Controversies in cardiology: Should patients with asymptomatic ventricular tachycardia be prophylactically treated with an implantable cardioverter-defibrillator?

Summary

For years, nonsustained ventricular tachycardia in the setting of structural heart disease has been identified as an independent predictor of sudden death. The MADIT trial showed a clear superiority of prophylactic defibrillator (ICD) compared to conventional drug therapy in patients with nonsustained ventricular tachycardia after myocardial infarction. This raised the clinical issue whether prophylactic ICD therapy should generally be recommended in high-risk patients with nonsustained ventricular tachycardia. The present article discusses the MADIT findings in the light of other randomised ICD trials with special focus on their applicability to a general infarct population and the question of how to safely identify patients at risk for sudden death.

Zusammenfassung

Nichtanhaltende Kammertachykardien (KT) bei zugrundeliegender struktureller Herzkrankheit wurden seit langem als unabhängiger Risikofaktor für einen plötzlichen Herztod erkannt. Die MADIT-Studie, die bei Patienten mit nichtanhaltenden Kammertachykardien nach Infarkt nun eindeutig die Überlegenheit eines implantierbaren Cardioverter-Defibrillators (ICD) im Vergleich zur konventionellen medikamentösen Therapie gezeigt hat, brachte die brennende Frage auf, ob die prophylaktische ICD-Therapie nicht generell bei Hochrisikopatienten mit nichtanhaltenden Kammertachykardien zu bevorzugen sei. Der vorliegende Artikel diskutiert die MADIT-Resultate im Licht anderer, randomisierter ICD-Studien mit speziellem Fokus auf deren Anwendbarkeit auf eine generelle Infarktpopulation und die Frage, wie Risikopatienten für einen plötzlichen Herztod sicher identifiziert werden sollen.

Résumé

La survenue de salves non soutenues de tachycardie ventriculaire en présence d'une cardiopathie constitue un marqueur indépendant du risque de mort subite. L'étude MADIT a montré récemment, chez les patients présentant des tachycardies ventriculaires non soutenus après infarctus du myocarde, une nette supériorité du défibrillateur implanté à titre prophylactique par rapport au traitement médical conventionnel. Cette étude pose donc la question de savoir si un défibrillateur prophylactique ne devrait pas être implanté de manière systématique chez les sujets à haut risque avec tachycardie ventriculaire non soutenue. Le présent article se propose de discuter les résultats de l'étude MADIT à la lumière des autres travaux randomisés concernant le défibrillateur interne, en essayant d'une part de définir la place de cette nouvelle thérapie dans une population globale post-infarctus et d'autre part de déterminer la meilleure façon d'identifier les patients à haut risque de mort subite.

Introduction

Identification and treatment of patients at risk for sudden death is certainly one of the major challenges for cardiologists today. For many years now, the presence of asymptomatic nonsustained ventricular tachycardia (defined as the presence of three or more consecutive complexes of ventricular origin at a rate greater than 100 beats per minute, but lasting less than 30 seconds) has been recognised as a hallmark for future serious arrhythmic events (sustained ventricular tachycardia or sudden death), especially in patients after myocardial infarction or presenting with various types of structural heart disease. The prognostic value of asymptomatic nonsustained ventricular tachycardia has been shown to be independent of other clinical variables such as left ventricular function for example [1,2,3].

The clinical and electrocardiographic spectrum of asymptomatic nonsustained ventricular tachycardia is relatively wide, including repetitive runs of strictly monomorphic ventricular tachycardia in the absence of structural heart disease, "torsades de pointes" and monomorphic or polymorphic nonsustained ventricular tachycardia in the presence of structural heart disease. In the present paper, only the last entity will be discussed with special regard to the question whether or not an ICD should be implanted in such patients. Nonsustained ventricular tachycardia in normal hearts carrying a benign prognosis, and “torsades de pointes“ having in most cases a correctable cause should not be treated with an ICD.

