- Email: [email protected]
Effects of enalapril on mortality in severe congestive heart failure: Results of the Cooperative North Scandinavian Enalapril Survival Study (CONSENSUS)
Effects of Enalapril on Mortality in Severe Congestive Heart Failure: Results of the Cooperative North Scandinavian Enalapril Survival Study (CONSENSUS) Karl Swedberg, MD, and John Kjekshus, MD, PhD, for the CONSENSUS Trial Study Group ~
To evaluate the influence of the angiotensin-convetting enzyme inhibitor, enalapril (2.5 to 40 mg/ day), on the prognosis of severe congestive heart failure, defined as New York Heart Association functional class iV, a double-blind study was undertaken in which 2 5 3 patients were randomized to receive either placebo (n = 126) or enaiaprii (n = 127) in addition to conventional treatment, including vasodilators. Follow-up averaged 188 days (range 1 day to 2 0 months). The reduction in crude mortality within 6 months (primary objective) was 40% in the enalapril-treated group (from 44 to 26%, p = 0.002) and within 1 year 31% (p = 0.001). By the end of the study, 68 subjects in the placebo group and 50 in the enalapril group had d i e d n a reduction of 27% (p = 0.003). The entire reduction in total mortality (50%) was found in patients dying from progressive heart failure, whereas no difference was seen in the incidence of sudden cardiac death. There was a significant improvement in New York Heart Association classification in the enalapril group, together with a reduction in heart size and a reduced requirement for other heart failure medication, it is concluded that the addition of enalapril to conventional therapy in patients with severe congestive heart failure can reduce mortality and improve symptoms. The effect seems to be due to a reduction in death from progression of heart failure. (Am J Cardiol 1988;62:60A-66A)
Address for reprints; Karl Swedberg, MD, Associate Professor, Department of Medicine, Gothenburg University, Ostra Hospital, S416 85 Gothenburg, Sweden. This work was supported by a grant from Merck Sharp & Dohme Research Laboratories, Rahway, New Jersey. * See Appendix.
60A
THE AMERICAN JOURNAL OF CARDIOLOGY VOLUME 62
he introduction of angiotensin-converting enzyme (ACE) inhibitors has been associated with both hemodynamic and symptomatic improvements in the long-term treatment of congestive heart failure (CHF).I To evaluate the effect of this therapy on survival, Furberg and Yusuf2 pooled the results of the placebocontrolled ACE inhibitor trials as well as other vasodilator trials (mentioned later in this issue). They found no overall effect of other vasodilators on survival, but suggested that ACE inhibitors might beneficially influence prognosis. Recently, a Veterans Administration (VA) trial showed that the combination of isosorbide dinitrate and hydralazine reduced mortality over a 3-year period) The Cooperative North Scandinavian Enalapril Survival Study was designed to study the effect on mortality of enalapril 4 compared with placebo, in addition to any conventional therapy, in patients with severe CHF. This paper presents the survival data from the 253 randomized patients (the main results have previously been published elsewhere) 5 and further analysis of clinically important factors in addition to mortality.
T
METHODS
This was a randomized, double-blind, placebo-controlled, parallel-group trial in patients with severe CHF (New York Heart Association [NYHA] functional class IV). Patients were randomized at each of 6 centers in Finland, 12 in Norway and 17 in Sweden. The study was organized by a Steering Committee. National monitors were in close contact with the Study Coordinator (KS), and this group had repeated meetings with the investigators. Questions could be discussed directly with the coordinator who was available on a 24-hour basis; every withdrawal was recommended for discussion with the coordinator. An international Ethical Review Committee monitored the progress of the study. At the start of the trial, patients were receiving optimal treatment with digitalis and diuretics and could also be treated with any other drug for heart failure, including vasodilators (e.g., nitrates, prazosin and hydralazine), but excluding ACE inhibitors. Diagnosis of CHF was based on clinical criteria; the patients had to be symptomatic at rest (NYHA functional class IV). Inclusion criteria required that the radiologic
heart size had to be >600 ml/m 2 body surface ~rea for men (normal <550 ml/m 2) or >550 ml/m 2 for women (normal <500 ml/m2). 6 Measurements of myocardial function were not required. A run-in period of not more than 14 days was allowed in order to stabilize the patients' conditions with digitalis and diuretics; patients who's conditions improved to class III or less were not randomized. Patients were excluded if any of the following criteria were present: acute pulmonary edema, hemodynamically important aortic or mitral valve stenosis, myocardial infarction within the previous 2 months, unstable angina, planned cardiac surgery, right heart failure due to pulmonary disease, or serum creatinine level >300/zmol/liter. Treatment with enalapril or identically matched placebo was started with 5 mg twice daily. At initiation and after dose increments, patients were observed for 6 hours for hypotension or any other adverse reactions. After 1 week, therapy was increased to I0 mg twice daily if the patient was not symptomatically hypotensive or had other adverse effects. According to clinical response, a further increase could be instituted up to a maximal dose of 20 mg twice daily. The patients were evaluated after 1, 2, 3, 6 and 16 weeks and thereafter at 6, 9 and 12 months and at the end of the study. In patients with worsening symptoms, additional vasodilator therapy with isosorbide dinitrate, hydralazine or prazosin, in that sequence, was recommended. Early in the trial, symptomatic hypotension led to a revision of the protocol after 67 patients had been randomized. No patients were unblinded. Patients considered to be at high risk (serum sodium <130 mmol/liter, serum creatinine 150 to 300 #mol/liter, an increased dose of diuretics within the last week, or treatment with potassium-sparing agents) were given an initial dose of 2.5 mg/day. If hypotension or an increase in serum creatinine levels did not occur after 3 or 4 days, the dose was increased to 2.5 mg twice daily for the remainder of the first week. Thereafter, the previous titration plan was followed. The main end points of the trial, as stated in the protocol, were 6-month mortality and mode of death. Based on detailed reports, mode of death was classified by 2 persons independently (KS, JK) before the code was broken. Death was described according to a modification of the American Heart Association's recommendations.7 Sudden death was defined as death within 1 hour from the onset of new symptoms. Unwitnessed deaths were classified according to the time from the last observation alive to the estimated time of death. Secondary objectives included 12-month and overall mortality during the whole trial period, according to the "intention-to-treat" principle. The study was performed under the auspices of the independent Ethical Review Committee, which every 3 months reviewed the number of inclusions, the number of deaths, and the instances of premature termination of test drug, separately for the enalapril group and placebo group.
TABLE I Baseline Characteristics Characteristic
Placebo (n = 126)
Enalapril (n = 127)
Mean
Mean
Age (years)
70
71
Weight (kg)
69
66
Heart size (ml/M2BSA) Systolic blood pressure (ram Hg) Diastolic blood pressure (ram Hg)* Heart rate (beats/rain)
853 121 76 80
875 118 74 79
Serum sodium (mmol/liter) Serum potassium (mmol/liter) Serum creatinine ~mol/liter) t
137 4.1 124 ' Percent
138 4.0 132 Percent
29 71
30 70
74 48 16 22 19 47 21
72 47 14 23 24 53 24
94 2
92 4
98 (200 mg) 55 (80 mg) 10
98 (210 rag) 50 (80 rag) 14
45 2 6 17 34
47 1 8
9 16 21 52 2
4 24 23 47 2
Sex
Female Male Etiologic factors Coronary artery disease Previous myocardial infarction Cardiomyopathy Valvular heart disease Hypertension Atrial fibrillation Diabetes mellitus Drug therapy Digitalis Beta blocker Diuretics Furosemide (mean dose) Spironolactone (mean dose) Any other diuretic Vasodilators Isosorbide dinitrate Hydralazine Prazosin Antiarrhythmics Anticoagulants Duration of heart failure <6 months 6-17 months 18-47 months >48 months Unknown
13
33
* p = 0.15; t p = 0.18. p > 0.25 if not otherwise stated. BSA = body surfacearea, Reprinted with permissionof N Engl J Med.s
STATISTICAL CONSIDERATIONS All randomized patients were included in the mortality analyses, and all patients with available data were included in the other analyses. All reported p values are 2tailed. Patient groups were compared with conventional methods. Differences in mortality between the treatment groups were analyzed using life-table methods. The intention-to-treat approach was used (i.e., the survival information for each patient from the date of randomization to the date of death or study termination was included in the analysis). The Kaplan-Meier survival curves were calculated, ~ and differences between the curves were analyzed using both the log-rank statistic9 and a Cox regression model, l° with treatment group as the only covariate. The 2 tests gave results that were virtually identical (p values were always within 0.002 of one
THE AMERICAN JOURNAL OF CARDIOLOGY JULY 11, 1988
6|A
A SYMPOSIUM: ADVANCES IN CONGESTIVE HEART FAILURE
0.80.7-
i1) E3
• .......
