Excitation-contraction coupling is a process providing the basis for muscle contraction, in which the key effector molecule is calcium ions. Calcium-induced calcium release (CICR) is a fundamental cellular mechanism for generating and amplifying intracellular calcium signals. The excitation-contraction coupling of cardiomyocytes depends on this process. PVCs lead to prolonged action potential time of ventricular cardiomyocytes and a decrease in the density of the transient outward K${}^{+}$ currents (Ito), inward rectifier K${}^{+}$ currents (IK1) and L-type Ca${}^{2+}$ currents (ICaL), accompanied by decreased expression of the associated ion channel subunits [21]. Changes in ion channels may lead to abnormal repolarization of cardiomyocytes, increasing the risk of malignant arrhythmias [28] (Table 1, Fig. 1). In the swine model of PVICM, decreased sarcoplasmic reticulum (SR) calcium ATPase (SERCA2a) and increased ryanodine receptor type 2 (RyR2), sodium-calcium exchanger (NCX1), calcium/calmodulin-dependent protein kinase II alpha (CaMKII-$\alpha$) and phospholamban (PLN) expression were observed [24] (Table 1). Alterations in protein expression and ion currents suggest that PVCs also lead to impaired CICR and abnormal excitation-contraction coupling, resulting in abnormal cardiac systolic function and smaller positive inotropic effects. Furthermore, voltage-gated calcium channel 1.2 (Cav1.2) protein expression and ICaL density were reduced, and Cav1.2 protein was relocated away from the t-tubules, leading to the decrease of Ca${}^{2+}$ entering the dyad gap through the opened L-type calcium channel (LTCC) resulting in the reduction in cytoplasmic Ca${}^{2+}$ concentration, reduced calcium spark events, and impairing E-C coupling in PVICM cardiomyocytes (Fig. 1). The decreased expression of SERCA2a, leading to reduced SR calcium reuptake in cardiomyocytes with PVICM, resulted not only in a reduction in Ca${}^{2+}$ concentration in the SR, but also in an increase in the cytosolic Ca${}^{2+}$ concentration, thereby inhibiting myocardial relaxation. Thus, dysfunction of the ventricular systolic and diastolic function results from decreased SERCA2a expression. Physiologically, dephosphorylated PLN interacts with SERCA and decreases SERCA calcium affinity, inhibiting calcium reuptake by the SR. In recent years it has become increasingly clear that several Ca${}^{2+}$-dependent proteins contribute to the fine tuning of E-C coupling. One of these is the Ca${}^{2+}$/calmodulin-dependent protein kinase (CaMK) of which CaMKII is the predominant cardiac isoform. PLN could be phosphorylated at Thr-17 by CaMKII, alleviating the PLN-mediated inhibition of SERCA activity and increasing the SR Ca${}^{2+}$ uptake [29]. The up-regulation of the inhibitory protein PLN aggravates the E-C coupling disorder of cardiomyocytes. The up-regulation of CaMKII-$\alpha$ can partially alleviate the abnormal calcium handling mediated by the up-regulation of PLN, which may be a compensatory mechanism to prevent the heart from complete systolic failure. Furthermore, CaMKII can enhance RyR2 activation, when it can both unload Ca${}^{2+}$ from the SR and induce arrhythmias in the setting of heart failure [29]. Additionally, increased NCX expression leads to increased intracellular calcium excretion, which may partially alleviate diastolic dysfunction due to reduced calcium reuptake via SR, but may lead to further reduced systolic function [30, 31, 32]. Reduced IK1 current disrupts resting membrane potentials and further enhances arrhythmias mediated by NCX upregulation [30].

Fig. 1.

premature ventricular contractions (PVCs) is the most common cardiac arrhythmia in patients, while some of them could develop cardiac dysfunction in the several years. According to different characteristics in patients, risk factors for premature ventricular contraction-induced cardiomyopathy (PVICM) are complex and diverse, including PVC burden, PVC QRS duration PVC origin and sexuality. Investigation of molecular mechanisms has predominantly been studied in animal models, primarily swine and canine. Based on those models, typical tissue alterations in PVICM are mild fibrosis and eccentric hypertrophy. Besides, frequent PVCs enhance sympathetic activity, further exacerbating structural alteration in models. Substructural remodeling includes reduced T-tubules, decreased dyad intensity and Z-line arrangement disarray, leading to reduced L-type calcium currents and decreased systolic calcium transient synchrony. All of the changes might function corporately or separately, resulting in cardiac dysfunction and malignant arrhythmia, increase the risk of sudden cardiac death. PESP, post-extrasystolic potentiation; JPH2, junctophilin-2; BIN1, bridging integrator-1; APD, action potential duration; Ito, transient outward K${}^{+}$ currents; IK1, inward rectifier K${}^{+}$ currents; ICaL, L-type Ca${}^{2+}$ currents. Created with https://biorender.com/.

