Patients with ICD indications (secondary and primary prevention of SCD) were enrolled in this nonrandomized, single-center, clinical, open, cross-sectional observational study. Individuals with myocardial infarction (MI) less than 3 months old, hypertrophic cardiomyopathy, severe concomitant pathology (end-stage renal failure and hepatic insufficiency, cancer of any location), cognitive dysfunction and indications for revascularization, and young patients (18 year) were not included in the study.

All patients underwent the full physical (6-minute walk distance test (6MWDT), transthoracic echocardiography (TTE), 12-lead electrocardiography (ECG), 24-hour Holter monitoring ECG, coronary angiography and blood analyses) and additional (mitochondrial nucleotide transition assessment using PBMC) examination. All patients received basic therapy in accordance with the guidelines for the management of ventricular tachycardia (VT) and heart failure (HF). All included patients were from a registered study (ClinicalTrials.gov, NCT03667989).

The echocardiography was performed using the Philips HD15 PureWave ultrasound machine (Philips Ultrasound, Inc., Bothell, WA, USA). The TTE was carried out from standard positions with the intracardiac hemodynamic parameters assessment and determination of the dimensional, volume and indexed indicators of the heart chambers and left ventricular ejection fraction (LVEF). The right and left ventricles contractility and cardiac valve function, with the exception of the pulmonary valve, were evaluated.

Blood sampling was performed before ICD implantation. DNA extraction from the whole ethylenediaminetetraacetic acid (EDTA) blood was carried out by phenol chloroform extraction. Isolated DNA samples were placed for storage at –20 °C until further stages of the study. The DNA sample quality and concentration were evaluated using a NanoDrop – 2000C spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA).

Three mtDNA polymorphisms were selected for the analysis, namely A2706G, G3010A and G9055A. The positions are numbered according corrected reference human mtDNA sequence [13]. Genotyping of the polymorphisms was carried out by restriction analysis. For this aim, we performed polymerase chain reaction (PCR) analysis using specific primers for each polymorphism [14]. Primer annealing conditions for PCR were identified individually for each primer pair. Resulting amplicons then underwent restriction fragment length polymorphism analysis using site-specific endonucleases (SibEnzyme Ltd., Novosibirsk, Russia). The mixture for the restriction procedure included PCR-product (10 µL), the restriction buffer supplied with the enzyme (1.2 µL), 1 U of the enzyme, and deionized water to a final volume of 12 µL. The mixture was incubated for 12–24 hours at the temperatures optimal for each enzyme, according to the manufacturer’s information. Separation of restriction products by size was carried out immediately after incubation with 2% agarose gel electrophoresis for 40 min at a voltage of 130 V with staining with ethidium bromide. Visualization and confirmation of the obtained results were carried out in ultraviolet light on the Gel Doc documentation system (BioRad, Hercules, CA, USA).

Information on clinical assessment was obtained for all of the 122 patients included in the study. All participants were divided into 2 cohorts in accordance with the _SVT presence. The _SVT criteria were 3 or more consecutive ventricular beats with rate of 120 or more beats per minute and duration for 30 or more sec. The 1st group included patients with _SVT, the 2nd group – without _SVT. The flow chart and study design are presented in Fig. 1.

Fig. 1.

Study flow chart and design. Abbreviations: HCM, hypertrophic cardiomyopathy; ICD, implantable cardioverter-defibrillator; mtDNA, mitochondrial DNA; PBMC, peripheral blood mononuclear cells; _SVT, sustained ventricular tachycardia.

All qualitative and categorical variables are presented as counts (n) and percentages (%). The continuous variables are presented as the mean (M) $\pm$ standard deviation (SD) for normally distributed variables or as the median (Me) and interquartile ranges [Q1; Q3] for non-normally distributed variables. Kolmogorov-Smirnov, Lilliefors and Shapiro-Wilk tests were used to test the normality of continuous data distribution. Differences between groups for continuous data with normal distribution were assessed using the two-sided Student’s t-test. The Mann–Whitney U-test was used for independent ordinal or non-normally distributed data. The significance of differences between the groups was determined using the nonparametric Chi² or Fisher’s exact test. A comparison between patients with and without sustained VT was performed for the complete dataset. Statistical analysis was performed with Statistica 10.0 (StatSoft Inc., Tulsa, OK, USA) and Medcalc 19.2.6 (MedCalc Software, Ostend, Belgium). Statistical significance was defined by a p-value 0.05.

