Riata and Riata ST implantable cardioverter–defibrillator leads are prone to structural and electrical failure (EF). Our objective was to evaluate Riata/ST lead performance over a long-term follow-up. Of 184 patients having undergone Riata/ST and Riata ST Optim lead implantation from September 2003 to June 2008, 154 patients were evaluated for EF and radiographic conductor externalization (CE). Survival analysis for EF was performed for Riata/ST leads, both for failure-free lead survival and cumulative hazard. Subanalysis on 7Fr leads was performed to evaluate EF and CE rates both for different Riata ST lead management (monitoring vs proactive) and between Riata ST and Riata ST Optim leads. During a mean follow-up of 7 years, Riata/ST lead EF rate was 13% overall. Similar failure-free survival rate was noted for 7Fr as for 8Fr leads (log-rank, p = 0.63). Of all failed leads, 64% failed only after 5 years of follow-up. Compared with the absolute failure rate of 1.84% per device year, cumulative hazard analysis for leads surviving past 5 years revealed an estimated failure rate of 7% per year. No clinical or procedural predictors for EF were found. The subanalysis on 7Fr leads showed an excellent outcome both for a proactive lead management approach as for Optim leads. In conclusion, long-term survival of the Riata/ST lead is impaired with an accelerating EF risk over time. An initial exponential trend was followed by a linear lead failure pattern for leads surviving past 5 years, corresponding to an estimated 7% annual EF rate. These findings may have repercussions on the lead management strategy in patients currently surviving with a Riata/ST lead to prevent significant clinical events like inappropriate shocks or failed device interventions.
Regarding the increased failure rates of the Riata (8Fr) and Riata ST (7Fr) silicone implantable cardioverter–defibrillator (ICD) shock lead (St Jude Medical, Symlar, California), a medical device advisory was issued on November 2011, followed by a Food and Drug Administration class I recall one month later. Previous studies, highlighting on the unique “inside–out” abrasion mechanism of conductor cables through its silicone insulation, were reporting rates between 11% and 22% of conductor externalization (CE). In addition to this structural failure, there is a higher rate of electrical failure (EF) in Riata leads compared with other ICD leads. As previously reported, the rate of electrical dysfunction (mean follow-up time ≤5.6 years) varies from 4.6% to 17.3%. To improve the understanding of lead failure trends over time, we aimed to assess the rates of Riata/ST lead failure (electrical and structural) over a long-term follow-up.
Methods
We retrospectively evaluated all 143 patients with a Riata/ST lead and 41 patients with a Riata ST Optim leads implanted from September 2003 to June 2008 at 2 high-volume implant centers. Thirty patients were excluded because they were lost to follow-up.
Most Riata/ST leads underwent a monitoring lead management approach consisting of routine device interrogation with lead revision performed in case of EF. On 7Fr Riata ST leads, a center-dependent proactive lead strategy was implemented consisting of systematic lead revision (implantation of new ICD lead and extraction of lead in situ) in case of battery replacement or radiographic detection of CE, regardless of electrical lead integrity. As the main focus of the study was to follow the natural behavior of Riata/ST leads in time, we considered first only Riata/ST leads who underwent a monitoring management. In addition, we performed a subanalysis on 7Fr leads comparing different lead management (monitoring vs proactive approach) and EF/CE rates between Riata ST and Riata ST Optim lead.
Records of the selected patients were reviewed. Demographic, clinical, and device data were obtained from the implantation reports and hospital records. Patients were followed 1 month after ICD implantation and at a 6-month interval thereafter in our device clinic and/or through remote monitoring when available. Electrical lead failure was defined as one of the following: (1) pace/sense conductor impedance out of range (outside 200 to 2000 Ω interval) or a sudden change >100% increase or >50% decrease of the stable baseline, (2) change in high-voltage impedance (>200 or <50 Ω), (3) increase in capture threshold >5V or >100%, (4) decrease in R-wave sensing (<3 mV or >50%), and (5) oversensing of nonphysiological electrical noise artifacts, not due to external interference such as electromagnetic interference. Lead dislodgments, physiological oversensing, T-wave oversensing, and header problems were not considered as EF. Inappropriate therapies associated with lead failure were recorded. For survival analysis, the date of lead failure was determined as the first occurrence of any lead abnormality conform the definition of EF as stated previously.
