Hasselt University and Heart Center, Jessa Hospital, Hasselt, Belgium
  1. Correspondence to Dr Frédéric Schnell, Department of Cardiology, University Hospitals Gasthuisberg, Herestraat 49, Leuven B-3000, Belgium; frederic.schnell{at}


Background Subepicardial delayed gadolinium enhancement (DGE) patches without underlying cardiomyopathy is poorly understood. It is often reported as the result of prior silent myocarditis. Its prognostic relevance in asymptomatic athletes is unknown; therefore, medical clearance for competitive sports participation is debated. This case series aims to relate this pattern of DGE in athletes to outcome.

Methods We report on seven young asymptomatic athletes with isolated subepicardial DGE detected during workup of abnormalities on their regular screening examination, that is, pathological T-wave inversions on ECG (n=4) or ventricular arrhythmias on exercise test (n=3). All underwent a comprehensive initial investigation in order to assess left ventricular (LV) function at rest and exercise (exercise cardiac MRI and/or exercise echocardiography) and occurrence of arrhythmias (exercise test, 24 h-ECG Holter, electrophysiological study). All underwent a careful follow-up with biannual evaluation.

Results All athletes had extensive subepicardial DGE (12.0±4.8% of LV mass), predominantly in the lateral wall. Three athletes had non-sustained ventricular arrhythmias, whereas two of them had LV ejection fraction <50% at rest with no contractile reserve at exercise. During a follow-up of 3.0±1.5 years in the four remaining athletes, two had symptomatic ventricular tachycardia and one demonstrated progressive LV dysfunction. Hence, six of seven athletes had to be excluded from competitive sports participation.

Conclusions Isolated large areas of subepicardial DGE in an asymptomatic athlete are not benign and require a careful evaluation at exercise and a strict follow-up. These findings question whether extreme exercise during silent myocarditis may facilitate fibrosis generation and adverse remodelling.

  • Athlete
  • MRI
  • Cardiology

Statistics from

Eligibility evaluation of asymptomatic competitive athletes is routinely performed in many European countries as recommended by the European Society of Cardiology guidelines and several international sport organisations.1–3 In cases in which abnormalities are identified on first-line screening, cardiac MRI (CMR) has emerged as a powerful diagnostic tool to detect underlying cardiomyopathy.4 ,5 However, the increased use of CMR also leads to findings that raise questions and even may increase the level of uncertainty about the correct management of the athlete.6 This is especially the case in the presence of isolated delayed gadolinium enhancement (DGE).

Although DGE is associated with a poor prognosis in patients with cardiomyopathy,7–9 the prognostic relevance of isolated DGE in the absence of overt structural cardiomyopathy is unclear.5 In past years, a number of series have reported on DGE in asymptomatic athletes; prevalence ranged from 12% to 50%.10–12 However, the major limitation of these studies was the inclusion of mixed forms of the DGE pattern (especially unrecognised myocardial infarction),12 mainly because the studies focused on veteran athletes. They also included small patches of DGE at insertion points,10 ,11 which may be benign in endurance athletes and related to the increased afterload of the right ventricle(RV).11

The present study focuses exclusively on subepicardial DGE. This DGE might be the result of prior silent myocarditis, but the exact pathophysiology and prognostic significance in athletes are still debated. The European and US recommendations on the management of athletes with cardiovascular abnormalities do not mention this entity.13 Therefore, deciding on medical clearance for competitive sport practice is very challenging.5 We draw attention to this enigmatic entity by reporting on the outcome of a series of well-phenotyped athletes with extensive subepicardial DGE without cardiovascular symptoms.


Study population

From December 2008 to November 2014, we included all consecutive athletes in whom subepicardial DGE was detected because of the workup of abnormalities on their regular screening examination. Athletes were included and followed in two sports cardiology centres, some after referral. They were defined as participants competing at the professional, international or national amateur level, practising at least 6 h/week in the past 5 years. We excluded athletes referred because of overt cardiac symptoms and athletes with any associated cardiac disease that might have explained the DGE, that is, hypertrophic cardiomyopathy (HCM),14 coronary artery disease and DGE when present at the insertion points only.11

The study was approved by the hospital ethics committee and conducted in accordance with the Declaration of Helsinki. All participants gave informed consent.

