Reassessment of clinical variables in cardiac resynchronization defibrillator patients at the time of first replacement: Death after replacement of CRT (DARC) score

Abstract Introduction Cardiac resynchronization defibrillator (CRT‐D) as primary prevention is known to reduce mortality. At the time of replacement, higher age and comorbidities may attenuate the benefit of implantable cardioverter‐defibrillator (ICD) therapy. The purpose of this study was to evaluate the progression of comorbidities after implantation and their association with mortality following CRT‐D generator replacement. In addition, a risk score was developed to identify patients at high risk for mortality after replacement. Methods and Results We identified patients implanted with a primary prevention CRT‐D (n = 648) who subsequently underwent elective generator replacement (n = 218) from two prospective ICD registries. The cohort consisted of 218 patients (median age: 70 years, male gender: 73%, mean left ventricular ejection fraction [LVEF]: 36 ± 11% at replacement). Median follow‐up after the replacement was 4.2 years during which 64 patients (29%) died and 11 patients (5%) received appropriate ICD shocks. An increase in comorbidities was observed in 77 patients (35%). The 5‐year mortality rate was 41% in patients with ≥2 comorbidities at the time of replacement. A risk score incorporating age, gender, LVEF, atrial fibrillation, anemia, chronic kidney disease, and history of appropriate ICD shocks at time of replacement accurately predicted 5‐year mortality (C‐statistic 0.829). Patients with a risk score of greater than 2.5 had excess mortality at 5‐year postreplacement compared with patients with a risk score less than 1.5 (57% vs. 6%; p < .001). Conclusion A simple risk score accurately predicts 5‐year mortality after replacement in CRT‐D patients, as patients with a risk score of greater than 2.5 are at high risk of dying despite ICD protection.


| INTRODUCTION
Cardiac resynchronization therapy (CRT) is part of the standard management in selected patients with chronic heart failure (HF), reduced left ventricular ejection fraction (LVEF), and electrical dyssynchrony. 1,2 The combination with a defibrillator, that is, cardiac resynchronization defibrillator (CRT-D), is supported by the lower sudden cardiac death rate due to defibrillator therapy in patients with HF and left ventricular dysfunction.
Several years after the initial implantation, elective device replacement will be necessary because of battery depletion.
However, at this point in time, some patients may face a limited prognosis due to advanced age and multiple comorbidities negating the benefit of defibrillator therapy. On the other hand, at least 20% of patients experienced appropriate shocks from their device before replacement. 3 Appropriate implantable cardioverter-defibrillator (ICD) shocks are associated with a subsequent 3-5-fold increased risk of death among patients with primary prevention ICDs. 3 Whether this association can be extended to survival postreplacement is unknown. Current guidelines for initial implantation state that patients should be expected to survive at least 1 year, but the issue of replacement is rarely covered. Recently, there is more debate to change this attitude also toward ICD replacement in every patient. 4,5 However, the paucity of data describing the characteristics and outcomes of patients receiving ICD replacements is a barrier to risk stratification and prediction and explains partly the lack of clear indications for replacement in practice guidelines. A few studies with heterogeneous study populations evaluated mortality and risk factors following device replacement. [6][7][8][9] The REPLACE registry included patients with pacemakers, ICDs, and CRT-Ds. 10 The study by Wuest et al. 8 included primary and secondary prevention patients implanted with ICDs and CRT-Ds. However, data in the setting of primary prevention patients who underwent CRT-D replacement are not available. Therefore, the objectives of the current study were to evaluate the progression of comorbidities in a cohort of primary prevention patients with a CRT-D between initial implant and replacement, their association with mortality, and to develop a mortality risk tool, designated as the death after replacement of CRT-D (DARC) risk score.