All available chest x-ray films (posteroanterior and lateral view) and fluoroscopic images of Riata/ST leads were analyzed. Only patients with a radiographic evaluation ≥1 annually and ≤1 year before the end of follow-up were selected. Externalization was defined as conductor(s) visible outside the lead body in any of the projected views. Possible externalization was determined in case of abnormal conductor spacing, that is, movement of one or more cables away from the main body of the lead.
Continuous variables are expressed as mean ± SD and categorical data as number and percentages. To identify the baseline characteristics associated with adverse outcome (EF), a Cox proportional hazard regression analysis was performed. Test of proportional hazard assumption was conducted. Survival and cumulative hazard were calculated using the Kaplan–Meier method and compared using the log-rank test. The change in EF risk over time was evaluated by calculation of the conditional survival probability and by a log–log plot of the cumulative hazard function (log H vs log t), as previously described. This analysis was repeated for leads surviving at 2, 4, and 5 years. If the slope of the conditional survival log H versus log t approximated 1 (i.e., constant lead failure rate over time), linear regression analysis was performed on the cumulative hazard plot (H vs t) to define the lead failure rate per year. Statistical analyses were conducted using the SPSS software, version 22, (SPSS, Chicago, Illinois).
Results
We analyzed 154 patients with a total of 72 (50%) model 1581 Riata, 52 (35%) model 7001 Riata ST leads, and 30 (15%) model 7021 Riata ST Optim leads. All screened leads had a dual-coil design. Leads under monitoring approach included 108 patients with a total of 72 (67%) 1581 Riata and 36 (33%) 7001 Riata ST leads. Baseline clinical characteristics are summarized in Table 1 . The mean follow-up was 7.0 ± 1.8 years.
| Variables at baseline | All ( n =108) |
|---|---|
| Age at implantation (years) | 63 ± 14 |
| Men | 87 (80%) |
| Underlying heart disease | |
| Ischemic cardiomyopathy | 60 (56%) |
| Dilated cardiomyopathy | 27 (24%) |
| Others (Brugada Syndrome, Hypertrophic cardiomyopathy, …) | 21 (20%) |
| Left ventricle ejection fraction (%) | 38 ± 14 |
| Persistant/permanent atrial fibrillation | 28 (27%) |
| Pacemaker dependancy | 11 (10%) |
| Primary prevention | 68 (63%) |
| Secondary prevention | 39 (37%) |
| Device intervention | |
| Anti-tachy pacing (appropriate) | 37 (35%) |
| Shock (appropriate) | 22 (21%) |
| Type of device | |
| Single-chamber | 49 (45%) |
| Dual-chamber | 46 (43%) |
| Biventricular | 13 (12%) |
| Distribution of venous access | |
| Left-sided implant | 104 (96%) |
| Cephalic | 68 (63%) |
| Subclavian | 40 (37%) |
| Right ventricle lead position | |
| Apex | 105 (97%) |
| Septal | 3 (3%) |
| Lead type and model | |
| 7 Fr – 7001 | 36 (33%) |
| 8 Fr – 1581 | 72 (67%) |
| Follow-up time (years) | 7,0±1,8 |
Cross-sectional analysis revealed a total of 14 (13%) Riata/ST electrical lead failures. Mean time to EF was 5.0 ± 2.7 years. The different presentations of EF in these leads are given in Table 2 . All abnormal impedance changes occurred on 8Fr leads, those with high voltage failure presented late appearance ≥72 months after implantation. In 4 (28%) leads, we noted >1 electrical abnormality, 1 lead presented 4 types of electrical dysfunction altogether. In an exploratory univariate Cox regression analysis, none of the baseline characteristics was withheld as a predictor for EF ( Table 3 ). No variable entered a subsequent multivariate model as no variable presented a p value <0.2 on the univariate analysis. Overall, 4 patients (4%) had lead-related inappropriate shocks, due to oversensing of nonphysiological noise on the ventricular sensing channel. One of these patients showed signs of cable externalization. No deaths were attributed to lead failure.
