Versus Low-Dose ACE Inhibition in Chronic Heart Failure a Double-Blind, Placebo-Controlled Study of Imidapril

JACC Vol. 32, No. 7 1811 December 1998:1811–8

High- Versus Low-Dose ACE Inhibition in Chronic Heart Failure A Double-Blind, Placebo-Controlled Study of Imidapril

DIRK J. VAN VELDHUISEN, MD, PHD, FACC, SABINE GENTH-ZOTZ, MD,* JAN BROUWER, MD, PHD, FRANS BOOMSMA, PHD,† TILO NETZER, PHD,‡ ARIE J. MAN IN ’T VELD, MD, PHD,† YIGAL M. PINTO, MD, PHD,࿣ K. I. LIE, MD, PHD,§ HARRY J. G. M. CRIJNS, MD, PHD Groningen, Rotterdam and Amsterdam, The Netherlands, and Mainz and Darmstadt, Germany

Objectives. To determine dose-related clinical and neurohu- dropped out: 3 patients died, and 9 developed progressive CHF moral effects of angiotensin-converting enzyme (ACE) inhibitors (3/182 patients on imidapril vs. 6/62 patients on placebo, p < 0.02 ؍ in patients with chronic heart failure (CHF), we conducted a 0.05). Exercise time increased 45 s in the 10-mg group (p double-blind, placebo-controlled, randomized study of three doses vs. placebo), but it did not significantly change in the 5-mg mg, 5 mg and 10 mg) of the long-acting ACE inhibitor (؉16 s), and 2.5-mg (؉11 s) imidapril group, compared to placebo 2.5) imidapril. (؉3 s). Physical working capacity also increased in a dose-related Background. The ACE inhibitors have become a cornerstone in manner. Plasma brain and atrial natriuretic peptide decreased the treatment of CHF, but whether high doses are more effective (p < 0.05 for linear trend), while (nor)epinephrine, aldosterone than low doses has not been fully elucidated, nor have the and endothelin were not significantly affected. Renin increased in mechanisms involved in such a dose-related effect. a dose-related manner, but plasma ACE activity was suppressed .Methods. In a parallel group comparison, the effects of three similarly (؎60%) on all three doses doses of imidapril were examined. We studied 244 patients with Conclusions. Already within 3 months after treatment initia- mild to moderate CHF (New York Heart Association class II–III: tion, high-dose ACE inhibition (with imidapril) is superior to who were stable on digoxin and diuretics. Patients low-dose. This is reflected by a more pronounced effect on exercise ,(20%/80%؎ were treated for 12 weeks, and the main end points were exercise capacity and some of the neurohormones, but it does not appear capacity and plasma neurohormones. to be related to the extent of suppression of plasma ACE. Results. At baseline, the four treatment groups were well- (J Am Coll Cardiol 1998;32:1811–8) matched for demographic variables. Of the 244 patients, 25 ©1998 by the American College of Cardiology

