Failure rates of leads from implantable cardioverter-defibrillator devices are traditionally published by manufacturers and are generally quite low.

Increasing Rates of ICD Lead Failure Noted During Long-Term Follow-Up CME

May 4, 2007 — The annual rate of implantable cardioverter-defibrillator (ICD) leads requiring intervention increases with time and reaches 20% in 10-year-old leads, a study has shown. Defective leads are a problem encountered in all lead models, and occurred in newer as well as older models, investigators report.

"The problem with the leads is known, but how many leads, or exactly when they become affected, is not entirely known," lead investigator Thomas Kleemann, MD, from Herzzentrum Ludwigshafen, Germany, told heartwire. "As more and more patients with longer life expectancies are implanted with ICDs, complications because of these defective leads are going to affect more and more patients."

The results of the study are published online in the April 30 Publish Ahead of Print issue of Circulation.
Lead Failure for 1 in 5 Patients at 10 Years

Lead failure is a long-term complication, said Dr. Kleemann, and with the number of patients with longer follow-up after implantation of an ICD increasing, investigators wanted to determine the long-term reliability of the leads, as well as identify important baseline clinical characteristics of patients who had lead defects.

To assess the annual rate of transvenous defibrillation lead defects after implantation, 990 consecutive patients who underwent first implantation of an ICD between 1992 and 2005 were analyzed. A lead defect was defined as a sensing flaw or fracture requiring surgery to correct it. Median follow-up time was 934 days.

Among the implantations, 15% of patients experienced a lead defect, with a median time to lead failure of just less than 5 years. The estimated lead survival rates at 5 and 8 years after implantation were 85% and 60%, respectively. The annual rate of lead failure increased over time and reached 20% after 10 years. Lead defects occurred in younger patients, female patients, and those with better left ventricular function at the time of implantation.

In a comparison between right ventricular leads implanted after 1997 and older models, investigators showed a trend to better 5-year lead survival with the older models vs a pooled group of newer leads. "We know that the older models had problems, particularly with insulation defects, and because of this there was a change to the newer models," Dr. Kleemann said. "Now we're seeing that we're having problems earlier with these newer leads."

An insulation defect, explained Dr. Kleemann, is a typical case of lead failure in the older polyurethane-insulated leads, a problem caused when then polyurethane breaks down because of oxidation. It appears now that the newer silicone-insulation leads are also prone to insulation defects, he said. Additionally, patients who received an older model typically were implanted with 1 lead, whereas more recent devices have led to multiple lead implantation, something that might explain the higher failure rate in newer models.

Dr. Kleemann told heartwire that leads should be carefully evaluated at the time of pulse generator replacement. Further study to identify predictors of lead failure and safe management strategies to minimize inappropriate shocks and lead revisions is needed, he added, as well as the future consideration of novel approaches, such as leadless ICDs.

Circulation. Published online April 30, 2007.

The complete contents of Heartwire, a professional news service of WebMD, can be found at www.theheart.org, a Web site for cardiovascular healthcare professionals.
Clinical Context

During the past decade, the use of ICDs has increased dramatically, along with expanding indications for ICD implantation. Not surprisingly, the number of patients with longer follow-up after ICD implantation is also increasing, as is recognition of complications associated with the pectoral defibrillator and its transvenous leads.

ICD lead failure may result in failure of the ICD to deliver therapy for ventricular tachycardia, resulting in syncope or sudden death. Other complications of lead dysfunction may include inappropriate shocks and subsequent psychological distress, need for operative revision, or removal, with resultant additional morbidity, mortality, and increased healthcare costs.

This study evaluated the annual rate of transvenous defibrillation lead defects, related to follow-up time after lead implantation, of 990 right ventricular defibrillation leads implanted between 1992 and May 2005.
Study Highlights

