1 Introduction

Outflow tract (OT) premature ventricular contractions (PVCs) are commonly observed in clinical practice, both in symptomatic and asymptomatic subjects. In most cases, this arrhythmia is benign, observed in structurally normal hearts and not requiring any therapeutic intervention. Occasionally, instead, PVCs may herald cardiac disease or induce cardiomyopathy [1,2,3,4,5]. OT PVCs can originate from the right ventricle (RVOT), more often anterior and superior septal site, just inferior to the pulmonic valve, or from the left ventricle (LVOT), with different occurrence rates [6,7,8,9]. Idiopathic PVCs may also originate from papillary muscle [10, 11].

Echocardiography is the most commonly used imaging method to evaluate the ventricular systolic function on the base of incomparable minor costs than cardiac magnetic resonance imaging (CMR) or computerized tomography (CT). However, CMR allows the assessment of cardiac structure and function with excellent resolution and accuracy compared with echocardiography and is indicated, according to the ESC 2015 guidelines on the prevention of sudden cardiac death, in cases where echocardiography is not able to evaluate these parameters optimally [12]. This is the case of some hereditary cardiomyopathies, such as arrhythmogenic right ventricular dysplasia (ARVD), which predispose to ventricular arrhythmias, ventricular dysfunction, and sudden cardiac death. Most of the studies that evaluated the prognosis of patients with PVCs in apparently normal hearts did not use advanced imaging to rule out the presence of cardiomyopathy [13,14,15]. Moreover, some studies in which patients underwent CMR have found the presence of substrate alterations whose clinical and prognostic significance has not yet been well evaluated [16, 17].

In this study, we therefore sought to evaluate patients with OT PVCs and apparently normal heart at echocardiography, who underwent CMR evaluation to identify any possible organic substrate potentially responsible for arrhythmias.

2 Methods

2.1 Study patients

We included in this study 33 consecutive patients with frequent PVCs originating from the ventricular OT (right and left), evaluated at the Cardiology Department of University of Foggia from September 2016 to February 2019. All patients had normal baseline electrocardiogram, without history of coronary artery disease, cardiomyopathy, and diabetes mellitus, or family history of sudden death.

All patients underwent clinical examination, 12-lead electrocardiogram, echocardiography, 24-h Holter monitoring, and CMR at Ospedali Riuniti University Hospital in Foggia, Italy. Patients with symptomatic arrhythmias were treated with an oral antiarrhythmic agent (beta-blockers, propafenone, flecainide) according to current guidelines. Moreover, patients with frequent PVCs or ventricular bigeminy during examination were treated with an oral antiarrhythmic agent before CMR examination in order to optimize ECG trigger and to obtain optimal image acquisition.

2.2 Twelve-lead ECG and site of origin

The place of origin of these PVCs was evaluated using the 12-lead ECG. Typically, the PVCs originating from RVOT have an inferior axis and a left bundle branch block (LBBB)-like configuration with precordial R/S transition at or later than lead V3, while those PVCs originating from LVOT have an inferior axis and a LBBB-like morphology with precordial R/S transition at or earlier than V3 [7, 18,19,20,21]. Patients with precordial R/S transition in V3, which may originate from both RVOT and LVOT, were evaluated according to the criteria by Betensky et al., based on the measure of the VT and SR precordial transition, the so called “V2 transition ratio”, which can predict with 95% sensitivity and 100% specificity the left or right side of origin [22].

2.3 Cardiac magnetic resonance

The CMR was performed with Philips Achieva 1.5 T apparatus. The standard protocol provided steady-state free precession sequences (SSFP) ECG gated, breath hold, in 4 chambers, 2 chambers, 3 chambers, and short axis for the study of kinetics of the left ventricle and of the right ventricle; T2 STIR sequences for the study of myocardial edema; and phase-sensitive inversion recovery sequences (PSIR) for the study of myocardial necrosis/fibrosis performed after 10–15 min from the administration of gadolinium (0.1 mmol/kg). The evaluation of the systolic function of both ventricles was carried out by means of a dedicated software (Segment version 1.9 R3878) on images SSFP ECG gated, which allows to obtain an excellent endocardium blood contrast. Manually delineating the endocardial edge on all the sections obtained in short axis, the software calculates the volume (product of the area for the thickness of each section) adding the partial volumes of the individual sections. This procedure is performed in the telediastolic and telesystolic frame, thus obtaining the telediastolic volume (VTD), the telesystolic volume (VTS), the ejection fraction (VTD-VTS/VTD), and the stroke volume (VTD-VTS). The cardiac mass was automatically calculated by using the software delineating the epicardial profile and calculating the epicardial volume in each slice in telediastole and telesystole. From this volume, the endocardial volume was subtracted to obtain the volume of the heart muscle. Multiplying this data by the specific weight of the heart muscle (1.05 g/mmq), the cardiac mass was obtained.

