The investigation of pathophysiological sex differences on a genetic or cellular level in ICM and NICM in humans is mainly limited by the number of available tissue samples. In healthy individuals, differences in gene expression are usually due to X- and Y-linked genes. Mechanisms that regulate sexual dimorphisms seem to be tissue-specific in the heart [20]. For example, genes encoding ion-channels and transporter subunits show reduced expressions of several potassium channel subunits, connexin43 and phospholamban in non-diseased female heart explants and an isoform switch in Na${}^{+}$/K${}^{+}$-ATPase compared to male hearts [21]. However, these differences might be attenuated in diseased hearts.

In a study of 70 men and 32 women with dilated cardiomyopathy and end-stage heart failure, $>$1800 genes were found to be differently expressed between sexes. Again, the majority of them was expressed on sex chromosomes. The biological processes related to these genes were mainly involved in ion transport activity, especially potassium and sodium channels, and G-protein coupled receptor signaling pathways [22]. These mechanisms lead to recognized sex differences: Potassium channel mutations are well known in long QT syndrome, a disease with known sex differences. Also, sex specific expressions of ventricular potassium currents cause differences in cardiac repolarization by lengthening of the action potential duration [23, 24]. Cardiac sodium channel abnormalities are associated with sudden cardiac death in women [25]. G-protein coupled receptor signaling pathways are involved in the regulation of cardiac inotropy and lusitropy [26]. These findings were corroborated by sex differences in a comprehensive analysis by Li et al. [27]: The authors investigated the transcriptome in patients with dilated cardiomyopathy and new-onset heart failure. They showed sex differences in mitochondrial oxidoreductase reactions and other mitochondrial substrates as well as upregulated nitric oxide synthase inhibitors and elevated trimethylamine N-oxide (TMAO) in males [27]. These differences in proteomics and metabolomics may explain different outcomes in structural heart diseases in men and women, as the pathways are involved in atherosclerosis, recovery after myocardial infarction and development of heart failure [28, 29, 30, 31]. Though not proven, it is conceivable that these sex differences in ICM and NICM may also translate into sex differences in VA mechanisms.

Sex hormones are thought to account for a large portion of sex differences in arrhythmia incidence and outcome. While the presence of estradiol receptors in cardiac tissue is known since over 40 years [32], their specific effects-especially in cardiomyopathies-are not yet fully understood.

Several experimental animal studies investigated the effects of estradiol on ventricular electrophysiology. Hara et al. [33] studied the effects of gonadal steroids on the cardiac action potential in rabbits. Estradiol-treated study animals had a significantly longer action potential duration and significantly more early afterdepolarizations, both of which are known to increase the likelihood of VA [33]. In dogs, cisapride in the presence of 17 beta-estradiol led to a greater increase of ventricular repolarization, a higher incidence of early afterdepolarizations and a subsequently higher rate of Torsades de Pointes [34]. In guinea pigs, James et al. [35] showed varying L-type Calcium currents at different time points of the estrus cycle. These findings might have translational implications to humans, as guinea pigs have a conventional estrus cycle similar to humans. These experimental findings support the role of estradiol in different VA risks between men and women and suggest a dynamic VA risk at different time points in the menstrual cycle.

Data on the influence of sex hormones on VA in humans consists mainly of observational cohort studies. In a cohort of 20 women with premature ventricular complexes (PVCs) and a control group of 18 healthy women, the number of PVCs in the menstruation period was 210 beats/day and decreased to 86 beats/day in the ovulation period (p 0.05). The average heart rate in the menstruation period was 81.4 $\pm$ 10 beats/min and increased significantly to 84.6 $\pm$ 8 beats/min in the ovulation period. There were no differences in cardiac repolarization parameters in the menstruation and ovulation periods between the study and control groups. However, when the authors compared the menstruation and the ovulation periods within subjects, J-Tpeak interval, which reflects early repolarization, was shorter in the ovulation period (193 $\pm$ 27.7 ms and 201.1 $\pm$ 28.6 ms, respectively; p 0.05) [36]. Further evidence of different arrhythmia risk along the estrus cycle stems from several case reports on paroxysmal VT during pregnancy [37, 38]. Brodsky et al. [39] reviewed reports on 26 pregnant women with VT, most of them without known structural heart disease. The VTs responded well to Beta-Blocker therapy and in 24 (92.3%) patients no more arrhythmia were recorded after delivery. In contrast, in a cohort of 44 pregnant women with implanted ICDs for ventricular fibrillation and long-QT syndrome , no increased VT rate was found during pregnancy or around labour [40]. Marchlinski et al. [41] investigated sex-specific triggers for RVOT-VT in 34 women and 13 men. In women, triggers depended on different hormonal states, including premenstrual, gestational, perimenopausal time points and the intake of birth control pills. In 20 (59%) of them hormonal states were the most common triggers and in 14 (41%) of them the only one. More support on the importance of female sex hormones on VA was actually found in men. The level of estradiol, but not testosterone, in male patients with outflow-tract VA was shown to be significantly lower compared to male controls (p 0.05) and a significant negative correlation was observed between the number of PVCs and the level of estradiol [42].