* 990 consecutive patients who underwent first implantation of an ICD between 1992 and May 2005 were studied. Patients with lead dislodgment during follow-up and those who required explantation of the device and lead system because of infection were excluded. All patients with coronary artery disease received cardiac catheterization, including coronary angiography and revascularization if needed, before ICD implantation.
* Lead implantation was performed successfully via a transvenous approach using nonthoracotomy lead systems. Most (95%) of the transvenous leads were implanted by puncture of the subclavian vein, 939 leads were connected to pectoral pulse generators, and 51 leads were tunneled to an abdominal pocket.
* Lead failure was diagnosed if there was oversensing unrelated to the cardiac cycle, lead impedance out of normal range with a surgical revision suggested by the manufacturer experts, fracture observed on x-ray, and/or evidence of a lead failure during electrical testing.
* Median follow-up time was 934 days (interquartile range, 368 - 1870 days); 34 (3%) patients were lost to follow-up. During follow-up, 148 (15%) defibrillation leads failed, with estimated lead survival rates of 85% at 5 years and 60% at 8 years after implantation. The annual failure rate continued to increase with time after implantation, and it was 20% in 10-year-old leads (P < .001). Median time to failure in failed leads was 1704 days (interquartile range, 578 - 2414). Insulation failure accounted for 70% of lead failures in leads older than 6 years.
* During follow-up, 207 (21%) patients died: 115 (55%) from congestive heart failure, 4 (2%) from sudden death, 18 (9%) from other cardiovascular death, 27 (13%) from noncardiac causes, and 45 (21%) from unknown causes. 7 patients underwent heart transplantation.
* Lead defects, defined as a severe lead failure that required surgical correction, affected both newer and older models. Compared with patients without lead defects, those with lead defects were 3 years younger at implantation (60 ±11 years vs 63 ± 0 years), more often female (28% vs 19%), and had better preserved left ventricular dysfunction. The indication for ICD implantation was not significantly different between groups. Multiple lead implantation was associated with a nonsignificant trend to a higher rate of defibrillation lead defects (P = .06).
* Lead complications included insulation defects (56%), lead fractures (12%), loss of ventricular capture (11%), abnormal lead impedance (10%), and sensing failure (10%).

Pearls for Practice

* The annual failure rate of ICD leads continues to increase with time after implantation, reaching 20% in 10-year-old ICD leads.
* Compared with patients without lead defects, those with lead defects were 3 years younger at implantation (60 ± 11 years vs 63 ± 0 years), more often female (28% vs 19%), and had better preserved left ventricular dysfunction.

Failure Rate of ICD Leads Are Greater Than Expected

Increased Failure Rate of a Polyurethane Implantable Cardioverter Defibrillator Lead

* WILLIAM H. MAISEL

Cardiac Arrhythmia Service, Cardiovascular Division, Brigham and Women's Hospital, Boston, Massachusetts

ICD
Patient AlertTM to detect ICD lead failure: efficacy, limitations, and implications for future algorithms
Dirk Vollmann1,*, Ali Erdogan2, Ewald Himmrich3, Jörg Neuzner4, Daniel Becker5, Christina Unterberg-Buchwald1, Johannes Sperzel6 for the SAFE Study Investigators

1 Abteilung Kardiologie und Pneumologie, Herzzentrum, Georg-August-Universität Göttingen, Germany; 2 Justus-Liebig-Universität Gießen, Germany; 3 Johannes-Gutenberg-Universität Mainz, Germany; 4 Klinikum Kassel, Germany; 5 Medtronic Bakken Research Center, Maastricht, The Netherlands; 6 Kerckhoff-Klinik Bad Nauheim, Germany
 6 August 2005

Aims An algorithm that alerts implantable cardioverter-defibrillator (ICD) patients, in case of abnormal lead impedance (Patient AlertTM, Medtronic), may help to recognize lead dysfunction. We aimed to determine the utility of Patient Alert for ICD lead-failure detection in a prospective study.

Methods and results Three hundred and sixty ICD patients were followed for 22±14 months. Patient Alert was active for pacing impedance <200 and >2000–3000 {Omega}, and high-voltage conductor impedance <10–20 and >200 {Omega}. Ten alert events and a total of 29 severe system complications occurred. Patient Alert detected three of 10 ICD lead failures, with a positive predictive value (PPV) of 77.8% for any severe system complication. Retrospective analysis identified 23 patients with a sensing integrity counter (SIC) >300 and revealed an additional four prior undetected lead defects. SIC detected ICD lead failure with 92.9% sensitivity and a PPV of 59.1%. Eight of nine patients with a false-positive SIC had an integrated bipolar lead. Patient Alert combined with SIC detected all ICD lead failures and 71.4% of all severe lead complications.