0.6-
J ............
I" • . . . . . . 2J"
"6
j"
,j- ---J
>,
~= 0.5J~
I - ....
0.4-
Z .....
:_.
~.1""
r.rJ~
:/ ' "
19 >
r_.r--J
"5 O.3"
=E
,---
O
!
O.2-
I
Placebo . . . . . . . . . Enalapril - -
O.1-
PIoc¢4~o, N t26 EnolopriI.N 127
I
!
|
I
I
I
I
I
l
1
2
3
4
5
6
7
8
9
10
11
12
102 111
78 98
63 88
59 82
53 79
Months 47 73
42 64
34 59
30 49
24 42
18 31
17 26
FIGURE 1. Cumulative probability of death in the 2 treatment groups. (Reprinted with permission from N Engl
Med. s)
TABLE II Crude Mortality from Any Cause Placebo (n = 126)
Mortality within 6 months (180 days) Mortality within 1 year (360 days) Total mortality
Enalapril (n = 127)
No.
%
No.
%
Percent Reduction in Relative Risk
55
44
33
26
40
0.002
66
52
46
36
31
0.001
68
54
50
39
27
0.003
p Values (Life-Table Analysis)
In the placebogroup, the mean number of daysof follow-upwas 237 daysamong the 58 survivorsand 93 daysamong the 68 deaths. (overallmean 160). In the enalaprilgroup, the mean number of days of follow-upwas 260 days among the 77 survivorsand 147 days among the bOdeaths (overall mean 215). Reprinted with permissionfrom N EnglJ Med)
TABLE
III M o r t a l i t y by M o d e of D e a t h
Mode of Death Any cardiac death Cardiac death within 24 hours Sudden cardiac death within 1 hour Progression of CHF Other cardiac death Stroke Other cardiovascular deaths* Noncardiovascular death (perforated ulcer) Total mortality
No. Taking Placebo (n = 126)
No. Taking Enalapril (n = 127)
p Values (Life-Table Analysis)
64 19
44 20
0.001 >0.25
14
14
>0.25
44 1 2 2 0
22 2 1 4 1
0.001 >0.25 >0.25 >0.25 >0.25
68
50
0.003
= includes patients dying from renal artery thrombosis, endocarditis, pulmonary emboli after leg amputation, bronchitis and concomitant heart failure, occlusion of femoral arterial graft, heart failure in relation to malena (ventricular ulcer). CHF = congestive heart failure. Reprinted with permission from N Engl J Med)
62A
THE AMERICAN JOURNAL OF CARDIOLOGY VOLUME 62
another); results from the Cox regression model are presented in this study. The study was approved by the Institutional Ethical Committee in each participating center. Informed consent was obtained from each patient. RESULTS
Because of a recommendation from the Ethical Review Committee (described later in this issue), the study was prematurely terminated. At that time, 253 patients had been randomized, 126 in the placebo group and 127 in the enalapril group. Baseline characteristics are presented in Table I. The groups were very similar at baseline. According to the entry criteria, all patients were in N Y H A functional class IV. Coronary artery disease (CAD) was the most frequently reported cause of heart failure (73%). In 49% of patients, vasodilator drugs, other than ACE inhibitors, were administered. The follow-up ranged from 1 day to 20 months (average 188 days). No
0.60.5-
p ........ I" ....
• .........
0.4FIGURE 2. Cumulative probability of death in patients dying from progressive heart failure (p <0.001).
.......
J
J
J
;j"
0.3-
.# 0.2-
/
I
s
i
i
0.1-
PLACEIO
..........
ENALAPRIL
-
-
!
i
!
i
,
i
i
i
i
i
!
.
1
2
3
4
5
6
7
8
9
10
11
12
MONTHS
0.7
0.7-
Group I ~-.Z
.......
Group II
J
0.6'
0.6"
#
,r- . . . . . . . . . . . . . . . .
0.5'
0.5" r- ............
f._J--~
0.4"
#
/
0.2. 0.1'
0
t
;-"
.,J
/
f--
.J
J
i
__J
r
04-
;J
r,'~ 0.3"
-
J
/
0.2-
.
.
.
.
.
.
.
.
J
.
,.-J-'
~ I
I
#
F~
z
~..t-'
0.3'
f
r- ....
S
S
I Placebo
..........
Enalapril
4
5 6 Months
;
8
9
0.1"
- -
1'0 11
1'2
Placebo
:
..........
Enalapril :3
4
5 6 Months
~'
8
9
1()
I1
1'2
FIGURE 3. Cumulative probability of death in patients not taking vasodilators (group I) and in those taking vasedilators at randomization (group II). (Reprinted with permission from N Engl J Med. s)
patient was lost to follow-up. Because of the premature termination, a 6-month follow-up was achieved in only 194 patients and a 12-month follow-up in 102 patients. The cumulative mortality rate up to 12 months is seen in Figure 1. Beyond 12 months, the sample size is too small for meaningful interpretation. The overall crude mortality at 6 months was 44% in the placebo group and 26% in the enalapril group--a reduction of 40% (p = 0.002; Table II). After 1 year, mortality was 52 and 36% (p = 0.001) in the 2 groups, respectively. At the end of the study, 68 patients had died in the placebo group and 50 patients in the enalapril group--a reduction of 27% (p ---
0.003). The mode of death is listed in Table III. Of the total 118 deaths in the study, 39 patients (33%) died suddenly and 66 patients (56%) died from progressive heart failure. There was no difference in the incidence of sudden cardi-
ac death between the 2 treatment groups; however, there was a 50% reduction in mortality due to progression of heart failure in the enalapril-treated group (p <0.001; Fig. 2). Autopsy was performed in 48% of the patients. In patients not treated with vasodilators at randomization, crude mortality was reduced from 60 to 37% (p <0.02). The corresponding 6-month life-table mortality figures were 50 and 34% (p -- 0.09; Fig. 3). Among the patients taking vasodilator therapy at baseline, the reduction was 48 to 42% (p = 0.11), and the 6-month life-table mortality figures were 46 and 24%, respectively (p = 0.009). Table IV presents the distribution of the N Y H A functional class at the end of the study. The distribution is significantly different between the treatment groups whether one considers all patients (with deaths included as the most severe category) or just the surviving patients THE AMERICAN JOURNAL OF CARDIOLOGY JULY 11, 1988
63A
A SYMPOSIUM" A D V A N C E S IN C O N G E S T I V E H E A R T F A I L U R E
real permitted dose level compared with 28 patients in the enalapril group (Table VII). In the enalapril group, there were significantly fewer initiations and more discontinuations of concomitant cardiovascular drugs than in the placebo group. The discontinued patients in the placebo group had been treated with a blinded drug before withdrawal for an average of 67 clays (median 28.5, range 0 to 311). The corresponding time for the patients treated with enalapril was 75 days (median 28.5, range 0 to 275). The mortality in the patients withdrawn was 78% in the placebo group and 68% in the enalapril group. The average survival after withdrawal in the placebo group was 27 days (median 28, range 2 to 75) and in the enalapril group 80 days (median 69, range 2 to 287).
