Article Text

Download PDFPDF
Original article
Impact of continuous intracardiac ST-segment monitoring on mid-term outcomes of ICD-implanted patients with coronary artery disease. Early results of a prospective comparison with conventional ICD outcomes
  1. Giovanni B Forleo1,
  2. Manfredi Tesauro1,
  3. Germana Panattoni1,
  4. Domenico G Della Rocca1,2,
  5. Lida P Papavasileiou1,
  6. Domenico Sergi1,
  7. Arianna Di Molfetta3,
  8. Luca Santini1,
  9. Francesco Romeo1
  1. 1Division of Cardiology, Department of Internal Medicine, University of Rome “Tor Vergata”, Rome, Italy
  2. 2Department of Cardiovascular Medicine, University of Florida, Gainesville, Florida, USA
  3. 3Institute of Clinical Physiology, Section of Rome, National Research Council (CNR), Rome, Italy
  1. Correspondence to Dr Giovanni B Forleo, Division of Cardiology, University Hospital of Tor Vergata, Viale Oxford 81, 00133 Rome, Italy; forleo{at}me.com

Abstract

Background Although myocardial ischaemia monitored by some implantable cardioverter-defibrillators (ICDs) might improve patient care, the clinical usefulness of this technology has not yet been validated.

Objective To investigate the potential impact of ICD-based ischaemia monitoring on clinical care and patient management of ICD recipients.

Design Prospective, controlled, non-randomised study.

Setting Single-centre, university hospital.

Patients Consecutive patients with known coronary artery disease, followed up for at least 6 months.

Interventions Patients implanted with either an ICD providing continuous intracardiac ST monitoring (n=53; ST group) or with an ICD without this capability (n=50).

Main outcome measures Major cardiovascular events, appropriateness of ST-shift episodes and unscheduled device-related visits.

Results During follow-up (15.4±8.4 months), one patient experienced ST-shift events confirmed by angiography to be related to myocardial ischaemia. Myocardial infarction was a rare event and occurred in one patient (ST group) who had an ST-elevation myocardial infarction 3 weeks after the implant, but at this time the algorithm had not yet been activated. In the ST group, seven patients had one or more episodes of false-positive ST events (median 9, range 1–90). The programmable features of the device helped overcome the problem in six patients. Among patients with a remote monitoring system, unscheduled outpatient visits were significantly increased in the ST group (17 vs 4; p=0.032).

Conclusions Although, this study was underpowered by the small number of acute ischaemic events, ICD-based ST monitoring failed to provide a benefit over ICDs without this capability and increased unscheduled evaluations in patients with remote follow-up. The sensitivity and specificity of the algorithm still require validation.

  • Coronary artery disease
  • continuous monitoring
  • intrathoracic electrogram
  • myocardial ischaemia
  • ST segment
  • Atrial fibrillation
  • implantable cardioverter defibrillator (ICD)
  • pacemakers
  • radiofrequency ablation (RFA)
  • cardiac resynchronisation therapy
  • implantable cardioverter defibrillator (ICD)
  • pacemakers
  • coronary physiology

Statistics from Altmetric.com

Introduction

Coronary artery disease (CAD) is a common comorbidity and a potential cause of unfavourable clinical events in implantable cardioverter-defibrillator (ICD) recipients. Early detection of myocardial ischaemia and prompt intervention may substantially improve clinical outcomes. However, available non-invasive assessment methods of myocardial ischaemia do not allow for continuous assessment or monitoring.

Intracardiac ST segment monitoring through implanted devices is emerging as a means of detecting early warning signs of myocardial ischaemia in order to reduce morbidity and mortality resulting from underlying CAD and its various complications.1 Comparative studies have shown that an intracardiac electrocardiogram (ECG) is more sensitive and specific than a surface ECG and early experiences have supported the use of intracardiac electrograms (IEGMs) for detecting acute ischaemic events.2–5 Recently, multi-programmable defibrillators that can carry out continuous IEGM-based ST monitoring through the ventricular lead have been developed. These devices (Analyst and Fortify-ST; St Jude Medical, Sylmar, CA, USA) continuously measure and store any change in ST segment that might be associated with myocardial ischaemia. This has implications for earlier and precise recognition and treatment of myocardial ischaemia, thereby potentially improving clinical outcomes in this subset of patients.