Conclusions Patient Alert, based on daily lead-impedance measurement, detected one-third of all ICD lead failures. Combined use with continuous lead integrity monitoring (SIC) increased sensitivity to 100%. Integrated bipolar leads may yield a false-positive SIC. Incorporating SIC and automated pace/sense threshold measurement may improve Patient Alert sensitivity for severe lead complications.

Key Words: Implantable cardioverter-defibrillator, Lead failure, Impedance measurement, Oversensing, Complication

The implantable cardioverter-defibrillator (ICD) has emerged as therapy of choice in patients at risk for ventricular tachyarrhythmias. For appropriate device therapy, the integrity of the implanted system is essential. Unfortunately, lead defects occur in a non-negligible proportion of ICD patients.1Go–4Go ICD lead failure can be caused by an insulation defect or conductor disruption, and can affect the high-voltage (HV) or the pace-sense circuit of the lead. Potential complications of ICD lead failure include oversensing of electrical noise, undersensing of ventricular tachyarrhyhtmias, inappropriate therapy, and lethal proarrhythmia.2Go,5Go–7Go With the growing number of prophylactic ICD implantations and generator replacements, concern about the long-term reliability of chronically implanted leads increases. Lead defects may be identified during routine follow-up, but early detection is essential to avoid complications. Newer ICDs provide lead-impedance monitoring for early detection of lead failure. An audible signal (Patient AlertTM) may notify the patient to contact the hospital if abnormal impedance indicates conductor fracture or insulation defect. Retrospective data from single centres suggest a potential clinical benefit of Patient Alert for ICD lead-failure detection,3Go,4Go,8Go but there is also evidence of limitations of the present algorithm.

The purpose of this study was to evaluate the utility of Patient Alert for the detection of ICD lead failure in a prospective multicentre study.

In the course of this investigation, the so-called sensing integrity counter (SIC) was introduced. The SIC continuously quantifies very short RR intervals that are typically caused by electrical noise. Recent reports suggest that the SIC may enhance the detection of ICD lead defects.9Go Thus, we performed a retrospective analysis after closure of the present study to determine the utility of the SIC for the detection of ICD lead failure.


This prospective observational study was conducted between March 2000 and November 2004 at 17 centres in Germany and Switzerland. The study protocol conformed to the guiding principles of the Declaration of Helsinki and was approved by the local ethics committees or institutional review boards.

Patients
Patients scheduled for ICD replacement were included if they received a Medtronic (Minneapolis, MN, USA) device that incorporates the Patient Alert feature and if the new generator was connected to a chronically implanted ICD lead. All enrolled patients gave informed written consent.

Impedance measurement and Patient Alert
All devices used in this study provide automatic, subthreshold lead-impedance measurement. The daily measurement of pacing impedance has been described elsewhere.9Go The HV impedance measurement differs between Medtronic Gem devices (Gem I, II, III, InSync) and more recent Medtronic ICD generations (Marquis, Maximo, Intrinsic, InSync Marquis). For the Gem ICD family, subthreshold HV impedance measurement has been described earlier.9Go In more recent Medtronic ICDs, HV impedance is determined with a subthreshold 1 V, 90 µs pulse delivered from right ventricular (RV) coil to can. If connected, the same pulse is also delivered from the superior vena cava coil to the can. In both cases, impedance is calculated by measurements within the same circuit. All impedances are determined once daily at 3.00 AM and are stored in the device memory.

The Patient Alert algorithm audibly warns the patient once daily at a programmable time if one of several defined conditions has been met. In this study, the alert was activated in all subjects for the following conditions: pacing impedance <200 or >2000 {Omega} (Marquis and more recent Medtronic ICDs: <200 or >3000 {Omega}) and HV impedance <10 or >200 {Omega} (Marquis and more recent Medtronic ICDs: <20 or >200). Optional alert features were low battery voltage, long charge time, >3 shocks delivered during one episode, and all therapies delivered in one detection zone. The ‘power on reset’ condition is a non-programmable, obligatory alert feature. All patients were informed about the alert function and were instructed to contact immediately the ICD outpatient clinic, if an alert should occur.

Sensing integrity counter
Lead failure is often associated with oversensing of electrical noise. Intervals between these signals (simulating RR intervals) are typically very short and their occurrence may be intermittent. The SIC continuously quantifies the cumulative number of RR intervals <140 ms between follow-ups. All ICDs used in this study stored SIC data in the device memory. In the Medtronic Gem III and in later Medtronic ICD generations, the SIC was also visible on the programmer at the time of device interrogation.