Studies to investigate scar patterns, electrograms and VT circuits usually underrepresent women by a ratio of 1:4 to men. Thus, sex specific analyses are often precluded [43, 44]. Moreover, only few studies investigated electrophysiological sex differences in patients with ICM or NICM. In an observational cohort study, Buxton et al. [45] reported a lower rate of inducible VT in female ICM patients compared to male. However, in their overall cohort of 1721 patients, only 14 (0.8%) were female and in the 549 inducible patients, only 9 (1.6%) were female [45]. Kuo et al. [46] analyzed 160 patients with NICM of whom 26 (16.3%) were female. The investigators compared endocardial uni- and bipolar and epicardial bipolar voltages, but did not find any significant differences. Similarly, they did not find sex differences in scar extent, distribution or transmurality based on cardiac MRI studies [46]. However, their analyses might have been limited by the small group of female patients. Yang et al. [47] analyzed 93 patients with idiopathic RVOT-VT of whom 63 (67.7%) were female: women had a lower RV voltage (3.0 $\pm$ 0.7 mV vs 3.7 $\pm$ 0.9 mV, p = 0.03) and about four times larger low voltage zones on the RVOT free wall compared to men (27.0% vs 6.7%, p = 0.02) [47]. In a different population of 70 patients with arrhythmogenic right ventricular dysplasia-34 (48.6%) of them being female - men had a larger extent of epicardial RV unipolar low-voltage zones and larger endo- and epicardial zones of late potentials [48]. In 83 patients with idiopatic VF, 44 (53%) of them being female, Purkinje related ectopics triggering VF were differently distributed between sexes. In women, these ectopics originated more commonly from the left or both ventricles, while in men, they originated mainly from the right ventricle [49].

Based on the paucity of data for sex differences in electrophysiological properties in structural heart disease, it is difficult to draw conclusions that can be generalized to larger patient populations. Regarding the definition of low-voltage zones, more research is needed to assess if current voltage cut-offs are similarly applicable to men and women, considering that women accounted only for 8–25% in the original studies defining the cut-offs [50, 51, 52].

Women were underrepresented in most of the major randomized controlled trials that investigated the outcome of catheter ablation for VA and subsequently influenced guideline recommendations (Fig. 1). The SMASH-VT (Substrate Mapping and Ablation in Sinus Rhythm to Halt Ventricular Tachycardia) trial enrolled 128 patients with ICM and investigated whether prophylactic radiofrequency catheter ablation of ventricular substrate would reduce the incidence of ICD therapy. Of those, only 17 (13.3%) were women [53]. The VTACH (Ventricular Tachycardia Ablation in Coronary Heart Disease) trial enrolled in total 107 patients with stable VT in the setting of ICM, of whom only 7 (6.5%) were female, and tested if catheter ablation for VT before implantation of an ICD conveys a potential benefit [54]. Of the 111 ICM patients enrolled in the SMS (Substrate Modification Study) trial investigating the effect of prophylactic catheter ablation of arrhythmogenic ventricular substrate on VA recurrence, 18 (16.2%) were female [55]. In the VANISH (Ventricular Tachycardia Ablation versus Escalation of Antiarrhythmic Drugs) trial, that compared VT catheter ablation versus escalated antiarrhythmic drugs in ICM patients, 18 (6.9%) out of 259 patients were women [56]. Similarly, the prematurely terminated BERLIN-VT (Preventive Ablation of Ventricular Tachycardia in Patients With Myocardial Infarction) study enrolled only 24 (14.7%) female patients out of 163 patients with ICM [57]. These small numbers of enrolled women preclude any meaningful subgroup analyses to investigate the efficacy and safety in female patients with ischemic cardiomyopathy. Moreover, they question the broad applicability of the trial findings to female patients. The recently presented, but not yet published, PAUSE-SCD (Pan-Asia United States PrEvention of Sudden Cardiac Death Catheter Ablation) trial (NCT02848781) investigated catheter ablation versus conventional medical therapy in patients with either ICM or NICM, monomorphic VT and an indication for ICD. The trial randomized 121 patients, of whom 23 (19.0%) were female, and met its primary composite endpoint of VT recurrence, cardiovascular rehospitalization and death. Importantly the investigators performed subgroup analyses and showed a significant interaction for sex, with less benefit in female compared to male patients.

