Europace Advance Access published June 10, 2015
CLINICAL RESEARCH
Europace doi:10.1093/europace/euv188
Incidence of and predictors for appropriate implantable cardioverter-defibrillator therapy in patients with a secondary preventive implantable cardioverter-defibrillator indication Beat Schaer*, Michael Ku¨hne, Tobias Reichlin, Stefan Osswald, and Christian Sticherling Department of Cardiology, University Hospital of Basel, Petersgraben 4, Basel 4031, Switzerland Received 13 February 2015; accepted after revision 1 May 2015
Aims
Incidence of implantable cardioverter-defibrillator (ICD) therapy in secondary prevention has been assessed in randomized trials and registries. However, results are considerably limited by short follow-up and hazy definition of treated arrhythmias. This study aimed to determine appropriate ICD therapy and to define predictors based on registry patients followed for up to 20 years. ..................................................................................................................................................................................... Methods All patients with a secondary prevention indication and ischaemic or dilated cardiomyopathy were identified. Arrhythmic endpoints were appropriate ICD therapies for any ventricular tachycardia (VT) .175 b.p.m. and appropriate ICD and results therapies in the ventricular fibrillation (VF) zone of .220 b.p.m. (potentially life-threatening). Predictors were determined by analysing 19 baseline characteristics. We included 357 patients, age 65 + 11 years, predominantly male (89%) with ischaemic cardiomyopathy (83%). During follow-up of 82 + 53 months, 156 (44%) patients died and 208 received any form of ICD therapies (59%), 71 of them (34%) in the VF zone. Forty-four patients (28%) died without experiencing any form of appropriate ICD therapy. Cumulative incidence of any form of ICD therapy at 10 years was 65%. Predictors for any form of ICD therapy were implantation for VT and age [VT: hazard ratio (HR) 1.45, 95% confidence interval (95% CI) 1.05 – 2.01, P ¼ 0.03; age (per year): HR 1.02, 95% CI 1.01 – 1.04, P ¼ 0.001]. For therapy in the VF zone, univariate analysis determined male gender (29 vs. 5%, P ¼ 0.01) as predictor. ..................................................................................................................................................................................... Conclusion The rate of appropriate ICD therapies in secondary prevention is high. No useful predictors for them, especially not for life-threatening arrhythmias could be identified.
----------------------------------------------------------------------------------------------------------------------------------------------------------Keywords
Implantable cardioverter-defibrillator † Predictor † Secondary prevention † Cumulative incidence
Implantable cardioverter-defibrillator (ICD) placement is a Class IA indication for survivors of cardiac arrest due to ventricular fibrillation (VF). Indication is Class IB in patients who present with sustained ventricular tachycardia (VT) and for those with syncope and inducible VT.1 These recommendations are based on randomized controlled trials.2,3 Implantable cardioverter-defibrillator implantation was labelled as ‘appropriate use’ in the recent Appropriate Use Criteria Task Force report.4 For such a secondary prevention setting, several cost-effectiveness calculations have been undertaken summarized in a review article by Boriani et al. published
in 2001.5 They all showed costs per life-year saved between US $12 000 and 54 000. The most recent analysis in a secondary preventive ICD population performed in 2001, using a Markov model, was mitigated through the inclusion of primary prevention trials as well.6 As then, corresponding research has focused on primary prevention or cardiac resynchronization therapy (CRT). The use of ICD [i.e. firing or antitachycardia pacing (ATP)] has been studied in registries7 – 11 but published results are considerably limited by short follow-up, inclusion of additional primary prevention patients or inhomogeneous definition of life-threatening
* Corresponding author. Tel: +41 61 328 62 22; fax: +41 61 265 45 98, E-mail address: [email protected] Published on behalf of the European Society of Cardiology. All rights reserved. & The Author 2015. For permissions please email: [email protected].
Page 2 of 5
B. Schaer et al.
What’s new?
tachycardias. As a consequence, predictors for ICD therapy are incongruous and largely heterogenic. The aims of this study were to determine the incidence of appropriate ICD therapy, to estimate ‘life years saved’, and to define predictors based on registry patients with a secondary preventive indication followed for up to 20 years.
100%
80%
Cumulative survival
† Long-term follow-up in secondary prevention ICD patients up to 20 years yields a cumulative incidence of appropriate ICD therapies of 65% and a mortality of 51% at 10 years. † Chances to predict ICD use seem limited, as after analysis of 19 baseline parameters only male gender emerged as a predictor for therapy in the VF zone .220 b.p.m. † The follow-up of up to 20 years and the high mean follow-up of almost 7 years present the longest observation period so far.
60%
40%
20%
0% 0
24
48
72
96 120 144 168 192 216 240 Follow-up in months
Figure 1 Kaplan – Meier curves of actual survival (right curve)
Methods All patients in whom an ICD was implanted at the University of Basel Hospital from 1994 onward are included in a registry that was started in 1999. At present, more than 1210 patients are incorporated. Besides specific cardiological data, multiple comorbidities and laboratory values are recorded at implant. The registry is continuously updated for all appropriate ICD therapies and death. From this registry, all patients with a secondary prevention indication (both Classes IA and IB and either ischaemic or dilated cardiomyopathy were extracted). Study endpoints were the time in months between implant and first appropriate ICD therapy and either patient death (n ¼ 156), transplantation (n ¼ 4), the loss of follow-up (n ¼ 14), or the administrative censoring date of the 31 January 2014 (n ¼ 183). The programming of the VT zone of the ICD was set to a value of 20 b.p.m. below the clinically documented VT, or in cases of fast VT or VF as index event to 175 – 185 b.p.m. The VF zone was set to 220 – 240 b.p.m. In the database, three different appropriate events can be recorded: a VT that is terminated by ATP or later by shock, a fast VT (i.e. a stable VT . 220 b.p.m.) treated by ATP during charging or shock, and finally true VF terminated by a shock. For the sake of this analysis, we simplified this to (i) any appropriate ICD therapy for VT . 175 – 185 b.p.m. and (ii) any appropriate ICD therapies in the VF zone of .220 b.p.m. (being ‘potentially life-threatening’). To be able to identify possible temporal trends, we formed four groups according to implant years (1: 1994 – 1999, 2: 2000 – 2004, 3: 2005 – 2009, and 4: 2010 – 2013) and calculated survival and rates of ICD interventions as shown above. To have meaningful and comparable results, only those at 3 years were determined, even though the observation period in the first three groups is obviously much longer. The first estimation of life-years saved was done by calculating the time difference in any individual patient between his first appropriate ICD therapy and his individual study endpoint. For the second estimation, we compared the actual survival of patients to a hypothetic survival of patients assuming they had ‘died’ at the moment of first ICD therapy. We thus constructed two Kaplan – Meier curves with the two arrhythmia settings and compared them to the actual survival at 7 years (¼ mean follow-up). This approach is shown in Figure 1. With this difference in %, we determined
and hypothetic survival equalling appropriate ICD therapy for VF to death. The horizontal bar shows the difference at the time point 7 years (≈ mean follow-up), the vertical bar the gain in median life expectancy.