Genome-Wide Association Studies (GWAS) confirm the association of genes coding for calcium handling proteins are at an increased risk of PVICM development in patients with frequent PVCs. It is hypothesized that mutations in calcium handling genes may affect calcium homeostasis, resulting in decreased sodium current and slow conduction, thereby prolonging the QRS duration [33].

Dyssynchronous ventricular contractions have long been considered to be the primary mechanism for the development of PVICM. In PVICM animal models, a linear relationship was found between the degree of LV dyssynchrony and the upregulation of CaMKII-$\alpha$-mediated RyR2 phosphorylation [24], suggesting that CaMKII-$\alpha$ may be a particularly important mediator and a potential therapeutic target for PVICM. In a PVICM swine model, LV intra mechanical remodeling persists despite the normalization of LV systolic function 4 weeks after stopping pacing [25], which suggests that PVICM patients recovering from cardiac dysfunction should continue to be closely followed.

The 24-hour electrocardiogram analysis of heart rate variability in patients with idiopathic PVCs found that autonomic activities were involved in the occurrence of PVCs [34, 35], and frequent PVCs lead to increased peripheral tissue and cardiac sympathetic activity and increased coronary sinus norepinephrine levels [26]. In a swine model of PVICM, neural remodeling was characterized by extracardiac sympathetic hyperinnervation and sympathetic neural hyperactivity [36]. The neural remodeling—stellate ganglia hyperinnervation—persisted despite normalization of LV systolic function [37]. Cardiac dysfunction in PVCs may be triggered and facilitated by chronic disruption of the sympathetic-vagal balance. At an early stage, depletion of cardiac transient receptor potential vanilloid-1 (TRPV1) afferents by resiniferatoxin (RTX) improved LV systolic function and alleviated cardiac fibrosis in PVICM animals, while this improvement was not apparent at late stages, and had no effect on autonomic activity [38]. These studies suggest that TRPV1 mainly plays a role in promoting myocardial fibrosis in the early stage, and then myocardial remodeling and dysfunction are mainly affected by sympathetic imbalance, indicating that neuromodulation may play distinct roles at different stages of PVICM.

The sliding filament theory is a classic model for explaining muscle contraction proposed by Andrew Huxley and Hugh Huxley in 1954, and is almost universally accepted [39, 40]. The cross-bridge cycling results in force production through cyclical conformational changes of myosin heads. The structure of myosin-actin interaction in the cross-bridge cycling has different characteristics and different spatial positions from thin filaments, as determined by X-ray diffractometry and scanning electron microscopy. The state of myosin heads close to the thin filament is defined as the disordered relaxed state; the super relaxed state is a low-energy metabolic state, where myosin head interacts with one another and the blocked head (BH) interacts with the lever of thick filaments, keeping the myosin head away from the filament and making it difficult to bind adenosine triphosphate (ATP) [41, 42]. It is currently believed that myosin super relaxed state (SRX) plays an important role in regulating energy utilization and cardiac contraction [43]. The cellular basis for the Frank-Starling mechanism is length-dependent activation (LDA). Calcium sensitivity increases when sarcomeres are stretched, causing increases in cardiac contractility. The mechanism of myofilament LDA derived from swine ventricular myocytes is that passive stretching converts more myosin SRX to disordered relaxed state (DRX) [44]. Changes in the balance between SRX/DRX may explain cardiac dysfunction occurring in cardiac diseases. Studies on familial hypertrophic cardiomyopathy have found that most variants are located in genes encoding myosin head and neck of filaments [45, 46, 47]. The structural changes of thick filaments change the DRX/SRX balance, and more myosin heads are in the DRX state resulting in cardiac hypercontractility and impaired diastolic function. Furthermore, Mavacamten, a myosin inhibitor that favors the closed conformation of myosin heads, achieved significant therapeutic effects in a Phase III clinical trial in obstructive hypertrophic cardiomyopathy [48]. In dilated cardiomyopathy, the aspartate-to-alanine substitution at position 94 in the regulatory light chain of myosin (RLC) encoded by the myosin regulatory light chain 2 (MYL2) gene results in an increased number of SRX heads and a subsequent reduction in myocardial contractility [45]. In addition to calcium handling, experimentally, myocardial force production is also modulated by alterations of conformations of the myosin head during ischemia, hypoxia or stretching. Therefore, it is important to explore the role of DRX/SRX in reversible cardiomyopathies such as PVICM and the efficacy of myosin inhibitors in PVICM.

