Reproducibility of exercise-induced premature ventricular beats predicts concealed non-ischemic left ventricular scar on cardiac magnetic resonance in athletes

Abstract

Abstract Background Non-ischemic left ventricular scar (NLVS) at cardiac magnetic resonance (CMR) is a common substrate of apparently idiopathic premature ventricular beats (PVBs) in athletes. PVBs morphology, complexity and exercise-behavior are known predictors of NLVS. It remains to be established whether reproducibility of PVBs at exercise testing correlates with CMR findings. Purpose We aimed to evaluate the possible role of PVBs reproducibility at repeated exercise testing in predicting the presence of an underlying NLVS in athletes who underwent CMR for apparently idiopathic ventricular arrhythmias. Methods We included all consecutive competitive athletes referred to our center for the evidence of PVBs during pre-participation screening, who underwent two maximal exercise testings within one month (the first at the time of pre-participation screening, the second at the time of outpatient evaluation at our sports cardiology clinic). Exclusion criteria were known heart disease, family history of cardiomyopathy or channelopathy, abnormal ECG and echocardiography. We also excluded athletes with “common/usually benign” PVBs (non-exercise-induced infundibular or fascicular PVBs) as well as those on ongoing antiarrhythmic therapy. Reproducibility was defined as the presence of PVBs with same morphology and exercise-behavior in two subsequent tests. LGE on CMR was quantified with 5-SD method, using a threshold of 5% to define the presence of pathological LVNS. Results A total of 64 apparently healthy competitive athletes (86% males, mean age of 33±14 years old) were included. On CMR, NLVS was identified in 26 (41%). PVBs reproducibility criteria was met in 32 athletes (50%). A statistically significant difference in PVBs reproducibility was found between patients with and without LVNS [26 (100%) versus 6 (16%), p<0,001], irrespective of other factors such as age, sex, biventricular volumes and function (Figure 1). Among the 26 patients with NLVS, 18 (69%) showed PVBs with right-bundle branch block (R-BBB) with superior axis configuration, alone or associated with other morphologies. These arrhythmic patterns were confirmed by a second test in all cases. Among the 38 patients with normal CMR, the first exercise test showed R-BBB/superior axis PVBs, alone or associated with other morphologies, in 22 (57%); the second test confirmed such pattern in only 6 (16%), showing instead different morphologies in 8 (21%) and no PVBs in 24 (63%). Figure 2 represents an example of two cases included in the study. Conclusion In apparently healthy athletes who underwent CMR for “uncommon” PVBs, the presence of underlying LVNS was predicted by ventricular arrhythmia reproducibility. Specifically, reproducibility of PVBs with RBBB/superior axis morphology was highly predictive of a positive CMR. This finding may be very relevant for appropriate prescription of CMR in athletes with apparently idiopathic ventricular arrhythmias. Funding Acknowledgement Type of funding sources: None.