Good morning, Dr Andreas Kucher – thank you for your over 10 years activity in our ICD TRAIN events. You are one of the most experienced specialist in programming and troubleshooting in ICD world-wide. Let me ask you a few questions regarding – present and the future of ICD.
What is your opinion about ICD programming according to guidelines? Is it really a good way for us? Maybe we should focus more on programming dedicated for each individual patient? Do you agree that it should be “the State of the Art”?
Implantable cardioverter-defibrillators (ICDs) have proven to be very effective for reducing mortality in patients at significant risk of sudden cardiac death. But also inappropriate ICD activations are frequent in both primary and secondary prevention of SCD. Various clinical studies have been performed in the last years with the goal, to optimize the device programming and to reduce inappropriate therapies. E.g. EMPIRIC (2006), PREPARE (2008), RELEVANT (2009), MADIT-RIT (2012), ADVANCE III (2013), PROVIDE (2013), and DECREASE (2015). In these studies, optimized parameter settings were tested vs. conventional parameters, either based on significantly increased intervention rates, prolonged detection durations or both together. The recently published new HRS Guidelines (Wilkoff et. al: Expert Consensus Statement on Optimal Implantable Cardioverter Defibrillator Programming and Testing, Heart Rhythm 2016;13:e50–e86) provide the newest manufacturer specific programming settings based on the current study statements. I am convinced that high programmed intervention rates in conjunc- tion with increased detection counters can significantly reduce inappropriate and unnecessary therapies, however I have a growing concern that this may happen “on the cost” of life-saving therapies. Over the last years I have frequently seen clinical cases with a very unfortunate patient outcome (more than 50 cases up to now). I can understand the wish to reduce inappropriate therapies. In my experience I can see more and more PDF-files where the device programming was done following HRS guidelines. Apparently an increasing number of cardiologists have become followers of the guidelines. As long as no episodes have been recorded, everything appears to go well. On regular basis a case with a fatal outcome is presented to me and the question arises why doctors do not see more of these cases?
It seems to be comparable with the iceberg theory where only a small part is visible, but a large part is not. In my experience, we obtain only information of such cases under the following three conditions:
1. If the patient died in the hospital (only then the physician has a chance to interrogate the ICD with the programming device).
2. If the patient had an OHCA, and if he was successfully resuscitated by the emergency doctors, so that he could be re-admitted in the hospital.
3. If the patient has died at night time in his bed, in the near vicinity of his Cardiomessenger and Home-monitoring was activated.
The probability for scenario 1 is extremely low, as it is for scenario 2. Approx 50% of the ICD-patients are connected to HM and not all patients die in their bed which means that the RF-connectivity to the Cardiomessenger is then not ensured. All collected cases with a fatal outcome are rather random observations and the absolute number of undiscovered cases is assumed to be much higher. Therefore, I advocate an obligatory post-mortem device interrogation (requested by law?) Once the patient has died, the attending cardiologist has typically no chance to position the programming wand anymore over the ICD of the deceased. These undiscovered fatal cases is certainly a major concern in the MADIT-RIT study.
The investigators could also not perform a post-mortem device interrogation in 75% of the 71 deceased patients. Even for the remaining 25% of interrogated ICDs after death, there is no conclusive proof based on absense of EGM data documenting the lethal arrhythmia. Because, the higher the programmed cut-off rate, the higher is the risk that the deadly tachycardia was neither detected nor treated, and most likely not even stored in the ICD memory. The physician remains unaware and is not able to conclude, that it was presumably wrong programming which inhibited therapy leading to patient death.
I am wondering why no case reports could be found in the journals over the last years (probably due to the iceberg effect?). In January this year, Montgomery et al. have published the first article (to my knowledge) with the title: „Sudden cardiac death despite a functional cardioverter-defibrillator: The case for early and agressive therapy for ventricular tachycardia in selected patients.” (J Cardiovasc Electrophysiol, Vol. 27, pp. 120-124, January 2016). The authors reported about three cases within 11 months at a single institution of sustained VT that fell below the programmed detection rate of the patients’ ICD, two of which continued until converting to an agonal VF that did not meet criteria for detection, and a third case that could not be successfully defibrillated after a prolonged period of VT.
Regarding your question, I believe that we should be careful with the guideline approach. Personally I would always prefer a patient-tailored programming. The doctor knows his patient best.
My friend, Prof Serge Barold said regarding this topic: “Luckily, the physician has always the liberty, to decide himself.”
My first impression regarding the manufacturer specific programming in the HRS guidelines was that it should apparently be the goal, to offer one „uniform” program (one setting for all). Contrary to this, I think today, that a patient-tailored programming must be taken into our considerations as the “Golden Standard”. I am sure that an individual device programming is the decisive factor in the success of ICD therapy.
A 30 years-young patient (sportive/active) needs certainly another setting for primary prevention, than a 75 years old patient with an already worsened cardiopulmonary disease, who is not able anymore to move away from the TV chair. Can we all agree upon that?