Angiotensin-converting enzyme (ACE) inhibitors have be- of all patients were on full-dose ACE inhibition (3). Still, there come a cornerstone in the treatment of chronic heart failure are many theoretical considerations why higher doses might be (CHF), as these drugs reduce morbidity and mortality in CHF more effective (4,5), and it is generally recommended to try to (1). In most CHF trials, however, relatively high doses of ACE achieve the target doses of ACE inhibitors (1,6). Assuming inhibitors were used, but in clinical practice much lower doses such a dose-dependent effect, it is important, however, to are often prescribed (2), and data from a recent large study in elucidate which mechanisms are involved. Because the effect of patients with moderate to severe CHF showed that only 25% ACE inhibitors is due to hemodynamic and neurohormonal factors, such studies should focus on both parameters, but so far only few data are available on this issue. From the Department of Cardiology/Thoraxcenter, University Hospital Imidapril hydrochloride is a long-acting, orally active, Groningen, Groningen, The Netherlands; *the II. Medical Clinic, Johannes Gutenberg-University Clinic, Mainz, Germany; †COEUR/Department of Inter- nonsulphydryl-containing ACE inhibitor that has been used in nal Medicine I, University Hospital Dijkzigt, Rotterdam, The Netherlands; hypertension, CHF and after myocardial infarction (7–12). ࿣ ‡Clinical Research, Merck KGaA, Darmstadt, Germany; the Institute of Imidapril is a pro-drug that is converted by the liver to its active Clinical Pharmacology, University of Groningen, Groningen, The Netherlands; and §the Department of Cardiology, Academic Medical Center, Amsterdam, metabolite imidaprilat. It has a potent hypotensive effect, and The Netherlands. Dr. Van Veldhuisen is a Clinical Scientific Investigator of the the inhibition of ACE lasts significantly longer in tissue than in Dutch Heart Foundation. This study was financially supported by Merck KGaA, plasma. In hypertensive patients, blood pressure was still Darmstadt, Germany. Presented in part at the 70th Scientific Sessions of the American Heart Association, Orlando, Florida, November 1997. decreased 24 h after imidapril administration in a dose- Manuscript received May 5, 1998; revised manuscript received July 13, 1998, dependent manner; plasma ACE was still 60% suppressed at accepted August 6, 1998. this time (8). Maximal reduction in blood pressure and plasma Address for correspondence: Dr. D.J. van Veldhuisen, Department of Cardiology/Thoraxcenter, University Hospital Groningen, PO Box 30 001, 9700 ACE was achieved with 10 mg imidapril once daily (8,9), and RB Groningen, The Netherlands. E-mail: [email protected]. with increasing doses no additional effect was found (8). When

©1998 by the American College of Cardiology 0735-1097/98/$19.00 Published by Elsevier Science Inc. PII S0735-1097(98)00464-1 1812 VAN VELDHUISEN ET AL. JACC Vol. 32, No. 7 HIGH- VS. LOW-DOSE ACE INHIBITION IN HEART FAILURE December 1998:1811–8