Studies of myocardial cytoskeletal proteins in heart failure (hypertrophic cardiomyopathy (HCM), dilated cardiomyopathy (DCM)) have suggested that cytoskeletal proteins, such as actin and desmin were significantly reduced [49, 50]. Similar changes in the actin cytoskeleton were observed in a swine model of PVICM, which was associated with Z-line arrangement disarray and played a crucial role in inefficient contractile function and cardiomyocyte remodeling [18]. This adaptive change may initially protect the heart from changes in mechanical stress to adapt to pressure changes, but long-term, there is a lack of adaptation and a decrease in the systolic and diastolic capacity of the heart. Similar alterations were observed in PVICM.

Dyads are subcellular structures of calcium signaling in cardiomyocytes, where the interconnection and connection of the plasma membrane network composed of the SR and the T tube, act as the intracellular calcium synapse of cardiomyocytes. In PVICM, the number of T-tubules is highly reduced, while the length of sarcomere and mitochondrial structure do not change significantly [22]. Decreased T-tubules diminish dihydropyridine receptor-ryanodine receptor (DHPR-RyR) coupling efficacy, which is responsible for reduced L-type calcium currents and decreased systolic calcium transient synchrony [22]. T-tubule remodeling occurs before the onset of ventricular dysfunction [51]. Previous studies suggest that the decreased expression of structural proteins such as junctophilin-2 (JPH2) and bridging integrator-1 (BIN1) is implicated in dyad formation [22, 52]. The N-terminal of JPH-2 is attached to the cell membrane through the MORN domain, and the C-terminal transmembrane structure anchors the SR, maintaining the stability of the SR and T tubular membrane structure. BIN1, as a membrane scaffold protein, plays an important role in the formation and structural maintenance of T-tubules. Decreased expression of both structural proteins will cause dyad abnormalities. It was hypothesized that increased stress on the local ventricular wall during PVCs was responsible for decreased expression of junctophilin-2 in cardiomyocytes and the remodeling of T-tubules [53] (Fig. 1).

PVC burden has been the most consistent parameter to demonstrate a relationship with the development of PVICM. Compared with PVC patients with normal ejection fraction, PVICM patients tend to have higher PVC burden (16–30% per day) [28, 54, 56, 61, 63, 64]. The lowest PVC burden resulting in cardiomyopathy was 10% [65]. Baman et al. [65] found that a PVC burden of $>$24% best separated the patient population with impaired as compared with preserved left ventricular function (sensitivity 79%, specificity 78%, area under the curve 0.89) and was independently associated with PVICM. Another study suggested that the burden threshold of $>$26% PVCs per day was independently associated with PVC-mediated LV dysfunction [66]. Echocardiographic results confirmed that cardiac function declines and structural remodeling of left ventricle is exacerbated with higher PVC burden. LVEF was negatively correlated with PVC burden, and cardiac diameters (left ventricular end-diastolic diameter (LVEDD) and left ventricular end-systolic diameter (LVESD)) were positively correlated [67]. Animal experiments have also indicated that canines developed PVICM when the PVC burden was over 25% [63]. The higher the PVC burden, the more severe the alterations of cardiac structure and function [63].

High PVC burden could serve as a predictor for the development of PVICM. However, it is still unclear why some patients do not develop cardiomyopathy despite a high PVC burden and why some patients appear to develop PVICM with a burden threshold of $\le$10% PVCs per day.

The PVC QRS duration in PVICM patients was significantly longer than that in patients without PVICM (164 $\pm$ 20 ms vs 149 $\pm$ 17 ms; p 0.0001) [58]. Yokokawa et al. [58] found that a PVC QRS width of 150ms had sensitivities of 80% and specificities of 52% for the diagnosis of PVICM, respectively. Patients with a longer PVC QRS duration ($>$150 ms) are more likely to experience a decline in cardiac function [55, 58]. Additionally, the sinus QRS duration was longer in the early stage of PVICM, and gradually prolonged with the development of the disease, although changes were minor [28]. Therefore, increased attention should be paid to the risk of heart failure in PVC patients, especially in those whose QRS duration of PVC and sinus beat is longer. Baseline QRS duration is inversely related to LVEF improvement after PVC ablation, especially in patients with a baseline QRS duration $>$130 ms [68].