Cardiovascular evaluation

Clinical examination and ECG

A cardiologist with experience in sport medicine conducted a personal and family history, symptoms and a physical examination. A resting 12-lead ECG was recorded and analysed using the Seattle Criteria.15

Resting transthoracic echocardiography

Transthoracic echocardiography was performed by a cardiologist and analysed according to the American Society of Echocardiography recommendations.16


CMR examinations were acquired using a Philips Achieva 3-T CMR or 1.5-T CMR (Philips Medical Systems, Best, The Netherlands). Cine-mode sequences were acquired in the short-axis, four chambers and LV long-axis views, whereas T2-weighted black-blood spin echo was carried out in the three typical planes before the gadolinium-chelate injection. First-pass perfusion images were acquired in the short-axis view during a first infusion of gadolinium-chelate (0.1 mmol/kg of gadolinium-chelate). About 5–10 min after the second injection (0.1 mmol/kg of gadolinium-chelate), late gadolinium images were performed in the three cardiac planes.17 Left (LV) and right ventricular (RV) volumes, mass, function and DGE quantification were assessed using customised analysis software by a blinded, single experienced investigator. The size of the enhanced myocardium was expressed in grams and as a percentage of LV mass.18

Maximal cardiopulmonary exercise test

Athletes performed a maximal exercise test with gas exchange analysis, and continuous 12-lead ECG monitoring. It was either performed on a cycle ergometer or on a treadmill in accordance with their sport specificity.

Evaluation of contractile reserve during exercise

Athletes underwent exercise CMR (n=3) and/or exercise echocardiography in order to assess global or regional LV dysfunction during a maximal exercise test. Exercise CMR was performed using a real-time CMR sequence; the subjects performed supine exercise during free breathing within the CMR bore.5 ,19 Exercise echocardiography was performed on a semisupine cycle ergometer. Absence of contractile reserve was defined as an improvement in LV ejection fraction (LVEF) of less than 5%.20

24 h Holter ECG

Athletes underwent 24 h Holter ECG including a training session. Non-sustained ventricular tachycardia (NSVT) was defined as ≥3 consecutive premature beats lasting less than 30 s.21

Electrophysiology study

An electrophysiology (EP) study was performed on the individual judgement of the cardiologist in charge of the athletes (n=3). It was performed before introduction of antiarrhythmic drugs (in the case of our patients only β-blockers were used) and without sedation. The ventricular stimulation protocols were standard, as previously described by our group.22 In all, stimulation protocol was applied during isoproterenol infusion if the baseline study was negative. Only the induction of monomorphic ventricular tachycardia (VT) was considered specific under these conditions.22

Longitudinal cardiovascular follow-up

In all athletes, a biannual cardiovascular evaluation was performed including the same non-invasive examinations. Athletes were carefully educated for ominous cardiac symptoms and were examined sooner if they developed symptoms.

Statistical analysis

Individual data are reported because of the small sample size. Data are expressed as mean±SD or percentage unless otherwise specified.

A paired-samples t test was used to compare CMR data before and after detraining (LV mass index and % of DGE/LV mass).