T A B L E I V NYHA Classifications at End of S t u d y
NYHA Classification
No. Taking Placebo (n = 126)
No. Taking Enalapril (n = 127)
I
0
3
II III IV Unknown Patient died
2 25 30 1 68
13 38 21 2 50
For the differencesbetweenthe groups, p < 0.001. NYHA= New York Heart Association. Reprinted with permission from N EnglJ Med.5
(p = 0.001). Improvement in N Y H A classification was seen in 22% of the placebo-treated patients and 42% of the enalapril-treated patients. Among patients surviving at the end of the study, heart size was reduced in the placebo group, from 867 ml/m 2 body surface area at baseline to 839 ml/m 2 body surface area (-3.2%) and in the enalapril group, from 910 to 823 ml/m 2 body surface area (-9.6%). The reduction in heart size was significantly greater in the enalapril group (p = 0.02). There was a significant decrease in systolic blood pressure in both groups after the first dose, but it was greater in the enalapril group (from 118 to 98 mm Hg [ - 2 0 mm Hg] vs -11 mm Hg in the placebo group; p <0.01). Heart rate was unchanged in the placebo group (mean 80 beats/rain), but decreased among the enalapril patients (from 80 to 77 beats/rain; p <0.05). The effect of enalapril was confined to patients in whom CAD was the cause of heart failure (Table V). The 6-month crude mortality in this patient group was reduced from 51 to 24% (p <0.001). Table VI lists some subgroups in relation to median values. Ages did not influence mortality. Patients with baseline serum creatinine > 120 #mol/liter or serum sodium < 137 mmol/liter seemed to benefit more from enalapril. The final mean dose level of enalapril was 18.4 mg and of matching placebo 27.3 mg (p <0.001). In the placebo group, 57 patients had been titrated to the maxi-
DISCUSSION This trial clearly demonstrated that treatment with enalapril in severe chronic C H F improves survival. A change in N Y H A classification from functional class IV to I or II was observed in 16 patients in the enalapril group compared with only 2 in the placebo group. The beneficial treatment effects were also reflected by reduction in use of concomitant cardiovascular drugs and reduction in heart size in enalapril-treated patients. The entire treatment effect was due to a reduced mortality from progression of heart failure. There was no indication of any effect on sudden cardiac death. C H F is a serious condition with a reported high mortality, ~ as also demonstrated by the present placebo group. The 6- and 12-month mortality rates calculated from the life-table analysis were 48 and 63%, respectively, in the placebo group. These figures are even higher than were anticipated at the begitming of the study, but are in agreement with survival in other functional class IV patients. 12One should keep in mind that N Y H A class IV is not a clearly delineated group. In this trial, the patients did not have to be dyspneic at rest all the time, but instead they could be mainly fatigued and marked by their heart failure at rest with tachypnea. If there were doubts whether a patient was in class IV, the investigators were encouraged not to include the patient. The mean age of the study patients was 70 years-somewhat older than those in several other trials in which
T A B L E V Crude Mortality (%) by Etiologic Cause of Heart Failure 6-Month
Overall
Etiology
Placebo
Enatapril
p
Placebo
Enalapril
p
Coronary artery disease (n = 192) Other causes (n = 61) Diabetes Yes (n = 56) No (n = 192) Atrial fibrillation Yes (n = 126) No (n = 127)
51
24
<0.001
61
40
<0.001
24
31
>0.25
33
37
>0.25
30 48
35 24
>0.25 0.001
52 55
55 35
>0.25 0.001
43 45
29 23
0.14 0.014
52 55
38 40
0.052 0.015
64A
THE AMERICAN JOURNAL OF CARDIOLOGY
VOLUME 62
the mean age was 60 to 65 years. In contrast, this study had no upper age limit. Our patients are, therefore, probably more representative of an unselected heart failure population. The older age may also contribute to the high mortality rate. The treatment effect, however, was independent of age (Table VII). The VA trial compared placebo with prazosin and with a combination of hydralazine and isosorbide dinitrate. 3 Prazosin had no effect on mortality, but on the basis of life-table estimates, the hydralazine-nitrate combination reduced mortality after 1 year from 19.5 to 12.1% (-38%) and after 3 years from 46.9 to 36.2% (-23%). In the present trial, a similar calculation of life-table estimates shows a mortality after 6 months of 48% reduced to 29% (-40%) and after 1 year of 63% reduced to 47% (-25%). The 6-month mortality in the present trial was similar to the mortality risk after 3 years in the VA trial, showing that we were studying a more severely ill group of patients. Patients in whom CAD was the cause of heart failure had a much higher mortality than those with other causes (Table V). This is in agreement with results from the VA trial. The effect of enalapril was only seen in those patients with ischemic heart disease. Table VI shows that the 6-month crude mortality in the placebo group varied between 36 and 51%. In all subgroups, there was a consistent reduction in mortality. It is interesting to note that mortality among the patients taking enalapril was fairly constant, between 23 and 29%. It has been argued that these patients are so ill and have such extensive myocardial damage that no important prolongation of survival by any medical treatment can be expected. 13The fact that the mortality curves are parallel after the initial divergence indicates a maintained effect of enalapril for at least 1 year, because there will be relatively more high-risk patients remaining in the enalapril group. The limited follow-up period does not permit an estimation of the duration of this beneficial effect or of the drug's effects in less severe forms of heart failure. The fact that death from progressive heart failure was greatly decreased may have important implications for the future understanding of myocardial dysfunction. Progression of heart failure could be due to a new myocardial infarction, continued myocardial fibrosis, or metabolically based deterioration of myocardial function. From this trial, we can only speculate as to how progression was influenced. Extracardiac mechanisms, such as neuroendocrine alterations, could be of major importance for the progression of heart failure. The reduction in mortality among patients treated with vasodilators at randomization was similar to that in those not taking vasodilators. However, in patients taking vasodilators, the treatment effect appeared to be more pronounced during the first 6 months. Vasodilation unloads the failing myocardium, but the pharmacologic means by which peripheral resistance is lowered may be a critical factor. Enalapril probably lowers peripheral resistance by reducing angiotensin-II-in-
TABLE V l Crude Mortality (6 month) in Some Subgroups Divided According to Median
Subgroup Age (yr) _<70 >70 Serum creatinine (#tool/liter) _<120 > 120 Serum sodium (mmol/liter) _<137 >137 Heart size (ml/m 2 BSA) _<800 >800 Heart rate (beats/rain) _<80 >8O
Placebo (%)
Enalapril (%)
Reduction in Mortality (%) p
42 45
23 28
45 38
0.03 0.02
38 50
29 24
24 52
0,12 0.002
51 38
25 27
51 24
0.004 0.12
36 49
26 26
28 47
>0.25 0.003
35 49
23 28
34 43
>0.25 <0.001
BSA= bodysurfacearea.