The study of myocardial fibrosis was performed using PSIR sequences performed 10–15 min after the administration of gadolinium (Gadovist 0.1 mmol/kg). Myocardial fibrosis has been identified as a hyperintensity of the signal in the PSIR sequences, in 4 chambers, 2 chambers, and short axis. The quantification of fibrosis in grams was carried out on all the slices obtained in the short axis, with a manual method by delineation of the area with signal hyperintensity, thus obtaining the volume to be multiplied by the specific weight of the heart muscle (1.05 g/mm2).

3 Results

The mean age of the patients was 43 years, 51% were male, 26% hypertensive, 84% symptomatic for palpitations, and 11 patients (33%) underwent transcatheter ablation in accordance with current guidelines (Table 1). Based on the electrocardiographic analysis, all patients had PVCs with an inferior axis, while the transition was in 4 patients (12%) ≤ V2, in 12 (36.3%) V3, and in 17 (51.5%) ≥ V4. Of the patients with a R/S transition in V3, 6 presented a V2 transition ratio ≥ 0.6, predicting an LVOT origin with high sensitivity and specificity according to the criteria by Betensky et al. [22]. The mean number of PVCs was 15,830/24 h (± 11,146/24 h). In 85% of cases, PVCs were monomorphic, in 15% polymorphic; 48.5% of patients had PVC couples, 30% non-sustained ventricular tachycardia (NSVT), and 9% sustained tachycardia (SVT).

Table 1 Population’s characteristics
Full size table

CMR showed anomalies in 5 patients out of 33 (15%), despite apparently normal electrocardiogram and echocardiography (Table 2).

Table 2 ECG intervals and PVC morphology in 5 patients out of 33 with CMR anomalies
Full size table

3.1 Patient #1

The first patient was a 53-year-old woman, symptomatic for palpitations. The ECG showed sinus rhythm, left axial deviation, and rSr’ aspect in V1–V2 with fractionated QRS in inferior and lateral leads and PVCs with two morphologies, LBBB-like with precordial R/S transition in V4 from RVOT, and right bundle branch block (RBBB)-like with upward axis compatible with an origin from the left posterior fascicle (Fig. 1a). The 24-h ambulatory electrocardiogram monitoring showed frequent (1440) PVCs (1.3% PVC burden), polymorphic (two morphologies), with couples and triplets; the prevalent morphology was from RVOT. The echocardiogram showed normal kinesis and R-LVOT diameters. CMR revealed increased LV end-diastolic diameters, increased indexed volumes, reduced ejection fraction, and accentuated ventricular wall trabeculation. PSIR-TFE sequences for the “late enhancement” study showed extensive hyperintensity with predominantly subepicardial distribution within the anterior, lateral, and inferior walls and mesocardial distribution within the interventricular septum compatible with areas of fibrosis (Fig. 1b and c). A therapy with controlled-release flecainide (200 mg) was started, with a significant reduction in the number of PVCs (640), in the absence of repetitive forms, but with a greater number of morphologies from right and left ventricles, compared with the previous Holter monitoring. This drug was then replaced with amiodarone after the finding of extended fibrosis at CMR with a further reduction of extrasystoles (123), without repetitive forms. Patient underwent genetic testing (desmocollin-2, desmoglein-2, desmoplakin, plakophilin-2, transmembrane protein 43, junction plakoglobin, desmin, SCN5A) at another hospital that was normal.

Fig. 1
figure 1

a Electrocardiogram. First morphology: LBBB-like with precordial R/S transition in V4 at origin from right outflow tract. Second morphology: RBBB-like and upper axis compatible with origin from the posterior fascicle of the left branch. b PSIR-TFE sequences for the “late enhancement” study showed extensive hyperintensity with predominantly subepicardial distribution of the anterior, lateral, and inferior walls and mesocardial of the interventricular septum compatible with areas of fibrosis. c T1-TSE sequences showed subepicardial adipose infiltration of the lateral wall of the left ventricle

Full size image

3.2 Patient #2

The second patient was a 52-year-old man, hypertensive with LBBB-like PVCs originating from LVOT and a R/S transition in V3. At ambulatory electrocardiogram monitoring, very frequent (12,134; 13.8% PVC burden) polymorphic RBBB-like PVCs (4 morphologies) were found, apparently from the left ventricle (Fig. 2a). The echocardiogram showed a slight accentuation of the apical trabeculation without evident anomalies in LV kinetics. CMR showed slightly increased left ventricular diameters and accentuated mid-apical trabeculation with a maximum non-compact/compact thickness of 2.6:1 (lower and lateral wall) indicative of non-compact myocardium (Fig. 2b and c). Also, this patient was initially prescribed drug therapy with controlled-release flecainide (200 mg), changed to amiodarone after the CMR findings with a significant reduction in the number of PVCs (655), during amiodarone therapy, in the presence of few repetitive forms (3 monomorphic couples).

