An implantable cardioverter-defibrillator (ICD) is a treatment associated with significant cost and risk of morbidity. Elderly patients (arbitrarily defined as age 75 or more) were under-represented in large clinical trials (e.g. MADIT-II1 and SCD-HeFT2) which showed a mortality benefit with ICD implantation in patients with reduced left ventricular function. Therefore, the net mortality benefit for this age group is unclear.Van Rees et al3 found that the cumulative incidences of appropriate therapy and appropriate shocks in patients age 75 or more with primary ICD implanted were 28% and 13% respectively. Their all-cause mortality was 29% with a cumulative incidence for death of 35% at 1 year following appropriate shock. Currently, there is no such New Zealand data available.The primary objective of our study was to ascertain survival in elderly patients (age 75 or more) who have had ICDs implanted together with device related morbidity. The secondary objective was to compare these outcomes with a younger subset of patients (age 70-74) receiving ICD.This was a descriptive study with retrospective data collection on 41 patients aged 75 or more living in the Auckland region who received ICDs between 1 January 2000 and 31 December 2010. Demographic, clinical and survival data were collected retrospectively from Auckland District Health Board (DHB), Counties Manukau DHB, Waitemata DHB and regional ICD databases.For comparison, data on 47 consecutive patients receiving ICDs at age 70-74 between 1 January 2001 and 31 December 2010 were retrieved from the National Wellington ICD registry. Mean follow-up was 50 months for the Auckland (age 75 or more) cohort (range 7 to 150 months) and 56 months for the Wellington (age 70-74) cohort (range 3 to 143 months).Apart from age, the baseline characteristics were comparable (Table 1). All recorded variables were defined according to literature or common practice. Ischaemic heart disease was defined as the presence of a diameter stenosis of at least 50% in at least one coronary artery.4 Table 1. Clinical characteristics Variables Auckland 75 or more cohort (n=41) Wellington 70-74 cohort (n=47) Male gender 34 (83%) 40 (85%) Mean age (range) 77 years (75-84) 72 years (70-74) Primary implants 10 (24%) 13 (28%) Secondary implants 31 (76%) 34 (72%) > for cardiac arrest 10 18 > for ventricular tachycardia 21 16 History of atrial fibrillation 18 (44%) 9 (19%) p=0.07 Ischaemic heart disease 29 (71%) 38 (81%) Non ischaemic cardiomyopathy 14 (38%) 7 (22%) p=0.11 Prior revascularisation 18 (44%) 25 (53%) > CABG alone 15 (83%) 12 (48%) > PCI alone 2 (11%) 8 (32%) > CABG & PCI 1 (6%) 5 (20%) NYHA functional class Class I 14 (34%) 9 Class II 18 (44%) 14 Class III 7 (17%) 5 Class IV 2 (5%) 0 Unknown 19 QRS>150ms 15 (37%) 17 (36%) LBBB 28 (68%) Mean LVEF (range) 35% (19-56) 35% (11-70) EF 35% or less 28 (68%) 26 (55%) ICD therapy (including shock and anti-tachycardia pacing (ATP)) was classified as appropriate when initiated by ventricular tachycardia (VT) or ventricular fibrillation (VF) that was still present when therapy was delivered. Twenty-two patients (54%) in the Auckland cohort and 26 patients (55%) in the Wellington cohort received appropriate ICD therapy. There was no significant difference between the two groups (Table 2). Table 2. Appropriate ICD therapy Variables Auckland 75 or more cohort (n=41) Wellington 70-74 cohort (n=47) Appropriate ICD therapy 22 (54%) 26 (55%) p=0.88 Shock 16 (39%) 18 (38%) p=0.95 > more than 3 episodes 8 4 > shock alone 4 3 ATP 18 (44%) 23 (49%) p=0.64 > 1-3 episodes 8 13 > more than 3 episodes 10 10 Device-related morbidity comprised inappropriate ICD therapy, ICD implant complications, subsequent ICD generator and lead complications including lead failure and lead dislodgement. ICD therapy was deemed inappropriate when triggered by supraventricular tachycardia, sinus rhythm, T-wave oversensing, lead dysfunction or occurred after spontaneous termination of a ventricular arrhythmia while lead dislodgement was defined as movement of a lead necessitating another procedure for repositioning.5 Overall, there were 16 deaths (39%) in the Auckland cohort and 15 deaths (32%) in the Wellington cohort. All cause mortality and device related morbidity were comparable between the two groups (Table 3). Mortality rates at 1 year (2.4% for Auckland cohort vs 4.2% for Wellington cohort, p=0.64) and 2 years (14.6% for Auckland cohort vs 8.5% for Wellington cohort, p=0.23) post ICD implantation were not significantly different between the two groups (Figure 1). Table 3. All-cause mortality and device-related morbidity Variables Auckland 75 or more cohort (n=41) Wellington 