Learning from implantable cardioverter-defibrillator leads in children.

I f you want something broken, give it to your kid. That truism applies to many items, from toys to technology. The article describing the results of the Pediatric Lead Extractability and Survival Evaluation (PLEASE) study by Atallah et al 1 in this issue of Circulation provides sobering but important evidence that this also applies to implantable cardioverter-defibrillator (ICD) leads. Most ICD systems are implanted in adults, but many are implanted in children. These pediatric patients with ICD systems now participate in larger ICD registries, 2 but an earlier specific pediatric registry is the target of this study. This study encompasses 874 patients and 965 ICD leads spanning 24 centers. The initial approximately one fifth of the patients were enrolled in a randomized study comparing 2 specific lead designs: the smaller Medtronic Fidelis lead and a larger but ePTFE (expanded polytetrafluoroethylene)–coated Boston Scientific alternative. This portion of the study was terminated when concerns about Fidelis lead performance surfaced, and these and all subsequent patients were entered into a registry with more patients and nearly inclusive entry for the latter portion of the enrollment. Although not prespecified, I think this flexibility was laudatory, allowing the salvage of useful and generalizable information, and does not appear to be an obvious source of bias or scientific compromise. The study design, however, is essentially that of a registry rather than a randomized study. ICDs work in children, 3 and transvenous ICD systems appear superior to epicardial systems. 4 Smaller patient size allows novel nonvascular lead configurations for leads usually placed transvenously, 5 but these were presumably not targeted in this study. The authors accurately summarize: " There were 139 ICD lead failures (14%) in 132 patients (15%) at mean lead age of 2.0±1.4 years, causing shocks in 53 patients (40%). " The failure rate in thin leads was significantly higher, but this was dominated by the failure of Fidelis leads. However, even non-Fidelis " good " leads showed disappointing performance: The actuarial yearly failure rate was 2.3%. The average time to failure was 2 years, although finite study follow-up will necessarily bias this estimate. Thus, although flawed leads failed, even our best available leads show a disconcertingly high failure rate in younger patients. Possibly a small lead other than a Fidelis lead might prove superior or inferior; for purposes of this study, thin is inescapably confounded with Fidelis. Smaller leads include the Riata and …