Data evaluation
Patient characteristics and data on the implanted system were collected upon enrolment and at the time of ICD replacement. Lead integrity and device performance were evaluated at implant, during routine follow-up, at unscheduled visits, after system modification, and in case of a patient death. Unscheduled visits were prompted by Patient Alert or by symptoms thought to be related to the implanted system. Appropriate system performance was evaluated according to the standard follow-up procedure of the participating centres. This included the interrogation of the device with retrieval of all stored events, episode data, and intracardiac electrograms. Included was also the measurement of sensing and pacing threshold and impedance, painless evaluation of the HV impedance, and recording of real-time electrograms. The SIC was evaluated if the corresponding value was visible to the investigator at the time of follow-up. The retrospective SIC analysis (after study closure) considered data that had been retrieved from the device memory at each follow-up in all patients. On the basis of prior reports,9Go,10Go an increment in the total SIC to >300 events since the last follow-up was considered as an indicator of potential lead failure. During each visit, the initial and final interrogations were documented by ‘save to disk’. All data were reviewed and classified by the local investigators and an independent study committee. A system-related severe complication was defined as an undesirable clinical occurrence related to the presence or the performance of the implanted system and resulting in (prolonged) hospitalization, an intervention (e.g. ICD reprogramming and lead replacement), or death. All system-related complications were classified as lead- or device-related. A Patient Alert was classified as appropriate if it detected a system-related severe complication. Lead failure was assumed if the occurrence of abnormal lead impedance, noise oversensing, or other clinical findings (e.g. chest X-ray) suggested conductor fracture or insulation defect, and if these findings resulted in lead replacement. In case of a patient death, physicians were asked to provide a death summary report and a final ICD interrogation. Death was classified as sudden cardiac if it occurred within 1 h of symptoms or if it occurred unwitnessed and unexpected without other apparent cause. Cardiac deaths not classified as sudden cardiac or cardiac deaths of hospitalized patients on inotropic support were classified as non-sudden cardiac death. Non-cardiac deaths were all deaths not classified as cardiac. All cardiac deaths were reviewed for a potential relationship to device dysfunction.

Study endpoints
Primary study endpoint was the proportion of ICD lead failures detected by Patient Alert. Prospectively defined secondary study endpoints were the number of inappropriate alerts and the incidence of system-related severe complications. A retrospective analysis evaluated the SIC sensitivity, specificity, and positive predictive value to detect ICD lead failure.

Statistics
Based on data described in prior product performance reports, the expectation was that in a time period of 3 years, 4.2% of the patients would experience ICD lead failure because of conductor fracture or an insulation defect. The null-hypothesis was that the proportion of ICD lead failures detected by Patient Alert was ≤50%. It was calculated that 300 patients and a follow-up period of 3 years were necessary to reach 80% power.

A total of 360 patients were enrolled and followed for 22±14 (range 2–56) months. Patient characteristics and implanted ICD systems are summarized in Table 1. None of the patients had a separate pacing system. Seventeen subjects had an abandoned RV lead. Complete follow-up information was available from 324 (90%) of the patients. Nine patients relocated to another hospital for follow-up and nine patients left the study for other reasons. Twenty-five deaths were reported and classified as sudden cardiac (n=2), non-sudden cardiac (n=14), and non-cardiac (n=8). In one patient, the mode of death remained unclassified but ICD interrogation provided no evidence of device or lead malfunction. In one case of sudden cardiac death, dysfunction of the implanted ICD lead (Endotak DSP, Guidant, St Paul, MN, USA) could not be excluded, because no final ICD interrogation was available. In all remaining cases, no evidence was found for a potential relationship between patient death and implanted system.


ICD lead failures and system-related complications
A total of 29 system-related severe complications occurred in 8% of the patients. Ten ICD lead failures were observed in 2.8% of all patients. Detailed information on clinical presentation, implanted lead model, and outcome of each event is given in Table. ICD lead failure caused inappropriate detection of ventricular fibrillation in four subjects and resulted in shock delivery in two patients. The majority of ICD lead failures (7/10) affected the Medtronic Transvene model 6936. On the basis of mode of failure and prior reports,3Go these events were attributed to insulation defects caused by metal ion oxidation.