Fig. 2
figure 2

a A 24-h 12-lead ambulatory monitoring. Four morphologies of ventricular extrasystoles with left ventricular morphology. The first morphology (on the left) originates from the LVOT. b, c Cardiac magnetic resonance showing slightly increased left ventricular diameters and accentuated mid-apical trabeculation with a maximum non-compact/compact thickness of 2.6:1 (lower and lateral wall) indicative of non-compaction myocardium

Full size image

3.3 Patient #3

The third patient was a 68-year-old hypertensive man, with 10,437 (8.9% PVC burden) monomorphic LBBB-like PVCs and transition in V4 compatible with RVOT origin. CMR showed meso-subepicardial fibrosis at medium basal segments of the interventricular septum (Fig. 3). The patient underwent a stress test ECG, before and after the antiarrhythmic therapy, without evidence of inducible myocardial ischemia and other ECG anomalies. After controlled-release flecainide therapy (150 mg), only few non-repetitive extrasystoles (5) were observed.

Fig. 3
figure 3

Cardiac magnetic resonance showing meso-subepicardial fibrosis at medium basal segments of the interventricular septum

Full size image

3.4 Patient #4

The fourth patient was a 58-year-old hypertensive woman with 10,220/day (9.9% PVC burden) monomorphic PVCs and episodes of monomorphic ventricular tachycardia of up to 64 beats with a LBBB-like appearance with R/S transition in V4 compatible with RVOT origin. Treatment with metoprolol 100 mg was found to be partially effective but associated with nocturnal AV 2:1 block episodes. Coronary angiography was normal, and CMR showed focal area of meso-subepicardial fibrosis of the basal segment of the inferior-lateral wall with an ischemic late gadolinium enhancement (LGE) distribution pattern. The patient underwent effective transcatheter ablation of the arrhythmogenic focus in RVOT anterior free wall.

3.5 Patient #5

The fifth patient was a 26 year-old man, referred for palpitations and 1178 (1.3% PVC burden) monomorphic LBBB-like PVCs, repetitive (120 pairs and phases of slow ventricular tachycardia, 110 bpm, with the same morphology), with transition in V2 compatible with LVOT origin. After administration of propafenone therapy 325 mg twice daily, there was a rapid significant reduction of PVCs. The echocardiogram showed normal diameters and ejection fraction, CMR minimal meso-subepicardial fibrosis of the basal segment of the inferior-lateral wall.

4 Discussion

In the present study, we show that CMR may carry an adjunctive value in the evaluation of patients affected by ventricular OT PVCs and apparently normal echocardiogram as able to further identify cases of structural heart disease, potentially leading to clinical complication when left not correctly treated. In our study, exclusively, subjects with history of OT PVCs, without other clinical, electrocardiographic, and echocardiographic criteria of cardiomyopathy, were enrolled. The data of the literature confirm in most cases the benign nature of these PVCs in the absence of structural cardiac anomalies (or with minor anomalies) [1, 23]. However, cases of benign PVCs from the right and left ventricular outflow tracts triggering malignant ventricular arrhythmias have also been described. These cases, yet, seem to have some distinctive electrocardiographic features compared with cases of benign PVCs such as shorter coupling interval, number of PVCs, and presence of non-sustained ventricular tachycardia run at high frequency at Holter monitoring [2, 24, 25].

The patients we evaluated presented structurally normal hearts with normal global and segmental ventricular systolic function on the echocardiogram. However, CMR-LGE showed in 3 of them areas of fibrosis limited in one case to the middle basal segments of the interventricular septum and in two patients to the middle basal segments of the inferior-lateral wall. In two of them, however, CMR-LGE showed significant alterations characterized, in one patient, by extensive areas of subepicardial fibrosis of the left ventricle compatible with arrhythmogenic left dominant dysplasia, while in another patient, it showed a marked trabeculation of the medium apical segments of the left ventricle compatible with non-compaction cardiomyopathy. Clinically, the 3 patients with limited alterations showed monomorphic outflow tract PVCs  that were significantly reduced by therapy with IC antiarrhythmic drugs, while the other two patients with significant structural alterations showed polymorphic PVCs that were not significantly reduced with IC antiarrhythmic drug therapy. Moreover, CMR-LGE data showed that the ejection fraction of both ventricles of patients with PVCs was slightly lower than those of patients without PVCs. It is already known that the frequent PVCs, even in the absence of sustained tachycardia, is associated with progressive dilatation and left ventricular dysfunction, thus representing a potential reversible cause of cardiomyopathy [3,4,5, 26,27,28].