In your very long clinical experience –are there more and more problems with inappropriate shocks? Or leads dysfunctions? Or others?
These are three questions.
a) More inappropriate shocks? No. The ICD developers have done a good job when they have introduced the new technology such as sensing amplifiers, new filter techniques and a new analog-digital converter. Since then, we do not see the typical problems anymore like T-wave oversensing or QRS double counting and other amplifier- or filter related problems. The new technolo- gy significantly reduced the number of inappropriate shocks. If we see inappropri- ate shocks, they are mostly linked to inappropriate SVT discrimination. I have high expectations that MorphMatch criteria will certainly help to also reduce these observations.
b) Lead problems? Yes. If I would have to draw a diagram with the percentages of troubleshooting queries, then the electrode-related problems would be almost the largest portion. But it is always a question of perspective, if we look into worldwide statistics, we observe that lead problems have a low incidence compared with the total number of implanted systems. Still, lead problems are always a serious matter because they always require a surgical intervention.
c) Under the rubric “others” I would rather classify not optimal programming of the ICD parameters.
S-ICD is one of the newest technology introducing in our clinical practice. However – S-ICD without ATP possibilities – should be really recommended in primary prevention in very high class IIA recommendations. Prevention of sudden cardiac death is one of the most important tasks of cardiology. Transvenous ICD-systems have impressively proven their effectiveness in numerous randomized trials. Transvenous ICD systems have their apparent limitations related to long-term lead complications. The S-ICD seems to be an interesting alternative for a selected group of patients in whom no pacemaker activity is needed (neither antibradycardia pacing nor ATP treatment). An important limitation of the S-ICD is, that it is not designed to treat ventricular arrhythmias at rates lower than 170 bpm. My experience is rather low in this field of S-ICD, as Biotronik does not offer S-ICDs in its product portfolio. But many case reports are meanwhile published, which show that the subcutane- ous signal sensing and VT/VF detection is not that easy/successful. The recorded EGM tracings are similar to far-field EGMs due to the distance of the sensing poles of approx. 10 cm and the large distance of the sensing poles to the active can. A far-field detection has some limitations and might be prone for several sensing problems, which can be small signal amplitudes during VT/VF (under- sensing), T-wave sensing (oversensing), QRS double counting (oversensing), myo-potentials (oversensing) and noise detection (oversensing). Apparently, the sophisticated sensing algorithms (INSIGHTTM Algorithm) of the S-ICD can currently not safely avoid such problems.
Burke et al. reported about the 2-Year Safety and Efficacy of the S-ICD in JACC VOL . 65, NO. 16, 2015, APRIL 28, 2015:1605 – 15. This report mentions the causes of inappropriate shocks. Inappropriate shocks were mainly caused by supraventricular arrhythmia (SVT) above the discrimination zone in 24%, T-wave oversensing in 39% and low amplitude signals in 21%. Frommeyer et al. published a letter (2015) to the editors of Clin Res Cardiol: “Limitations in S-ICD therapy: reasons for system explantation.” They reported about 6 cases, where 2 had pocket infections, 1 patient with a combined system (S-ICD and an AAI pacemaker) who developed then a slow VT of 140 bpm; the case 4 patient had episodes of unexplained syncopes and an ILR revealed sinus arrest; the case 5 patient showed inappropriate shocks due to T-wave and P-wave oversensing; and the case 6 patient was observed with ineffective shocks due to an insufficient DFT safety margin. All six S-ICDs had to be replaced by transvenous ICD systems.
Maybe that possible candidates for an S-ICD could be patients with structural heart disease, patients who lack venous access for transvenous lead placement, patients with channelopathies that confer risk of sudden cardiac death and who do not need anti-bradycardia pacing, some patients awaiting cardiac transplantation, and those primary prevention patients who are best treated without a transvenous lead due to tricuspid valve concerns or a previously infected transvenous system. S-ICD is – at the moment – certainly no alternative to the transvenous system due to limited programming options and the lack of stimulation, however, it is an interesting supplement of ICD therapy for a limited number of candidates. Medtronic has recently announced a new EV-ICD System, which currently is in development and not yet available. The EV-ICD system is a new approach to ICD therapy with leads placed outside the heart and veins, but under the rib cage and the leads substernal.
During ESC congress a DANISH study was presented, which shows, that there is no reduction in mortality in primary prevention in non-ischemic cardiomyopathy. Are you not afraid, that in current modern cardiology – with optimal pharmacological treatment (Statins, B-blockers, ACEI inhibitors) and a wide range of possibilities offered by today’s vascular intervention – “MADIT population” just does not exist anymore and as a result ICD won’t be used in a primary prevention of SCD in the future?
