83 ECG changes in Fabry disease (e.g. P-wave duration, PR-interval, QRS- duration, Cornell index, spatial QRS-T angle and frontal QRS-axis) (figure 1-2). Since not all female patients will develop cardiac events of FD, it is especially important to be able to detect development of cardiac disease in these patients. As can be deducted from supplemental figure 2-5 changes in ECG parameters do not occur in all patients and a proportion of the female patients follows a trajectory over time that is similar to control subjects. This suggests that these parameters may indeed be suitable to differentiate between those patients who develop progressive cardiac disease and those who do not. The parameter that best distinguishes patients with from those without LGE is the spatial QRS-T angle, which is likely to be a much more reproducible measurement compared to other ECG parameters [41] to quantify the interaction between ventricular depolarisation and repolarisation. This study shows, for the first time, the increment in spatial QRS-T angle in FD patients. Previous studies have shown that this vectorcardiographic parameter can independently predict cardiac events in the general population and patients with other heart disease (e.g., heart failure and myocardial infarction) [31, 39, 40]. More longitudinal studies are required to investigate whether the spatial QRS-T angle is a useful parameter to predict cardiovascular events in FD patients. What is needed to firmly establish these parameters as prognostic biomarkers is an established link between early ECG changes and clinical outcomes. This study shows that men with classical FD showed differences in ECG parameters compared to control subjects from age 20 years onwards, with the Cornell index as the earliest marker to change. In women with classical FD this occurs from age 30 years onwards (figure 4A). Whether these ECG alterations predict cardiac complications or the occurrence of LVH and LGE, that are known to occur approximately 20 years later [14, 20] could not be determined because of: a) the relatively small number of cardiac events in this cohort and the absence of cardiac events in younger patients and b) the missing long-term ECG data prior to an event for patients who were diagnosed after their first cardiovascular event (especially women without a positive FD family history or classical FD symptoms). To investigate the predictive utility of ECG markers in FD patients, it is essential to conduct multicenter studies that can provide these missing data. The current study provides insight which parameters can be studied to assess their predictive value for the development of cardiovascular complications in FD patients. This study has several limitations. First, survival (premature death) and treatment bias (exclusion of ECGs of patients with externally paced rhythm) attribute to selection bias. This is in particular relevant for older men with classical FD, since it has led to under-recording of ECGs of these severely affected patients. The unavoidable selection of ECGs from patients over the age of 50 with a relatively 3
RkJQdWJsaXNoZXIy MTk4NDMw