Results from primary investigations

Clinical data

The demographics of the seven athletes are summarised in table 1. The mean age was 26.1±4.9 years. All were involved in high-level competitive sport (6 professional athletes and 1 national level amateur), mostly practising endurance sports (n=5). Athletes were referred to CMR because of the suspicion of cardiomyopathy due to pathological T-wave inversions (PTWI; n=4) or ventricular arrhythmias on a screening exercise test (ventricular couplets (n=2) or NSVT (n=1)); figure 1). None of the athletes reported overt cardiovascular symptoms; but three reported hindsight decrement in their usual sporting performance (athletes n°1, 3, 4). Two of these athletes had suffered from acute infections in the preceding months (acute Toxoplasma infection in athlete n°3; and Mycoplasma pneumonia in athlete n°4), both with apparently full clinical recovery. No athlete reported a familial history of cardiac disease, and all denied any recreational or performance-enhancing drug use. Physical examination was normal in all athletes. The average peak VO2 reached expected values (62.9±15.6 mL/min/kg; 161.7±35.9% of expected peak VO2).

Table 1

Characteristics of the population at inclusion

Figure 1

Abnormal regular screening examinations. (A) ECG with pathological T-wave inversion and (B) rhythm strips from exercise testing.

DGE pattern and LV function

Figure 2 shows the DGE findings in the seven athletes (see online supplementary figures 1 and 2). DGE was localised predominantly to the lateral wall. The size of the DGE region was 20.3±7.7 g (12.0±4.8% of LV mass, table 2). DGE was exclusively subepicardial (n=4) or associated with transmural (athlete n°1) or with intramural patches (athletes n°2 and 3). No athlete had evidence of myocardial oedema on T2-weighted imaging; also, there was no pericardial effusion on cine mode and no microvascular obstruction on first-pass perfusion imaging. Therefore, all lesions were considered likely to reflect chronic scarring. In two athletes, global LV function was borderline at rest (LVEF 47% and 45% in athlete n°1 and 2) with no contractile reserve during exercise. The athlete with transmural DGE (athlete n°1) had focal hypokinesia in the corresponding segments, which worsened during exercise. Based on the measures of wall thickness and LV mass, none of the athlete was suspicious of HCM (table 2).

Table 2

CMR characteristics at inclusion

Figure 2

Delayed gadolinium enhancement (DGE) pattern in the seven athletes at presentation. Cardiac MRI showing subepicardial DGE (arrows).


Three athletes had exercise-induced NSVT (athletes n°1, 2, 3). Three athletes presented with less significant ventricular arrhythmias: one had a nocturnal slow NSVT but only couplets during exercise (athlete n°6); two other athletes demonstrated exercise-related couplets (athletes n°4, 7). One athlete had no evidence of ventricular arrhythmias (athlete n°5). An EP study was performed in three athletes (athletes n°2, 4, 7); the study was negative in each of them.

Overall decision after primary investigations

After this initial evaluation, three athletes were excluded from competitive sport (athletes n°1, 2, 3), two on the basis of LV dysfunction at rest and during exercise, combined with malignant exercise-related ventricular arrhythmias (NSVT). The third was stopped because of NSVT during exercise. All received medical therapy with β-blockers and ACE inhibitors in case of LV dysfunction. An implantable cardioverter defibrillator (ICD) was not implanted because they were asymptomatic. However, all athletes were followed up closely to detect any deterioration.

Longitudinal follow-up results

Mean follow-up (FU) was of 2.6±2.1 years (table 3).

Table 3

Findings during follow-up

Athletes excluded from competitive sport after initial evaluation (n=3)

In athlete n°1, global LV function recovered under medical therapy, although focal hypokinesia remained. Her peak VO2 increased, although she had stopped intense training (FU 5.4 years). Athlete n°2 had no recovery of LV function, although FU was only 6 months. None had recurrence of ventricular arrhythmia under medical therapy.

Athletes not excluded from competitive sport after initial evaluation (n=4)

The mean FU was 3.0±1.5 years. Two athletes developed symptomatic palpitations. Athlete n°4 developed palpitations related to polymorphic NSVT, whereas athlete n°5 developed fast sustained ventricular tachycardia during exercise with dizziness. Both athletes were excluded from competitive sport and received β-blocker therapy. Athlete n°5 had recurrence of VT under medical therapy leading to implantation of an ICD and a VT ablation procedure. Another athlete (n°6) demonstrated a decrease in resting LVEF from 60% to 40% with focal hypokinesia on exercise echocardiography for which he was disqualified. These three athletes also had a decrease in peak VO2 relative to the initial cardiopulmonary exercise test (table 3). Athlete n°7 remains asymptomatic, and has no LV function dysfunction and no arrhythmia. He ended his career as a professional cyclist but is still competing at an amateur level without any adverse outcome (FU of 3.4 years).

