Malini Madhavan, MBBS, Paul A. Friedman, MD, FACC, FHRS
A 69-year-old woman received a single chamber Medtronic ICD for primary prevention of sudden death. The device was programmed to detect VF at 320 ms if 30/40 intervals met criteria. “Wavelet” morphology discrimination was turned on so that therapy would be withheld if tachycardia electrogram (EGM) matchedwith that of the normal rhythm template by 70% or more. The supraventricular tachycardia (SVT) limit was set at 280 ms, and high rate time out was off. The patient received shocks, and device interrogation is shown in Figure 68-1.
FIGURE 68-1 Inappropriate shock for supraventricular tachycardia. Misclassification by the morphology discrimination algorithm is seen due to clipping (inset) of the electrogram. (Reproduced with permission from Hayes DL, Asirvatham SJ, Friedman PA. Cardiac pacing, defibrillation and resynchronization: a clinical approach. 3rd ed.: Wiley-Blackwell; 2012.)
The arrhythmia seen is an SVT. The device compares the morphology of the far field EGM during tachycardia with that of a template stored during normal rhythm and finds that eight of the last eight complexes do not “match.” The far field electrogram records the EGM between the pulse generator and the right ventricular coil, although other vectors are programmable. In contrast, the near field electrogram records the signal between the right ventricular tip and ring (or tip and coil) and generally contains less morphology information.
Since the morphology during tachycardia does not match the template, the device classifies the rhythm as ventricular fibrillation (VF) and delivers therapy. On closer examination, the top of the EGM appears to be “cut-off,” or in other words, truncated. Figure 68-2 shows clipped and unclipped EGMs from the same patient. Saturation of the amplifiers results in clipping of the EGM that distorts the EGM resulting in misclassification of the SVT by the morphology algorithm as a VT with subsequent inappropriate shock delivery.
FIGURE 68-2 Morphology error due to clipping. Unclipped electrograms are shown on the left and top. (Left) The top electrogram is the near field channel, the bottom electrogram is far field, and the markers are at bottom. The letters “WV” indicate that SVT is diagnosed by the wavelet algorithm and therapy withheld. The electrogram matches with that of the template, and the rhythm is classified appropriately as SVT. (Right) Shown is the clipped electrogram. Clipping results in distortion of the electrograms, which do not match with that of the template (bottom) despite no physiologic change in the signal itself. It can be prevented by reprogramming the dynamic gain. (Reproduced with permission from Hayes DL, Asirvatham SJ, Friedman PA. Cardiac pacing, defibrillation and resynchronization: a clinical approach. 3rd ed.: Wiley-Blackwell; 2012.)
SVT-VT discrimination algorithms are designed to prevent inappropriate therapy for SVT. Morphology algorithms compare EGMs during tachycardia to a template acquired during normally conducted rhythm. If the tachycardia EGM differs from the template by more than a programmed threshold, the rhythm is classified as VT. Morphology is the only noninterval based single chamber algorithm for SVT-VT discrimination, and it is generally the most accurate. However, misclassification can occur due to truncation of EGM as seen in this case. This can be avoided by adjusting the amplitude gain so that the sensed EGM falls within 25% to 75% of the dynamic range. Other causes of morphology misclassification include:
• SVT with rate related aberrancy. To prevent this, a template can be acquired during atrial pacing at a rapid rate (eg, 120 bpm) and template updates disabled.
• Errors in alignment of EGM.
• Myopotential distortion of the EGM. This is a unique situation in which noise on the far field signal can result in inappropriate shock.
• Changes in morphology over time due to lead maturation or bundle branch block. These can be avoided by periodic automatic updating of the template.
• Recurrent arrhythmia shortly after shock delivery. The EGM is often distorted in the minutes immediately following shock delivery. While morphology is not applied during redetection, if an episode terminates and a new arrhythmia develops, the algorithm may be applied before physiologic recovery of the EGM.
