How To Troubleshoot ICD Problems

In EP Lab Digest’s second edition of the new “How To” section, author Craig Swygman reviews a few of the clinical scenarios that may be encountered during implantable cardioverter defibrillator (ICD) troubleshooting. Results of clinical trials conducted in the past several years have led to a dramatic increase in the number of ICDs implanted in the United States and around the world. ICDs have been shown to be superior to anti-arrhythmic medications for primary and secondary prevention of sudden death. Additionally, ICDs with the capability of providing cardiac resynchronization therapy have been shown to provide benefit in patients with congestive heart failure who are also at risk of lethal tachyarrhythmias. ICD technology has progressed rapidly with sophisticated features available both for the treatment of bradycardia and tachycardia. However, the increasingly complex features and capabilities of ICDs can make the follow-up of these devices very challenging. This article will discuss how to troubleshoot some common ICD problems encountered in the ICD follow-up clinic.

ICD Therapy: Appropriate vs. Inappropriate

The ICD follow-up clinic allows the clinician to assess device function and the clinical status of the patient. ICD interrogation reveals whether any therapy has been delivered by the device since the previous follow-up session. ICD therapies can be classified as either appropriate or inappropriate. Appropriate therapies are for ventricular tachycardia (VT) or ventricular fibrillation (VF). ICDs may inappropriately deliver therapy for supraventricular tachycardias (SVT) that occur in the programmed VT or VF zones of the device. SVTs that can result in therapy are sinus tachycardia, atrial fibrillation (AF), atrial flutter, atrial tachycardia, atrioventricular nodal reentrant tachycardia (AVNRT), and atrioventricular reentrant tachycardia (AVRT). Therapy may also be inappropriately delivered for oversensing of either physiologic or non-physiologic signals. A careful review of the stored information from the episode resulting in device therapy can assist the clinician in interpreting the ICD therapy as either appropriate or inappropriate. When ICD interrogation shows that anti-tachycardia pacing or shocks have been delivered since the last follow-up, the clinician should review the stored data from each episode to determine the appropriateness of the therapy. Stored data from single-chamber defibrillators is limited to information from the ventricle; this can make interpretation of the rhythm difficult (i.e., differentiation between VT and SVT). However, single-chamber ICDs have a number of automated features that can be used to discriminate between SVTs and VT. For example, ventricular tachycardia typically has a sudden onset, unlike sinus tachycardia, which usually has a gradual onset. Therefore, the “onset” feature on a single-chamber ICD can be used to reduce therapy for sinus tachycardia and increase the specificity of the device. The “onset” feature will result in therapy being withheld if the tachycardia has a gradual onset. VT is also usually very regular, unlike AF, with typically irregular intervals. On a single-chamber ICD, the “stability” discriminator can be used to differentiate between VT and AF. When programmed on, the “stability” feature will result in no therapy being delivered if the tachycardia is irregular. The morphology of the bipolar intracardiac electrogram can also be used to differentiate between VT and SVT. The specific electrogram morphology in sinus rhythm is compared to the tachycardia electrogram morphology that resulted in ICD therapy. The tachycardia electrogram morphology will usually be different than the sinus morphology if the tachycardia mechanism was VT, due to a different direction of ventricular depolarization. The tachycardia electrogram will be similar to the sinus morphology if the rhythm was supraventricular in origin, as conduction would have been through the normal conduction system. Careful attention should also be paid to the direction of deflections to determine if the electrogram morphology is similar or dissimilar to sinus rhythm. Many defibrillators have an automated SVT discriminator that saves a sinus rhythm electrogram “template,” and in the event of a tachycardia occurring in the VT zone, uses a “matching” algorithm to compare the tachycardia electrogram morphology to the stored sinus rhythm electrogram “template.” The device then determines whether the morphology is similar to the sinus template, indicating SVT, or different than the template, indicating VT. Dual-chamber defibrillators allow for storage of both atrial and ventricular intracardiac electrograms, in the event of a tachycardia in the VT or VF zones. This can make differentiation between SVT and VT much easier for the ICD follow-up clinician. VT usually results in VA dissociation or variable retrograde conduction. Therefore, if the stored intracardiac electrogram from the tachycardia episode shows more ventricular deflections than atrial deflections, the diagnosis of VT can be made. Dual-chamber ICDs have a programmable SVT discriminator that can determine if the ventricular rate is greater than the atrial rate (V>A), and deliver therapy if VT is indicated. Occasionally, patients can have dual tachycardias (e.g., VT in the setting of ongoing or paroxysmal AF). The clinician can then use the characteristics mentioned above, such as sudden or gradual onset, interval stability, and electrogram morphology to determine the appropriateness of device therapy.