Ͼ60 years. A 20% difference in exercise was allowed between Abbreviations and Acronyms the two baseline tests (the shorter exercise duration serving as ACE ϭ angiotensin-converting enzyme the 100% value), and both tests were limited by dyspnea and/or ANP ϭ atrial natriuretic peptide fatigue. ϭ ATLAS Assessment of Treatment of Lisinopril and Survival Patients were excluded from participation if they had BNP ϭ brain natriuretic peptide CHF ϭ chronic heart failure hemodynamically significant valvular disease, active myocardi- NYHA ϭ New York Heart Association tis, thyroid disease or hypertrophic cardiomyopathy, a history PWC ϭ physical (or pulse) working capacity of myocardial infarction or open heart surgery Ͻ3 months, severe hypertension (requiring treatment other than diuretics) or hypotension (supine systolic blood pressure Ͻ100 mm Hg), angina pectoris (limiting exercise capacity as well as unstable administered to patients with acute myocardial infarction, complaints), atrial fibrillation (requiring other than digitalis imidapril was found to improve left ventricular ejection frac- therapy), supraventricular or ventricular arrhythmias requiring tion and to reduce plasma brain natriuretic peptide (BNP) treatment (antiarrhythmic drugs and pacemakers not allowed), levels (11). In patients with CHF (New York Heart Associa- right-sided CHF, significant electrolyte abnormalities, any tion [NYHA] functional class II–III), 2.5 mg and 5 mg other relevant disease (including anemia) and significant over- imidapril lowered plasma ACE to a similar extent, whereas the or underweight from normal. Patients with known or suspicion effect on blood pressure was more pronounced on 5 mg of bilateral renal artery stenosis, and those with significant imidapril (12). impaired renal function (serum creatinine Ͼ1.8 mg/dl or Ͼ160 In the present study, we examined the dose-related clinical ␮mol/liter) or renal transplantation were also excluded. Pa- and neurohormonal effects of three doses (2.5 mg, 5 mg and tients who had a history of substance abuse were also not 10 mg) of imidapril during 12 weeks of treatment, compared to included. The following drugs were not allowed: vasodilators placebo in patients with stable mild to moderate CHF, who (including calcium channel blockers) except for nitrates, were treated with digoxin and diuretics. ␤-blockers, potassium-retaining diuretics, lithium, allopurinol, cytostatic or immunosuppressive agents, neuroleptics, imipra- mine, and nonsteroidal anti-inflammatory drugs except aspirin. Methods Also, ACE inhibitor therapy was not allowed within 4 weeks Study design. The study was a double-blind, placebo- prior to the start of the study (ϭ6 weeks prior to randomiza- controlled, randomized, parallel group comparison of three tion), and patients with known intolerance were also excluded. different doses of imidapril once daily. It was carried out in 25 Exercise protocol. During the run-in phase, exercise testing centers in The Netherlands, Germany and Belgium (see Ap- was performed twice (see above), and it was repeated after 8 pendix). The study started with a single-blind, placebo run-in and 12 weeks of double-blind treatment. Exercise testing was period of 2 weeks, during which the clinical stability of patients performed in the morning between 9 and 10 AM, before intake was confirmed and the bicycle exercise test was repeated. The of study medication; it was intended that this test was carried second part was a double-blind phase of 12 weeks’ duration, out 24 Ϯ 4 h after intake of the previous study medication, and and patients were randomized to one of the four treatment 2 to 4 h after a light breakfast, without coffee or tea. Exercise arms. During this phase, the imidapril dose was gradually testing was carried out in the sitting position, using an electri- increased in a double-blind manner during the first 3 weeks: all cally braked bicycle ergometer. The protocol started with a patients who were randomized to active treatment (three of workload of 50 W, and was increased with steps of 10 W; each the four groups) received 2.5 mg during the first week. The step was maintained for 1 min. Heart rate and blood pressure dose was then increased to 5 mg in the 5- and 10-mg groups in were measured before and throughout the test, and the the second week, and in the third week it was uptitrated to electrocardiogram was monitored continuously. 10 mg in the 10-mg group only. After randomization, patients Exercise tolerance was assessed by determining total exer- were seen on weeks 1, 2, 3, 8 and 12. The protocol was cise time and physical (or pulse) working capacity (PWC) conducted in accordance with the revised declaration of Hel- determined at 110 beats/min during exercise (PWC110). The sinki, and was approved by the Ethics Committee of each PWC allows us to measure submaximal exercise tolerance and hospital. All patients gave their written informed consent is not dependent on patients’ motivation (13–15). The PWC is before entering the study. determined by examining the relationship between heart rate Patients. Subjects were eligible if they met the following and power output (in watts). By using PWC, one gets infor- criteria: 1) signs and symptoms of mild to moderate CHF mation on how many watts (W) can be reached at a certain (NYHA class II–III); 2) aged 21 to 75 years (females were heart rate. Because of the limited maximal pulse reached by postmenopausal, or taking contraceptive medication); 3) a left CHF patients during exercise testing, we (prospectively) chose ventricular ejection fraction Ͻ0.45; and 4) a minimal exercise 110 beats/min, but in healthy subjects higher heart rates are Ͼ duration 2 min, and a maximal exercise duration of 12 and often used (13). The PWC110 is calculated as follows: 10 min, respectively, for men and women Ͻ46 years, of 10 and ϭ ϩ ϫ ͑ Ϫ ͒ ͑ Ϫ ͒ 8 min for subjects 46 to 60 years, and 8 and 6 min for subjects PWC110 W1 10 110 P1 / P2 P1 JACC Vol. 32, No. 7 VAN VELDHUISEN ET AL. 1813 December 1998:1811–8 HIGH- VS. LOW-DOSE ACE INHIBITION IN HEART FAILURE

in which W1,P1 and P2 are: by including baseline data from patients who were subse- ● ϭ Ͻ Ϯ W1 highest stage in watts with a pulse 110 beats/min quently withdrawn from the study. Values are mean SD ● ϭ ϭ ϩ P1 pulse at W1, and P2 pulse at W2, which is W1 unless indicated otherwise. All p values reported are for 10 W. two-tailed test, and a two-tailed alpha of Ͻ0.05 was considered