Prognosis of asymptomatic nonsustained ventricular tachycardia

Table 1. Prognostic significance of nonsusrained ventricular tachycardia. HD = heart disease post MI = post myocardial infarction DCM= dilated cardio-myopathy HCM = bypertrophic cardiomyopathy nonsust. VT= nonsusrained ventricular tachycardia + = present − = absenr.

author	year	type of HD	number of patients c/pos. VT	mortality (%)
				nonsust. VT+	nonsust. VT−
Anderson et al. [1]	1978	post MI	915	16%	8%
Kleiger et al. [4]	1981	post MI	289	17%	6%
Bigger et al. [5]	1981	post MI	430	54%	19%
Bigger et al. [2]	1984	post MI	766	25%	6%
Denes et al. [6]	1991	post MI	755	15%	8%
Maggioni et al. [7]	1993	post MI	8676	5%	3%
Huang et al. [8]	1983	DCM	35	14%	7%
Olshausen et al. [9]	1988	DCM	73	36%	5%
Unverferth et al. [10]	1984	DCM	69	45%	0%
Doval et al. [11]	1996	post MI/DCM	516	50%	31%
Savage et al. [12]	1979	HCM	100	10%	0%
McKenna et al. [13]	1981	HCM	86	21%	3%
total			12710	11.9%	4.3%

summarises data from the literature concerning the prognostic significance of asymptomatic nonsustained ventricular tachycardia detected during ambulatory monitoring in various types of heart disease. The presence of nonsustained ventricular tachycardia clearly identifies a group of patients at high risk of serious arrhythmic events or sudden death not only after myocardial infarction but also in dilated or hypertrophic cardiomyopathy. Overall mortality in this group is almost three times higher than in patients without ventricular tachycardia (11.9 versus 4.3%). However, the prevalence of asymptomatic ventricular tachycardia ranges from 1 to 22% in patients after myocardial infarction and from 50 to 100% in patients with dilated cardiomyopathy. Thus, the positive predictive value of the presence of nonsustained ventricular tachycardia alone is relatively low and identification of high-risk patients requires additional information such as ventricular function, the presence/absence of ventricular late potentials during signal-averaging, preservation/reduction of heart rate variability and arrhythmia inducibility/non-inducibility during programmed ventricular stimulation.

Risk stratification in patients with nonsustained ventricular tachycardia

Table 2. Prognostic value of various noninvasive tests to identify patients at high risk of serious arrhythmic events and/or sudden death after myocardial infarction. LP = presence of ventricular late potentials during signal-averaging EF = reduced left ventricular ejection fraction Holter = presence of nonsustained ventricular tachycardia during Holter monitoring HRV = reduced heart rate variability (ref. [14,15,16,17,18,19,20]).

	sensitivity	specificity	positive predictive value	negative predictive value
LP + EF	25–100%	59–94%	19–37%	94–100%
LP + Holter	65–100%	45–89%	27–35%	99–100%
EF + Holter	33–92%	44–93%	19–37%	93–94%
LP + EF + Holter	20–100%	53–97%	28–50%	97–100%
HRV + LP + Holter	29%	99%	58%	95%

summarises results obtained by several noninvasive techniques used in combination to identify high-risk patients after myocardial infarction [14,15,16,17,18,19,20]. The negative predictive value of all these tests is relatively high but the positive predictive value remains low whatever combination is used. Programmed ventricular stimulation has been used to further refine risk stratification in patients with asymptomatic nonsustained ventricular tachycardia. This invasive approach has proven useful in postmyocardial infarction patients but questionable in patients with dilated or hypertrophic cardiomyopathy (

Table 3. Results of programmed ventricular stimulation in patients with nonsustained ventricular tachycardia. sust. mono. VT = sustained monomorphic ventricular tachycardia CAD = coronary artery disease.

