Implantable cardioverter defibrillators (ICD) are indicated for heart failure (NYHA II or III) with a reduced left-ventricular ejection fraction (LV-EF) ≤ 35% despite optimal medical treatment for more than 3 months, as well as patients who have recovered from a hemodynamic unstable ventricular arrhythmias without a reversible cause (1, 2). Patients are also required to have a life expectancy of more than one year with a good quality survival. Patients without the need for bradycardia pacing, anti-tachycardia pacing or cardiac resynchronization therapy (CRT) may receive either a subcutaneous implantable cardioverter defibrillator (S-ICD) or a transvenous implantable cardioverter defibrillator (TV-ICD) (1).

With progressive chronic kidney disease (CKD), patients are at an increased risk of cardiovascular mortality and morbidity, which still exists after correction for typical concomitant risk factors (3). A main trigger is believed to be an inflammatory reaction in CKD (4) which has been identified as an independent risk factor for cardiac implantable electronic device (CIED) infections and complications (5). Patients with a glomerular filtration rate (GFR) < 30 ml/min/1.73 m² are a CKD subgroup exposed to a risk for infections and lead complications (6) whilst those with a GFR 15–29 ml/min/1.73 m² for at least three months are defined as stage 4 CKD and at even higher risk (7).

The S-ICD has been associated with lower rates for lead related complications during long-term follow-up (FU) while demonstrating comparable safety and efficacy compared to TV-ICDs (8). S-ICD device-related infections are less frequent as there are no leads on or in the heart. If they do occur, S-ICD infections can be handled conservatively without the need for urgent removal as is the case with TV-ICDs, where transvenous lead extraction (TLE) is associated with high mortality and morbidity (9, 10). Very limited long-term data on infection and complication rates in patients with progressive CKD receiving an ICD have been published. There is however no comparative data in this compromised cohort of CKD patients between TV-ICD and S-ICD recipients. The aim of the present study was to compare long-term device associated safety and outcomes in patients with stage 4 CKD who received either an TV-ICD or S-ICD with a single right ventricular lead.

Out of the population of patients undergoing TV-ICD or S-ICD implantation, those with stage 4 CKD, GFR 15–29 ml/min/1.73 m² for at least 3 months, were included in this retrospective analysis from six experienced centers in Germany (University Hospital Bergmannsheil Bochum, University Hospital Katholische Kliniken St. Josef Bochum, Marienhospital Gelsenkirchen, Augusta Hospital Bochum, Marien-Hospital Luenen and Elisabeth Hospital Recklinghausen). Patients on hemodialysis were excluded from this study as hemodialysis is an independent risk factor for device infection and complication (11). Patients with stage 5 CKD without dialysis were excluded as well, as there is a very high likelihood for hemodialysis during FU compared to stage 4 CKD (12).

The devices were implanted between 2012 and 2020. We obtained informed consent from all patients. The study protocol conforms to the ethical guidelines of the 1964 Declaration of Helsinki and its later amendments. It was approved by the local ethics committees of the Ruhr-University, Bochum as the leading ethics committee (Register 22-7593-BR). All patients had indications for a cardiac defibrillator implantation following the guidelines and a life-expectancy >1 year (1, 2).

The aim of the present study was to investigate whether there was a difference in TV-ICD and S-ICD associated infections, complications and hospitalizations in patients with stage 4 CKD who received a prophylactic ICD.

Patients with progressive CKD exhibit an increased risk for heart failure, ventricular arrhythmias and sudden cardiac death (SCD) (3). This group of patients has been excluded from most studies demonstrating a reduced mortality for ICD compared with drug therapy (1, 14). Patients with progressive CKD (GFR <30 ml/min/1.73 m²) were excluded from the S-ICD Investigational Device Exemption Trial (15). Unsurprisingly, there are no specific recommendations in the ICD therapy guidelines for patients with progressive CKD.

The role of ICDs in this seriously compromised group of patients remains unclear as large studies failed to show a survival benefit from implanting ICDs in patients with CKD (16). Higher CIED infection rates are assumed to be a trigger for higher hospitalization rates in patients with CKD who received an ICD compared with control groups with CKD who did not receive an ICD (17). We therefore postulated that in a similar group of compromised patients, there would be a lower CIED infection and device related hospitalization rate in S-ICD recipients as there are no lead on or in the heart.

Baseline demographics are comparable to a cohort study of 5,877 patients with CKD and heart failure reported by Bansal et al. (16). However, the number of patients with coronary artery disease, history of myocardial infarction and ischemic cardiomyopathy is higher in our cohort whilst the age in our S-ICD group was significantly younger. Implanting physicians are more likely to choose an S-ICD in younger patients, as they expect fewer complications in patients with CKD due to preservation of the vasculature. In addition, there was a higher likelihood in younger patients for a further progression of the CKD with the need for hemodialysis (HD). The feasibility of creating an arteriovenous fistula in patients requiring HD can be complicated by the presence of a CIED in the vasculature. Venous occlusion or central venous stenosis has been reported in 11%–36% patients with transvenous CIED (18).