the ‘number-needed-to-treat’ (NNT). The gain in median life expectancy was determined as proposed by Wright and Weinstein12 as the difference in months between the two Kaplan – Meier curves at the 50% survival point of the corresponding y-axes. Again this is shown in Figure 1.
Statistics In all baseline parameters depicted in Table 1, univariate analyses were performed both for any ICD therapy and ICD therapy in the VF zone alone. Cox regression analysis was done with the five significant parameters (P , 0.05) of the univariate analysis for any VT therapy. Kaplan – Meier curves were drawn for the determination of the incidence of appropriate ICD therapy and mortality at 5 and 10 years. Values are given as mean + 1SD in parenthesis. Other comparisons were done with the x 2 test and the Fisher’s t-test where appropriate. All analyses were performed with SPSS Version 21.
Results We included 357 patients, with a mean age of 65 + 11 years, predominantly male (89%) and with ischaemic cardiomyopathy (83%). Further baseline data are shown in Table 1. The mean followup was 82 + 53 months and median 77 months (interquartile range 49 – 120). Overall, 579 ICDs were implanted (213 patients with 1, 106 with 2, 38 with 3, and 10 with 4 ICDs). A CRT defibrillator (CRT-D) was implanted in 25 of the 33 patients (76%) with a corresponding indication of left bundle branch block, New York Heart Association (NYHA) Class . 2 and left-ventricular ejection fraction ≤ 35%. Over time 156 (44%) patients died and 208 patients received ICD therapies (59%). A first ICD therapy in the VF zone
Page 3 of 5
Predictors for appropriate ICD therapy
Table 1 Baseline characteristics of the 357 patients, overall and according to presenting arrhythmia Overall
Implant for VF (n 5 108)
Implant for VT (n 5 249)
............................................................................................................................................................................... Male gender
318 (89%)
93 (86%)
225 (90%)
Ischaemic cardiomyopathy Body mass index
295 (83%) 27 + 4
91 (84%) 26 + 4
204 (82%) 27 + 4
Ejection fraction (in %, +1SD)
33 + 10
34 + 12
32 + 9
0.03
Ejection fraction . 35% Creatinine clearance (mL/min)
120 (33%) 70 + 25
43 (40%) 74 + 26
77 (31%) 68 + 25
0.04
QRS width (ms)
121 + 33
Left bundle branch block NYHA Class . 2
121 + 34
121 + 33
94 (26%) 48 (13%)
28 (26%) 9 (8%)
66 (27%) 49 (20%)
Bypass surgery
132 (37%)
46 (42%)
86 (35%)
Percutaneous coronary intervention Hypertension
159 (45%) 226 (63%)
48 (44%) 62 (57%)
111 (45%) 164 (66%)
Diabetes
74 (21%)
20 (19%)
54 (22%)
Vascular disease Atrial fibrillation at implant
60 (17%) 29 (8%)
17 (16%) 6 (6%)
43 (17%) 23 (9%)
ACEs/ARBs therapy
324 (91%)
94 (87%)
230 (93%)
Beta-blocker therapy Diuretic therapy
292 (82%) 228 (64%)
83 (77%) 65 (60%)
210 (85%) 163 (66%)
Amiodarone therapy
131 (37%)
31 (29%)
100 (40%)
Index arrhythmia Ventricular tachycardia
0.008
0.04
207 (58%)
Syncope and inducible ventricular tachycardia
42 (12%)
VVI-ICD DDD-ICD
237 (66%) 84 (24%)
82 (76%) 17 (16%)
155 (63%) 67 (27%)
CRT-ICD
36 (10%)
9 (8%)
27 (10%)
0.04
Table 2 Breakdown of patient characteristics and events according to implant periods Time period
1: 1994–1999
2: 2000–2004
3: 2005–2009
4: 2010– 2013
Number of patients
69
128
98
62
Index arrhythmia VF (%) Ejection fraction (%, +1SD)
36 33 + 11
28 33 + 10
23 33 + 10
39 33 + 10
0.12 0.92
Any ICD therapy at 3 years
65%
50%
51%
42%
0.2 (0.48 if period 1 excluded)
ICD therapy in VF zone at 3 years Survival at 3 years
87% 94%
18% 89%
20% 81%
11% 82%
0.001 (0.17 if period 1 excluded) 0.25
was delivered in 71 of them (34%, or 20% of all patients, respectively). The overall number of shocks in these 71 patients was 562 (range 1 – 174), the overall number of antitachycardia pacing in 187 patients was 9 616 (range 1–1 700). Forty-four patients (28%) died without experiencing any form of appropriate ICD therapy. In Table 2, the temporal trend of important baseline parameters and events is shown. No significant differences with regard to index arrhythmia and mean ejection fraction at time of implant were seen. ICD therapies in the VF zone were significantly different, but this was due to an excess in time group 1. No differences were seen among time groups 2, 3, and 4. There was no significant difference regarding cumulative mortality at 3 years. However, patients
P
implanted before 2005 (time groups 1 and 2) had better survival than those implanted later (6 and 11% vs. 19% and 185). In univariate analysis, predictors for any ICD therapy were male gender, ICD implantation for VF, history of hypertension, advancing age, and poorer ejection fraction. After multivariate analysis, only implantation for VT and age remained significant [implant for VT: hazard ratio (HR) 1.450, 95% confidence interval (95% CI) 1.047–2.008, P ¼ 0.025; age (per year): HR 1.024, 95% CI 1.010–1.038, P ¼ 0.001]. For therapy in the VF zone, univariate analysis determined only male gender (29% in males vs. 5% in females, P ¼ 0.01) as a predictor, apart from the clinically not meaningful slightly higher creatinine clearance (75 + 25 vs. 69 + 26, P ¼ 0.03). Cumulative