Efficacy and limitations of Patient Alert
Ten Patient Alert events occurred in 10 patients. Event classification is illustrated in Figure 2. As shown, 70% (7/10) of all alert events were appropriate. Patient Alert detected three ICD lead failures. Patient Alert also revealed one case of atrial lead failure and three other severe device-related complications.


Two inappropriate Patient Alert events were observed (Figure 2). In one patient, the alert was triggered because the algorithm was not capable of measuring atrial impedance during atrial fibrillation. This algorithm limitation has been resolved in newer devices. Another alert occurred after the lower threshold for RV pacing impedance had been programmed to <300 {Omega} (nominal: <200 {Omega}). Impedance ranged between 200 and 300 {Omega} in this patient, but no evidence for system malfunction was found. The issue resolved by re-programming the nominal threshold value. One Patient Alert remained unclassified because of missing follow-up information. The alert was triggered by a pacing impedance >2000 {Omega} and resulted in surgical revision. Intraoperative lead values were normal. The ICD was replaced but no defect was found upon technical examination, suggesting that a loose connector screw or an incomplete lead fracture may have caused the alert.

Table 3 summarizes the utility of Patient Alert for the detection of ICD lead failure and system-related severe complications. Seven ICD lead failures were not detected by the Patient Alert. These cases presented with episodes of noise oversensing and/or an increased SIC (Table 2). Also undetected was one case of increased ICD lead pacing threshold, a reduction in ICD lead R-wave amplitude sensing and one atrial lead dislodgement. These cases were all revealed by altered pacing and sensing thresholds during the follow-up testing. One case of phrenic nerve stimulation and two cases of lead dislodgement were associated with a left ventricular lead (Patient Alert not applicable).


Of note, two patients did not hear the alarm tone. In patients who heard the alarm, the average time from the first alert sounding until ICD interrogation was 5.3 days. Five patients reported having heard an alarm tone without a documented event in the device memory.

Sensing integrity counter
The individual SIC increased to a total of >300 (7508±1783) events since the last follow-up in 23 patients. In 11 patients with a Medtronic Gem/Gem II device, this increase was not visible to the investigator at the time of device interrogation. Retrospective analysis showed that the SIC had increased in 9 of 10 cases with confirmed ICD lead failure. In four patients, ICD lead failure was detected during routine follow-up, only because of an increase in the SIC (Table 2). In three of these patients (no. 57, no. 147, and no. 261), a decrease in ring-coil impedance confirmed an insulation breach of the Medtronic 6936 lead. The change in ring-coil impedance only became evident during retrospective analysis. In another patient, lead failure was confirmed at the time of follow-up by a loss of sensing during arm movement. Two patients with ICD lead failure in which the SIC had not been visible on the programmer presented with episodes of ventricular fibrillation (VF) detection (oversensing), causing inappropriate shock delivery in one patient. One case of lead failure within the HV circuit was not detected by the SIC (Patient no. 12). Of note, retrospective analysis of all data revealed the early onset of ICD lead failure in four additional patients with a Medtronic Transvene model 6936. Early onset of lead failure was indicated by a combined increase in SIC and change in ring-coil impedance. These patients were scheduled for lead replacement after study closure.

In nine patients with a SIC>300, no other evidence of lead failure was found. The average SIC increase per day was somewhat lower in these subjects (86±36) than in those with confirmed lead failure (266±120, P=0.12). Eight of the nine patients with a false-positive SIC had an ICD lead for integrated bipolar sensing (Guidant model 0125: n=4; Ventritex SPL: n=2; Guidant Endotak C: n=1; Guidant model 0144: n=1; Medtronic model 6932: n=1). In three of these patients, intermittent T-wave oversensing had been documented. A false-positive SIC increase affected 4.7% of the patients with integrated bipolar leads and 0.6% of the patients with true bipolar leads. None of these patients had an abandoned RV lead. In one patient with a Medtronic Transvene model 6936 lead, the SIC increased 2 days after the ICD had been replaced. Classification was not possible because of missing follow-up information.

In summary, an SIC>300 identified ICD lead failure with 92.9% sensitivity (13/14), 97.1% specificity (336/346), and a positive predictive value of 59.1% (13/22). The combined use of Patient Alert and SIC identified ICD lead failure with 100% sensitivity and detected 71.4% (15/21) of all lead-related severe complications. These data include those lead failures that were detected after retrospective analysis of the SIC.

