A Limitation of Tachycardia Discrimination Algorithm in a Patient with Hypertrophic Cardiomyopathy and Intermittent Complete Heart Block

A 30-year-old man presented to the outpatient department in early 2004 with history of three episodes of syncope. He had a family history of sudden cardiac death. His father and paternal uncle had died suddenly at the age of 38 and 40 years, respectively. An electrocardiogram revealed left ventricular hypertrophy. On evaluation, he was found to have nonobstructive hypertrophic cardiomyopathy. He had two brothers, aged 25 and 32 years. They were both asymptomatic. On echocardiographic screening, they were found to be normal. They are on regular follow-up. He underwent single-lead implantable cardioverter defibrillator (ICD) (Gem 7227, Medtronic Inc., Minneapolis, MN, USA) in 2004. He received both appropriate and inappropriate shocks over the next 3 years. There was only one episode of ventricular tachycardia (VT) at a cycle length of 370 ms, which had received appropriate therapy. There were two episodes when he received inappropriate therapy for sinus tachycardia (ST). In 2007, he had significant exertional fatigue and was noted to have intermittent complete heart block and the device started to pace at the backup rate of 40/minute. The patient was receiving metoprolol 25 mg twice a day which was stopped and his ventricular pacing rate was increased from 40 to 60/minute. Subsequent electrocardiogram revealed left bundle branch block with prolonged PR of 240 ms. An echocardiogram at this time revealed a left ventricular outflow tract obstruction (LVOTO) of 20 mmHg at rest with no provokable increase in gradient. In view of the inappropriate shocks and intermittent complete heart block, the device was upgraded, in 2007, to a dual chamber device (EnTrust D154ATG, Medtronic Inc.). Detection was programmed to two zones (VT 150- 188 and VF >188 with antitachycardia pacing ATP during charging). Measured P/R wave were 3.3/14.2 mV. P/R sensitivity were programmed at 0.3/0.3 mV respectively. PR Logic™ was programmed on for supraventricular tachycardia (SVT)/VT discrimination. Enhancements for atrial fibrillation and ST were programmed on at the time of implant. His symptoms of exertional fatigue resolved. The algorithm for minimizing ventricular pacing (Managed Ventricular Pacing MVP) was programmed on. He was pacing the ventricle 70% of the time. In April 2008, he received ATP therapy from the device (Figs. 1A,B, 2). Is the therapy appropriate? If not, what is the cause and what measures should be taken? From top to bottom: atrial, ventricular near-field, and marker channels, respectively. (A) Sinus tachycardia/supraventricular tachycardia with high-degree AV block with intermittent pacing (due to MVP programmed on). Sudden spontaneous 1:1 conduction leads to ventricular rate in the VT detection zone. Note that the ventricular EGMs after resumption of 1:1 AV conduction are the same as the sensed ventricular EGMs prior to 1:1 AV conduction. (B) Second sequence of ATP delivered. No effect on rhythm of ATP. The heart rate gradually slows down. Interval plot shows atrial rate in the VT zone and ventricular rate initially slow and irregular and then suddenly becomes regular with the atrial and ventricular rate becoming the same. There was detection of tachycardia and two sequences of ATP therapies are delivered. The heart rate gradually slows to below the VT detection zone. Spontaneous complete heart block (unrelated to Alcohol Septal Ablation) is rare in hypertrophic cardiomyopathy, with very few cases reported in literature. Some of these cases have been familial. The cause is more or less unknown with one case report describing discontinuity of the bundle of His as it crosses the central fibrous body and another case report describing interstitial fibrosis and myocardial necrosis of the conduction system.1,2 Complete heat block is also known to occur in the PRKAG2 phenotype.3 It is a storage disorder and diagnosis can be made with biopsy and genetic studies. A biopsy was not done in our case. In Figure 1A, atrial cycle length was 360 ms. In the initial part, there was ST with high-degree atrioventricular (AV) block and intermittent ventricular pacing. Then suddenly there was one to one AV conduction with prolonged PR interval. It is worth noting that there was no change in the ventricular electrogram (EGM) after the initiation of 1:1 AV conduction. This further confirms that the rhythm was supraventricular and the interval plot (Fig. 2) suggested gradual offset, suggesting it was likely to be sinus. However, the possibility of an atrial tachycardia (AT) with variable block cannot be completely ruled out. In Figure 1B, the second sequence ATP therapy was delivered. There was no change in rhythm after ATP therapy and the rate gradually slowed down. The EnTrust ICD uses the Adaptive ST rule for ST arm of PR Logic.4 The same rule also applies for ST and SVT with 1:1 conduction with atrial EGM not falling in the junctional zone (50 and 80 ms on either side of ventricular EGM, respectively). The device incorrectly identified the rhythm as VT and delivered ATP therapy. The automatic detection and discrimination algorithm detected tachycardia in VT zone as soon as the AV conduction became 1:1. Due to the sudden transition to 1:1 conduction, the “recent history” was not consistent with 1:1 AV conduction and the device interpreted the rhythm as “not consistent with ST” and hence as VT. ST is responsible for a sizable percentage of inappropriate shocks. The earlier ICD models of Medtronic, for example, Gem DR, used the original ST rule for differentiating ST from VT.5,6 However, ST with prolonged PR and ST/AT with intermittent far-field R wave oversensing were responsible for 56% of inappropriate shocks with the original ST rule.5 Subsequently, the rule was modified to have a programmable ST/VT boundary.7 Gem III DR and Marquis DR used this modified “Enhanced ST rule.” It had its own limitation with underdetection of VT with retrograde 1:1 conduction. This algorithm was replaced with the “Adaptive ST rule.” This rule looks for 1:1 AV conduction and checks whether the RR and PR fall in the expected range and uses these data to generate a cumulative ST evidence counter. ST with prolonged PR was no longer an issue with the Adaptive ST rule. This rule has a similar sensitivity with improved specificity compared to the original ST rule. However, sudden onset AT, high density of atrial ectopic, AV Wenckebaching in the VT detection zone may be incorrectly classified as VT with this rule. In our case, high-degree AV block during ST in the VT detection zone, suddenly conducting 1:1 led to incorrect classification as VT. We would like to point out that during 1:1 conduction the PR was prolonged and would also have been classified inappropriately as VT by the original ST rule. This also underlines the need to incorporate wavelet in the dual chamber algorithms for VT/SVT discrimination. In view of documented VT at a cycle length of 370 ms, we did not reprogram the VT zone. We restarted β-blockers and increased the dose to further impair the AV conduction. Ventricular pacing was not an issue in this case as the underlying disease was hypertrophic cardiomyopathy with mild LVOTO. He has not received any inappropriate therapy since then.