Analysis of complications of transvenous extraction of electrodes of cardiac implantable devices

Introduction. The widespread use of cardiac implantable electronic devices (CIEDs) has led to a natural increase in the number of patients requiring transvenous lead extraction (TLE) due to infectious complications and system dysfunction. Despite refined techniques, TLE remains a procedure with a risk of life-threatening complications, such as cardiac perforation or major vein rupture, with an associated mortality rate of up to 50%. Several predictive risk scores (MB Score, LED Score, SAFeTY Score, EROS) have been proposed for risk stratification; however, there is no unified consensus regarding their applicability, diagnostic value, or the most significant predictors. Objective. To conduct a comparative analysis of the prognostic value of existing risk scores (MB Score, LED Score, SAFeTY Score, EROS) and to identify independent predictors of major intraoperative complications during transvenous lead extraction (TLE). Materials and Methods. A multicenter retrospective study included 300 patients who underwent TLE between 2019 and 2023. All patients underwent preoperative risk stratification using four prognostic risk scores. Statistical analysis included correlation analysis (Spearman's coefficient) and ROC analysis. To assess the independent effect of predictors on the development of major complications, binary logistic regression with forward stepwise selection was used. Results. Major complications were recorded in 11 (3.7%) patients. A statistically significant positive correlation was found between the development of complications and scores on the MB Score (r=0.156; p=0.007), LED Score (r=0.155; p=0.007), and SAFeTY Score (r=0.13; p=0.024), and EROS Score (r=0.134; p=0.021). ROC analysis demonstrated that the MB Score (AUC 0.737; p=0.008) and LED Score (AUC 0.739; p=0.007) had the best diagnostic performance; SAFeTY Score showed moderate predictive quality (AUC 0.699; p=0.025), while EROS Score showed borderline statistical significance (AUC 0.673; p=0.051). ROC analysis identified additional factors associated with complications: lead age > 126 months, infection duration > 4.5 months, eGFR < 73 mL/min, female sex, diabetes mellitus, passive lead fixation, and the presence of cut or fractured leads. In stepwise logistic regression of the four risk scores, only LED Score was included in the final model (OR=1.087; 95% CI: 1.019–1.159; p=0.011). When the additional factors were tested in the multivariate model, only infection duration > 4.5 months retained independent significance (OR=4.114; 95% CI: 1.120–15.109; p=0.033). Female sex showed a trend toward significance (p=0.056). The remaining factors lost their prognostic significance after adjustment. Conclusion: TLE is a safe and effective method with a major complication rate of 3.7%. LED Score is the most informative prognostic scale, and infection duration > 4.5 months is an additional independent predictor of complications, supporting the need for timely referral of patients for surgical treatment. Consideration of these factors improves perioperative risk stratification and optimizes the choice of surgical strategy.