Four athletes (n°1, 4, 5, 6) underwent repeated CMR after a detraining period of 65, 4, 6 and 9 months, respectively. LV mass index decreased by −16.4±4.8% (from 81.2±9.4 to 67.5±4.7 g/m2, p=0.015), whereas DGE persisted and remained stable (% of DGE/LV mass remained unchanged, 14.3–14.6%, p=0.863; table 3).


The major result of the study is that manifest subepicardial DGE detected during the workup of asymptomatic young elite athletes presenting with PTWI or ventricular arrhythmia on screening exercise test is not benign even in the absence of thick walls (ie, no evidence for HCM) or RV structural abnormalities (ie, no evidence for arrhythmogenic right ventricular cardiomyopathy).

Despite being clinically asymptomatic at inclusion, all but one of the athletes presented with LV dysfunction and/or significant ventricular arrhythmias either at initial workup or during FU. This is more of a concern than previous studies may have suggested,10–12 certainly because the size of the DGE regions was much larger in the present series (12.0±4.8% vs 2.3%12 or 4.5%10 in other series). Moreover, the distribution of DGE was different, as these studies also included DGE of ischaemic origin or due to mechanical stress at the hinge point of the RV free wall and septum.10–12

It is difficult to determine the exact aetiology of DGE in the present series. The most frequently mentioned hypothesis is that subepicardial DGE is a scar from a previous myocarditis.23 None of our athletes had a prior diagnosis of acute perimyocarditis, but two athletes had an infection with agents that can cause myocarditis several months before the discovery of DGE.24 Myocarditis can be asymptomatic25 and athletes can develop a compromise resistance to common minor infectious illnesses as intensive sport can weaken immune function.26 Myocarditis could explain why the DGE was predominantly present in the lateral wall.27 One of the proposed explanations is that cardiotropic viruses can cause pericarditis after initial viraemia. Since the left lateral free wall is in direct contact with the pericardium, it might be the prime location for propagation via direct contact of viral infestation and/or inflammatory processes.28 Alternatively, in some disease states such as Fabry’s disease, the basal lateral wall has a predilection for early involvement, and it has been hypothesised that the lateral wall is subject to greater wall stress.29

An alternative explanation would be that the athletes had an inherited cardiomyopathy. There was no evidence of HCM in any of the athletes. However, it is more difficult to definitively exclude left dominant arrhythmogenic cardiomyopathy (LDAC).30 Indeed, this disease has several features in common with our cohort: PTWI, ventricular arrhythmia of LV origin, mild LV dilation and/or systolic impairment and non-ischaemic LV DGE. However, RV wall motion abnormalities and RV DGE, frequent findings in LDAC, were not observed in our case series. Furthermore, no familial history of cardiac disease was reported in our athletes, although this does not completely rule out the diagnosis as penetrance is variable and family members may not have exercised to the same level.31

Another cause to consider in athletes is use of anabolic/androgenic steroids.32 ,33 We had no indication that the present athletes had taken or had been exposed to recreational or performance-enhancing drug use. Moreover, there is no known pathophysiological data to explain focal fibrosis with such exposure.