TABLE VII Dosage Levels of Test Drug at End of Study Dose (rag/day)
Placebo (n : 126)
Enalapril (n : 127)
2.5 5 10 20 40 Unknown
1 7 13 41 57 7
16 10 33 30 28 10
duced vasoconstriction. 14Conversely, conventionally acting vasodilators may actually stimulate the renin-angiotensin system. ~5 Could the effects of enalapril on survival in severe heart failure be found in patients with less pronounced symptoms? This important question cannot be answered from the present study. However, it is reasonable to believe that the degree of myocardial depression is a most important prognostic indicator. Although we did not measure myocardial function directly, it could be presumed that severe myocardial dysfunction was present in these patients based on clinical history and x-ray findings. It is possible that among patients with a similar degree of myocardial dysfunction, a similar effect on progression of heart failure could be found, even if they are not in NYHA class IV. However, that question must be evaluated further, and studies are underway.
CONCLUSIONS In summary, this study has shown that enalapril given to patients with severe CHF is associated with a marked reduction in mortality. This treatment was well tolerated; however, it is important to initiate therapy with low doses of enalapril, particularly in high-risk patients. THE AMERICAN JOURNAL OF CARDIOLOGY JULY 11, 1988
65A
A SYMPOSIUM: ADVANCES IN CONGESTIVE HEART FAILURE
REFERENCES
!. Lipkin D, Poole-WilsonP. Treatment of chronic heart failure." A review of
recent drug trials. Br Med J 1985;291:993-996. 2. FurbergCD, YusufS. Effect of vasodilators on survival in chronic congestive heart failure. Am J Cardiol 1985;55.'11I0-1113.
CohnJN, ArchibaldDG,ZiescheS, FrancioscaJA, HarstonWE,Tristani FE, DunkmanWB, JacobsW, FrancisGS, FlohrKH, GoldmanS, Cobb FR, Shah PM, Saunders R, Fletcher RD, Loeb HS, Hughes VC, Baker B. Effect of
3.
vasodilator therapy on mortality in chronic congestive heart failure. N Engl J Med 1986;314:1547-1552. 4.. ToddPA, HeelRC. Enalapril: a reviewof its pharmacodynamic and pharmacokinetic properties, and therapeutic use in hypertension and congestive heart failure. Drugs 1986;31:198-248. 5. The CONSENSUSTrial StudyGroup.Effects ofenalapril on mortality in severe congestive heart failure: results of the Cooperative North Scandinavian EnalaprilSurvival Study (CONSENSUS). N Engl J Med 1987;316:1429-1435. 6. JoosellS. A method for the determination of the heart size by teleroentgenography (a heart volume index). Acta Radiol 1939;20:325-340. 7. Gillum RF, FortmannSP, Prineas RJ, KonkeTE. International diagnostic criteria for acute myocardial infarction and acute stroke. Am Heart J 1985; 108:150-154. 8. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc 1958;53.'457-481. 9. Mantel N. Chi-square test with one degree of freedom: Extension of the ManteI-Haenzel procedure. J Am State Assoc 1963;58.'690-700. 10. Cos DR. Regression methods and life tables. J Roy Stat Soc 1972;34:187220. 11. PackerM. Sudden unexpected death in patients with congestive heart failure: a second frontier. Circulation 1985;72:681-685. 12. McFateSmith W. Epidemiology of congestive heartfailure. Am J Cardiol 1985;55.'3A -SA. 13. PfefferMA.PfefferJM, SteinburgC. Finn P.Surt,ival afteran experimental myocardial infarction." beneficial effects of long-term therapy with captopril. Circulation 1985;72:406-412.
14. BaylissJ, Norell MS, Canepa-AnsonR, ReidC, Poole-WilsonP, SuttonG. Clinical importance of the renin-angiotensin system in chronic heart failure: double-blind comparison of captopril and prazosin. Br Med J 1985;290;18111865. 15. UedaK,Sakai M, MatsushitaS, Kuwajima1,MurakamiM. Effect of orally administered hydralazine on neurohumoral factors and hemodynamic response in aged patients with chronic congestive heartfailure. Jpn Heart J 1983;29:711721.
APPENDIX The CONSENSUS Trial Study group consisted of the following investigators (names are followed by locations and numbers of patients): Finland." U. Id~np~i~n-Heikkil~i, MD (Espoo, 2); J. Remes, MD, T. Hirvonen, MD, and K. Py6r/il5., MD (Kuopio, 10); U. Korhonen, MD, J. Takkunen, MD, and L. H~.m~il~inen, MD (Oulu, 10); A. Pasternack, MD, and M. Nikkil/i, MD (Tampere, 10); M. Arstila, MD (Turku, 3); and M. Lindroos, MD, and B. Heikius, MD (Vaasa, 9). N o r w a y : T .
66A
THE AMERICAN JOURNAL OF CARDIOLOGY VOLUME62
Gundersen, MD, and B. Aslaksen, MD (Arendal, 11); L. Gullestad, MD, and E. Soyland, MD (Baerum, 11); G. Tjonnfjord, MD, and A. Ose, MD (Gjovik, 3); P. Aukrust, MD, and J. Haerem, MD (Hamar, 7); K. Waage, MD, and F. Riddervold, MD (Haugesund, 5); J. Offstad, MD, and E. G. Aksnes, MD (Lillehammer, 10); T. Pedersen, MD, and P. Sirnes, MD (Oslo, Aker Hospital, 8); A. Westheim, MD, E. Sivertssen, MD, and J. Eritsland, MD (Oslo, Ullev~l Hospital, 11); K. Overskeid, MD, and O. Brubakk, MD (Sarpsborg, 4); J. Naesgaard, MD, and A. Hallargker, MD (Skien, 6); K. Dickstein, MD, and V. Bonarjee, MD (Stavanger, 10); H. Wang, MD, and E. S. P. Myhre, MD (Tromso, 3). Sweden: S. A. Forsberg, MD, and L. Karelds, MD (BorS.s, 6); K. Lindvall, MD, and S. Eriksson, MD (Danderyd, 10); G. Frithz, MD, and A. Stjerna, MD (Eskilstuna, 6); G. Ahlmark, MD, and H. Saetre, MD (Falun, 7); B. Andersson, MD, and M. Hartford, MD (Gothenburg, Sahlgrenska Hospital, 8); R. Bergstrand, MD, P. Held, MD, and G. Ulvenstam, MD (Gothenburg, Ostra Hospital, 17); L. Lundkvist, MD, and P. Andersson, MD (Hudiksvall, 4); U. Dahlstr6m, MD, and M. Broquvist, MD (Link6ping, 17); A. Henriksson, MD, and A. S6derling, MD (Lule~t, 7); S. Persson, MD (Lund, 1); N. Stobaeus, MD (Motala, 2); A. Sjfgren, MD, and E. Loogna, MD (Nacka, 5); J. Frid6n, MD, and P. O. Andersson, MD (Norrkfping, 5); L. Ryd6n, MD, and P. Smedg~trd, MD (Sk6vde, 5); I. Liljefors, MD, and C. Hofman-Bang, MD (Stockholm, Sabbatsberg Hospital, 3); M. Lycksell, MD, and C. Ringqvist, MD (Sundsvall, I 1); O. Johnson, MD, and L.-O. Hemmingson, MD (Ume~t, 5). The members of the CONSENSUS Steering Committee were as follows: J. Kjekshus MD, Sandvika, Norway (Chairman); H. Frick MD, Helsinki, Finland; K. Swedberg MD, Gothenburg, Sweden; and L. Wilhelmsen MD, Gothenburg, Sweden. Coordinating Office: K. Swedberg MD, Ostra Hospital, Gothenburg, Sweden (Coordinator); G. Andersson RN, Ostra Hospital, Gothenburg, Sweden. The members of the Ethical Review Committee were as follows: J. Lubsen, MD, The Netherlands (Chairman); D. Julian, MD, Great Britain; B. Levine, MD, United States; and W. McFate Smith, MD, United States. Administration Office: K. Kristianson, PhD, Bromma, Sweden (Head). Statistical Analysis: S. Snapinn, PhD, Merck Sharp & Dohme Research Laboratories, Brussels, Belgium.