The binary logistic regression analysis was used to distinguish variables associated with the _SVT. The multicollinearity was excluded using the non-parametric Spearman analysis. Based on pairwise correlation coefficients, the presence of collinear factors was revealed. Factors were considered collinear if R $>$0.7. Only one of the interrelated variables was added to the model. The multivariable logistic regression analysis was performed using a stepwise elimination technique. Only indicators with p 0.05 were added in the final model. Goodness-of-fit was performed by the Hosmer-Lemeshow test. The Pearson correlation and t-test were used to examine the significant correlations between the predictors. Regression analysis results were presented as odds ratios (ORs) with 95% confidence intervals (CIs). Odds ratios and the beta coefficients used in the logistic equation could be converted into each other: OR = e$\beta$.

Finally, three variables (age, estimated glomerular filtration rate (_eGFR) and absence of mtDNA A2706G transition), independently associated with _SVT, were added to the predictive model. The area under the curve (AUC) was calculated to assess the discriminatory ability of the risk stratification score.

In 141 patients with an ICD implantation indication, we excluded 19 patients with failed blood samples (n = 17) and hypertrophic cardiomyopathy (n = 2). For the remaining 122 (100.0%) participants, the mean age was 63.7 $\pm$ 10.8 years and 94 (77.0%) patients were males. The 1st group included 70 (56.6%) patients with _SVT. The 2nd group included 52 (43.4%) patients without _SVT. The baseline characteristics of participants are presented in Table 1. Patients with _SVT were significantly older than patients without _SVT (66.9 $\pm$ 9.9 year vs. 59.5 $\pm$ 10.6 year, p 0.001). The subgroups of patients also differed significantly by the portion of patients with I (7.1% vs. 25.0%, p = 0.006) and II (65.7% vs. 38.4%, p = 0.002) New York Heart Association (NYHA) class, but the 6MWDT indicator didn’t differ between groups (p = 0.635). Individuals with _SVT were more likely to have ventricular fibrillation (8.6% vs. 0.0%, p = 0.030), lower _eGFR (65.7 $\pm$ 19.7 mL/min/1.73 m² vs. 77.9 $\pm$ 16.1 mL/min/1.73 m², p 0.001) and left ventricular end-systolic volume (74.5 [42.0; 106.0] mL vs. 147.5 [114.0; 205.0] mL, p 0.001), and higher LVEF (47.5 [41.0; 62.0]% vs. 32.5 [28.0; 36.0]%, p 0.001). Significant differences in echocardiographic indicators were due to the fact that all patients without _SVT had primary SCD prevention indications for ICD implantation. Also, patients without _SVT were more likely to have prescribed loop diuretics (59.6% vs. 27.1%, p 0.001), mineralocorticoid receptor antagonists (84.6% vs. 50.0%, p 0.001), angiotensin receptor/neprilysin inhibitors (34.6% vs. 8.6%, p 0.001) and sodium glucose co-transporter 2 inhibitors (44.2% vs. 10.0%, p 0.001) due to the fact that patients have HF with reduced LVEF and received the therapy in accordance with the current guidelines for the HF management. Patients with _SVT were more likely to have longer baseline corrected QT interval (QTc) (429.3 $\pm$ 33.8 ms vs. 419.8 $\pm$ 31.3 ms, p = 0.031). This significant difference was due to the fact that the majority of patients with _SVT were treated with amiodarone (62.8% vs. 38.4%, p = 0.007). Also, patients with _SVT were more likely to have lipid-lowering treatment (92.8% vs. 80.7%, p = 0.044).