	year	number of patients	population	% inducible sust. mono.VT	follow-up (months)	number of arrhythmic events
						inducible	non-inducible
Zheutlin et al. [21]	1986	88	mixed	37%	22	4/33 (12%)	0/55 (0%)
Buxton et al. [22]	1987	62	CAD	45%	28	7/28 (25%)	4/34 (12%)
Klein et al. [23]	1989	40	CAD	43%	14	9/17 (53%)	0/23 (0%)
Turitto et al. [24]	1990	90	mixed	24%	30	4/22 (18%)	5/68 (7%)
Hammill et al. [25]	1990	110	mixed	19%	15	2/21 (10%)	4/89 (3%)
Kowey et al. [26]	1990	205	CAD	33%	18	12/68 (18%)	27/137 (20%)
Wilber et al. [27]	1990	100	CAD	37%	16	8/37 (22%)	2/63 (3%)
Manolis et al. [28]	1990	40	CAD	12%	21	0/5 (0%)	1/35 (3%)

). Patients at highest risk are those with a low left ventricular ejection fraction and inducible, non-suppressible sustained ventricular tachycardia during programmed ventricular stimulation. In this subgroup, mortality at 2 years is almost 50% according to Wilber et al. [26].

Management

The treatment of patients presenting with asymptomatic ventricular tachycardia remains controversial, although data from several recent large scale randomised trials have helped clinicians to find the right solution for selected cases. Findings from the CAST trial [27] have shown that class Ia antiarrhythmic drugs were associated with an increased cardiac mortality in coronary artery disease patients at risk for sudden death. In the CAMIAT [28] and EMIAT [29] trials, amiodarone has been shown to reduce arrhythmic death but not total mortality in high-risk patients after myocardial infarction. However, it has to be mentioned that these two studies were not designed (CAMIAT) or underpowered (EMIAT) to detect a modest but still important reduction in total mortality. In the recently published ATMA trial [36], a meta analysis of all large scale amiodarone trials analysing the results of more than 6500 individual patients' files to investigate the effect of prophylactic amiodarone in MI and heart failure patients, amiodarone clearly reduced total mortality by 13%. Also, betablockers have shown a consistent reduction in total mortality mainly by reducing cardiac and arrhythmic death in post myocardial infarction patients with or without ventricular premature beats, with the greatest benefit in patients with depressed left ventricular function [30].

Based on the poor results obtained by antiarrhythmic drugs on the one hand, the high efficacy in preventing arrhythmic death by the implantable cardioverter-defibrillator (ICD) on the other hand, and the possibility to identify high-risk patients by the various tests cited above, several large scale trials looking at the effect of prophylactic ICD therapy have been conducted. One of those was the MADIT trial [31], which showed that implantation of an ICD could save lives in high-risk patients after myocardial infarction with left ventricular dysfunction ( ejection fraction <35 % ), nonsustained ventricular tachycardia and inducible non-suppressible sustained ventricular tachycardia during programmed ventricular stimulation. In total 196 patients were randomised to either prophylactic ICD implantation (n = 95) or conventional antiarrhythmic therapy (n = 191). During an average of 27 months, there were 15 deaths in the ICD group and 39 deaths in the conventional therapy group (most of them being treated with amiodarone), resulting in a 54% mortality reduction in the ICD group. Apart from survival benefit obtained by the ICD, this study clearly demonstrated the power of electrophysiological testing to identify patients with ICD-preventable death. MAD IT was the first randomised primary prevention trial to demonstrate such a dramatic survival benefit from prophylactic ICD therapy. However, it should be stressed that the MADIT results are only of limited clinical impact, since they apply to a highly selected patient population only. In fact, Andresen et al. [32] have recently tried to evaluate the clinical significance of the MAD IT study by looking at the question, to how many infarct survivors the MADIT inclusion criteria would apply; they found that less than 0.2 % of an unselected population after myocardial infarction does fulfil the MADIT selection criteria, and thus, may benefit from prophylactic ICD therapy.

This important limitation raises the question, whether the results of MADIT are applicable to other patient populations at all?