However, no clinical experience with ICD-based ischaemia monitoring has yet been reported. The purpose of this prospective, controlled study was to evaluate the potential impact of ICD-based ischaemia monitoring on clinical care and patient management of ICD recipients with known CAD.

Patients and methods

Patients and study protocol

Consecutive patients with known CAD, scheduled for implantation or elective device replacement of an ICD, were prospectively enrolled. All patients had either primary or secondary prevention indications according to guidelines at the time of implantation.6 Patients dependent on a pacemaker were excluded from the study.

Over a period of 36 months, eligible patients were implanted with an ICD with a continuous IEGM-based ST monitoring system (n=53; ST group) or with a device without this capability (n=50; control group). The choice of implanted system was not related to patient clinical characteristics but to device availability on the day of implantation. Devices were individually programmed according to arrhythmia and patient characteristics.

Device characteristics and implant procedure

Commercially available transvenous ICDs were used in this study. Analyst and FortifyST ICDs are single- or dual-chamber devices equipped with a new system designed to detect any ST change suggestive of ischaemic events. After device implantation, patients returned within 1 month (3–5 weeks) to programme heart rate ranges and respective ischaemia detection thresholds, as suggested by the algorithm. Doctors can programme the positive and negative ST-segment deviation threshold beyond which ST-segment changes are recorded as an event. At first follow-up, the automatically chosen ischaemia thresholds were commonly used in all patients. In case of false-positive (FP) events, ischaemia thresholds were set based upon clinical judgement.

The recording vector used for ST-segment monitoring is the unipolar ICD to right ventricular tip. Every 90 s the device monitors and compares eight consecutive ST segments with a reference obtained at least 24 h earlier. Both positive and negative ischaemia detection thresholds are defined for an ST shift, and if at least six out of eight ST segments are shifted, the set is declared shifted. If three consecutive shifted sets are analysed, an ST-shift episode is considered detected. The device will then store the episode date/time of onset, the heart rate zone in which the episode was detected, the duration and the maximum ST shift recorded. Each ST-shift episode stored has three accompanying electrograms: the baseline IEGM obtained at least 24 h earlier, an IEGM taken at the time of detection and an IEGM which records the largest measured ST shift during the episode.

Standard techniques were used for device implantation. All devices were implanted in a left subcutaneous pocket: 84 leads were placed in the ventricular apex and 19 leads were placed in the interventricular septum. Pacing threshold ≤1 V and R wave ≥5 mV were required for successful implantation.

Follow-up and end points

Postoperative follow-up evaluation was scheduled within 1 month after ICD implantation, at 3 months and every 6 months thereafter, except for acute events. For the purpose of this study, patients were followed up for a minimum of 6 months after implantation. At each visit patients were clinically assessed and a complete device interrogation was performed. The ICD-based ischaemia monitoring system was activated in all patients within 1 month after device implantation. Patients with a remote home monitoring (RM) system were taught how to use the monitor during their first visit to the clinic.

Device-related issues, as well as the assessment of the appropriateness of the ST-shift episode and unscheduled visits to the electrophysiology clinic for a device-related issue were evaluated. Based on analysis of IEGMs and the patient's clinical history, occurrence of an ST-shift episode was classified as appropriate or inappropriate by three independent experienced cardiologists. If an ST episode occurred a dedicated clinical assistant reviewed the event and contacted the referring electrophysiologist and, if necessary, the patient for device interrogation and/or further diagnostic evaluations.

Particular emphasis was given to verifying the appropriateness of ST-segment shifts. All potential episodes of transient ischaemia were printed on paper and the magnitude of the ST-segment shift was measured. The automatic trend analysis was also examined for any episode. Significant ST depression was defined as horizontal or downsloping ST-segment deviation below the baseline at the J point plus 80 ms. ST elevation was defined as the development of an upward ST-segment deviation above the isoelectric line at the J point. The difference between the heart rate before ischaemia and the rate at the onset of ischaemia was evaluated. Changes in the T-wave vector were also evaluated. In addition, the duration of each episode was recorded and its association with chest pain or cardiac arrhythmias. Programming adjustments and pharmacological treatment were left to the discretion of the doctor performing the follow-up.