To evaluate the influence of the angiotensin-convetting enzyme inhibitor, enalapril (2.5 to 40 mg/ day), on the prognosis of severe congestive heart failure, defined as New York Heart Association functional class iV, a double-blind study was undertaken in which 2 5 3 patients were randomized to receive either placebo (n = 126) or enaiaprii (n = 127) in addition to conventional treatment, including vasodilators. Follow-up averaged 188 days (range 1 day to 2 0 months). The reduction in crude mortality within 6 months (primary objective) was 40% in the enalapril-treated group (from 44 to 26%, p = 0.002) and within 1 year 31% (p = 0.001). By the end of the study, 68 subjects in the placebo group and 50 in the enalapril group had d i e d n a reduction of 27% (p = 0.003). The entire reduction in total mortality (50%) was found in patients dying from progressive heart failure, whereas no difference was seen in the incidence of sudden cardiac death. There was a significant improvement in New York Heart Association classification in the enalapril group, together with a reduction in heart size and a reduced requirement for other heart failure medication, it is concluded that the addition of enalapril to conventional therapy in patients with severe congestive heart failure can reduce mortality and improve symptoms. The effect seems to be due to a reduction in death from progression of heart failure. (Am J Cardiol 1988;62:60A-66A)
Address for reprints; Karl Swedberg, MD, Associate Professor, Department of Medicine, Gothenburg University, Ostra Hospital, S416 85 Gothenburg, Sweden. This work was supported by a grant from Merck Sharp & Dohme Research Laboratories, Rahway, New Jersey. * See Appendix.
60A
THE AMERICAN JOURNAL OF CARDIOLOGY VOLUME 62
he introduction of angiotensin-converting enzyme (ACE) inhibitors has been associated with both hemodynamic and symptomatic improvements in the long-term treatment of congestive heart failure (CHF).I To evaluate the effect of this therapy on survival, Furberg and Yusuf2 pooled the results of the placebocontrolled ACE inhibitor trials as well as other vasodilator trials (mentioned later in this issue). They found no overall effect of other vasodilators on survival, but suggested that ACE inhibitors might beneficially influence prognosis. Recently, a Veterans Administration (VA) trial showed that the combination of isosorbide dinitrate and hydralazine reduced mortality over a 3-year period) The Cooperative North Scandinavian Enalapril Survival Study was designed to study the effect on mortality of enalapril 4 compared with placebo, in addition to any conventional therapy, in patients with severe CHF. This paper presents the survival data from the 253 randomized patients (the main results have previously been published elsewhere) 5 and further analysis of clinically important factors in addition to mortality.
T
METHODS
This was a randomized, double-blind, placebo-controlled, parallel-group trial in patients with severe CHF (New York Heart Association [NYHA] functional class IV). Patients were randomized at each of 6 centers in Finland, 12 in Norway and 17 in Sweden. The study was organized by a Steering Committee. National monitors were in close contact with the Study Coordinator (KS), and this group had repeated meetings with the investigators. Questions could be discussed directly with the coordinator who was available on a 24-hour basis; every withdrawal was recommended for discussion with the coordinator. An international Ethical Review Committee monitored the progress of the study. At the start of the trial, patients were receiving optimal treatment with digitalis and diuretics and could also be treated with any other drug for heart failure, including vasodilators (e.g., nitrates, prazosin and hydralazine), but excluding ACE inhibitors. Diagnosis of CHF was based on clinical criteria; the patients had to be symptomatic at rest (NYHA functional class IV). Inclusion criteria required that the radiologic
heart size had to be >600 ml/m 2 body surface ~rea for men (normal <550 ml/m 2) or >550 ml/m 2 for women (normal <500 ml/m2). 6 Measurements of myocardial function were not required. A run-in period of not more than 14 days was allowed in order to stabilize the patients' conditions with digitalis and diuretics; patients who's conditions improved to class III or less were not randomized. Patients were excluded if any of the following criteria were present: acute pulmonary edema, hemodynamically important aortic or mitral valve stenosis, myocardial infarction within the previous 2 months, unstable angina, planned cardiac surgery, right heart failure due to pulmonary disease, or serum creatinine level >300/zmol/liter. Treatment with enalapril or identically matched placebo was started with 5 mg twice daily. At initiation and after dose increments, patients were observed for 6 hours for hypotension or any other adverse reactions. After 1 week, therapy was increased to I0 mg twice daily if the patient was not symptomatically hypotensive or had other adverse effects. According to clinical response, a further increase could be instituted up to a maximal dose of 20 mg twice daily. The patients were evaluated after 1, 2, 3, 6 and 16 weeks and thereafter at 6, 9 and 12 months and at the end of the study. In patients with worsening symptoms, additional vasodilator therapy with isosorbide dinitrate, hydralazine or prazosin, in that sequence, was recommended. Early in the trial, symptomatic hypotension led to a revision of the protocol after 67 patients had been randomized. No patients were unblinded. Patients considered to be at high risk (serum sodium <130 mmol/liter, serum creatinine 150 to 300 #mol/liter, an increased dose of diuretics within the last week, or treatment with potassium-sparing agents) were given an initial dose of 2.5 mg/day. If hypotension or an increase in serum creatinine levels did not occur after 3 or 4 days, the dose was increased to 2.5 mg twice daily for the remainder of the first week. Thereafter, the previous titration plan was followed. The main end points of the trial, as stated in the protocol, were 6-month mortality and mode of death. Based on detailed reports, mode of death was classified by 2 persons independently (KS, JK) before the code was broken. Death was described according to a modification of the American Heart Association's recommendations.7 Sudden death was defined as death within 1 hour from the onset of new symptoms. Unwitnessed deaths were classified according to the time from the last observation alive to the estimated time of death. Secondary objectives included 12-month and overall mortality during the whole trial period, according to the "intention-to-treat" principle. The study was performed under the auspices of the independent Ethical Review Committee, which every 3 months reviewed the number of inclusions, the number of deaths, and the instances of premature termination of test drug, separately for the enalapril group and placebo group.
TABLE I Baseline Characteristics Characteristic
Placebo (n = 126)
Enalapril (n = 127)
Mean
Mean
Age (years)
70
71
Weight (kg)
69
66
Heart size (ml/M2BSA) Systolic blood pressure (ram Hg) Diastolic blood pressure (ram Hg)* Heart rate (beats/rain)
853 121 76 80
875 118 74 79
Serum sodium (mmol/liter) Serum potassium (mmol/liter) Serum creatinine ~mol/liter) t
137 4.1 124 ' Percent
138 4.0 132 Percent
29 71
30 70
74 48 16 22 19 47 21
72 47 14 23 24 53 24
94 2
92 4
98 (200 mg) 55 (80 mg) 10
98 (210 rag) 50 (80 rag) 14
45 2 6 17 34
47 1 8
9 16 21 52 2
4 24 23 47 2
Sex
Female Male Etiologic factors Coronary artery disease Previous myocardial infarction Cardiomyopathy Valvular heart disease Hypertension Atrial fibrillation Diabetes mellitus Drug therapy Digitalis Beta blocker Diuretics Furosemide (mean dose) Spironolactone (mean dose) Any other diuretic Vasodilators Isosorbide dinitrate Hydralazine Prazosin Antiarrhythmics Anticoagulants Duration of heart failure <6 months 6-17 months 18-47 months >48 months Unknown
13
33
* p = 0.15; t p = 0.18. p > 0.25 if not otherwise stated. BSA = body surfacearea, Reprinted with permissionof N Engl J Med.s
STATISTICAL CONSIDERATIONS All randomized patients were included in the mortality analyses, and all patients with available data were included in the other analyses. All reported p values are 2tailed. Patient groups were compared with conventional methods. Differences in mortality between the treatment groups were analyzed using life-table methods. The intention-to-treat approach was used (i.e., the survival information for each patient from the date of randomization to the date of death or study termination was included in the analysis). The Kaplan-Meier survival curves were calculated, ~ and differences between the curves were analyzed using both the log-rank statistic9 and a Cox regression model, l° with treatment group as the only covariate. The 2 tests gave results that were virtually identical (p values were always within 0.002 of one
THE AMERICAN JOURNAL OF CARDIOLOGY JULY 11, 1988
6|A
A SYMPOSIUM: ADVANCES IN CONGESTIVE HEART FAILURE
0.80.7-
i1) E3
• .......