According to the ESC guidelines [12], echocardiography is the first-line imaging method in the evaluation of patients with PVCs, whereas CMR should be considered if echocardiography is not able to provide an accurate assessment of the function of the ventricles and in the evaluation of structural alterations. In our case series, the only echocardiographic evaluation, especially in cases with subepicardial fibrosis, would have provided satisfactory information that would not make us consider the execution of the CMR. That raises important questions from the clinical and prognostic points of view. Clinically, the failure to identify an endocardial fibrosis pattern may put the patient at greater risk in case of administration of IC-class antiarrhythmic drugs. It is known that these drugs may slow conduction within a crucial portion of a potential circuit until re-entrant excitation is established, exacerbating a ventricular tachycardia [29]. As demonstrated in some studies, this effect is particularly evident when the heart rate increases, as during a physical effort, due to their frequency-dependent block [30]. Moreover, primary transcatheter ablation is recommended in patients with ventricular arrhythmias originating from the RVOT, while in those originating from the LVOT, this should be considered only if the antiarrhythmic therapy is ineffective, due to of the close anatomical proximity between the left ventricular outflow tract, the coronary arteries, and the great cardiac veins [12]. From a prognostic point of view, the lack of identification of a relevant cardiomyopathy, such as a pattern of arrhythmogenic dysplasia, can put such patients at greater risk of sudden death, especially in young subjects, and may require an ICD implantation. In contrast, patients with idiopathic PVCs have a favorable prognosis and do not require an ICD.

The favorable role of CMR in the identification of areas of fibrosis has been already considered. Two previous study evaluated that in subjects with frequent PVC of LBBB morphology, CMR allows risk stratification, and identification of RV abnormalities was associated with worse outcome, in the absence of definite diagnosis of arrhythmogenic right ventricular cardiomyopathy/dysplasia (ARVC/D) [31, 32]. By contrast, a study of CMR in 20 patients with idiopathic RVOT tachycardia showed no differences in the incidence of qualitative CMR findings in patients compared with controls [33]. However, all these studies did not use LGE images that provide better differentiation between normal and diseased myocardium and are able to provide a potential fibrotic arrhythmic substrate, as demonstrated in a recent work in three-quarters of patients that underwent sudden cardiac arrest [34]. Therefore, these findings are unlikely to be specific, and some of these studies may have included patients with different pathological substrates. A more recent study in patients with the clinical diagnosis of RVOT tachycardia, DE-CMR reveals RV structure, function, and myocardial tissue characteristics similar to normal controls in the absence of myocardial scar in all 46 patients with tachycardia and 16 normal controls [35], suggesting that the vast majority of patients with RVOT arrhythmias have a primary electrical disorder that is not a forme-fruste of ARVC. In contrast, a study that included forty-six patients with monomorphic PVCs of LV origin and negative routine diagnostic work-up detected myocardial structural changes with LGE-CMR in a significant proportion of patients (41%) and a worse outcome [36]. In our series, however, we identified 5 patients (15%) out of 33 with structural abnormalities, 4 of which are referable to areas of fibrosis. In these patients, scar location and PVC morphology do not correlate, and the causal association between underlying structural heart disease and PVCs remains uncertain. A recent study suggested a pathogenic link between occult inflammation and PVCs that are considered to be idiopathic in patients with both reduced and preserved LVEF [37]; 51% out of 107 patients with frequent symptomatic PVCs (> 5000/24 h) and no history of ischemic heart disease had evidence of underlying inflammation suggested by a positive 18F-fluorodeoxyglucose (FDG) uptake on PET imaging, confirming a low sensitivity for LGE-CMR, since only 6 patients out of 41 had concordant findings of LGE-CMR and the localized FDG uptake on PET imaging.

Finally, in our study, one of these patients was diagnosed with arrhythmogenic “left dominant” cardiomyopathy, a pathology described for the first time in 2008 by Sen-Chowdhry et al. [38], which shows a propensity for electrical instability. It is possible that its prevalence has been underestimated to date, because it is confused with dilated cardiomyopathy. In this patient, despite normal echocardiogram, the ECG revealed a fragmentation of the QRS in the inferior and lateral leads and in V2, which partly agreed with the abnormalities detected by MRI.

Correct myocardial assessment with CMR of subject with outflow tract PVCs and apparently normal echocardiogram may be particularly relevant from a therapeutic and prognostic point of view also for risk stratification and assessment of family members.

5 Conclusions

CMR may show an added value in the evaluation of patients affected by ventricular OT PVCs and apparently normal electrocardiogram and echocardiogram examinations. A preliminary screening with CMR may be considered before any further invasive electrophysiology assessment and therapeutic planning.

6 Limitations

Principal limitations of this study are represented by the observational and monocentric nature of the study and the relatively small population enrolled. The study is very small and descriptive and does not provide any long-term outcome information.