Indeed, the DANISH trial revealed interesting results. The authors figured out that prophylactic ICD implantation in patients with symptomatic systolic heart failure not caused by coronary artery disease was not associated with a significantly lower long-term rate of death from any cause than was usual clinical care. This new finding is certainly a reason to restart the debate about prohylactic ICD indications. The long-term data from the MADIT- II trial indicate that clinical risk scores are used to identify the patients with ischemic heart disease who will benefit most from ICD implantation. Such a method for identifying patients with nonischemic heart failure who are at high risk for death from arrhythmia would be very useful.
Perhaps, this study has somehow confirmed, what we have already thought in the last years. I think that there is still enough room for ICD implantations for primary prevention of SCD. The criteria for the primary prevention however, must be redefined. There is a necessity to differentiate better between the patients with symptomatic systolic heart failure from those with an ischemic heart disease. This will be a challenging task for the future.
Even with the previous indicators for the prophylactic ICD implantation one can have doubts whether an old moycadial infarction and a reduced LVEF < 30% as the alone criteria are specific enough to figure out only those patients with the highest risk for SCD. If we compare in prophylactic ICD patients the number of patients with an empty episode list (free of tachyarrhythmic events) versus the number of appropriate therapies in the presence of a real VT/VF and versus the number of inappropriate shocks, the finding is a disappointing disbalance. The ICD group in the DANISH trial has also shown a higher percentage of inappropriate shocks (5.9%) compared with the percentage of SCD (4.3%).
It is an assumption that an increasingly number of prophylactic ICDs has been implanted over the last years and subsequently more patients have suffered from side effects, in comparison to those patients who have really profited from their prophylactic ICD.
Further research is needed, perhaps, one or two further indicators would be useful in addition to the old MI and LVEF < 30%, in order to improve the outcome. Programming higher intervention rates and a longer detection duration might certainly help to reduce inappropriate therapies, but herewith is also associated an increased risk of underdetection and non-treatment of malignant VTs, and one might overlook a VT because IEGMs will not be stored if they occur below the programmed intervention rate.
Additionally if we are still increasing VT/VF zones – it would be more and more difficult to prove reduction in mortality after ICD implantations.
Please allow me to put some sarcasm into the answer, if someone would create a new study design, where a VT zone is programmed at 250 bpm and a VF zone at 300 bpm, I am sure that the number of inappropriate shocks can be significantly reduced to almost 0 %. The survival of SCD is another question. The overall mortality could be assessed with the number of patients who are lost for follow-up. But the reason of death remains unclear, because the life-threatening VT/VF is not stored in the memory of the ICD. One could classify the death as a non-arrhytmic death, although this diagnosis is definitely wrong.
In your opinion- what are really clinical benefits of Home Monitoring – for which pts? – ICD? CRT? Or both? In which pts HM should be obligatory?
HM is certainly a benefit for most of the patients irrespective whether with an ICD or CRT-D. The special advantage is the Early Detection. Early Detection of both technical peculiarities (hardware or software problems) as well as unexpected clinical changes.
I remember very well that we could figure out the beginning of potential lead problem which gave us the opportunity for a timely intervention, before any inappropriate therapies have occurred.
Especially in patients with a CRT-D, the trend curves of the heart rate at rest and the complete pacemaker statistics enable to assess the CRT success.
I am always glad if immediate and detailed HM information is available for troubleshooting questions. From the perspective of technical service queries, I would appreciate if much more patients with cardiac implantable electronic devices (CIED) would have been connected to the HM system. This is valid for all devices ICDs, CRT devices and PMs. I see major benefits for the devices with resynchronization therapy.
CRT– we proved that it reduces mortality – but we are not able to achieve more than 60% of responders. Which algorithms should be developed to improve this result?
The success of the resynchronization therapy is depending on various factors. Here I want to mention the implantation/lead handling and the optimal programming. My impression is that in too many patients the LV electrode is not optimally positioned in the area of the latest electrical LV activation. The LV lead is probably not on the best place if the interventricular conduction delay – measured between the RVs and the LVs event – is not sufficient Occasionally the LV-lead might be positioned unintendedly in the near vicinity of scar tissue, due to a former myocardial infarction in that region, which can result in a latency or a locally slower conduction. Quattro-polar LV-leads and the CRT-Opt algorithm could be new approaches to overcome such limitations.
I know, that since beginning You are great enthusiastic of DX system. Do you think that in chronic AF pts such ICD might be useful?
That is correct. Due to the atrial information, the DX system allows us much better to figure out the true origin of an unknown tachyarrhythmia, than with the limited possibilities of a standard single-chamber ICD. I think, that the diagnostic advantage of the atrial EGM channel in a DX device remains valuable.
Thank for your comments, education for us and over 10 years of participa- tion in ICD TRAIN unique project – I think one of the longest ICD education activity.
Thank you for the opportunity to speak about all topics which are of outmost interest to me and were I like to share my opinion. I thank you for the good relationship and the many years of trust, I have learned a lot more from the doctors who have shared the most interesting cases with me, it was always a two-sided learning. I am looking forward to further cooperation in the coming years.