Patient Alert, based on daily measurement of lead impedance, identified 30% of all ICD lead failures in our prospective observational study. This sensitivity is below the anticipated value of 50% and lower than the sensitivity of 69% reported by Becker et al.4Go Patient Alert did not detect lead dysfunction in seven patients. These subjects presented with an elevated SIC or with episodes of inappropriate VF detection, both caused by intermittent oversensing of electrical noise. One can anticipate that an underlying insulation defect or conductor breach would also eventually produce abnormal low or high lead impedance. However, if the structural lead defect is discrete at first, electrical integrity may be lost only for brief moments (e.g. during arm movement). Current devices measure lead impedance only once daily, and it is unlikely that such discrete measurements will reveal abnormal impedance if lead failure causes sporadic dysfunction. Thus, multiple impedance measurements per day may be necessary to enhance the sensitivity of the Patient Alert. Furthermore, others noted that it might not be sufficient to define abnormal impedance by absolute upper and lower threshold values. Gunderson et al.9Go retrospectively analysed a clinical database to predict ICD lead failure with different algorithms. Using fixed thresholds (e.g. 2000 {Omega}), impedance measurement identified 6.9% of the lead failures. Using thresholds defined in relation to the baseline (e.g. <50% of the minimum baseline), they found that abnormal impedance detected 41.4% of the lead failures with 99.7% specificity.

In contrast to the daily measurement of lead impedance, the SIC continuously quantifies the cumulative amount of short RR intervals (<140 ms). This also allows the detection of very short, intermittent, and clinically silent episodes of electrical noise oversensing. Accordingly, we found that continuous surveillance of lead integrity by the SIC was a more sensitive tool for the detection of ICD lead failure than the present Patient Alert algorithm. Compared with our results, Becker et al.4Go reported a higher Patient Alert sensitivity for the detection of ICD lead failure (69 vs. 30%). In their analysis, however, older device generations than in the present investigation were studied, and the SIC was probably visible in a lower proportion of patients during follow-up. Thus, more lead defects with sporadic dysfunction may have remained undetected in the Becker study. This could explain why their relative number of ICD lead failures detected by Patient Alert was larger than in our study.

Retrospective analysis of SIC data from all our patients revealed four additional cases of lead failure in the Medtronic Transvene model 6936. The resulting cumulative failure rate of this coaxial, polyurethane lead design was 24%. Dorwarth et al.2Go and Ellenbogen et al.3Go reported comparable high failure rates and found that metal ion oxidation and polyurethane breakdown are common causes for insulation breach in this lead design.

Several authors have suggested that integration of the SIC may enhance the sensitivity of Patient Alert.4Go,9Go Gunderson et al.9Go found that a combined algorithm of abnormal impedance trend and increase in SIC detected 86% (25/29) of all ICD lead failures. Our analysis supports these findings, but also elucidates potential limitations of the SIC. In ~40% of the patients with a total SIC >300 since the last interrogation, system dysfunction could not be confirmed. Almost all false-positive cases occurred with leads designed for integrated bipolar sensing. It has been shown that integrated bipolar leads are more susceptible than true bipolar electrodes for oversensing of P- and T-waves,11Go and diaphragmatic myopotentials.12Go Furthermore, double-sensing of premature ventricular complexes has been reported as a cause of false-positive SIC.10Go This event also occurred in a patient with an integrated bipolar lead (Medtronic model 6945, B. Gunderson, personal communication). A causal relationship between oversensing and increase in SIC could not be established in our study, but we observed intermittent T-wave oversensing in three patients with integrated bipolar leads and false-positive SIC. Thus, the SIC may be a sensitive indicator for intermittent oversensing, but one should be aware that the specificity for lead failure is limited in patients with integrated bipolar sensing electrodes.