Page 4 of 5
B. Schaer et al.
Table 3 Cumulative incidences of mortality, first-ever ICD therapy, and in the VF zone only after 5, 10, and 15 years, respectively At 5 years (%)
At 10 years (%)
At 15 years (%)
Mortality Any ICD therapies
23.9 59.2
50.8 65.3
64.6 69.7
ICD in the VF zone only
20.4
22.3
23.7
................................................................................
first ICD shock for a therapy in the VF zone, 44 patients lived on for at least 6 years, and 24 of them for at least 10 years and more. At the time point 7 years (≈ ¼ mean follow-up), the difference between the actual and the hypothetic survival was 18% with the assumption that any ICD therapy equalled ‘death’. This resulted in a NNT of 6 (95% CI 4.5 – 7.2). The corresponding figures were 10% with the assumption that only ICD therapy in the VF zone equalled ‘death’, resulting in a NNT of 10 (95% CI 6.6 –14.5). The gain in life expectancy was 24 months for the setting ‘all ICD therapies’ and 11 months for the setting ‘life threatening arrhythmias’.
Discussion Table 4 Occurrence of ICD therapies in the VT zone and the VF zone according to the index arrhythmia Presenting arrhythmia
VF
VT
P
................................................................................ Any ICD therapy
48 (44%)
160 (64%)
ICD therapy in the VT zone ICD therapy in the VF zone
20 (18%) 28 (26%)
117 (47%) 43 (17%)
0.0001
100
Survival-free from ICD therapy
80
60
40
20
300 239 188 138
86
55
30
8
3
0 0
24
48
72
96 120 144 168 192 216 240 Follow-up in months
Figure 2 Occurrence of first-ever ICD therapy over time (x-axis follow-up in months, y-axis cumulative incidence), including number of patients at risk.
incidences of mortality, first-ever ICD therapy, and for therapy in the VF zone only after 5, 10, and 15 years, respectively, are shown in Table 3. The distribution of ICD therapies according to the index arrhythmia is shown in Table 4. Figure 2 highlights the occurrence of first-ever ICD therapy over time. Of note, the incidence was 63% after 7 years and ascended on slightly during further 8 years to a total of 70%. Survival after first ICD therapy is shown in Table 5. After a
In line with all randomized controlled trials, this retrospective study indicates that ICD implantation in all secondary prevention patients is recommended, as no predictors for ICD therapy could be identified for long-term ICD use. Incidence of ICD therapies in a pure secondary prevention population has been studied in only two other studies, both with some relevant limitations. Freedberg et al.9 followed 125 patients with ischaemic cardiomyopathy for a mean of only 13 months with ICD use of 43% after 1 year and 62% after 2 years. Cumulative incidences of ICD therapies after 5 and 10 years in the Leiden registry8 were 52 and 61% and thus comparable with the rates reported here. Huge differences exist with regard to the incidence of potentially lifethreatening arrhythmias with incidences after 5 and 10 years of 52 and 61%8 compared with 20 and 22% in our study. They are explained by the different definition of life-threatening, i.e. cut-off rates of .188/min (in Ref. 8) and .220/min (in the present study), respectively. The majority of first-ever therapy occurred within the first 5 years with only minimal increases after 10 and 15 years, in our study, e.g. from 20.4 to 23.7% after 15 years for life-threatening arrhythmias. The risk after 5 years is thus low, but obviously not zero, as has also been shown in other studies with very long-term follow-up.13 Predictors for ICD therapies have been determined in several studies,7 – 9 but results are very heterogenic and even contradictory. Implantation for sustained monomorphic VT e.g. has been established as a predictor with a hazard ratio of 1.51,8 that was confirmed in our population with a hazard ratio of 1.45. In general, prediction of ICD therapies seems to be rather limited. A weakness of the Kaplan –Meier curves applied to determine the effect of ICDs is that they disregard the amount of years gained after an effective ICD treatment. We tried to overcome this by calculating survival after first ICD therapy. Wright and Weinstein12 highlight this by declaring that the mean gain reflects a small gain for many members of a population but a very large gain for those who might have died prematurely without the intervention. However, quality of life and the impact of severe comorbidity burden on survival are disregarded. As quality of care (drug therapy, revascularization strategy, etc.) has remarkably changed since 1994 and the first ICD implants in the study population, one might expect an impact on both delivery of ICD therapy and mortality. Albeit this seems to be true for ICD therapies in the VF zone, no difference was seen in patients implanted after 2000. Similarly, mortality at 3 years was not different, even somewhat lower in patients implanted in the
Page 5 of 5
Predictors for appropriate ICD therapy
Table 5 Years gained after first-ever ICD therapy to death in intervals of 2 years for a first-ever ICD therapy (above) and for life-threatening arrhythmias (below) 0 –2 years
2 –4 years
4 –6 years
6–8 years
8–10 years
10– 12 years
12 –14 years
14– 16 years
16 –18 years
14 12%
24 21%
24 21%
17 15%
12 11%
12 11%
4 4%
1 1%
7 6%
1
0
2
3%
0%
6%
...............................................................................................................................................................................