Statistical analysis

Categorical data are summarised using absolute values (percentage). Continuous data are presented as mean (SD) or, where shown, as median (IQR). Non-continuous variables expressed as proportions were compared using χ2 analysis or Fisher exact test. Comparison between groups was performed with either the Student t test or, when data were not normally distributed, the Wilcoxon rank-sum or signed-rank test. A p value <0.05 was considered to indicate statistical significance.

Results

A total of 103 consecutive patients with at least 6 months of follow-up of were analysed. Of the 103 patients, 53 received an ICD with a continuous intracardiac ST monitoring system and 50 received an ICD without this capability. Baseline clinical characteristics were similar between the two treatment groups (table 1).

Table 1

Baseline characteristics, procedural data and outcomes

Coronary angiography data were available for all patients. In patients who had undergone coronary angiography more than 12 months before ICD placement, a symptom-limited exercise or pharmacologic stress test was used to screen for newly developed coronary disease. Seventy-six patients (73.8%) underwent coronary revascularisation (CR) before enrolment. Median time between the last CR procedure and enrolment was 42 months.

A satisfactory pacing and sensing position for ventricular leads was achieved in all patients, and 65.0% of patients received a RM system. Of the enrolled patients, nine were receiving an ICD as a device replacement or to upgrade an existing pacemaker system.

Clinical outcomes

Follow-up was obtained for all patients: 62.1% completed 12 months of follow-up and 39.8% completed 18 months of follow-up. The average follow-up was 15.4±48.4 months without significant difference between the two groups. By Kaplan–Meier analysis, time to appropriate ICD therapy or all-cause death was similar in the two groups. A schematic representation of the outcomes of the two study groups is depicted in table 1.

Documented myocardial infarction (MI) was rare and occurred in one patient in the ST group. This patient had an ST-elevation MI, which occurred 3 weeks after the implant, but at this time the algorithm had not yet been activated. Twenty-three patients in the ST group had residual stenosis known from previous angiography. Thirteen of these patients had provocative tests, showing residual myocardial ischaemia in three patients, of whom two underwent percutaneous revascularisation during follow-up.

Eleven months after ICD implantation one patient with previous coronary bypass surgery underwent angiography because of unstable angina preceded by typical chest pain in the previous 3 months. The examination showed a critical stenosis in the mid-portion of the right coronary artery venous graft that was treated with stent implantation. Device interrogation, performed after the revascularisation procedure, showed frequent ST-shift events (n=141) detected during elevated heart rates. These events had started approximately 5 months after ICD implantation but remained unrecognised until the device interrogation because the patient did not have an RM system. Analysis of the five IEGMs stored showed ST changes suggestive of ischaemic events (figure 1). After revascularisation, the patient remained asymptomatic and free of ST events during 8 months of follow-up.

Figure 1

(A–D) Stored intracardiac electrograms showing ischaemia-related ST-segment depression. The first episode (A) occurred approximately 5 months after implantable cardioverter-defibrillator implantation but remained unrecognised until the device interrogation. After percutaneous coronary revascularisation, the patient remained asymptomatic and free of ST events.

False-positive ST-shift events

False-positive ST events occurred in 7/53 patients (13.2%), for an episode rate of 2.46 a year (163 episodes/66.3 patient-years). One patient had one episode and six had four or more episodes. These seven patients had a median of nine episodes (range 1–90) after ICD implantation. The longest interval to the first ST episode was 17 months. The majority of episodes occurred in patients with insignificant CAD (no stenosis >50% on coronary angiography, or positive stress test). Four of these patients had episodes without a heart rate rise, and the remaining three had episodes both with and without a preceding heart rate increase. None of the FP alarms occurred as a result of ventricular pacing or cardiac arrhythmias. A detailed description of the ST-shift events is given in table 2.