LV fibrosis induced by chronic exposure to repetitive bouts of endurance exercise has been demonstrated in animal models,34 ,35 and mentioned in previous human studies.10–12 ,36 ,37 However, since the athletes in this study were younger than those reported in the literature, it may be expected that the cumulative exercise dose may be less, thus making it difficult to explain the more extensive DGE. Moreover, the distribution of DGE in these reported studies was much smaller and usually confined to the site of RV attachment, as observed in patients with pulmonary hypertension.11

Therefore, we do not consider that exercise alone could have induced such a huge amount of DGE. However, we wonder whether and to what extent strenuous, repeated exercise may affect negative remodelling and scar formation in the presence of another trigger, such as inflammation due to myocarditis or to a genetic cardiomyopathy like LDAC.30 ,31 Moreover, exercise is known to have an adverse effect on myocarditis. Indeed, mice forced to exercise during the initial days of coxsackie B3 infections developed a replacement fibrous scar and had a poorer outcome.38 Thirty years ago, sudden cardiac death among young orienteers in Sweden were reduced after recommendations not to train while infected.39 Nevertheless, athletes frequently train and compete during viral illnesses.40 Moreover, intense exercise can aggravate late negative remodelling, leading to myocardial dysfunction and trigger arrhythmias.

The results of this study are not sufficient to recommend exclusion from competitive sport practice solely based on manifest subepical DGE. At present, it is difficult to predict in whom negative remodelling will occur. As some cases show, LV remodelling may be progressive, leading to clinical arrhythmias despite initial negative EPS. Thus, a close FU of all participants with large DGE patches is warranted even if the initial presentation seems benign.


The study was not designed to estimate the prevalence of subepicardial DGE in athletes. Nevertheless, in a previous study, we demonstrated that isolated subepical DGE was associated with PTWI in 2.6% of cases.4

We have no comparison data in non-athletic cohorts to prove the deleterious effect of exercise in participants with subepicardial DGE. Nevertheless, in these athletes, strenuous exercise might not only have a deleterious effect on remodelling but also trigger ventricular arrhythmia.

We did not perform endomyocardial biopsies or genetic testing to exclude LDAC. However, given the location of the DGE lesions, a right-sided biopsy would almost certainly be negative.

With regard to the small sample size, it is difficult to deduct prognostic factors, making it difficult to speculate on the reasons for interindividual variability in negative remodelling.


Subepicardial DGE in asymptomatic athletes is not a benign finding. We wonder whether its development and poor outcome are due to the negative remodelling aggravated by intense exercise. A very comprehensive initial evaluation and a close FU are mandatory in such athletes. We call for a large prospective registry in order to better characterise the prevalence and outcome of athletes with subepicardial DGE.

What are the new findings?

  • Subepicardial delayed gadolinium enhancement (DGE) in asymptomatic athletes evaluated for suspected cardiomyopathy (in case of pathological T-wave inversions or ventricular arrhythmias during exercise) is not a benign finding, but is associated with ventricular arrhythmias and left ventricular dysfunction.

  • Adverse cardiac events can occur even several years after discovery of DGE.

  • Our findings raise the question whether extreme exercise during silent myocarditis may facilitate fibrosis generation and/or promote adverse remodelling.

How might it impact on clinical practice in the near future?

  • In case of subepicardial DGE in an asymptomatic athlete, medical clearance for competitive sport practice should be given only after a comprehensive initial evaluation.

  • This evaluation should focus on the occurrence of arrhythmias during exercise and exercise-induced left ventricular dysfunction. The place of electrophysiological studies needs further study, although (non-)inducibility has been related to outcome in other forms of structural heart disease.

  • In case of initial medical clearance, a close follow-up is mandatory throughout the athletic career, and the athlete has to be educated regarding ominous cardiac symptoms.

  • Implantation of a subcutaneous loop recorder should be considered in athletes who resume training and competition.


Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.


  • Contributors FS, GC, FC and HH contributed to the planning. FS, GC, ALG, JB, P-AL, PC, PM, FC and HH contributed to the conduct. FS, GC, ALG, JB, P-AL, PC, PM, FC and HH contributed to the reporting.

  • Funding FS is funded by a grant from the French Federation of Cardiology.

  • Competing interests None declared.

  • Ethics approval Hospital ethics committee.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • ▸ References to this paper are available online at

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles