Results of a Pilot Study

Date: 3/18/2001
Presented by Karl-Heinz Kuck, M.D.
St. George Hospital, Hamburg
At the 50^th Annual Scientific Sessions of theAmerican College of Cardiology

Sponsored CPI Guidant GmgH &Co.
Medizintechnik KG

The purpose of the trial was to determine the natural courseof idiopathic dilated cardiomyopathy, the efficacy of ICD in the treatment ofthese patients, and the prognostic value of EP studies, Holter monitoring, stress testing, signal ECG, and echocardiograms. The purpose of the trial was to determine the outcome of ICD patients vs control with angiographically proven idiopathic dilated cardiomyopathy ina randomized, multi-center, prospective pilot study involving approximately 20 centers.

Inclusion criteria

• Angiographically documented dilated cardiomyopathy diagnosed within less than or equal to 9 months
• Left ventricular ejection fraction less than or equal to 30%
• New York Heart Association Class II or III
• Age 18-70 years

Exclusion criteria

• Sustained ventricular tachycardia
• Resuscitation
• Structural heart disease including valvular heart disease,coronary artery disease with stenoses > 70%
• Heart transplantation expected within 6 months
• Symptomatic cardiac arrhythmias requiring treatment

Study design
Patients meeting the inclusion criteria were randomized to two groups: control group or ICD therapy group.All patients were maintained on medical therapy, usually including digoxin, diuretics, and ACE inhibitors.

Patients underwent EP testing utilizing a protocol including at least 2premature ventricular stimuli, Holter monitoring, and left and right heart catheterization. At 3 and 6 months and 1year follow-up, an ECG was obtained, and an ICD check was performed. Holter, stress test, SAECG, Echocardiograms,and laboratory tests were performed.

At 24 months, aright heart catheterization and full clinical evaluation were performed.

Endpoints
Primary endpoint of the study was total mortality. When the main study was designed, the expected difference in mortality in the ICD group was estimated to be at least 6% less than that in the control group at 1 year. It was estimated that there would be a 30% 1 year mortality with40% due to sudden death and 60% due to heart failure. With an alpha error of 5%and a beta error of 20%, power calculations determined that 670 patients were required to prove the hypothesis.

Secondary endpoints included heart transplantation ,survival following cardiac arrest, occurrence of sustained ventricular arrhythmias, and symptomatic arrhythmias

An interim analysis was performed for the first 100 patientswith at least 1 year follow-up to see if the estimated mortality was correctand to see if at least 6% difference was observed in the two study groups.

The results of the interim analysis were presented.

 Baseline Characteristics

	ICD	Control
Number	N=50	N=54
Age	52+12	52+10
Gender	43M	40M
	7F	14F
EF	24%	25%
NYHA III	33%	36%

Medications

	ICD	Control
Diuretics	88%	76%
Digoxin	88%	76%
ACE Inhibitors	94%	98%
Calcium Channel blockers	16%	7%

Using an intention to treat analysis, the total mortality at1 year in 104 patients was 5.6%. Thiswas much less than the 30% expected total mortality. There was a 2.6% difference in mortality in favor of the control group but with no significant difference in the two groups. Based on this 1 year mortality, the trial was stopped and the original patients have been followed for a period of 7 years. After 5 years, there was a 30% mortality and after 7 years a 33%mortality. There was no difference inmortality in the two groups.

Years Follow-up: Number of patients

	1	2	3	4	5	6	7
ICD N=50	46	46	45	41	32	24	15
Control N=54	52	50	44	40	35	25	13

The overall mortality was made significantly worse by the presence of nonsustained ventricular tachycardia. There was almost no mortality in patients without nonsustained ventricular tachycardia and a 35% 5-year mortality inpatients with nonsustained ventricular tachycardia. Nonsustained ventricular tachycardia was present in approximately two-thirds of patients. There was approximately a 33% mortality in the group with nonsustained ventricular tachycardia

Years Follow-up

	0	1	2	3	4	5	6	7
Nonsustained VT	55	50	48	42	36	30	25	14
No NSVT	30	30	30	30	29	22	17	8

There were 30 deaths reported during the follow-upperiod. There were 19 cardiac deathswith 6 sudden deaths and 13 non sudden deaths due to heart failure. There were 2 non cardiac deaths and 8 unknown deaths.

There were 7 patients that crossed over from one group toanother. Four patients randomized to the ICD arm refused implantation of an ICD (3 patients) or had removal of the ICDs due to infection (1 patient).There were 3 patients randomized to the control arm that developed acardiac arrest and had an ICD implanted.

In the patients without NSVT, there was no difference inmortality between the ICD and control groups.

Years Follow-up

	0	1	2	3	4	5	6	7
ICD	16	16	16	16	16	11	8	4
Control	14	14	14	14	13	11	9	4

Similarly there was no benefit in the patients with nonsustained VT.

Years Follow-up

	0	1	2	3	4	5	6	7
ICD	26	22	22	21	18	14	13	8
Control	29	28	26	21	18	16	12	6

In summary, at 1 year, the mortality was 5.6% rather than the 30% mortality that was projected The 30% mortality was reached only after 5 years of follow-up. A 33.1% mortality was reached by 7 years. There was no survival benefit with the use ICDs in this patient population. Patients with NSVT had poorer survival.

Review written by Paul J. Wang, M.D.