| Study population
Patients for this retrospective observational cohort study were obtained from two prospective ICD registries of the cardiology departments of Erasmus MC and the University Hospital of Basel.
In these registries, we identified all patients in whom a CRT-D was implanted for chronic HF and primary prevention of sudden

| Clinical variables, comorbidities, and drug treatment
Baseline data on clinical variables, comorbidities, laboratory values, and drug treatment are prospectively collected in both ICD registries.
Those parameters were reassessed at the time of the first generator replacement. Both ICD registries and medical records were reviewed to obtain data on these parameters at baseline and replacement.
For the current study, we investigated the progression of non-ICD indication-related comorbidities between implantation and first elective replacement, and their association with mortality postreplacement. Non-ICD indication-related comorbidities were defined as atrial fibrillation (AF), diabetes mellitus, anemia, chronic obstructive pulmonary disease (COPD), peripheral vascular disease (PVD), cerebrovascular disease, cancer, and chronic kidney disease (CKD).
Diabetes mellitus was defined as HbA 1c > 6.5% or the use of oral hypoglycemic agents or the use of insulin; anemia as a serum hemoglobin concentration of less than 12 g/dl (female) or less than 13 g/dl (male). The renal function was assessed by estimating the glomerular filtration rate (eGFR) using the Chronic Kidney Disease Epidemiology Collaboration equation. 11

| Follow-up and ICD therapy event analysis
Follow-up started at the time of ICD implantation. Device interrogation was performed on scheduled regular visits and after symptomatic events. At each visit, arrhythmic events with stored electrograms (EGMs) were retrieved from the device's memory.
Appropriate ICD therapy was defined as antitachycardia pacing (ATP) or shock delivered for ventricular tachyarrhythmia; ventricular fibrillation (VF) or ventricular tachycardia (VT). The presence of atrioventricular dissociation (ventricular rate greater than atrial rate) was used to diagnose ventricular tachyarrhythmia when the baseline atrial rhythm is sinus rhythm.
In the case of AF baseline atrial rhythm, ventricular tachyarrhythmias were defined as events with a sudden increase in rate combined with a change in the ventricular near-field and far-field EGM morphology from the baseline rhythm without biventricular pacing.

| Endpoint
The clinical endpoint for this study was all-cause mortality after replacement; patients who underwent cardiac transplantation or who received a ventricular assist device were censored on the day of surgery. The secondary endpoint was the association of appropriate ICD shock within the VF zone (being "potentially lifethreatening") and mortality. In addition, the association of appropriate ATP within the VT zone and mortality were also evaluated.

| Statistical analysis
The normality of distribution was assessed by using the Shapiro-Wilk test. Continuous variables are presented as mean ± SD or as median with 25th and 75th percentiles, where appropriate.

Data were compared by the paired Student's t test or
Mann-Whitney U test, as appropriate. Categorical data are expressed as percentages and compared with the McNemar test. The mortality rate was calculated using the Kaplan-Meier method and differences between the groups were evaluated by the log-rank test.
Univariate logistic regression analyses were used to determine potential clinical predictors of mortality, with the calculation of odds ratio (OR) with 95% confidence intervals (CIs). Any variable with a

| RESULTS
During the study period, a total of 648 patients received a CRT-D for the primary prevention of sudden cardiac death. Of these, 218 underwent at least one replacement and comprised the study cohort.
The study population was predominantly male (73%) with a median The prevalence of comorbidities at implantation and replacement is presented in Table 2. Overall, an increase in non-ICD indication-related comorbidities was observed in 77 patients (35%). The proportion of patients with at least one non-ICD indication-related comorbidity increased between implantation and replacement; 63% of patients at implantation versus 76% at replacement (p < .001). Development of new CKD was observed in 19% of patients followed by diabetes mellitus (12%) and AF (11%). Following CRT implantation, renal function remained unchanged in 93 patients (43%) and worsening was observed in 98 patients (45%). In patients with CKD at replacement (n = 108), a worsening of CKD stage was observed in 77 patients (71%) whereas it remained unchanged in 25 patients (23%) when compared to their CKD stage at implantation (p < .001).
Increasing comorbidity burden was associated with an increased risk of mortality ( Figure 1). For patients without any non-ICD indication-related comorbidity, the 5-year mortality rate was 11%. It was 20% in patients with one comorbidity, and 41% in patients with at least two comorbidities at the time of replacement. AF, anemia, and CKD, adjusted for age and LVEF at replacement and gender, were independently associated with increased risk of mortality postreplacement (Table 3).
Appropriate ICD shock before replacement occurred in 27 patients (12%) yielding an incidence rate of 2.6 per 100 personyears no difference between the different implant groups (p = .48).
Appropriate ICD shocks before replacement was associated with an increased risk for mortality after replacement (OR: 9.6, 95% THEUNS ET AL. | 1689 CI: 4.0-23.2; p < .001). Considering appropriate ATP before replacement which occurred in 40 patients (18%), no association with mortality after replacement was found (p = .50).
Using age, gender, LVEF, AF, anemia, CKD, and history of appropriate ICD shocks at replacement, a risk score was developed to predict mortality after replacement of CRT-D (Table SI).  After generator replacement, 11 patients (5%) received appropriate ICD shocks, yielding an incidence rate of 1.5 per 100-person years. Appropriate ICD shock rate was not different between patients with LVEF ≤ 35% versus those with LVEF > 35% (p = .22). The cumulative appropriate shock rates stratified by three risk groups are presented in Figure 4. At 5-year postreplacement, appropriate shock rate ranged from 2% (low risk), 4% (medium risk) to 13% (high risk). In the high-risk group, 85% of the deceased patients experienced no appropriate ICD shocks after replacement.