UNNECESSARY VERSUS INAPPROPRIATE SHOCK
Delivery of ICD shock for rhythms other than ventricular tachycardia (VT) or VF is termed “inappropriate shock.” Recent studies have shown that programming the device to delay shocks for nonsustained VT or VF and programming antitachycardia pacing (ATP) can improve outcomes. The term “unnecessary shock” encompasses inappropriate shocks and therapy delivered for VT that would have responded to ATP or self-terminated given time. Optimal programming of ICDs prevents unnecessary shocks.
PROGRAMMING TO PREVENT UNNECESSARY SHOCKS
The detection and treatment of arrhythmia by an ICD involves a series of steps, each of which provides an opportunity to minimize shocks (Figure 68-3).
FIGURE 68-3 Overview of detection and treatment of ventricular arrhythmia by implantable cardioverter-defibrillator (ICD). The detection and treatment of ventricular arrhythmia by an ICD involves a sequence of events that provide opportunities to prevent unnecessary shocks using appropriate programming. (Reproduced with permission from Madhavan et al. Optimal programming of implantable cardiac-defibrillators, Circulation. 2013;Aug 6;128(6):659-672.)
(1) Rate and Duration of Initial Detection
The ICD sense amplifier processes local EGMs to create discrete cardiac events. The time interval between these events defines the heart rate. Tachycardia is detected if the heart rate crosses the detection threshold for a programmable duration or number of intervals. Programming the ICD to delay detection will allow termination of nonsustained VT/VF; programming a higher cutoff rate avoids therapy for slower, better-tolerated rhythms. In contrast to secondary prevention patients, patients with a primary prevention ICD have VT rates that tend to be faster than SVT, so that programming a higher detection rate serves to discriminate VT from SVT, and thus prevent inappropriate detection of SVTs.1
The MADIT-RIT study (Multicenter Automatic Defibrillator Implantation Trial to Reduce Inappropriate Therapy) prospectively randomized 1500 primary prevention ICD recipients to 1 of 3 groups: 1) High-rate therapy (detection at ≥200 bpm with a 2.5-second delay totherapy), 2) delayed therapy (60-second delay at 170-199 bpm, 12-second delay at 200-249 bpm, and 2.5-second delay at ≥250 bpm), and 3) conventional therapy (2.5-second delay at 170-199 bpm and 1-second delay at ≥200 bpm).2 The high-rate and delayed-therapy groups had a lower risk of inappropriate therapy for SVT and death (Figure 68-4) with a similar incidence of syncope compared with conventional programming. These data indicate that programming primary prevention ICDs with a detection zone ≥200 bpm or with delayed therapy at >170 bpm is a preferred strategy. Although nuanced differences in determination of heart rate and duration exists among manufacturers, the fundamental concept of rate and duration triggering detection is similar across devices and therefore is hardware independent and broadly applicable.
FIGURE 68-4 Extension of the intervals to detection and delaying time to detection of VT/VF in the MADIT-RIT trial resulted in reduction in number of inappropriate therapies (A) and mortality (B). (Reproduced with permission from Moss et al. Reduction in inappropriate therapy and mortality through ICD programming, NEJM. 2012; Dec 13;367(24):2275-2283.)
(2) Detection Enhancements: SVT/VT Discrimination
Following initial tachycardia detection, the device utilizes algorithms to distinguish true VT from an SVT. These discriminators are hardware specific and differ between single and dual chamber devices and manufacturers. The algorithms used in single chamber devices are presented in Figure 68-5. Dual chamber algorithms incorporate atrial timing information with ventricular intervals and morphology to improve accuracy. These algorithms differ significantly between manufacturers. While dual chamber algorithms generally have improved accuracy, atrial under- or oversensing can lead to misclassification.
FIGURE 68-5 Single chamber SVT-VT discrimination algorithms. (Reproduced with permission from Madhavan et al. Optimal programming of implantable cardiac-defibrillators, Circulation, 2013;Aug 6;128(6):659-672.)
(3) Optimizing Ventricular Sensing
ICDs have to sense reliably both normal R waves and small fibrillatory waves while avoiding sensing of T waves and extracardiac signals. ICDs utilize dynamic sensitivity or gain wherein the sensitivity is at the minimum after each R wave and progressively increased until a subsequent ventricular event. These prevent sensing of physiological events such as T waves and yet maintain sufficient sensitivity to detect small fibrillatory waves.