4
9
5
3
4
4
12%
28%
16%
10%
12%
12%
early days. This suggests that in spite of the wide time range of the study, results are still relevant for a contemporary population. The most important limitation of the present study was the necessity to take a surrogate endpoint to determine benefit of ICD therapy due to the absence of a control group of patients without an ICD, but this approach was chosen in the two other studies8,9 as well. While it is obvious that termination of a stable VT of e.g. 180 b.p.m. after a few seconds by a single ATP burst can hardly be equalled to aborted sudden cardiac death, this probably might not be true for a fast VT of 260 b.p.m. We tried to overcome this by presenting results in two different scenarios of ‘any ICD therapy’ and ‘ICD therapy in the VF zone’. The paper from Ellenbogen et al. 14 is often cited to indicate that ICD therapies are thought to be not valid surrogates for sudden cardiac death, but these data stem from a trial in primary prevention and the ICD was programmed as a shock box with a cut-off of 180 – 200 b.p.m. However, all this does not get over this limitation. Another limitation is the heterogeneity of ICD programming with regard to detection periods. Only during the course of the study period, the importance of longer detection intervals and the consecutive reduction in ATP and shock numbers was recognized.15 According to this, the reported incidence of appropriate ICD use in this study is probably rather overestimated than underestimated. Features such as e.g. ‘ATP during charging’ implemented over time to avoid shocks do not represent a limitation, as the heart rate and not the application of a shock was used to define ‘life threatening arrhythmias’. A smaller limitation is that not all patients who had a contemporary indication for CRT-D were implanted with such a device.
Conclusions The incidence of appropriate ICD therapy in secondary prevention is high with a cumulative incidence at 10 years of 65%. No predictors to guide the clinician in decision-making especially with regard to the occurrence of potentially life-threatening arrhythmias in the VF zone .220 b.p.m. during long-term follow-up could be identified. Conflict of interest: B.S. has served on the speakers’ bureau for Medtronic and Sorin. M.K. has served on the speakers’ bureau for Boston Scientific, St. Jude Medical, and Biotronik. S.O. has served on the speakers’ bureau for Medtronic, Boston Scientific, Biotronik, and St. Jude Medical and has received unrestricted grants from Medtronic, Boston Scientific, Biotronik, and St. Jude Medical. C.S.
has served on the speakers’ bureau for Medtronic, Biotronik, and Sorin and had scientific support from Medtronic, Biotronik, Boston Scientific, St. Jude Medical, and Sorin.
References 1. Zipes DP, Camm AJ, Borggrefe M, Buxton AE, Chaitman B, Fromer M et al. ACC/ AHA/ESC 2006 guidelines for management of patients with ventricular arrhythmias and the prevention of sudden cardiac death. Europace 2006;8:746 – 837. 2. The Antiarrhythmics versus Implantable Defibrillators (AVID) Investigators. A comparison of antiarrhythmic-drug therapy with implantable defibrillators in patients resuscitated from near-fatal ventricular arrhythmias. N Engl J Med 1997; 337:1576 –83. 3. Connolly SJ, Gent M, Roberts RS, Dorian P, Roy D, Sheldon RS et al. Canadian implantable defibrillator study (CIDS): a randomized trial of the implantable cardioverter defibrillator against amiodarone. Circulation 2000;101:1297 – 302. 4. Russo AM, Stainback RF, Bailey SR, Epstein AE, Heidenreich PA, Jessup M et al. ACCF/ HRS/AHA/ASE/HFSA/SCAI/SCCT/SCMR 2013 appropriate use criteria for implantable cardioverter-defibrillators and cardiac resynchronization therapy. J Am Coll Cardiol 2013;61:1318–68. 5. Boriani G, Biffi M, Martignani C, Gallina M, Branzi A. Cost-effectiveness of implantable cardioverter-defibrillators. Eur Heart J 2001;22:990 –6. 6. Sanders GD, Hlatky MA, Owens DK. Cost-effectiveness of implantable cardioverter-defibrillators. N Engl J Med 2005;353:1471 –80. 7. Blumer J, Wolber T, Hellermann J, Holzmeister J, Binggeli C, Duru F et al. Predictors of appropriate implantable cardioverter-defibrillator therapy during long-term follow-up of patients with coronary artery disease. Int Heart J 2009;50:313 –21. 8. Borleffs CJ, van Erven L, Schotman M, Boersma E, Kies P, van der Burg AE et al. Recurrence of ventricular arrhythmias in ischaemic secondary prevention implantable cardioverter defibrillator recipients: long-term follow-up of the Leiden out-of-hospital cardiac arrest study (LOHCAT). Eur Heart J 2009;30:1621 – 6. 9. Freedberg NA, Hill JN, Fogel RI, Prystowsky EN, Group C. Recurrence of symptomatic ventricular arrhythmias in patients with implantable cardioverter defibrillator after the first device therapy: implications for antiarrhythmic therapy and driving restrictions. CARE Group. J Am Coll Cardiol 2001;37:1910 –5. 10. Stockburger M, Krebs A, Nitardy A, Habedank D, Celebi O, Knaus T et al. Survival and appropriate device interventions in recipients of cardioverter defibrillators implanted for the primary versus secondary prevention of sudden cardiac death. Pacing Clin Electrophysiol 2009;32(Suppl 1):S16 –20. 11. Tandri H, Griffith LS, Tang T, Nasir K, Zardkoohi O, Reddy CV et al. Clinical course and long-term follow-up of patients receiving implantable cardioverterdefibrillators. Heart Rhythm 2006;3:762 –8. 12. Wright JC, Weinstein MC. Gains in life expectancy from medical interventions-standardizing data on outcomes. N Engl J Med 1998;339:380 –6. 13. Yap SC, Schaer BA, Bhagwandien RE, Kuhne M, Dabiri Abkenari L, Osswald S et al. Evaluation of the need of elective implantable cardioverter-defibrillator generator replacement in primary prevention patients without prior appropriate ICD therapy. Heart 2014;100:1188 – 92. 14. Ellenbogen KA, Levine JH, Berger RD, Daubert JP, Winters SL, Greenstein E et al. Are implantable cardioverter defibrillator shocks a surrogate for sudden cardiac death in patients with nonischemic cardiomyopathy? Circulation 2006;113:776 – 82. 15. Wilkoff BL, Williamson BD, Stern RS, Moore SL, Lu F, Lee SW et al. Strategic programming of detection and therapy parameters in implantable cardioverterdefibrillators reduces shocks in primary prevention patients: results from the PREPARE (Primary Prevention Parameters Evaluation) study. J Am Coll Cardiol 2008;52:541 –50.
CLINICAL RESEARCH
Europace doi:10.1093/europace/euv188
Incidence of and predictors for appropriate implantable cardioverter-defibrillator therapy in patients with a secondary preventive implantable cardioverter-defibrillator indication Beat Schaer*, Michael Ku¨hne, Tobias Reichlin, Stefan Osswald, and Christian Sticherling Department of Cardiology, University Hospital of Basel, Petersgraben 4, Basel 4031, Switzerland Received 13 February 2015; accepted after revision 1 May 2015
Aims
Incidence of implantable cardioverter-defibrillator (ICD) therapy in secondary prevention has been assessed in randomized trials and registries. However, results are considerably limited by short follow-up and hazy definition of treated arrhythmias. This study aimed to determine appropriate ICD therapy and to define predictors based on registry patients followed for up to 20 years. ..................................................................................................................................................................................... Methods All patients with a secondary prevention indication and ischaemic or dilated cardiomyopathy were identified. Arrhythmic endpoints were appropriate ICD therapies for any ventricular tachycardia (VT) .175 b.p.m. and appropriate ICD and results therapies in the ventricular fibrillation (VF) zone of .220 b.p.m. (potentially life-threatening). Predictors were determined by analysing 19 baseline characteristics. We included 357 patients, age 65 + 11 years, predominantly male (89%) with ischaemic cardiomyopathy (83%). During follow-up of 82 + 53 months, 156 (44%) patients died and 208 received any form of ICD therapies (59%), 71 of them (34%) in the VF zone. Forty-four patients (28%) died without experiencing any form of appropriate ICD therapy. Cumulative incidence of any form of ICD therapy at 10 years was 65%. Predictors for any form of ICD therapy were implantation for VT and age [VT: hazard ratio (HR) 1.45, 95% confidence interval (95% CI) 1.05 – 2.01, P ¼ 0.03; age (per year): HR 1.02, 95% CI 1.01 – 1.04, P ¼ 0.001]. For therapy in the VF zone, univariate analysis determined male gender (29 vs. 5%, P ¼ 0.01) as predictor. ..................................................................................................................................................................................... Conclusion The rate of appropriate ICD therapies in secondary prevention is high. No useful predictors for them, especially not for life-threatening arrhythmias could be identified.
----------------------------------------------------------------------------------------------------------------------------------------------------------Keywords
Implantable cardioverter-defibrillator † Predictor † Secondary prevention † Cumulative incidence
Implantable cardioverter-defibrillator (ICD) placement is a Class IA indication for survivors of cardiac arrest due to ventricular fibrillation (VF). Indication is Class IB in patients who present with sustained ventricular tachycardia (VT) and for those with syncope and inducible VT.1 These recommendations are based on randomized controlled trials.2,3 Implantable cardioverter-defibrillator implantation was labelled as ‘appropriate use’ in the recent Appropriate Use Criteria Task Force report.4 For such a secondary prevention setting, several cost-effectiveness calculations have been undertaken summarized in a review article by Boriani et al. published
in 2001.5 They all showed costs per life-year saved between US $12 000 and 54 000. The most recent analysis in a secondary preventive ICD population performed in 2001, using a Markov model, was mitigated through the inclusion of primary prevention trials as well.6 As then, corresponding research has focused on primary prevention or cardiac resynchronization therapy (CRT). The use of ICD [i.e. firing or antitachycardia pacing (ATP)] has been studied in registries7 – 11 but published results are considerably limited by short follow-up, inclusion of additional primary prevention patients or inhomogeneous definition of life-threatening
* Corresponding author. Tel: +41 61 328 62 22; fax: +41 61 265 45 98, E-mail address: [email protected] Published on behalf of the European Society of Cardiology. All rights reserved. & The Author 2015. For permissions please email: [email protected].
Page 2 of 5
B. Schaer et al.
What’s new?
tachycardias. As a consequence, predictors for ICD therapy are incongruous and largely heterogenic. The aims of this study were to determine the incidence of appropriate ICD therapy, to estimate ‘life years saved’, and to define predictors based on registry patients with a secondary preventive indication followed for up to 20 years.
100%
80%
Cumulative survival
† Long-term follow-up in secondary prevention ICD patients up to 20 years yields a cumulative incidence of appropriate ICD therapies of 65% and a mortality of 51% at 10 years. † Chances to predict ICD use seem limited, as after analysis of 19 baseline parameters only male gender emerged as a predictor for therapy in the VF zone .220 b.p.m. † The follow-up of up to 20 years and the high mean follow-up of almost 7 years present the longest observation period so far.
60%
40%
20%
0% 0
24
48
72
96 120 144 168 192 216 240 Follow-up in months
Figure 1 Kaplan – Meier curves of actual survival (right curve)
Methods All patients in whom an ICD was implanted at the University of Basel Hospital from 1994 onward are included in a registry that was started in 1999. At present, more than 1210 patients are incorporated. Besides specific cardiological data, multiple comorbidities and laboratory values are recorded at implant. The registry is continuously updated for all appropriate ICD therapies and death. From this registry, all patients with a secondary prevention indication (both Classes IA and IB and either ischaemic or dilated cardiomyopathy were extracted). Study endpoints were the time in months between implant and first appropriate ICD therapy and either patient death (n ¼ 156), transplantation (n ¼ 4), the loss of follow-up (n ¼ 14), or the administrative censoring date of the 31 January 2014 (n ¼ 183). The programming of the VT zone of the ICD was set to a value of 20 b.p.m. below the clinically documented VT, or in cases of fast VT or VF as index event to 175 – 185 b.p.m. The VF zone was set to 220 – 240 b.p.m. In the database, three different appropriate events can be recorded: a VT that is terminated by ATP or later by shock, a fast VT (i.e. a stable VT . 220 b.p.m.) treated by ATP during charging or shock, and finally true VF terminated by a shock. For the sake of this analysis, we simplified this to (i) any appropriate ICD therapy for VT . 175 – 185 b.p.m. and (ii) any appropriate ICD therapies in the VF zone of .220 b.p.m. (being ‘potentially life-threatening’). To be able to identify possible temporal trends, we formed four groups according to implant years (1: 1994 – 1999, 2: 2000 – 2004, 3: 2005 – 2009, and 4: 2010 – 2013) and calculated survival and rates of ICD interventions as shown above. To have meaningful and comparable results, only those at 3 years were determined, even though the observation period in the first three groups is obviously much longer. The first estimation of life-years saved was done by calculating the time difference in any individual patient between his first appropriate ICD therapy and his individual study endpoint. For the second estimation, we compared the actual survival of patients to a hypothetic survival of patients assuming they had ‘died’ at the moment of first ICD therapy. We thus constructed two Kaplan – Meier curves with the two arrhythmia settings and compared them to the actual survival at 7 years (¼ mean follow-up). This approach is shown in Figure 1. With this difference in %, we determined
and hypothetic survival equalling appropriate ICD therapy for VF to death. The horizontal bar shows the difference at the time point 7 years (≈ mean follow-up), the vertical bar the gain in median life expectancy.