| Electrical lead failure | All leads ( n = 14) |
|---|---|
| R-wave sensing reduction | 3 (21%) |
| Increased pacing threshold | 7 (50%) |
| Pacing impedance out of range | 3 (21%) |
| High voltage failure | 3 (21%) |
| Oversensing/noise/shock | 4 (28%) |
| More than 1 electrical abnormality | 4 (28%) |
| Hazard Ratio | 95% Confidence Interval | P – value | |
|---|---|---|---|
| Univariate analysis | |||
| Age | 1,01 | 0,96-1,06 | 0,67 |
| Female gender | 1,65 | 0,26-10,32 | 0,6 |
| Non ischaemic cardiomyopathy | 0,65 | 0,15-2,72 | 0,56 |
| Left ventricle ejection fraction | 1,01 | 0,96-1,06 | 0,81 |
| Atrial fibrillation | 1,25 | 0,24-6,44 | 0,8 |
| Secondary prevention | 0,92 | 0,25-3,47 | 0,91 |
| Pacemaker dependancy | 0,55 | 0,05-6,76 | 0,65 |
| Ant-tachy pacing (appropriate) | 1,21 | 0,22-6,52 | 0,82 |
| Shock (appropriate) | 2,75 | 0,21-35,4 | 0,44 |
| ∗ Left-sided implant | 7,77 | 0,70-86,4 | 0,1 |
| Subclavian access | 0,582 | 0,15-2,17 | 0,42 |
| 8 French lead | 1,48 | 0,27-8,01 | 0,65 |
| 2/3-chamber device | 1,97 | 0,57-6,76 | 0,28 |
The overall Kaplan–Meier survival curve for the Riata/ST leads, 7Fr and 8Fr pooled is shown in Figure 1 . The 3- and 6-year survival rates for the Riata leads were 95% and 92%, respectively. Of 14 leads in total, 9 leads (64%) failed only after 5 years of follow-up. Survival analysis, plotted out separately for 7Fr and 8Fr is shown in Figure 1 . Both models showed a similar failure-free survival rate (log-rank, p = 0.63).
Using the averaging method of calculating EF rates with an absolute failure rate of 13% involving 14 leads and 760 total device years of follow-up, we calculated an EF rate of 1.84% per device year. Log–log plot analysis of cumulative hazard versus time, performed for leads after 5 years after implantation, revealed a straight line with a slope of 1.002 (R 2 0.92, 95% CI 0.93 to 1.06; p <0.0001). By performing a linear regression analysis on the cumulative hazard plots obtained for leads surviving past 5 years, we found an estimated EF rate of 7% per year (conditional survival analysis method). An illustration of the 2 approaches to estimate lead failure rates over time is represented in Figure 2 .
We selected 89 patients for the evaluation of CE. Most patients (75; 84%) were evaluated by standard chest x-ray films (posteroanterior and lateral view), only patients with EF (14; 16%) were evaluated by fluoroscopy imaging. Clear externalized conductors were identified in 12 patients (13%) overall, of which 5 leads with (i.e., 5 of 14; 36%) and 7 leads without additional EF (i.e., 7 of 75; 9%), whereas 9 of the nonexternalized leads had EF (Fisher’s exact test p = 0.05). CE and EF had a slightly positive correlation (Phi’s coefficient +0.22; p = 0.026). Almost all externalized conductors were seen on 8Fr leads (11; 91%). Possible externalization (abnormal cable spacing) was present in another 18 patients (20%) and contributed positively to the correlation with EF if combined with clear externalization (Phi’s coefficient +0.36; p <0.001).
Revision data were available for 12 of 14 (85%) electrically failed Riata/ST leads. These leads presented integrity of the lead under the can, with inclusion of the 2 failed leads that underwent previous battery replacement. Other forms of failure were not considered in this study as lead EF: 5 lead dislodgments and 1 subclavian crush syndrome. All leads with EF were revised by inserting a new high-voltage lead. Most leads were capped and abandoned (71%). In the 4 remaining cases, there were 2 lead removals by traction and 2 laser lead extractions. No major complications or deaths were reported because of the lead revision.
Considering only 7Fr leads, 4 leads presented EF in the Riata ST group under monitoring, whereas no EF was observed in the Riata ST leads with a proactive management approach ( Table 4 ). Latter more aggressive approach consisted in a premature lead replacement, as was the case during elective battery replacement in 3 patients (19%), and after detection of radiographic CE in 1 patient (6%). Lead removal was performed by means of laser extraction in 3 cases and right atriotomy in one case (concomitant with a coronary artery bypass surgery). These procedures were uneventful and without any postprocedural complications. In our study, also Riata ST Optim leads showed an excellent outcome, without occurrence of EF or CE during a follow-up of >5 years ( Table 4 ). These results of the subanalysis must be interpreted carefully given the shorter follow-up compared with Riata/ST leads in our study, both for Riata ST leads with proactive approach (inherent to a shorter lead dwell time) and Riata ST Optim leads (more recently implanted). Moreover, this subanalysis is merely descriptive as the number of patients and events was too low to perform associative statistics or survival analysis.