It should also be noted that both the impedance measurement and the SIC failed to detect complications resulting from a decrease in R-wave amplitude sensing or an increase in pacing threshold. In addition, our study highlights the fact that present algorithms do not monitor the function of the left ventricular lead. In the light of the increasing number of cardiac resynchronization devices being implanted, this problem could become more relevant in the near future. Monitoring of left ventricular pacing impedance and automated measurements of atrial, right, and left ventricular sensing and pacing thresholds could help to overcome the limitations of the present Patient Alert feature. Performing multiple measurements per day may increase the chance to detect intermittent system dysfunction. Last, a significant proportion of patients do not hear the alert signal, and the time between alert and ICD interrogation may be substantial, even in subjects who immediately notice the alarm. Therefore, remote monitoring systems combined with advanced algorithms could be a promising way to enhance early detection of ICD system dysfunction and to prevent severe complications in the future.

Limitations
The effect of Patient Alert on ICD-related morbidity and mortality cannot be concluded from our findings and has to be evaluated in a prospective, randomized study. The main objective of the present investigation is to determine the efficacy and potential limitations of the Patient Alert algorithm for the detection of ICD lead failure. The development and use of novel lead models may yield other failure modes and thus provide different results. We addressed only one lead-monitoring algorithm from a single manufacturer, because no comparable algorithm existed in devices from other companies at the time the study was initiated. Last, this investigation was sponsored by industry, but the manufacturer had no role in the interpretation of the results.

Patient Alert, based on daily measurement of lead impedance, detects approximately one-third of all ICD lead failures. Combined use with continuous lead-integrity monitoring by the SIC increased sensitivity to 100%. Leads for integrated bipolar sensing may yield a false-positive SIC, most likely due to intermittent far-field oversensing of cardiac or diaphragmatic potentials. Integration of the SIC and automated pacing and sensing threshold measurements could provide high Patient Alert sensitivity for all lead-related complications.

Transvenous Implantable Cardioverter-Defibrillator Leads: The Weakest Link.

Editorial

Lead failures constitute a major risk for patients with an implantable cardioverter defibrillator (ICD).However, data about the incidence and patterns of ICD-lead failures in a larger population are lacking. We analyzed the short-term and midterm performance of 27 epicardial and 103 nonthoracotomy ICD-lead systems during a follow-up period of 36 plus/minus 21 months and 22 plus/minus 10 months, respectively (p < 0.05). The failure rate was 5 (19%) of 27 in the epicardial and 6 (6%) of 103 in the nonthoracotomy group (p < 0.05). The most common symptom was erroneous detection of ventricular fibrillation from artifact sensing in five patients. Two patients had to be resuscitated because of failure to defibrillate. Loss of pacing and loss of sensing were seen in two patients. Only two asymptomatic lead fractures could be diagnosed on routine radiograph. In conclusion, there was a considerable rate of lead failures, especially in epicardial systems. Long-term studies addressing the longevity of ICD leads, mechanisms of their failures, and improved diagnostic facilities are important to further increase the safety of this therapeutic approach. (AM HEART J 1995;130:1040-4.)

High spatial resolution measurements of specific absorption rate around ICD leads.

Beard BB, Mirotznik MS, Chang IA.

Food and Drug Administration, Center for Devices and Radiological Health, Electrophysics Branch, HFZ-133, 9200 Corporate Boulevard, Rockville, MD 20850, USA. bbb@cdrh.fda.gov

INTRODUCTION: It has been shown that strong electric shocks can cause local refractoriness in the heart. This is of particular concern if the region of refractoriness is the area sensed by an implant to determine cardiac rhythm, as is the case with many Implantable Cardioverter Defibrillator (ICD) leads which use the same electrodes for shocking and sensing. Failure to sense the true cardiac rhythm can cause application of unnecessary shocks and potential induction of arrhythmias. We developed a system to accurately map the areas where local refractoriness is most probable. We measured the Specific Absorption Rate (SAR) around typical ICD leads. Current density (J), a parameter that determines defibrillation effectiveness, is proportional to the square root of SAR. METHODS AND RESULTS: SAR measurements were performed in a homogeneous saline media using a variety of ICD leads. Gated 60 Hz shocks were used to produce heating, which was measured by thermistor probes. The temperature-rate-of-change is directly proportional to the SAR. Measurement techniques were developed that produced accurate SAR results at high spatial resolutions. Multiple polarities and configurations of ICD leads were tested. CONCLUSIONS: We confirmed the spatial distribution of the SAR and corresponding current density possessed sharp peaks and were highly localized around the leads' electrodes. Scans with a resolution of 1 mm or less are required in the area of peak SAR in order to capture the peak's value.