Inappropriate ICD Therapy Due to Oversensing

Heart rate is the main parameter used by ICDs to detect VT or VF. Therefore, it is essential that sensing of each intracardiac electrogram is accurate and consistent. ICDs automatically adjust the gain or sensitivity on a beat-to-beat basis to ensure appropriate sensing of each R-wave. The devices must be able to differentiate between asystole and fine VF. An inherent limitation of increased sensitivity is that the device may sense cardiac or extracardiac signals in addition to the ventricular depolarization. This oversensing can result in the device determining the heart rate to be in the VT or VF zone and result in ICD therapy. Cardiac signals that can result in oversensing are P-waves, T-waves, and uncommonly double-counting of R-waves. A careful review of the stored intracardiac electrogram from the episode can demonstrate that the sensed events correspond in time to the P-wave, T-wave, or late in the R-wave after the appropriate R-wave sensed event. T-wave oversensing can occur if R-wave amplitude decreases intermittently or permanently. Smaller R-waves and T-wave oversensing can occasionally be related to postural changes in the patient. A careful review of the patient’s activity at the time of the inappropriate shock can often be helpful to determine the specific scenario that led to oversensing. Interventions that may be necessary to eliminate oversensing of cardiac signals include: reprogramming the sensitivity to a less sensitive setting, increasing the ventricular blanking period, or repositioning the lead. Some ICDs allow for programming of the timing and rate of increase of the sensitivity adjustment after a sensed event. Oversensing of extracardiac signals can also result in inaccurate measurement of the heart rate and lead to inappropriate therapy from the ICD. The most common source of extracardiac signals resulting in device oversensing is the diaphragm. Diaphragmatic myopotentials are low-amplitude, high-frequency signals. Oversensing can lead to inhibition of pacing and inappropriate shocks. Oversensing of diaphragmatic myopotentials is more common if the ventricular lead is implanted at the right ventricular apex, and if brady pacing is present, as ICD sensitivity is higher after a paced event. Oversensing of diaphragmatic myopotentials can often be reproduced in the follow-up setting; it can be demonstrated by recording the telemetered electrogram while having the patient perform various postural changes and respiratory maneuvers. Valsalva maneuver can be especially helpful to recreate diaphragmatic myopotential oversensing in the clinical setting. Interventions that may be necessary to eliminate oversensing of diaphragmatic myopotentials are adjusting the device sensitivity, repositioning the lead away from the apex, or implanting a separate rate-sensing lead. Patients might also receive inappropriate ICD therapy due to sensing of electromagnetic interference (EMI) from an external source. Oversensing of EMI typically results in high-frequency “noise” on all stored intracardiac electrograms. As with other causes of inappropriate therapy, a careful review of the patient’s activity at the time of the shock can often be helpful to determine the specific source of the EMI. The best approach to avoid further inappropriate therapy is for the patient to avoid the source of EMI if possible. Patients should be reassured that distance from the EMI source is the main determinant of whether device oversensing occurs.

Device and Lead Problems

As with any electronic device, medical or otherwise, ICDs and leads can be subject to random failure or inherent design flaws. Device or lead failures can result in inappropriate therapy or loss of necessary therapy. However, most lead and device failures are recognized at routine ICD follow-up visits. Lead failure may be due to fracture or insulation defect. Evidence of lead failure can be high or low pacing impedance, abnormal high-voltage impedance, high pacing threshold, non-physiologic short intervals, and noise on the electrogram resulting in oversensing and ICD therapy or aborted therapy. It is important to identify the specific sensing vector if oversensing consistent with lead failure is present. This will allow the clinician to determine if the lead defect is related to the sensing portion of the lead only, or if the high-voltage component is also involved. If there is evidence of oversensing early after device implant, the cause may be a loose set-screw in the header of the device. Patients with multiple leads may have oversensing due to mechanical “chatter” between the leads being sensed as intracardiac signals. In the clinic setting, it may be possible to reproduce the oversensing with pocket manipulation, arm movements, or isometric maneuvers. Intervention for lead failure is implantation of a new lead, with or without extraction of the old lead.

Conclusion

ICD and lead technology continue to improve as indications for device implantation increase. It is imperative that clinicians performing ICD follow-ups are current with techniques to determine the appropriateness of device therapy and to identify evidence of device or lead problems.