Subcutaneous implantable cardiac defibrillator - A personalised approach

Abstract

Sudden cardiac death (SCD) remains one of the leading causes of death in the modern world with most of these deaths being attributed to ventricular arrhythmias. Implantable cardiac defibrillators (ICDs) are well established treatment and are recommended by the international guidelines for prevention of sudden cardiac death triggered by ventricular arrhythmias in high-risk populations. But they are not risk-free, and traditional transvenous ICDs are associated with long-term complications with potentially fatal consequences. The subcutaneous ICD (S-ICD) was designed utilising a totally extra-thoracic approach avoiding the complications which have been associated with transvenous ICDs (TV-ICD). The results of clinical trials demonstrated the efficacy of the S-ICD systems in recognising and treating ventricular arrythmias with fewer lead-related complications when compared to the TV-ICDs. However, the downside of the S-ICDs is that unlike TV-ICDs, they are unable to provide bradycardia pacing or Anti-tachycardia pacing (ATP) therapy to terminate ventricular arrythmias painlessly without the need to deliver a shock. Also, they exhibit a relatively higher rate of inappropriate shocks when compared with TV-ICDs. Most of these shocks can be attributed to Twave oversensing (TWO), an inherent risk to the sensing mechanism of the S-ICD. Not all patients are eligible for S-ICDs and mandatory screening of all potential candidates following device manufacturer guidelines helps identify eligible patients based on their underlying ECG morphology. Variable rates of screening success and S-ICD eligibility are reported in the literature. In this thesis I will start by reporting a retrospective analysis of S-ICD eligibility using current recommended screening practices at a tertiary centre for cardiac devices (University Hospital of Southampton). I will then proceed to suggest adopting a different approach towards screening of S-ICD candidates which considers the dynamicity of the ECG signal. I will explore the role of applying prolonged screening using Holter monitors in a wide range of patients’ cohorts and prove there is variation in the S-ICD eligibility overtime which can explain oversensing and inappropriate shocks in S-ICDs despite current screening practices. I will also introduce and explain a novel technique utilising artificial intelligence and deep learning methods which has the potential to be applied to clinical practice to help identify S-ICD eligible patients as well as guide vector selection in S-ICD recipients. I will justify choosing R:T ratio as the main determinant of S-ICD eligibility and validate the novel deep learning methodology used in my studies by comparing the outcomes to those produced by the “gold standard” S-ICD simulator. I will then proceed to propose targeting less strict R:T ratios in S-ICD screening in vectors that prove to be stable with prolonged screening. I will then compare the eligibility rates for S-ICD using different R:T ratios, proposing that it is reasonable to revisit the S-ICD Screening thresholds if we adopt prolonged screening approaches. Afterwards, I will address the inability of S-ICDs to provide pacing therapy and the need to coimplant pacemaker devices to cover pacing therapy when clinically indicated. I will highlight that the effect of pacing on the S-ICD sensing has not been well studied before. I will proceed by first introducing a simple radiological method to define the LP position, then demonstrate that there is no effect of LP position on short- and long-term LPs performances. Afterwards I will show that pacing regardless of the pacing site has a significant effect on the R:T ratio, one of the main determinants of S-ICD eligibility and increases the risk of S-ICD oversensing. However, I will conclude that with adoption of personalised approach towards device therapy, it is theoretically feasible to utilise concomitant device therapies in most patients, without increasing the risk of adverse clinical events. In summary, I believe that extravascular cardiac devices are soon to establish themselves as the new standard of care to provide defibrillation protection and pacing therapy. Current issues with S-ICDs can be overcome by incorporating efficient artificial intelligence methods to help with more accurate and yet efficient screening for better patient selection. Together with adopting a personalised approach, higher S-ICD eligibility can be achieved, and the risk of inappropriate shocks can be mitigated.

More information

Identifiers

Catalogue record

Export record