T-wave oversensing (TWOS) can occur in the setting of low R wave amplitude, large T waves and long QT interval. TWOS can lead to double counting and inappropriate therapy. Figure 68-6A shows TWOS during sinus rhythm leading to inappropriate shock. In this case the R- to T- wave ratio was sufficiently large to allow correction of the problem by reduction of ventricular sensitivity. This may however increase the risk of undersensing VF. However, decreasing ventricular sensitivity is not advisable if the T-wave amplitude is large in comparison to the R wave as in Figure 68-6B. Changing the sensing vector (dedicated versus integrated bipolar sensing) may sometimes provide a more favorable R to T ratio. Lead revision may sometimes be required.
FIGURE 68-6 (A) Inappropriate shock due to T-wave oversensing. The R- to T-wave amplitude ratio is sufficiently large to allow programming of lower ventricular sensitivity. This was sufficient to prevent shocks in this patient. (B) Inappropriate detection of SVT as ventricular arrhythmia due to T-wave oversensing. In contrast to the tracing shown in Figure 68-6A, the R- to T-wave amplitude ratio is less than 1. Reprogramming of ventricular sensitivity will not be successful in avoiding T-wave oversensing and increases risk of VF undersensing. Hence this patient underwent revision of the ICD lead.
(4) Antitachycardia Pacing
Antitachycardia pacing (ATP) is pacing at a cycle length shorter than VT that terminates VT by depolarizing the excitable gap to block reentry. ATP can reduce shocks, improve quality of life, and lengthen pulse generator life without significantly increasing the risk of VT acceleration or syncope. Although ATP has traditionally been used to terminate slow VT, it has also been subsequently shown to be effective for fast monomorphic VT (>180-200 bpm).3,4 ATP during charge for fast VT reduces the time to shock if ATP is not successful.
(5) Reconfirmation and Noncommitted Shocks
If ATP fails to terminate VT, the device charges to deliver shock. VT may sometimes terminate during the charge. First generation ICDs delivered the shock regardless of arrhythmia termination. Hence the shock was “committed.” Contemporary devices “reconfirm” the persistence of arrhythmia during and after charge. If arrhythmia terminates spontaneously the charge is painlessly dissipated and the shock is “noncommitted.” Shocks should be programmed as noncommitted when possible to avoid unnecessary shocks. The process of reconfirmation is less specific than initial detection and has been reported to result in shocks despite VT termination.5
ICD programming to nonnominal settings prevents unnecessary shocks. Several strategies are available to minimize shocks. The extension of detection intervals and duration, application of SVT-VT discriminators, and use of ATP have been shown to reduce shocks and should be routinely employed.
1. Wilkoff BL, Hess M, Young J, et al. Differences in tachyarrhythmia detection and implantable cardioverter defibrillator therapy by primary or secondary prevention indication in cardiac resynchronization therapy patients. J Cardiovasc Electrophysiol. 2004;15(9):1002-1009.
2. Moss AJ, Schuger C, Beck CA, et al. Reduction in inappropriate therapy and mortality through ICD programming. N Engl J Med. 2012;367(24):2275-2283.
3. Wathen MS, DeGroot PJ, Sweeney MO, et al. Prospective randomized multicenter trial of empirical antitachycardia pacing versus shocks for spontaneous rapid ventricular tachycardia in patients with implantable cardioverter-defibrillators: pacing fast ventricular tachycardia reduces shock therapies (PainFREE Rx II) trial results. Circulation. 2004;110(17):2591-2596.
4. Wilkoff BL, Williamson BD, Stern RS, et al. Strategic programming of detection and therapy parameters in implantable cardioverter-defibrillators reduces shocks in primary prevention patients: results from the prepare (primary prevention parameters evaluation) study. J Am Coll Cardiol. 2008;52(7):541-550.
5. Bernier M, Essebag V. Inappropriate shock despite successful termination of supraventricular tachycardia by antitachycardia pacing during charging. Pacing Clin Electrophysiol. 2010;33(9):e81-83.