the ‘number-needed-to-treat’ (NNT). The gain in median life expectancy was determined as proposed by Wright and Weinstein12 as the difference in months between the two Kaplan – Meier curves at the 50% survival point of the corresponding y-axes. Again this is shown in Figure 1.
Statistics In all baseline parameters depicted in Table 1, univariate analyses were performed both for any ICD therapy and ICD therapy in the VF zone alone. Cox regression analysis was done with the five significant parameters (P , 0.05) of the univariate analysis for any VT therapy. Kaplan – Meier curves were drawn for the determination of the incidence of appropriate ICD therapy and mortality at 5 and 10 years. Values are given as mean + 1SD in parenthesis. Other comparisons were done with the x 2 test and the Fisher’s t-test where appropriate. All analyses were performed with SPSS Version 21.
Results We included 357 patients, with a mean age of 65 + 11 years, predominantly male (89%) and with ischaemic cardiomyopathy (83%). Further baseline data are shown in Table 1. The mean followup was 82 + 53 months and median 77 months (interquartile range 49 – 120). Overall, 579 ICDs were implanted (213 patients with 1, 106 with 2, 38 with 3, and 10 with 4 ICDs). A CRT defibrillator (CRT-D) was implanted in 25 of the 33 patients (76%) with a corresponding indication of left bundle branch block, New York Heart Association (NYHA) Class . 2 and left-ventricular ejection fraction ≤ 35%. Over time 156 (44%) patients died and 208 patients received ICD therapies (59%). A first ICD therapy in the VF zone
Page 3 of 5
Predictors for appropriate ICD therapy
Table 1 Baseline characteristics of the 357 patients, overall and according to presenting arrhythmia Overall
Implant for VF (n 5 108)
Implant for VT (n 5 249)
............................................................................................................................................................................... Male gender
318 (89%)
93 (86%)
225 (90%)
Ischaemic cardiomyopathy Body mass index
295 (83%) 27 + 4
91 (84%) 26 + 4
204 (82%) 27 + 4
Ejection fraction (in %, +1SD)
33 + 10
34 + 12
32 + 9
0.03
Ejection fraction . 35% Creatinine clearance (mL/min)
120 (33%) 70 + 25
43 (40%) 74 + 26
77 (31%) 68 + 25
0.04
QRS width (ms)
121 + 33
Left bundle branch block NYHA Class . 2
121 + 34
121 + 33
94 (26%) 48 (13%)
28 (26%) 9 (8%)
66 (27%) 49 (20%)
Bypass surgery
132 (37%)
46 (42%)
86 (35%)
Percutaneous coronary intervention Hypertension
159 (45%) 226 (63%)
48 (44%) 62 (57%)
111 (45%) 164 (66%)
Diabetes
74 (21%)
20 (19%)
54 (22%)
Vascular disease Atrial fibrillation at implant
60 (17%) 29 (8%)
17 (16%) 6 (6%)
43 (17%) 23 (9%)
ACEs/ARBs therapy
324 (91%)
94 (87%)
230 (93%)
Beta-blocker therapy Diuretic therapy
292 (82%) 228 (64%)
83 (77%) 65 (60%)
210 (85%) 163 (66%)
Amiodarone therapy
131 (37%)
31 (29%)
100 (40%)
Index arrhythmia Ventricular tachycardia
0.008
0.04
207 (58%)
Syncope and inducible ventricular tachycardia
42 (12%)
VVI-ICD DDD-ICD
237 (66%) 84 (24%)
82 (76%) 17 (16%)
155 (63%) 67 (27%)
CRT-ICD
36 (10%)
9 (8%)
27 (10%)
0.04
Table 2 Breakdown of patient characteristics and events according to implant periods Time period
1: 1994–1999
2: 2000–2004
3: 2005–2009
4: 2010– 2013
Number of patients
69
128
98
62
Index arrhythmia VF (%) Ejection fraction (%, +1SD)
36 33 + 11
28 33 + 10
23 33 + 10
39 33 + 10
0.12 0.92
Any ICD therapy at 3 years
65%
50%
51%
42%
0.2 (0.48 if period 1 excluded)
ICD therapy in VF zone at 3 years Survival at 3 years
87% 94%
18% 89%
20% 81%
11% 82%
0.001 (0.17 if period 1 excluded) 0.25
was delivered in 71 of them (34%, or 20% of all patients, respectively). The overall number of shocks in these 71 patients was 562 (range 1 – 174), the overall number of antitachycardia pacing in 187 patients was 9 616 (range 1–1 700). Forty-four patients (28%) died without experiencing any form of appropriate ICD therapy. In Table 2, the temporal trend of important baseline parameters and events is shown. No significant differences with regard to index arrhythmia and mean ejection fraction at time of implant were seen. ICD therapies in the VF zone were significantly different, but this was due to an excess in time group 1. No differences were seen among time groups 2, 3, and 4. There was no significant difference regarding cumulative mortality at 3 years. However, patients
P
implanted before 2005 (time groups 1 and 2) had better survival than those implanted later (6 and 11% vs. 19% and 185). In univariate analysis, predictors for any ICD therapy were male gender, ICD implantation for VF, history of hypertension, advancing age, and poorer ejection fraction. After multivariate analysis, only implantation for VT and age remained significant [implant for VT: hazard ratio (HR) 1.450, 95% confidence interval (95% CI) 1.047–2.008, P ¼ 0.025; age (per year): HR 1.024, 95% CI 1.010–1.038, P ¼ 0.001]. For therapy in the VF zone, univariate analysis determined only male gender (29% in males vs. 5% in females, P ¼ 0.01) as a predictor, apart from the clinically not meaningful slightly higher creatinine clearance (75 + 25 vs. 69 + 26, P ¼ 0.03). Cumulative