The age difference between our TV-ICD and S-ICD groups could be an explanation for a higher rate of atrial arrhythmia in the TV-ICD group and hence a higher proportion of patients receiving anti-arrhythmic medication and cardiac glycosides. With a higher incidence of atrial arrhythmias and anti-arrhythmic medication there can be concerns regarding a higher drug accumulation due to lower renal clearance (19–21). This could be an important issue regarding morbidity and mortality data in future investigations. The significant age difference in our study between both groups could favor the beneficial results for the S-ICD, however data from a TV-ICD meta-analysis and a S-ICD study suggest that device associated complications are not associated with higher age (22, 23).

Implantation duration of our groups was comparable with other large studies comparing TV-ICD and S-ICD therapies (24). Patients with progressive CKD experience a significantly higher proportion of intraoperative complications, especially bleeding, vascular injury and in-hospital mortality according to an analysis of a nationwide database from the United States of 40,075 cases with ICD implantation (25). Minor procedural complications occurred significantly more often in our TV-ICD cohort. The procedural complication rate of 4.8% in our S-ICD cohort was comparable to a 3.8% procedural complication rate of 429 S-ICD procedures reported by Knops et al. (24). The rate of procedural complications associated with TV-ICD implantation was higher in our study (12.2%) compared with a rate of 4.7% in 423 procedures reported by Knops et al. (24). A high rate of oral anti-coagulation in our TV-ICD cohort and a high rate of complications in patients with CKD receiving a TV-ICD are explanations (16). However, no difference between our groups for procedural complications requiring surgical or other interventions has been reported.

Patients with progressive CKD are at an exponentially increased risk for death with decreasing kidney function and increasing cardiovascular mortality being the leading cause of death, rather than renal failure (3, 26). Progressive CKD is associated with a wide range of cardiac and non-cardiac features providing a substrate for vulnerable myocardium and increased risk of arrhythmias and consequently SCD (27, 28). The incidence of ventricular arrhythmias together with adequate and inadequate therapy in both ICD groups, and their lack of significance, is comparable with a large cohort reported by Russo et al. (29). Device lifetime infection rate is 2%–3% for TV-ICD systems in a general population according to the Danish ICD registry (30). CKD is an independent risk factor for device infections (31). In a general population, CIED infections occur significantly more often in TV-ICD patients compared with S-ICD patients (29). If infections with S-ICD systems occur, these can frequently be handled conservatively with antibiotics without the need for removal, whereas TV-ICD systems need to be explanted as soon as possible, often with complicated TLE procedures (15, 32). Our data is in line with these reports as there were no infections in the intermuscularly implanted S-ICD group and a significantly higher proportion of 16.3% CIED infections in the TV-ICD group. The high prevalence of device associated infections in the TV-ICD group can be explained by the progressive CKD. In S-ICD patients with progressive CKD, even those with HD, do not have an increased incidence of CIED infections (33, 34). If however a device removal of an S-ICD would be required, the procedure is safe and easy to perform (35).

Patients with a TV-ICD system and CKD are not only exposed to a high risk of CIED infection, but also other device associated complications. Jukema et al. reported an overall TV-ICD associated complication rate in 80 (31.3%) patients with end-stage CKD, of which 4 (5.0%) patients required ICD removal for bacteremia during a median FU of 6.8 ± SD years (36). A hospitalization rate for device-associated complications of 36.7% in our TV-ICD groups seems reasonable for a high risk population considering a pooled complication-rate, excluding inappropriate therapy, of 9.1% in a large TV-ICD meta-analysis by Ezzat et al. during an average FU of 17.9 months and an annual rate of 12.0% for complications requiring surgical intervention in prospective registry reported by Hawkins et al. of 3410 ICD recipients in a general population during a median FU of 34 months (22, 37). A meta-analysis of 2,387 patients by Fong et al. reported that compared with TV-ICD patients, S-ICD patients experienced significantly fewer (RR: 0.14, 95% CI: 0.07–0.29, p < 0.0001) lead related complications (8). These findings are supported by our data showing significantly fewer CIED-associated complications, especially lead related complications, in the S-ICD group during a long-term FU.

During a long-term follow-up of patients with stage 4 chronic kidney disease at high risk for ICD related complications, the intermuscularly implanted S-ICD has significantly fewer device-associated infections and complications compared with TV-ICDs.