| DISCUSSION
In this study, we present the evolution of comorbidities and association with mortality after CRT-D replacement in primary prevention patients from two international tertiary centers. Furthermore, we developed a dedicated mortality risk tool, the DARC risk score, for patients at the time of CRT-D replacement. In the present study, 29% of patients died at a median interval of 2.7 years after replacement. We demonstrate that an increasing non-ICD indication-related comorbidity burden has a cumulative effect on mortality. The presence of atrial AF and anemia, level of CKD, LVEF, and past appropriate ICD shocks before replacement was highly predictive of mortality after elective replacement. In the high-risk group, the majority of patients die without appropriate ICD shock therapy after replacement. A few studies with heterogeneous study populations reported mortality rates following device replacement. A substudy of the REPLACE Registry reported an overall 6 months all-cause mortality rate of 4%. 6 The study by Kramer et al. 7 reported a 1-year mortality rate of 9.8% following ICD and CRT-D replacement. The reported 5-year mortality rates ranged from 25% to 41% in previous studies. 7-9, 13 We found similar overall mortality rates following CRT-D replacement in a patient group with primary preventive ICD

| LIMITATIONS
Although the analysis was retrospective, data in both registries, including mortality and appropriate ICD shocks were all collected prospectively. The study cohort included patients over a 12-year period, during which guidelines for the implantation of defibrillators and treatment of HF changed. In the same period, the programming of devices with respect to the detection and treatment of ventricular arrhythmias changed. We accounted for this by defining three groups according to the date of the implant. Although the DARC risk score accurately identified patients at high risk for mortality after replacement, external validation could not be performed. Internal validation has been performed by bootstrap analysis. We encourage further studies to validate our findings of the DARC risk score in a larger cohort. In addition, the study included only patients who underwent CRT-D replacement.
Patients who did not undergo CRT-D replacement and those who were downgraded to a CRT-P were not included in the current analysis.

| CONCLUSION
In this real-life cohort of primary prevention CRT-D patients, we observed a significant increase in comorbidity burden between initial implantation and elective generator replacement. A high comorbidity burden was associated with increased mortality after replacement.
Age, gender, LVEF, comorbidity, such as CKD and anemia, and prior appropriate ICD shock therapy were identified as contributors to mortality after generator replacement. A simple risk score accurately predicts 5-year mortality after replacement, as patients with a risk score of greater than 2.5 are at high risk of dying despite ICD protection. In the future, clinical trials are necessary to evaluate the clinical benefits of CRT-D replacement or downgrade to CRT-P.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.

ORCID
Dominic A. M. J. Theuns https://orcid.org/0000-0002-5160-9700 Sing-Chien Yap http://orcid.org/0000-0003-4520-2725 F I G U R E 3 Cumulative mortality rates after cardiac resynchronization defibrillator (CRT-D) replacement stratified by risk group according to DARC risk score. Low risk (black line), medium risk (blue line), and high risk (red line). p Value for logrank less than .001. DARC, death after replacement of CRT-D F I G U R E 4 Cumulative appropriate shock rates after cardiac resynchronization defibrillator (CRT-D) replacement stratified by risk group according to DARC risk score. Low risk (black line), medium risk (blue line), and high risk (red line). p Value for logrank .015. DARC, death after replacement of CRT-D