0.6-
J ............
I" • . . . . . . 2J"
"6
j"
,j- ---J
>,
~= 0.5J~
I - ....
0.4-
Z .....
:_.
~.1""
r.rJ~
:/ ' "
19 >
r_.r--J
"5 O.3"
=E
,---
O
!
O.2-
I
Placebo . . . . . . . . . Enalapril - -
O.1-
PIoc¢4~o, N t26 EnolopriI.N 127
I
!
|
I
I
I
I
I
l
1
2
3
4
5
6
7
8
9
10
11
12
102 111
78 98
63 88
59 82
53 79
Months 47 73
42 64
34 59
30 49
24 42
18 31
17 26
FIGURE 1. Cumulative probability of death in the 2 treatment groups. (Reprinted with permission from N Engl
Med. s)
TABLE II Crude Mortality from Any Cause Placebo (n = 126)
Mortality within 6 months (180 days) Mortality within 1 year (360 days) Total mortality
Enalapril (n = 127)
No.
%
No.
%
Percent Reduction in Relative Risk
55
44
33
26
40
0.002
66
52
46
36
31
0.001
68
54
50
39
27
0.003
p Values (Life-Table Analysis)
In the placebogroup, the mean number of daysof follow-upwas 237 daysamong the 58 survivorsand 93 daysamong the 68 deaths. (overallmean 160). In the enalaprilgroup, the mean number of days of follow-upwas 260 days among the 77 survivorsand 147 days among the bOdeaths (overall mean 215). Reprinted with permissionfrom N EnglJ Med)
TABLE
III M o r t a l i t y by M o d e of D e a t h
Mode of Death Any cardiac death Cardiac death within 24 hours Sudden cardiac death within 1 hour Progression of CHF Other cardiac death Stroke Other cardiovascular deaths* Noncardiovascular death (perforated ulcer) Total mortality
No. Taking Placebo (n = 126)
No. Taking Enalapril (n = 127)
p Values (Life-Table Analysis)
64 19
44 20
0.001 >0.25
14
14
>0.25
44 1 2 2 0
22 2 1 4 1
0.001 >0.25 >0.25 >0.25 >0.25
68
50
0.003
= includes patients dying from renal artery thrombosis, endocarditis, pulmonary emboli after leg amputation, bronchitis and concomitant heart failure, occlusion of femoral arterial graft, heart failure in relation to malena (ventricular ulcer). CHF = congestive heart failure. Reprinted with permission from N Engl J Med)
62A
THE AMERICAN JOURNAL OF CARDIOLOGY VOLUME 62
another); results from the Cox regression model are presented in this study. The study was approved by the Institutional Ethical Committee in each participating center. Informed consent was obtained from each patient. RESULTS
Because of a recommendation from the Ethical Review Committee (described later in this issue), the study was prematurely terminated. At that time, 253 patients had been randomized, 126 in the placebo group and 127 in the enalapril group. Baseline characteristics are presented in Table I. The groups were very similar at baseline. According to the entry criteria, all patients were in N Y H A functional class IV. Coronary artery disease (CAD) was the most frequently reported cause of heart failure (73%). In 49% of patients, vasodilator drugs, other than ACE inhibitors, were administered. The follow-up ranged from 1 day to 20 months (average 188 days). No
0.60.5-
p ........ I" ....
• .........
0.4FIGURE 2. Cumulative probability of death in patients dying from progressive heart failure (p <0.001).
.......
J
J
J
;j"
0.3-
.# 0.2-
/
I
s
i
i
0.1-
PLACEIO
..........
ENALAPRIL
-
-
!
i
!
i
,
i
i
i
i
i
!
.
1
2
3
4
5
6
7
8
9
10
11
12
MONTHS
0.7
0.7-
Group I ~-.Z
.......
Group II
J
0.6'
0.6"
#
,r- . . . . . . . . . . . . . . . .
0.5'
0.5" r- ............
f._J--~
0.4"
#
/
0.2. 0.1'
0
t
;-"
.,J
/
f--
.J
J
i
__J
r
04-
;J
r,'~ 0.3"
-
J
/
0.2-
.
.
.
.
.
.
.
.
J
.
,.-J-'
~ I
I
#
F~
z
~..t-'
0.3'
f
r- ....
S
S
I Placebo
..........
Enalapril
4
5 6 Months
;
8
9
0.1"
- -
1'0 11
1'2
Placebo
:
..........
Enalapril :3
4
5 6 Months
~'
8
9
1()
I1
1'2
FIGURE 3. Cumulative probability of death in patients not taking vasodilators (group I) and in those taking vasedilators at randomization (group II). (Reprinted with permission from N Engl J Med. s)
patient was lost to follow-up. Because of the premature termination, a 6-month follow-up was achieved in only 194 patients and a 12-month follow-up in 102 patients. The cumulative mortality rate up to 12 months is seen in Figure 1. Beyond 12 months, the sample size is too small for meaningful interpretation. The overall crude mortality at 6 months was 44% in the placebo group and 26% in the enalapril group--a reduction of 40% (p = 0.002; Table II). After 1 year, mortality was 52 and 36% (p = 0.001) in the 2 groups, respectively. At the end of the study, 68 patients had died in the placebo group and 50 patients in the enalapril group--a reduction of 27% (p ---
0.003). The mode of death is listed in Table III. Of the total 118 deaths in the study, 39 patients (33%) died suddenly and 66 patients (56%) died from progressive heart failure. There was no difference in the incidence of sudden cardi-
ac death between the 2 treatment groups; however, there was a 50% reduction in mortality due to progression of heart failure in the enalapril-treated group (p <0.001; Fig. 2). Autopsy was performed in 48% of the patients. In patients not treated with vasodilators at randomization, crude mortality was reduced from 60 to 37% (p <0.02). The corresponding 6-month life-table mortality figures were 50 and 34% (p -- 0.09; Fig. 3). Among the patients taking vasodilator therapy at baseline, the reduction was 48 to 42% (p = 0.11), and the 6-month life-table mortality figures were 46 and 24%, respectively (p = 0.009). Table IV presents the distribution of the N Y H A functional class at the end of the study. The distribution is significantly different between the treatment groups whether one considers all patients (with deaths included as the most severe category) or just the surviving patients THE AMERICAN JOURNAL OF CARDIOLOGY JULY 11, 1988
63A
A SYMPOSIUM" A D V A N C E S IN C O N G E S T I V E H E A R T F A I L U R E
real permitted dose level compared with 28 patients in the enalapril group (Table VII). In the enalapril group, there were significantly fewer initiations and more discontinuations of concomitant cardiovascular drugs than in the placebo group. The discontinued patients in the placebo group had been treated with a blinded drug before withdrawal for an average of 67 clays (median 28.5, range 0 to 311). The corresponding time for the patients treated with enalapril was 75 days (median 28.5, range 0 to 275). The mortality in the patients withdrawn was 78% in the placebo group and 68% in the enalapril group. The average survival after withdrawal in the placebo group was 27 days (median 28, range 2 to 75) and in the enalapril group 80 days (median 69, range 2 to 287).