Abnormalities of Calcium handling related proteins interrupt calcium homeostasis and are associated with prolonged QRS duration [33]. Therefore, it is thought that changes of QRS duration in PVCs suggest abnormal calcium handling in ventricular cardiomyocytes, ultimately resulting in PVICM.

The site of the PVC origin has an impact on the PVC QRS width and the degree of cardiac asynchrony. LV dysfunction was observed with PVCs from all common anatomic regions of origin. In contrast, PVCs originating from the epicardium appear to cause a more pronounced LV dyssynchrony and to induce LV systolic dysfunction, with longer QRS duration [58, 69]. PVCs originating from the RV were more likely to induce cardiac enlargement and compromise cardiac function [54]. However, results in other studies were contradictory, negating the association of PVC origin with the PVICM [66, 70]. It is commonly assumed that cardiac dyssynchrony is related to the PVC origin, but some studies have found that the degree of dyssynchrony is mainly dependent on the coupling interval rather than the origin [70].

In a previous study, a PVC coupling interval $>$500 ms was associated with abnormal LV remodeling [60]. According to a cohort study of 108 patients with frequent monogenic PVCs (of whom 22 were diagnosed with PVICM), the coupling interval in PVCIM was significantly higher than that in the control group. A PVC coupling interval of 486 ms had sensitivities of 0.789 and specificities of 0.738 for the diagnosis of PVICM [71]. Another study has shown that PVC coupling interval heterogeneity, rather than coupling interval duration, is an independent risk factor for PVICM [15]. Compared to PVC patients without cardiomyopathy, PVICM patients had significantly longer coupling interval dispersion (CI-dispersion: 115 $\pm$ 25 ms vs. 94 $\pm$ 19 ms; p 0.001) [56, 72]. Longer coupling interval induced cardiac and hemodynamic dysfunction, and caused sympathetic nervous system excitation to trigger the renin-angiotensin-aldosterone system (RAAS) system, thereby causing ventricular remodeling. In addition, increase in the coupling interval duration exacerbates ventricular systolic dyssynchrony, and increasing afferent neuronal activity was observed as the PVC coupling interval increased [70]. Additionally, the variability of PVC coupling interval might exacerbate hemodynamic abnormalities and the compensatory response of the neuroendocrine system.

The burden of interpolated PVCs was higher in the PVICM patients compared with other PVC patients. Interpolation of PVCs can independently predict PVICM-induced cardiomyopathy (odds ratio 4.43, 95% confidence interval 1.06–18.48, p = 0.04) [61].

PESP is defined as a physiological phenomenon of the increase in contractility following an extrasystole, and was first proposed by Oscar Langendoff in 1885. The cellular mechanism of PESP is that calcium release from intracellular stores is increased during the post-extrasystolic heartbeat. During the premature heartbeat, the transient decrease in calcium is caused by the refractoriness of RyRs. During the post-extrasystolic beat, RyRs have recovered from inactivation, and then increased intracellular calcium stores are released from these channels, resulting in increased contractility [73]. Previous studies found a significant increase in PESP in heart failure patients, and suggested that PESP could serve as a risk predictor of cardiac dysfunction and a prognostic indicator for patients with myocardial infraction [73, 74]. Increased PESP is associated with abnormal calcium cycling induced by heart failure. Similarly, patients with PVICM had a significantly higher PESP compared to controls [62]. In animal experiments, PESP was higher than at baseline after PVICM developed in canine models [75]. Furthermore, the level of PESP at baseline had a negative correlation with LVEF, suggesting that baseline PESP at the early stage of PVCs might be a predictor for PVICM [75].

In accordance with the multi-year follow-up results of multiple retrospective and prospective studies, males are at a greater risk of PVICM than females [76]. The reasons for the sex disparity in PVICM are still unclear. Female patients with frequent PVCs are more likely to experience symptoms such as palpitation and chest tightness than males, and they more often seek medical attention [77]. Furthermore, a longer history of palpitations and asymptomatic PVCs are independent risk factors for PVICM [59].