The AVID trial [33] has clearly demonstrated that survivors of ventricular fibrillation or poorly tolerated ventricular tachycardia have a better prognosis with ICD than with antiarrhythmic drugs (reduction of mortality of 39% at one year and 31 % at 3 years). However, in contrast to MADIT, the AVID trial was a secondary prevention trial in high-risk patients who have clinically proven to suffer from life-threatening sustained arrhythmia. In the other primary prevention trial, the CABG-PATCH trial [34], there was no benefit of prophylactic use of an ICD in patients at high risk for ventricular arrhythmias (left ventricular ejection fraction <36% and an abnormal signal-averaged ECG) after coronary artery bypass graft surgery.

These results leave the question open whether nonsustained ventricular tachycardia alone or in combination with other criteria is an appropriate selection criterion to identify patients who should receive prophylactic ICD therapy? It can be argued that the presence of inducible non-suppressible ventricular tachycardia in the setting of poor left ventricular were the true indicators of high risk in the MAD IT population, and that nonsustained ventricular tachycardia, though an official inclusion criterion, may have played a secondary role with regard to risk stratification. In that respect it would be interesting to see, what a similar trial to MADIT, including post-MI patients with bad ejection fraction and inducible non-suppressible ventricular tachycardia (but without spontaneous non-sustained ventricular tachycardia) would have looked like?

To answer such and other questions, several prospective, randomised studies are currently in progress. They will certainly redefine the indications for prophylactic ICD use in high-risk patients [35]:

a)

The Cardiomyopathy Trial (CAT) evaluates the benefit of an ICD in patients with dilated cardiomyopathy, a left ventricular ejection fraction <30% and asymptomatic ventricular arrhythmias.

b)

The Sudden Cardiac Death in Heart Failure Trial (SCD-HeFT) will randomise patients with a left ventricular ejection fraction <36%, NYHA functional class II or III, and nonsustained ventricular tachycardia, to placebo, amiodarone or ICD implantation.

c)

The MADIT-2 Trial is designed to randomise patients after myocardial infarction with a left ventricular ejection fraction <30% and >10 premature ventricular complexes per hour or couplets, either to ICD or conventional therapy.

d)

The Cardiac Arrest Survivors Hamburg Trial (CASH) is currently evaluating the benefit of ICD implantation versus amiodarone and beta-blockers in cardiac arrest survivors. Initially, patients were also randomised to propafenone, but this arm has been discontinued after a few months because of an excess of mortality with this agent, emphasising the deleterious effect of class Ic antiarrhythmic drugs in patients with severe arrhythmias.

e)

The Multicenter Unsustained Tachycardia Trial (MUSTT) will examine the strategy of electrophysiologically guided therapy versus conventional therapy in post myocardial infarction patients with a left ventricular ejection fraction <40% and nonsustained ventricular tachycardia. This trial is not designed to specifically compare ICD therapy to conventional treatment; however, it will certainly provide important information with regard to the efficacy of these two forms of therapy in high-risk patients early after myocardial infarction.

f)

Finally, the Defibrillator IN Acute Myocardial Infarction Trial (DINAMIT) will randomise 525 infarct survivors with an ejection fraction <35% and depressed heart rate variability to either prophylactic ICD or no active treatment. Endpoint is all cause mortality.

Conclusion

Should every asymptomatic patient with nonsustained ventricular tachycardia be prophylactically treated with an ICD? The answer is certainly: no. On the other hand, implantation of an ICD in patients after myocardial infarction who have nonsustained ventricular tachycardia, a left ventricular ejection fraction of less than 35% and inducible non-suppressible ventricular tachycardia during programmed ventricular stimulation has clearly proven to save lives. However, the major drawback of this stringent indication is that it applies to only 0.2% of all infarct survivors leaving the far greater proportion of patients unrecognised at risk. This situation emphasises both, the need of better risk stratification in post MI patients and also the need of a low cost ICD allowing to widen its prophylactic use not only from the patient's perspective, but also from a cost-effectiveness point of view. At the moment, specific recommendations for other high-risk patients should await publication of ongoing trials.