T A B L E I V NYHA Classifications at End of S t u d y
NYHA Classification
No. Taking Placebo (n = 126)
No. Taking Enalapril (n = 127)
I
0
3
II III IV Unknown Patient died
2 25 30 1 68
13 38 21 2 50
For the differencesbetweenthe groups, p < 0.001. NYHA= New York Heart Association. Reprinted with permission from N EnglJ Med.5
(p = 0.001). Improvement in N Y H A classification was seen in 22% of the placebo-treated patients and 42% of the enalapril-treated patients. Among patients surviving at the end of the study, heart size was reduced in the placebo group, from 867 ml/m 2 body surface area at baseline to 839 ml/m 2 body surface area (-3.2%) and in the enalapril group, from 910 to 823 ml/m 2 body surface area (-9.6%). The reduction in heart size was significantly greater in the enalapril group (p = 0.02). There was a significant decrease in systolic blood pressure in both groups after the first dose, but it was greater in the enalapril group (from 118 to 98 mm Hg [ - 2 0 mm Hg] vs -11 mm Hg in the placebo group; p <0.01). Heart rate was unchanged in the placebo group (mean 80 beats/rain), but decreased among the enalapril patients (from 80 to 77 beats/rain; p <0.05). The effect of enalapril was confined to patients in whom CAD was the cause of heart failure (Table V). The 6-month crude mortality in this patient group was reduced from 51 to 24% (p <0.001). Table VI lists some subgroups in relation to median values. Ages did not influence mortality. Patients with baseline serum creatinine > 120 #mol/liter or serum sodium < 137 mmol/liter seemed to benefit more from enalapril. The final mean dose level of enalapril was 18.4 mg and of matching placebo 27.3 mg (p <0.001). In the placebo group, 57 patients had been titrated to the maxi-
DISCUSSION This trial clearly demonstrated that treatment with enalapril in severe chronic C H F improves survival. A change in N Y H A classification from functional class IV to I or II was observed in 16 patients in the enalapril group compared with only 2 in the placebo group. The beneficial treatment effects were also reflected by reduction in use of concomitant cardiovascular drugs and reduction in heart size in enalapril-treated patients. The entire treatment effect was due to a reduced mortality from progression of heart failure. There was no indication of any effect on sudden cardiac death. C H F is a serious condition with a reported high mortality, ~ as also demonstrated by the present placebo group. The 6- and 12-month mortality rates calculated from the life-table analysis were 48 and 63%, respectively, in the placebo group. These figures are even higher than were anticipated at the begitming of the study, but are in agreement with survival in other functional class IV patients. 12One should keep in mind that N Y H A class IV is not a clearly delineated group. In this trial, the patients did not have to be dyspneic at rest all the time, but instead they could be mainly fatigued and marked by their heart failure at rest with tachypnea. If there were doubts whether a patient was in class IV, the investigators were encouraged not to include the patient. The mean age of the study patients was 70 years-somewhat older than those in several other trials in which
T A B L E V Crude Mortality (%) by Etiologic Cause of Heart Failure 6-Month
Overall
Etiology
Placebo
Enatapril
p
Placebo
Enalapril
p
Coronary artery disease (n = 192) Other causes (n = 61) Diabetes Yes (n = 56) No (n = 192) Atrial fibrillation Yes (n = 126) No (n = 127)
51
24
<0.001
61
40
<0.001
24
31
>0.25
33
37
>0.25
30 48
35 24
>0.25 0.001
52 55
55 35
>0.25 0.001
43 45
29 23
0.14 0.014
52 55
38 40
0.052 0.015
64A
THE AMERICAN JOURNAL OF CARDIOLOGY
VOLUME 62
the mean age was 60 to 65 years. In contrast, this study had no upper age limit. Our patients are, therefore, probably more representative of an unselected heart failure population. The older age may also contribute to the high mortality rate. The treatment effect, however, was independent of age (Table VII). The VA trial compared placebo with prazosin and with a combination of hydralazine and isosorbide dinitrate. 3 Prazosin had no effect on mortality, but on the basis of life-table estimates, the hydralazine-nitrate combination reduced mortality after 1 year from 19.5 to 12.1% (-38%) and after 3 years from 46.9 to 36.2% (-23%). In the present trial, a similar calculation of life-table estimates shows a mortality after 6 months of 48% reduced to 29% (-40%) and after 1 year of 63% reduced to 47% (-25%). The 6-month mortality in the present trial was similar to the mortality risk after 3 years in the VA trial, showing that we were studying a more severely ill group of patients. Patients in whom CAD was the cause of heart failure had a much higher mortality than those with other causes (Table V). This is in agreement with results from the VA trial. The effect of enalapril was only seen in those patients with ischemic heart disease. Table VI shows that the 6-month crude mortality in the placebo group varied between 36 and 51%. In all subgroups, there was a consistent reduction in mortality. It is interesting to note that mortality among the patients taking enalapril was fairly constant, between 23 and 29%. It has been argued that these patients are so ill and have such extensive myocardial damage that no important prolongation of survival by any medical treatment can be expected. 13The fact that the mortality curves are parallel after the initial divergence indicates a maintained effect of enalapril for at least 1 year, because there will be relatively more high-risk patients remaining in the enalapril group. The limited follow-up period does not permit an estimation of the duration of this beneficial effect or of the drug's effects in less severe forms of heart failure. The fact that death from progressive heart failure was greatly decreased may have important implications for the future understanding of myocardial dysfunction. Progression of heart failure could be due to a new myocardial infarction, continued myocardial fibrosis, or metabolically based deterioration of myocardial function. From this trial, we can only speculate as to how progression was influenced. Extracardiac mechanisms, such as neuroendocrine alterations, could be of major importance for the progression of heart failure. The reduction in mortality among patients treated with vasodilators at randomization was similar to that in those not taking vasodilators. However, in patients taking vasodilators, the treatment effect appeared to be more pronounced during the first 6 months. Vasodilation unloads the failing myocardium, but the pharmacologic means by which peripheral resistance is lowered may be a critical factor. Enalapril probably lowers peripheral resistance by reducing angiotensin-II-in-
TABLE V l Crude Mortality (6 month) in Some Subgroups Divided According to Median
Subgroup Age (yr) _<70 >70 Serum creatinine (#tool/liter) _<120 > 120 Serum sodium (mmol/liter) _<137 >137 Heart size (ml/m 2 BSA) _<800 >800 Heart rate (beats/rain) _<80 >8O
Placebo (%)
Enalapril (%)
Reduction in Mortality (%) p
42 45
23 28
45 38
0.03 0.02
38 50
29 24
24 52
0,12 0.002
51 38
25 27
51 24
0.004 0.12
36 49
26 26
28 47
>0.25 0.003
35 49
23 28
34 43
>0.25 <0.001
BSA= bodysurfacearea.
TABLE VII Dosage Levels of Test Drug at End of Study Dose (rag/day)
Placebo (n : 126)
Enalapril (n : 127)
2.5 5 10 20 40 Unknown
1 7 13 41 57 7
16 10 33 30 28 10
duced vasoconstriction. 14Conversely, conventionally acting vasodilators may actually stimulate the renin-angiotensin system. ~5 Could the effects of enalapril on survival in severe heart failure be found in patients with less pronounced symptoms? This important question cannot be answered from the present study. However, it is reasonable to believe that the degree of myocardial depression is a most important prognostic indicator. Although we did not measure myocardial function directly, it could be presumed that severe myocardial dysfunction was present in these patients based on clinical history and x-ray findings. It is possible that among patients with a similar degree of myocardial dysfunction, a similar effect on progression of heart failure could be found, even if they are not in NYHA class IV. However, that question must be evaluated further, and studies are underway.