70-74 cohort (n=47) All-cause mortality 16 (39%) 15 (32%) p=0.46 > cardiac 9 8 Device related morbidity 16 (39%) 20 (43%) p=0.79 > Inappropriate therapy 5 (12%) 9 (19%) p=0.38 - 1 to 2 times 3 8 - more than 2 times 2 1 > Implant complications 5 (12%) 3 (6%) p=0.35 > Generator complications 5 (12%) 2 (4%) p=0.17 > Lead complications 1 (2%) 6 (13%) p=0.08 Figure 1. Survival in first 24 months post ICD implantation We compared the survival and device related morbidity of a group of elderly patients (age 75 or more) receiving ICDs in Auckland with a younger subset of patients (age 70-74) receiving ICDs in Wellington and found no significant difference in terms of outcomes. However, the small sample sizes could have prevented any significant differences from being detected. Compared to the cohort of patients age 75 or more receiving primary ICDs by van Rees et al3, the Auckland cohort had a longer median follow up, higher percentage of appropriate therapy and appropriate shocks but lower 1-year mortality post appropriate shock only. However, a majority of patients in the Auckland cohort had a secondary ICD indication which may have been a factor in the better outcome (Table 4). Table 4. Comparison between Auckland cohort with Van Rees et al Variables Auckland 75 or more cohort Van Rees et al Median follow up 4.25 years 1.6 years Appropriate therapy 54% 28% Appropriate therapy 39% 13% 1 year mortality post appropriate shock only 25% 35% Primary ICD implant 24% 100% The incidence of inappropriate shocks for the Auckland cohort was 9.8% which is higher than van Rees et al\u2019s 8%.3 However, it is comparable with those reported in MADIT-II (11.5%)6 and SCD-HeFT (9.9%).7 In conclusion, while the total number of patients is small, our data support the value of ICD implantation in carefully selected elderly patients as outcomes for patients of age 75 or more were similar to a subset of patients of age 70-74. Ming Han Lim Registrar Greenlane Cardiovascular Unit, Auckland City Hospital, Auckland District Health Board Auckland, New Zealand lmhan88@hotmail.com Warren Smith Associate Professor and Cardiologist Greenlane Cardiovascular Unit, Auckland City Hospital, Auckland District Health Board Auckland, New Zealand Peter Larsen Professor and Cardiologist Wellington Cardiovascular Research Group Wellington, New Zealand

An implantable cardioverter-defibrillator (ICD) is a treatment associated with significant cost and risk of morbidity. Elderly patients (arbitrarily defined as age 75 or more) were under-represented in large clinical trials (e.g. MADIT-II1 and SCD-HeFT2) which showed a mortality benefit with ICD implantation in patients with reduced left ventricular function. Therefore, the net mortality benefit for this age group is unclear.Van Rees et al3 found that the cumulative incidences of appropriate therapy and appropriate shocks in patients age 75 or more with primary ICD implanted were 28% and 13% respectively. Their all-cause mortality was 29% with a cumulative incidence for death of 35% at 1 year following appropriate shock. Currently, there is no such New Zealand data available.The primary objective of our study was to ascertain survival in elderly patients (age 75 or more) who have had ICDs implanted together with device related morbidity. The secondary objective was to compare these outcomes with a younger subset of patients (age 70-74) receiving ICD.This was a descriptive study with retrospective data collection on 41 patients aged 75 or more living in the Auckland region who received ICDs between 1 January 2000 and 31 December 2010. Demographic, clinical and survival data were collected retrospectively from Auckland District Health Board (DHB), Counties Manukau DHB, Waitemata DHB and regional ICD databases.For comparison, data on 47 consecutive patients receiving ICDs at age 70-74 between 1 January 2001 and 31 December 2010 were retrieved from the National Wellington ICD registry. Mean follow-up was 50 months for the Auckland (age 75 or more) cohort (range 7 to 150 months) and 56 months for the Wellington (age 70-74) cohort (range 3 to 143 months).Apart from age, the baseline characteristics were comparable (Table 1). All recorded variables were defined according to literature or common practice. Ischaemic heart disease was defined as the presence of a diameter stenosis of at least 50% in at least one coronary artery.4 Table 1. Clinical characteristics Variables Auckland 75 or more cohort (n=41) Wellington 70-74 cohort (n=47) Male gender 34 (83%) 40 (85%) Mean age (range) 77 years (75-84) 72 years (70-74) Primary implants 10 (24%) 13 (28%) Secondary implants 31 (76%) 34 (72%) > for cardiac arrest 10 18 > for ventricular tachycardia 21 16 History of atrial fibrillation 18 (44%) 9 (19%) p=0.07 Ischaemic heart disease 29 (71%) 38 (81%) Non ischaemic cardiomyopathy 14 (38%) 7 (22%) p=0.11 Prior revascularisation 18 (44%) 25 (53%) > CABG alone 15 (83%) 12 (48%) > PCI alone 2 (11%) 8 (32%) > CABG & PCI 1 (6%) 5 (20%) NYHA functional class Class I 14 (34%) 9 Class II 18 (44%) 14 Class III 7 (17%) 5 Class IV 2 (5%) 0 Unknown 19 QRS>150ms 15 (37%) 17 (36%) LBBB 28 (68%) Mean LVEF (range) 35% (19-56) 35% (11-70) EF 35% or less 28 (68%) 26 (55%) ICD therapy (including shock and anti-tachycardia pacing (ATP)) was classified as appropriate when initiated by ventricular tachycardia (VT) or ventricular fibrillation (VF) that was still present when therapy was delivered. Twenty-two patients (54%) in the Auckland cohort and 26 patients (55%) in the Wellington cohort received appropriate ICD therapy. There was no significant difference between the two groups (Table 2). Table 2. Appropriate ICD therapy Variables Auckland 75 or more cohort (n=41) Wellington 70-74 cohort (n=47) Appropriate ICD therapy 22 (54%) 26 (55%) p=0.88 Shock 16 (39%) 18 (38%) p=0.95 > more than 3 episodes 8 4 > shock alone 4 3 ATP 18 (44%) 23 (49%) p=0.64 > 1-3 episodes 8 13 > more than 3 episodes 10 10 Device-related morbidity comprised inappropriate ICD therapy, ICD implant complications, subsequent ICD generator and lead complications including lead failure and lead dislodgement. ICD therapy was deemed inappropriate when triggered by supraventricular tachycardia, sinus rhythm, T-wave oversensing, lead dysfunction or occurred after spontaneous termination of a ventricular arrhythmia while lead dislodgement was defined as movement of a lead necessitating another procedure for repositioning.5 Overall, there were 16 deaths (39%) in the Auckland cohort and 15 deaths (32%) in the Wellington cohort. All cause mortality and device related morbidity were comparable between the two groups (Table 3). Mortality rates at 1 year (2.4% for Auckland cohort vs 4.2% for Wellington cohort, p=0.64) and 2 years (14.6% for Auckland cohort vs 8.5% for Wellington cohort, p=0.23) post ICD implantation were not significantly different between the two groups (Figure 1). Table 3. All-cause mortality and device-related morbidity Variables Auckland 75 or more cohort (n=41) Wellington 70-74 cohort (n=47) All-cause mortality 16 (39%) 15 (32%) p=0.46 > cardiac 9 8 Device related morbidity 16 (39%) 20 (43%) p=0.79 > Inappropriate therapy 5 (12%) 9 (19%) p=0.38 - 1 to 2 times 3 8 - more than 2 times 2 1 > Implant complications 5 (12%) 3 (6%) p=0.35 > Generator complications 5 (12%) 2 (4%) p=0.17 > Lead complications 1 (2%) 6 (13%) p=0.08 Figure 1. Survival in first 24 months post ICD implantation We compared the survival and device related morbidity of a group of elderly patients (age 75 or more) receiving ICDs in Auckland with a younger subset of patients (age 70-74) receiving ICDs in Wellington and found no significant difference in terms of outcomes. However, the small sample sizes could have prevented any significant differences from being detected. Compared to the cohort of patients age 75 or more receiving primary ICDs by van Rees et al3, the Auckland cohort had a longer median follow up, higher percentage of appropriate therapy and appropriate shocks but lower 1-year mortality post appropriate shock only. However, a majority of patients in the Auckland cohort had a secondary ICD indication which may have been a factor in the better outcome (Table 4). Table 4. Comparison between Auckland cohort with Van Rees et al Variables Auckland 75 or more cohort Van Rees et al Median follow up 4.25 years 1.6 years Appropriate therapy 54% 28% Appropriate therapy 39% 13% 1 year mortality post appropriate shock only 25% 35% Primary ICD implant 24% 100% The incidence of inappropriate shocks for the Auckland cohort was 9.8% which is higher than van Rees et al\u2019s 8%.3 However, it is comparable with those reported in MADIT-II (11.5%)6 and SCD-HeFT (9.9%).7 In conclusion, while the total number of patients is small, our data support the value of ICD implantation in carefully selected elderly patients as outcomes for patients of age 75 or more were similar to a subset of patients of age 70-74. Ming Han Lim Registrar Greenlane Cardiovascular Unit, Auckland City Hospital, Auckland District Health Board Auckland, New Zealand lmhan88@hotmail.com Warren Smith Associate Professor and Cardiologist Greenlane Cardiovascular Unit, Auckland City Hospital, Auckland District Health Board Auckland, New Zealand Peter Larsen Professor and Cardiologist Wellington Cardiovascular Research Group Wellington, New Zealand