The PWC110 was only determined in patients with sinus statistically significant. rhythm, as the chronotropic response to exercise in atrial fibrillation and CHF is markedly enhanced and difficult to quantify (16). Results Plasma neurohormones. Measurements were determined A total of 244 patients fulfilled the entry criteria and were in a subset of patients and performed at three core laboratories included, and the four treatment groups were well-matched at (see Appendix). All samples were taken from an indwelling baseline (Table 1). Most patients were males, the majority had venous canula, after Ͼ30 min of supine rest, and at peak rather mild disease (NYHA class II), and coronary artery exercise. Samples were taken at baseline and at the end of the disease was the most common underlying disorder. Of the 244 study. Plasma norepinephrine (normal value in our laboratory: patients, 39 patients (16%) had previously used an ACE 100 to 500 pg/ml) and epinephrine (normal: 10 to 70 pg/ml) inhibitor, and they were normally divided among the four were determined by high performance liquid chromatography groups. During the 12-week study period, three patients died with electrochemical detection (17). Active plasma renin (nor- (sudden death n ϭ 2, myocardial infarction n ϭ 1, p ϭ NS); mal:5to50␮U/ml) was determined with immunoradiometric another patient survived an episode of ventricular fibrillation assay (Nichols Institute, Bad Nauheim, Germany) (18). Plasma (Table 2). An additional 21 patients dropped out; therefore, aldosterone (normal: 50 to 250 pg/ml) was determined by 219 patients completed the protocol. Nine of these 21 patients radioimmunoassay (Coat-a-Count, Diagnostic Products, Los developed progressive CHF and discontinued the study; this Angeles, California) (19). Plasma endothelin concentrations drop-out was higher among patients taking placebo (n ϭ 6of (normal: 1 to 5 pg/ml) were measured by using radioimmuno- 62, 10%) than imidapril (n ϭ 3 of 182, 2%, p Ͻ 0.05). The assay (Amersham Buchler, Braunschweig, Germany) (20). NYHA class was not significantly affected. Four of the 21 Plasma ACE (normal: 18 to 55 U/I) was determined radiomet- patients discontinued the study because of (probably related) rically by using (phenyl-4-3H)-hippuryl-glycyl-glycine as a sub- side effects, including one with dry cough, and one with strate (21,22). Plasma atrial natriuretic peptide (ANP; normal: angioedema. No other significant side effects were reported, 50 to 110 pg/ml) and BNP (normal: 25 to 55 pg/ml) were and no significant biochemical or hematological changes were measured (after SepPak extraction) with kits from the Nichols observed. All patients in the three imidapril groups were Institute, Nijmegen, The Netherlands, and Peninsula Labora- uptitrated to their maximal dose. Imidapril did not significantly tories, Belmont, California, respectively (23,24). affect heart rate or blood pressure, neither at rest nor during Statistical analysis. The primary end point of the study exercise, and only small changes in blood pressure were was change in maximal exercise time after 12 weeks. Secondary observed (largest change in systolic pressure Ϫ6 Ϯ 15 mm Hg end points were changes in submaximal exercise tolerance on 10 mg imidapril, p ϭ NS). using PWC110, CHF symptomatology, progression of CHF and Exercise capacity. After 8 weeks, no significant improve- neurohormones. In addition, various safety parameters were ment in exercise time was observed (imidapril 10 mg: ϩ26 s, assessed. It was estimated that 60 patients per group would be P ϭ NS). After 12 weeks, exercise time (intention to treat sufficient to detect a true difference between the medication analysis) (Fig. 1A) increased 45 Ϯ 82 s in patients treated with groups, with a probability of 90% at an ␣ ϭ 0.05 significance 10 mg imidapril (p ϭ 0.018 vs. placebo, and p Ͻ 0.05 vs. 2.5 mg level (2-sided, normal distribution). imidapril). Changes in exercise duration in the 5- and 2.5-mg The primary analysis was performed with the intention-to- imidapril groups were not significantly different from placebo. treat population, to ensure that all patients were included in The dose-response relation was significant for linear-trend the analysis, and to avoid bias from missing data, as previously (p ϭ 0.019). Per protocol analysis (n ϭ 219) showed similar, described (19). In short, we used a nonparametric approach in nonsignificant trends (10 mg imidapril: ϩ54 Ϯ 78 s; p ϭ 0.06 vs. which all patients who dropped out were assigned lowest rank, placebo). and missing values were substituted by the lowest exercise After 12 weeks, small but statistically significant differences ϭ results of that particular group (worst-case analysis). In addi- in PWC110 (Fig. 1B) were observed between groups (p 0.005, tion, only patients who completed the protocol were analyzed and p ϭ 0.02 for 10 and 5 mg imidapril vs. placebo, respec- (per protocol population). tively), with a significant dose-related effect by linear trend. Distribution of the baseline variables was compared with With regard to maximal workload, small and nonsignificant analysis of variance (ANOVA) and the t test (for continuous changes were observed (10-mg imidapril group ϩ9 Ϯ 13 W, variables) and chi-square test (for categoric variables). Differ- p Ͻ 0.1 vs. placebo and vs. 2.5 mg imidapril). ences among groups after randomization were analyzed at Plasma neurohormones. These were available in 82 pa- each scheduled follow-up visit by comparing the mean change tients (placebo n ϭ 19, imidapril 2.5/5/10 mg, n ϭ 22/17/24, from baseline with use of the ANOVA and the t test. Use of respectively). This subgroup was not different from the total this procedure avoided any bias that would have been caused study group, and the four groups were also well balanced. At 1814 VAN VELDHUISEN ET AL. JACC Vol. 32, No. 7 HIGH- VS. LOW-DOSE ACE INHIBITION IN HEART FAILURE December 1998:1811–8