Page 4 of 5
B. Schaer et al.
Table 3 Cumulative incidences of mortality, first-ever ICD therapy, and in the VF zone only after 5, 10, and 15 years, respectively At 5 years (%)
At 10 years (%)
At 15 years (%)
Mortality Any ICD therapies
23.9 59.2
50.8 65.3
64.6 69.7
ICD in the VF zone only
20.4
22.3
23.7
................................................................................
first ICD shock for a therapy in the VF zone, 44 patients lived on for at least 6 years, and 24 of them for at least 10 years and more. At the time point 7 years (≈ ¼ mean follow-up), the difference between the actual and the hypothetic survival was 18% with the assumption that any ICD therapy equalled ‘death’. This resulted in a NNT of 6 (95% CI 4.5 – 7.2). The corresponding figures were 10% with the assumption that only ICD therapy in the VF zone equalled ‘death’, resulting in a NNT of 10 (95% CI 6.6 –14.5). The gain in life expectancy was 24 months for the setting ‘all ICD therapies’ and 11 months for the setting ‘life threatening arrhythmias’.
Discussion Table 4 Occurrence of ICD therapies in the VT zone and the VF zone according to the index arrhythmia Presenting arrhythmia
VF
VT
P
................................................................................ Any ICD therapy
48 (44%)
160 (64%)
ICD therapy in the VT zone ICD therapy in the VF zone
20 (18%) 28 (26%)
117 (47%) 43 (17%)
0.0001
100
Survival-free from ICD therapy
80
60
40
20
300 239 188 138
86
55
30
8
3
0 0
24
48
72
96 120 144 168 192 216 240 Follow-up in months
Figure 2 Occurrence of first-ever ICD therapy over time (x-axis follow-up in months, y-axis cumulative incidence), including number of patients at risk.
incidences of mortality, first-ever ICD therapy, and for therapy in the VF zone only after 5, 10, and 15 years, respectively, are shown in Table 3. The distribution of ICD therapies according to the index arrhythmia is shown in Table 4. Figure 2 highlights the occurrence of first-ever ICD therapy over time. Of note, the incidence was 63% after 7 years and ascended on slightly during further 8 years to a total of 70%. Survival after first ICD therapy is shown in Table 5. After a
In line with all randomized controlled trials, this retrospective study indicates that ICD implantation in all secondary prevention patients is recommended, as no predictors for ICD therapy could be identified for long-term ICD use. Incidence of ICD therapies in a pure secondary prevention population has been studied in only two other studies, both with some relevant limitations. Freedberg et al.9 followed 125 patients with ischaemic cardiomyopathy for a mean of only 13 months with ICD use of 43% after 1 year and 62% after 2 years. Cumulative incidences of ICD therapies after 5 and 10 years in the Leiden registry8 were 52 and 61% and thus comparable with the rates reported here. Huge differences exist with regard to the incidence of potentially lifethreatening arrhythmias with incidences after 5 and 10 years of 52 and 61%8 compared with 20 and 22% in our study. They are explained by the different definition of life-threatening, i.e. cut-off rates of .188/min (in Ref. 8) and .220/min (in the present study), respectively. The majority of first-ever therapy occurred within the first 5 years with only minimal increases after 10 and 15 years, in our study, e.g. from 20.4 to 23.7% after 15 years for life-threatening arrhythmias. The risk after 5 years is thus low, but obviously not zero, as has also been shown in other studies with very long-term follow-up.13 Predictors for ICD therapies have been determined in several studies,7 – 9 but results are very heterogenic and even contradictory. Implantation for sustained monomorphic VT e.g. has been established as a predictor with a hazard ratio of 1.51,8 that was confirmed in our population with a hazard ratio of 1.45. In general, prediction of ICD therapies seems to be rather limited. A weakness of the Kaplan –Meier curves applied to determine the effect of ICDs is that they disregard the amount of years gained after an effective ICD treatment. We tried to overcome this by calculating survival after first ICD therapy. Wright and Weinstein12 highlight this by declaring that the mean gain reflects a small gain for many members of a population but a very large gain for those who might have died prematurely without the intervention. However, quality of life and the impact of severe comorbidity burden on survival are disregarded. As quality of care (drug therapy, revascularization strategy, etc.) has remarkably changed since 1994 and the first ICD implants in the study population, one might expect an impact on both delivery of ICD therapy and mortality. Albeit this seems to be true for ICD therapies in the VF zone, no difference was seen in patients implanted after 2000. Similarly, mortality at 3 years was not different, even somewhat lower in patients implanted in the
Page 5 of 5
Predictors for appropriate ICD therapy
Table 5 Years gained after first-ever ICD therapy to death in intervals of 2 years for a first-ever ICD therapy (above) and for life-threatening arrhythmias (below) 0 –2 years
2 –4 years
4 –6 years
6–8 years
8–10 years
10– 12 years
12 –14 years
14– 16 years
16 –18 years
14 12%
24 21%
24 21%
17 15%
12 11%
12 11%
4 4%
1 1%
7 6%
1
0
2
3%
0%
6%
...............................................................................................................................................................................
