Several limitations of this study warrant mention. First, the number of patients positive for LGE was relatively small. However, we believe that the sample is representative of patients with extracardiac sarcoidosis. Second, the estimation of IVS thinning was based on qualitative assessment performed by two technicians and, as such, may have suffered from inaccuracy. Therefore, the use of standard assessment methods will be necessary in future studies investigating IVS thinning. Third, because patients were not monitored by 24-h ECG, it is possible that several tachyarrythmias and bradyarrhythmias were missed. For this reason, one of the study end points regarding arrhythmic events was defined as symptomatic arrhythmia including ventricular tachyarrhythmia necessitating admission and bradycardia leading to pacemaker implantation. Fourth, 18F-fluorodeoxyglucose PET, in addition to LGE-CMR, for detecting inflammatory activity was not performed, and there may be several uncertainties on the precise pathophysiologic interpretation of myocardial hyperenhancement. Fifth, we did not completely exclude CAD as a cause of LGE by coronary angiography, because the American Heart Association guidelines state that neither CT angiography nor magnetic resonance angiography should be used to screen for CAD in patients who have no signs or symptoms suggestive of CAD,25 and no patients in the current cohort had such signs or symptoms. In addition, no patient in the LGE-positive group had typical subendomyocardial damage (CAD pattern), as reported in the Results section. Sixth, cardiac involvement may be present in the absence of any abnormality on standard cardiac testing, may manifest as an abnormal ECG alone, or may be recognized as an asymptomatic abnormality on echocardiogram, or even on LGE-CMR. Considering these facts and the findings of our study, it would be difficult to distinguish between an earlier and clinically silent phase of cardiac sarcoidosis with detection of myocardial damage by multimodalities including LGE-CMR. Finally, the sample size was small, thereby limiting the ability to generalize the findings and the statistical power for detecting differences in negative data. Regarding the follow-up period, although it was longer than that of previous studies, it might be still inadequate to conclude that asysmptomatic patients in the LGE-positive and LGE-negative groups had similar outcomes; namely, the chance of a type 2 error is a possibility. Therefore, further prospective studies with a larger asymptomatic population and longer follow-up are warranted.