Machine learning-derived major adverse event prediction of patients undergoing transvenous lead extraction: Using the ESC EHRA EORP European lead extraction ConTRolled ELECTRa registry

Enhancing transvenous lead extraction risk prediction: Integrating imaging biomarkers into machine learning models

BACKGROUND

Machine learning (ML) models have been proposed to predict risk related to transvenous lead extraction (TLE).

OBJECTIVE

The purpose of this study was to test whether integrating imaging data into an existing ML model increases its ability to predict major adverse events (MAEs; procedure-related major complications and procedure-related deaths) and lengthy procedures (≥100 minutes).

METHODS

We hypothesized certain features-(1) lead angulation, (2) coil percentage inside the superior vena cava (SVC), and (3) number of overlapping leads in the SVC-detected from a pre-TLE plain anteroposterior chest radiograph (CXR) would improve prediction of MAE and long procedural times. A deep-learning convolutional neural network was developed to automatically detect these CXR features.

RESULTS

A total of 1050 cases were included, with 24 MAEs (2.3%) . The neural network was able to detect (1) heart border with 100% accuracy; (2) coils with 98% accuracy; and (3) acute angle in the right ventricle and SVC with 91% and 70% accuracy, respectively. The following features significantly improved MAE prediction: (1) ≥50% coil within the SVC; (2) ≥2 overlapping leads in the SVC; and (3) acute lead angulation. Balanced accuracy (0.74-0.87), sensitivity (68%-83%), specificity (72%-91%), and area under the curve (AUC) (0.767-0.962) all improved with imaging biomarkers. Prediction of lengthy procedures also improved: balanced accuracy (0.76-0.86), sensitivity (75%-85%), specificity (63%-87%), and AUC (0.684-0.913).

CONCLUSION

Risk prediction tools integrating imaging biomarkers significantly increases the ability of ML models to predict risk of MAE and long procedural time related to TLE.

Lead Extraction-Indications, Procedure, and Future Directions

Cardiac implantable electronic device (CIED) implantation has steadily increased in the United States owing to increased life expectancy, better access to health care, and the adoption of updated guidelines. Transvenous lead extraction (TLE) is an invasive technique for the removal of CIED devices, and the most common indications include device infections, lead failures, and venous occlusion. Although in-hospital and procedure-related deaths for patients undergoing TLE are low, the long-term mortality remains high with 10-year survival reported close to 50% after TLE. This is likely demonstrative of the increased burden of comorbidities with aging. There are guidelines provided by various professional societies, including the Heart Rhythm Society, regarding indications for lead extraction and management of these patients. In this paper, we will review the indications for CIED extraction, procedural considerations, and management of these patients based upon the latest guidelines.

Transvenous Lead Extraction: Outcomes From a Single Centre Providing a National Service for New Zealand

BACKGROUND

With increasing demand for cardiac implantable electronic devices there is a parallel increase in the need for transvenous lead extraction (TLE). Due to its small population, all TLE procedures in New Zealand are currently performed in a single centre, Auckland City Hospital. We analysed the clinical characteristics and outcomes of those undergoing TLE since this service was established.

METHODS

We performed a retrospective, single-centre cohort study of all TLE procedures between October 2015 and December 2021. Definitions from the European Lead Extraction Controlled study, Heart Rhythm Society, European Heart Rhythm Association consensus documents were used.

RESULTS

A total of 247 patients had 480 leads extracted, averaging 40 TLE procedures annually. Patients had a median lead dwell time of 6 (interquartile range [IQR] 3-11) years, 60 (13%) of leads had been in-situ >15 years, median age 61 (IQR 48-70) years, 73 (30%) female, 28 (11%) Māori, 23 (9%) Pasifika. Lead dysfunction (115 patients, 47%) and infection (90 patients, 37%) were the most common indications for TLE. Complete clinical and radiological success was achieved for 96% and 95%, respectively. Procedure-related complications occurred in 16 (7%) patients. Major intra-procedure complications occurred in 5 patients (2%), including 2 (1%) deaths. Death within one year of TLE occurred in 13 (26%) with systemic infection, 5 (3%) with local infection, and 5 (3%) with non-infection indications for TLE, p <0.01.

CONCLUSIONS

TLE is associated with high radiographic and clinical success, low complication, and low mortality rate. At our single centre providing a national service, TLE outcomes are comparable with those achieved internationally.

Predictors for adverse events during cardiac lead extraction - Experience from a large single centre

BACKGROUND

As the use of cardiac implantable electronic devices (CIED) has increased in recent years, the need for transvenous lead extraction (TLE) has also steadily increased. However, the TLE procedure could lead to serious complications and even death. Clinical decision-making tools are necessary for predicting these adverse events, but the appropriate tools have not yet been developed.

OBJECTIVE

To explore the possible predictors and develop a clinical model to predict TLE related adverse events.

METHODS

All the patients who were admitted to our cardiac center for TLE from January 2014 to January 2021 were included in this study. The patient information, device baseline characteristics, procedure-related information, complications and outcomes were recorded. Independent predictors of TLE related adverse events were identified by univariate, LASSO and multivariate analysis. A nomogram for predicting these adverse events was developed based on these independent predictors. Calibration and decision curve analysis were conducted to evaluate the nomogram.

RESULTS

One thousand and one hundred patients were included in this study, 778 (70.7%) were male and the median age was 68 years old. A total of 2,208 leads were extracted and 2.01±0.74 leads were extracted per procedure. Fifty-five patients (5%) developed adverse events including minor complications (2.4%), major complications (2.3%) and death (0.27%). Seven independent predictors for TLE related adverse events were identified and selected to establish the nomogram including BMI, female gender, hypoalbuminemia, number of extracted leads>3, longest dwell time of the extracted leads and manual traction. The area under the receiver operating characteristic (ROC) curve (AUC) for the prediction model was 0.774. Calibration curve and decision curve analysis showed that the nomogram had good prediction performance.

CONCLUSION

TLE related adverse events are some of the key issues that concern clinicians. We have identified seven independent factors and established a predictive model that may help clinicians identify at-risk patients and create better plans for lead extraction.