4
9
5
3
4
4
12%
28%
16%
10%
12%
12%
early days. This suggests that in spite of the wide time range of the study, results are still relevant for a contemporary population. The most important limitation of the present study was the necessity to take a surrogate endpoint to determine benefit of ICD therapy due to the absence of a control group of patients without an ICD, but this approach was chosen in the two other studies8,9 as well. While it is obvious that termination of a stable VT of e.g. 180 b.p.m. after a few seconds by a single ATP burst can hardly be equalled to aborted sudden cardiac death, this probably might not be true for a fast VT of 260 b.p.m. We tried to overcome this by presenting results in two different scenarios of ‘any ICD therapy’ and ‘ICD therapy in the VF zone’. The paper from Ellenbogen et al. 14 is often cited to indicate that ICD therapies are thought to be not valid surrogates for sudden cardiac death, but these data stem from a trial in primary prevention and the ICD was programmed as a shock box with a cut-off of 180 – 200 b.p.m. However, all this does not get over this limitation. Another limitation is the heterogeneity of ICD programming with regard to detection periods. Only during the course of the study period, the importance of longer detection intervals and the consecutive reduction in ATP and shock numbers was recognized.15 According to this, the reported incidence of appropriate ICD use in this study is probably rather overestimated than underestimated. Features such as e.g. ‘ATP during charging’ implemented over time to avoid shocks do not represent a limitation, as the heart rate and not the application of a shock was used to define ‘life threatening arrhythmias’. A smaller limitation is that not all patients who had a contemporary indication for CRT-D were implanted with such a device.
Conclusions The incidence of appropriate ICD therapy in secondary prevention is high with a cumulative incidence at 10 years of 65%. No predictors to guide the clinician in decision-making especially with regard to the occurrence of potentially life-threatening arrhythmias in the VF zone .220 b.p.m. during long-term follow-up could be identified. Conflict of interest: B.S. has served on the speakers’ bureau for Medtronic and Sorin. M.K. has served on the speakers’ bureau for Boston Scientific, St. Jude Medical, and Biotronik. S.O. has served on the speakers’ bureau for Medtronic, Boston Scientific, Biotronik, and St. Jude Medical and has received unrestricted grants from Medtronic, Boston Scientific, Biotronik, and St. Jude Medical. C.S.
has served on the speakers’ bureau for Medtronic, Biotronik, and Sorin and had scientific support from Medtronic, Biotronik, Boston Scientific, St. Jude Medical, and Sorin.
References 1. Zipes DP, Camm AJ, Borggrefe M, Buxton AE, Chaitman B, Fromer M et al. ACC/ AHA/ESC 2006 guidelines for management of patients with ventricular arrhythmias and the prevention of sudden cardiac death. Europace 2006;8:746 – 837. 2. The Antiarrhythmics versus Implantable Defibrillators (AVID) Investigators. A comparison of antiarrhythmic-drug therapy with implantable defibrillators in patients resuscitated from near-fatal ventricular arrhythmias. N Engl J Med 1997; 337:1576 –83. 3. Connolly SJ, Gent M, Roberts RS, Dorian P, Roy D, Sheldon RS et al. Canadian implantable defibrillator study (CIDS): a randomized trial of the implantable cardioverter defibrillator against amiodarone. Circulation 2000;101:1297 – 302. 4. Russo AM, Stainback RF, Bailey SR, Epstein AE, Heidenreich PA, Jessup M et al. ACCF/ HRS/AHA/ASE/HFSA/SCAI/SCCT/SCMR 2013 appropriate use criteria for implantable cardioverter-defibrillators and cardiac resynchronization therapy. J Am Coll Cardiol 2013;61:1318–68. 5. Boriani G, Biffi M, Martignani C, Gallina M, Branzi A. Cost-effectiveness of implantable cardioverter-defibrillators. Eur Heart J 2001;22:990 –6. 6. Sanders GD, Hlatky MA, Owens DK. Cost-effectiveness of implantable cardioverter-defibrillators. N Engl J Med 2005;353:1471 –80. 7. Blumer J, Wolber T, Hellermann J, Holzmeister J, Binggeli C, Duru F et al. Predictors of appropriate implantable cardioverter-defibrillator therapy during long-term follow-up of patients with coronary artery disease. Int Heart J 2009;50:313 –21. 8. Borleffs CJ, van Erven L, Schotman M, Boersma E, Kies P, van der Burg AE et al. Recurrence of ventricular arrhythmias in ischaemic secondary prevention implantable cardioverter defibrillator recipients: long-term follow-up of the Leiden out-of-hospital cardiac arrest study (LOHCAT). Eur Heart J 2009;30:1621 – 6. 9. Freedberg NA, Hill JN, Fogel RI, Prystowsky EN, Group C. Recurrence of symptomatic ventricular arrhythmias in patients with implantable cardioverter defibrillator after the first device therapy: implications for antiarrhythmic therapy and driving restrictions. CARE Group. J Am Coll Cardiol 2001;37:1910 –5. 10. Stockburger M, Krebs A, Nitardy A, Habedank D, Celebi O, Knaus T et al. Survival and appropriate device interventions in recipients of cardioverter defibrillators implanted for the primary versus secondary prevention of sudden cardiac death. Pacing Clin Electrophysiol 2009;32(Suppl 1):S16 –20. 11. Tandri H, Griffith LS, Tang T, Nasir K, Zardkoohi O, Reddy CV et al. Clinical course and long-term follow-up of patients receiving implantable cardioverterdefibrillators. Heart Rhythm 2006;3:762 –8. 12. Wright JC, Weinstein MC. Gains in life expectancy from medical interventions-standardizing data on outcomes. N Engl J Med 1998;339:380 –6. 13. Yap SC, Schaer BA, Bhagwandien RE, Kuhne M, Dabiri Abkenari L, Osswald S et al. Evaluation of the need of elective implantable cardioverter-defibrillator generator replacement in primary prevention patients without prior appropriate ICD therapy. Heart 2014;100:1188 – 92. 14. Ellenbogen KA, Levine JH, Berger RD, Daubert JP, Winters SL, Greenstein E et al. Are implantable cardioverter defibrillator shocks a surrogate for sudden cardiac death in patients with nonischemic cardiomyopathy? Circulation 2006;113:776 – 82. 15. Wilkoff BL, Williamson BD, Stern RS, Moore SL, Lu F, Lee SW et al. Strategic programming of detection and therapy parameters in implantable cardioverterdefibrillators reduces shocks in primary prevention patients: results from the PREPARE (Primary Prevention Parameters Evaluation) study. J Am Coll Cardiol 2008;52:541 –50.