Table 2

Characteristics of ST-shift events

In two patients, the ST episodes were classified as FP by analysing IEGM and clinical data. Among the other five patients, two underwent coronary angiography, one underwent a CT coronary angiography and two an ischaemia stress test. In six patients (85.7%) reprogramming of the device resolved the problem (free from any FP detections after 3 months). Examples of false-positive ST events are shown in figure 2.

Figure 2

Stored intracardiac electrograms from four case subjects showing examples of false-positive ST events. In some cases it was very easy to recognise that they were false-positive events since no significant ST shift was evident at visual inspection of the intracardiac electrograms stored. In panel (A) the event was associated with an ST-segment depression. Intracardiac ST shifts, wrongly considered by the algorithm as significant ST-segment elevations, are shown in panels (B–D). In panel (B) the ST event was related to a rate-dependent aberrancy on QRS complex.

Baseline characteristics were compared between patients with FP episodes and those without. Surprisingly, ECG abnormalities were not identified as predictor of future inappropriate ST-segment episodes and the baseline characteristics of the two groups were similar. (table 3).

Table 3

Clinical characteristics and outcomes of patients with or without false-positive events

Among patients with the RM system (n=67), 21 unscheduled outpatient visits were recorded during the follow-up period, and the calculated visit rate was 0.26/patient-years of follow-up. The number of unscheduled visits was significantly higher in the ST group (17 vs 4; p=0.032) (table 1). The most frequent cause of visits among patients in the ST group was an ST event, which accounted for 80% of total unscheduled visits.

Discussion

This report is the first study to date describing in detail the clinical utility of ICD technology to detect myocardial ischaemia in high-risk individuals as identified by a history of CAD. It resulted in the following important findings: (1) in the mid-term follow-up, a small acute coronary syndrome (ACS) event rate was found in ICD recipients; (2) ICD-based continuous ST monitoring failed to provide a benefit over standard ICDs without this capability; (3) a significant number of FP events was found in patients in the ST group increasing the unscheduled evaluations in patients with remote follow-up.

The clinical relevance of the observations made in this study may be partly viewed in the context of our current failure to achieve a significant number of acute ischaemic events during follow-up. Ventricular function affords additional prognostic information to coronary anatomy,7 therefore ACS incidence was expected to be higher in our series. However, published studies report that in patients with left ventricular dysfunction and a history of MI, sudden cardiac death comprises about 50% of the causes of cardiac death, with the remainder attributed to progressive heart failure.8 9 Importantly, there are no reliable data on the incidence of ACS in ICD recipients with chronic ischaemic heart disease, therefore this subset of patients cannot be easily compared with others. Generally, ICD recipients are stable at implant, should be at least 40 days after MI and we performed a coronary angiography with the intention of revascularising all patients before implantation. It has been shown that coronary revascularisation in conjunction with ICD implantation is associated with improved outcomes among patients at high risk after MI,10 and approximately 40% of our patients had a revascularisation procedure within 1 year before ICD implant. One further reason for the low incidence of ischaemia in this study might have been the effectiveness of maximal antianginal drugs in reducing myocardial ischaemia as 93% of patients were receiving β blockers. These aspects are likely to be representative of the overall population presently considered for ICD treatment. Therefore, although a benefit might be seen with a longer follow-up, the low ACS event rate in this study is likely to be typical of ICD recipients, at least in the short- and mid-term follow-up. The implication of this observation is that it remains questionable whether the additional costs required to treat a large number of people to prevent only a small number of events would be justified. Larger sample sizes are needed to provide a definitive answer.