Fig. 1.

Sex distribution in randomized-controlled trial investigating catheter ablation for ventricular tachycardia.

Based on the female under-representation in randomized VT ablation trials, their results may not be fully applicable to women [58, 59]. As shown in the subgroup analysis of the PAUSE-SCD trial, women may confer no or less benefit from ablation compared to men, though these results should not be over-interpreted as they are based on small sample sizes.

VT ablation approaches based on activation and entrainment mapping should be equally effective in female and male patients as each individual VT is delineated. This may, however, not hold true for purely substrate based ablation approaches during sinus rhythm due to sex differences in scar and isthmus distribution. We therefore reviewed sex distribution in the pivotal studies that tested these ablation strategies. Unfortunately, similarly to the randomized controlled trials on VT ablation and cardiovascular outcomes, women were markedly underrepresented in studies on substrate based ablation strategies. A linear ablation approach within the scar, guided by pacemapping, was investigated in a ICM and NICM population with 1 out of 4 patients being female [60]. Similarly, scar dechanneling in NICM was tested in a population including 18% women [61]. Endo-epicardial scar homogenization approaches in ICM included 19% women in the initial study [62]. Scar core isolation was tested in the initial study in ICM and NICM with only 5% being female [63]. The elimination of local abnormal ventricular activities in ICM and NICM was tested in a population with a female proportion of 10% [64]. Studies investigating the ablation of late potentials included 8, 6 and 13% of female patients [65, 66, 67]. Ablation based on isochronal late activation maps ICM and NICM was tested in a population with 15% female patients [68].

Similarly to the randomized trials mentioned above, these small proportions of female participants do not allow sex specific analyses of these ablation approaches and can not ascertain similar outcomes inbetween sexes.

As women were generally underrepresented in the above mentioned trials and studies, registry data may be more informative on outcomes and complications of VT ablation in females compared to males.

In a report from the International Ventricular Tachycardia Ablation Center Collaborative Group Study, women had significantly higher rates of VT recurrence in the year following ablation than men (30.5% vs 25.3%) despite being younger, having a higher left ventricular ejection fraction and fewer medical comorbidities (Fig. 2, Ref. [69]). Again only 262 (12.7%) out of 2062 patients with structural cardiomyopathy were women [69]. Hosseini et al. [70] analyzed $>$500.000 performed ablations, including also atrial ablations, and found female sex as an independent predictor for complications. In contrast, Baldinger et al. [16] did not find differences in success and complication rates between female and male patients in patients with and without structural heart disease (Fig. 3, Ref. [16]). In a cohort with both ICM and NICM reported by Santangeli et al. [71], sex was not a predictor for periprocedural acute hemodynamic decompensation after VT ablation. However, also in this study only 12% out of 193 patients were female.

Fig. 2.

Ventricular tachycardia–free and transplant-free survivals by sex. (A) Women had worse 1-year ventricular tachycardia–free survival than men following ablation (log-rank p = 0.03). (B) Women with ischemic cardiomyopathy had worse 1-year ventricular tachycardia–free survival than men with ischemic cardiomyopathy following ablation (log-rank p = 0.03). (C) Women and men with nonischemic cardiomyopathy had similar 1-year ventricular tachycardia–free survival following ablation (log-rank p = 0.55). (D) Women and men had similar long-term survival free of death and heart transplant (log-rank p = 0.24). Adapted from Frankel et al. [69] with permission.

Fig. 3.

Comparison of women and men undergoing catheter ablation for sustained monomorphic ventricular tachycardia in different subgroups. Estimated recurrence free survival at 1 year was 87 $\pm$ 5% for men and 76 $\pm$ 6% for women without SHD, 70 $\pm$ 3% for men and 52 $\pm$ 10% for women with CAD, 66 $\pm$ 4% for men and 63 $\pm$ 8% for women with NICM, and 89 $\pm$ 5% for men and 71 $\pm$ 17% for women with ARVC. ARVC, arrhythmogenic right ventricular cardiomyopathy; CAD, coronary artery disease; NICM, nonischemic cardiomyopathy; SHD, structural heart disease. Adapted from Baldinger et al. [16] with permission.

Taken together, some but not all observational studies point to an inferior outcome and higher complication rate in female patients compared to males. These differences might be related to the investigated study populations.