CONCLUSIONS In summary, this study has shown that enalapril given to patients with severe CHF is associated with a marked reduction in mortality. This treatment was well tolerated; however, it is important to initiate therapy with low doses of enalapril, particularly in high-risk patients. THE AMERICAN JOURNAL OF CARDIOLOGY JULY 11, 1988
65A
A SYMPOSIUM: ADVANCES IN CONGESTIVE HEART FAILURE
REFERENCES
!. Lipkin D, Poole-WilsonP. Treatment of chronic heart failure." A review of
recent drug trials. Br Med J 1985;291:993-996. 2. FurbergCD, YusufS. Effect of vasodilators on survival in chronic congestive heart failure. Am J Cardiol 1985;55.'11I0-1113.
CohnJN, ArchibaldDG,ZiescheS, FrancioscaJA, HarstonWE,Tristani FE, DunkmanWB, JacobsW, FrancisGS, FlohrKH, GoldmanS, Cobb FR, Shah PM, Saunders R, Fletcher RD, Loeb HS, Hughes VC, Baker B. Effect of
3.
vasodilator therapy on mortality in chronic congestive heart failure. N Engl J Med 1986;314:1547-1552. 4.. ToddPA, HeelRC. Enalapril: a reviewof its pharmacodynamic and pharmacokinetic properties, and therapeutic use in hypertension and congestive heart failure. Drugs 1986;31:198-248. 5. The CONSENSUSTrial StudyGroup.Effects ofenalapril on mortality in severe congestive heart failure: results of the Cooperative North Scandinavian EnalaprilSurvival Study (CONSENSUS). N Engl J Med 1987;316:1429-1435. 6. JoosellS. A method for the determination of the heart size by teleroentgenography (a heart volume index). Acta Radiol 1939;20:325-340. 7. Gillum RF, FortmannSP, Prineas RJ, KonkeTE. International diagnostic criteria for acute myocardial infarction and acute stroke. Am Heart J 1985; 108:150-154. 8. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc 1958;53.'457-481. 9. Mantel N. Chi-square test with one degree of freedom: Extension of the ManteI-Haenzel procedure. J Am State Assoc 1963;58.'690-700. 10. Cos DR. Regression methods and life tables. J Roy Stat Soc 1972;34:187220. 11. PackerM. Sudden unexpected death in patients with congestive heart failure: a second frontier. Circulation 1985;72:681-685. 12. McFateSmith W. Epidemiology of congestive heartfailure. Am J Cardiol 1985;55.'3A -SA. 13. PfefferMA.PfefferJM, SteinburgC. Finn P.Surt,ival afteran experimental myocardial infarction." beneficial effects of long-term therapy with captopril. Circulation 1985;72:406-412.
14. BaylissJ, Norell MS, Canepa-AnsonR, ReidC, Poole-WilsonP, SuttonG. Clinical importance of the renin-angiotensin system in chronic heart failure: double-blind comparison of captopril and prazosin. Br Med J 1985;290;18111865. 15. UedaK,Sakai M, MatsushitaS, Kuwajima1,MurakamiM. Effect of orally administered hydralazine on neurohumoral factors and hemodynamic response in aged patients with chronic congestive heartfailure. Jpn Heart J 1983;29:711721.
APPENDIX The CONSENSUS Trial Study group consisted of the following investigators (names are followed by locations and numbers of patients): Finland." U. Id~np~i~n-Heikkil~i, MD (Espoo, 2); J. Remes, MD, T. Hirvonen, MD, and K. Py6r/il5., MD (Kuopio, 10); U. Korhonen, MD, J. Takkunen, MD, and L. H~.m~il~inen, MD (Oulu, 10); A. Pasternack, MD, and M. Nikkil/i, MD (Tampere, 10); M. Arstila, MD (Turku, 3); and M. Lindroos, MD, and B. Heikius, MD (Vaasa, 9). N o r w a y : T .
66A
THE AMERICAN JOURNAL OF CARDIOLOGY VOLUME62
Gundersen, MD, and B. Aslaksen, MD (Arendal, 11); L. Gullestad, MD, and E. Soyland, MD (Baerum, 11); G. Tjonnfjord, MD, and A. Ose, MD (Gjovik, 3); P. Aukrust, MD, and J. Haerem, MD (Hamar, 7); K. Waage, MD, and F. Riddervold, MD (Haugesund, 5); J. Offstad, MD, and E. G. Aksnes, MD (Lillehammer, 10); T. Pedersen, MD, and P. Sirnes, MD (Oslo, Aker Hospital, 8); A. Westheim, MD, E. Sivertssen, MD, and J. Eritsland, MD (Oslo, Ullev~l Hospital, 11); K. Overskeid, MD, and O. Brubakk, MD (Sarpsborg, 4); J. Naesgaard, MD, and A. Hallargker, MD (Skien, 6); K. Dickstein, MD, and V. Bonarjee, MD (Stavanger, 10); H. Wang, MD, and E. S. P. Myhre, MD (Tromso, 3). Sweden: S. A. Forsberg, MD, and L. Karelds, MD (BorS.s, 6); K. Lindvall, MD, and S. Eriksson, MD (Danderyd, 10); G. Frithz, MD, and A. Stjerna, MD (Eskilstuna, 6); G. Ahlmark, MD, and H. Saetre, MD (Falun, 7); B. Andersson, MD, and M. Hartford, MD (Gothenburg, Sahlgrenska Hospital, 8); R. Bergstrand, MD, P. Held, MD, and G. Ulvenstam, MD (Gothenburg, Ostra Hospital, 17); L. Lundkvist, MD, and P. Andersson, MD (Hudiksvall, 4); U. Dahlstr6m, MD, and M. Broquvist, MD (Link6ping, 17); A. Henriksson, MD, and A. S6derling, MD (Lule~t, 7); S. Persson, MD (Lund, 1); N. Stobaeus, MD (Motala, 2); A. Sjfgren, MD, and E. Loogna, MD (Nacka, 5); J. Frid6n, MD, and P. O. Andersson, MD (Norrkfping, 5); L. Ryd6n, MD, and P. Smedg~trd, MD (Sk6vde, 5); I. Liljefors, MD, and C. Hofman-Bang, MD (Stockholm, Sabbatsberg Hospital, 3); M. Lycksell, MD, and C. Ringqvist, MD (Sundsvall, I 1); O. Johnson, MD, and L.-O. Hemmingson, MD (Ume~t, 5). The members of the CONSENSUS Steering Committee were as follows: J. Kjekshus MD, Sandvika, Norway (Chairman); H. Frick MD, Helsinki, Finland; K. Swedberg MD, Gothenburg, Sweden; and L. Wilhelmsen MD, Gothenburg, Sweden. Coordinating Office: K. Swedberg MD, Ostra Hospital, Gothenburg, Sweden (Coordinator); G. Andersson RN, Ostra Hospital, Gothenburg, Sweden. The members of the Ethical Review Committee were as follows: J. Lubsen, MD, The Netherlands (Chairman); D. Julian, MD, Great Britain; B. Levine, MD, United States; and W. McFate Smith, MD, United States. Administration Office: K. Kristianson, PhD, Bromma, Sweden (Head). Statistical Analysis: S. Snapinn, PhD, Merck Sharp & Dohme Research Laboratories, Brussels, Belgium.