Fischell et al5 have recently reported a clinical experience with an intracardiac ischaemia monitoring system in 37 patients at high risk of ACS. The implanted monitor proved to be useful for detecting and alerting patients to ischaemic events, thereby reducing alert-to-door time for patients at high risk of recurrent coronary syndromes. Since ICD patients with known CAD are at a high risk of adverse events, it was reasonable to postulate that ICD-based ischaemia monitoring would improve survival among these patients. However, available ICDs do not have any patient alert system, therefore it remains questionable whether a system without the patient alert might be useful in this subset of patients. In this study, both acute and demand-related myocardial ischaemia were detected in one patient but remained unrecognised until the following device interrogation. With devices that offer the potentials of an internal alert mechanism and the possibility of sending automatic alert messages, both the doctor and patient could be notified at the onset of myocardial ischaemia so that appropriate treatment might be promptly instituted. The next generation of system will deal with the aforementioned limitations; in this way critical events may be reported earlier, allowing prompt and effective treatment.

Long-term ST-segment monitoring by permanently implanted devices may be of substantial clinical value. However, its practical application may, in part, be limited by the prevalence of FP ST-segment events. A relatively high incidence of FP events was detected in our series, increasing unscheduled evaluations in patients with remote follow-up. False-positive detections occurred with the device programmed using either the manually or automatically chosen ischaemia thresholds and it is noteworthy that we were unable to identify any predictors of FP detections. As suggested by two phase 1 clinical studies,5 the automatically chosen thresholds were adopted in our patients. Of note, manual reprogramming of ischaemia thresholds resolved any FP detections in 85.7% of patients. Manual programming of these devices is time consuming and requires experienced electrophysiologists, nevertheless our results suggest that it is required in all patients, shortly after the ST-algorithm activation. Similarly, a higher specificity of the algorithm is necessary and would avoid unnecessary clinical evaluations once the alert system is activated. Future research will focus on optimal device programming that will offer maximal benefit, increasing the clinical utility of this algorithm. A large European registry investigation is being conducted to assess the performance and optimal programming of this system and to assess sensitivity for ACSs.

Unscheduled patient evaluation after ICD implantation has a major impact on quality of life and cost-effectiveness in defibrillator recipients. A remote monitoring system is associated with reduced outpatient consultations for ICD monitoring11 and may also save costs. In our small series, the number of unscheduled patient evaluations was significantly higher in the ST group, with a 0.36 visit rate per patient-year of follow-up. This is important because the increase in unscheduled visits slightly offsets the benefit of reduced scheduled appointments in patients with an RM.

The determinants of myocardial ischaemia are likely to differ in patients with ACS as the underlying pathological substrate usually consists of plaque rupture with intracoronary thrombus formation. Whereas the value of intracardiac ST-segment monitoring has been demonstrated in patients with ACS,5 no published study has analysed its possible effect in patients with stable CAD. Of note, no true-positive ST-shift episodes were detected in patients with significant stenosis on coronary angiography, and ischaemia induced by provocative testing was documented in three of these patients. These findings raise questions about the sensitivity of ICD-based ischaemia monitoring in patients with stable CAD. Further, larger trials are required to settle this issue.

Conclusions

This study represents the first consecutive series of ICD implantation with ST capability so far published. After a follow-up of 15.4±8.4 months, this innovative system failed to provide a benefit over standard ICDs without this capability. Furthermore, a relatively high incidence of FP ST events was detected in our series, increasing unscheduled evaluations in patients with remote follow-up. However, this study was underpowered by the small number of ACS events, therefore additional supporting data, including longer follow-up, greater number of patients and further analysis aimed at optimal programming of the algorithm, are required.

Limitations

This was a single-centre, non-randomised study and has the limitations inherent to this design. However, it is important to note that there was no selection bias between the systems, and the initial choice of implanted system was based on availability at the time of implantation rather than patient-specific variables. The most important limitations are the low ACS event rate, the length of follow-up and the number of patients enrolled. Although a benefit might be seen with a longer follow-up, the low ACS event rate in this study is likely to be representative of ICD recipients. Therefore our data do not suggest a high likelihood that a larger study would demonstrate a clinical benefit from ST-segment monitoring in patients with an ICD.

Acknowledgments

The authors thank Cecilia Rubaudo for her assistance in the preparation of this manuscript.

References

Footnotes

  • Competing interests None.

  • Patient consent Obtained.

  • Ethics approval Ethics approval was provided by the ethics committee of the University of Tor Vergata.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.