Comparison of the mtDNA polymorphisms abundance between the groups revealed that patients with _SVT less frequently had mtDNA A2706G substitution (55.7% vs. 76.9%, p = 0.015). The other two polymorphisms G3010A and G9055A did not show statistically significant differences. A detailed comparison of mitochondrial nucleotide transition assessment parameters between groups is shown in Table 2.

Based on the assessed indicators, age was the parameter most strongly correlated with _SVT. Its discriminative ability was evaluated by receiver operating characteristic (ROC) analysis showing an AUC of 0.698 (95% CI 0.608–0.778) (Fig. 2B). The ability of _eGFR to distinguish for the _SVT was analogous with an AUC of 0.688 (95% CI: 0.598–0.769) (Fig. 2C). Concerning the mitochondrial A2706G polymorphism, the ROC curve calculation showed an AUC of 0.606 (95% CI 0.514–0.693) (Fig. 2A).

Fig. 2.

Receiver operating characteristic (ROC) curves. ROC curves to assess the ability of (A) mtDNA A2706G polymorphism presence, (B) age, (C) estimated glomerular filtration rate (_eGFR) and (D) the whole risk stratification model to discriminate patients with and without sustained ventricular tachycardia (_SVT). mtDNA, mitochondrial DNA; AUC, area under the curve.

According to the univariable logistic regression analysis age (OR = 1.073; 95% CI 1.032–1.116; p 0.001), _eGFR (OR = 0.964; 95% CI 0.943–0.985; p 0.001), absence of mtDNA A2706G polymorphism (OR = 0.377; 95% CI 0.169–0.839; p = 0.014) and lipid-lowering treatment (OR = 3.095; 95% CI 0.988–9.692; p = 0.045) were independently associated with the _SVT (Fig. 3).

Fig. 3.

Forest plot for the univariable logistic regression results. Abbreviations: 95% CI, 95% confidence interval; _eGFR, estimated glomerular filtration rate; G, guanine; QTc, corrected QT interval; mtDNA, mitochondrial DNA; OR, odds ratio; _SVT, sustained ventricular tachycardia.

Only three variables (age, _eGFR and absence of mtDNA A2706G substitution) remained considerable in multivariable regression analysis, even after amendment for male gender, QTc and presence of MI in anamnesis (OR = 1.055, 95% CI 1.009–1.103, p = 0.017; OR = 0.974, 95% CI 0.949–0.999, p = 0.041; OR = 0.335, 95% CI 0.141–0.797, p = 0.013; respectively). The beta coefficients in the logistic equation match the OR natural logarithm (ln(OR) = $\beta$; OR = e^β).

A risk score for _SVT was established using logistic regression according to collected data. Three variables (age, _eGFR and absence of mtDNA A2706G polymorphism) were added in the final model syne these indicators remained significant in multivariable logistic regression, even after amendment for male gender, QTc and presence of MI in anamnesis. The AUC was calculated to assess the discriminatory potential of the risk stratification score. The risk model showed the ability to discriminate evidenced by an AUC of 0.761 (Fig. 2D). At a cut-off point of $>$0.6; the model demonstrates a sensitivity of 65.71% and a specificity of 76.92% in discriminating against patients with ICD implantation indications regarding _SVT. Our risk stratification model showed a positive prognostic value of 80.00% and a negative prognostic value of 53.85%.

The age and _eGFR values were inserted in the score equation (in year and mL/min/1.73 m², respectively). The substitution of adenine to guanine in mtDNA A2706G was indicated in the score equation by assigning a value of 1 (apparent) or 0 (in-apparent). The result of the logistic equation below is the probability (p) of _SVT development. If the score is $>$0.6, this risk score allows identification of individuals with ICD implantation indications at an increased risk for _SVT and following intensified therapy.

Eqn. 1: Probability (p) of _SVT development.

$$\begin{array}{c}\hfill p=\frac{1}{1+e^{-z}}\hfill \\ \hfill z=-0.52-1.09\times \text{mtDNA A2706G presence}+0.05\times \text{age}[\text{year}]-0.02\times {}_{\text{e}}\text{GFR}[\mathrm{mL}/\min /1.73 {\mathrm{m}}^2]\hfill \end{array}$$