Table 1. Baseline Characteristics of the Four Treatment Groups Placebo Imidapril 2.5 mg Imidapril 5 mg Imidapril 10 mg (n ϭ 62) (n ϭ 65) (n ϭ 57) (n ϭ 60) Age (yr) 61 Ϯ 961Ϯ 961Ϯ 962Ϯ 9 % Male 71% 82% 81% 75% LV ejection fraction 0.33 Ϯ 0.07 0.34 Ϯ 0.10 0.33 Ϯ 0.08 0.34 Ϯ 0.08 Duration of CHF (months) 31 Ϯ 40 32 Ϯ 40 36 Ϯ 40 40 Ϯ 47 NYHA class (%) II 76 74 82 77 III 24 26 18 23 Etiology CHF (%) Ischemic (coronary disease) 69 60 61 62 Idiopathic cardiomyopathy 23 31 28 18 Other 8 9 11 20 Exercise capacity Time (s) 406 Ϯ 128 447 Ϯ 140 461 Ϯ 141 462 Ϯ 127 Ϯ Ϯ Ϯ Ϯ PWC110 (watts) 57 20 59 22 59 18 63 19 Maximal work load (watts) 85 Ϯ 21 92 Ϯ 24 94 Ϯ 23 93 Ϯ 22 Blood pressure (mm Hg) Systolic 138 Ϯ 14 139 Ϯ 21 134 Ϯ 18 135 Ϯ 18 Diastolic 84 Ϯ 783Ϯ 881Ϯ 10 82 Ϯ 9 Heart rate (beats/min) 78 Ϯ 15 75 Ϯ 12 80 Ϯ 15 75 Ϯ 14 Rhythm (%) Sinus rhythm 97 86 93 87 Atrial ﬁbrillation 3 14 7 13 Biochemistry (serum) Sodium (mmol/liter) 141 Ϯ 3 141 Ϯ 3 141 Ϯ 3 141 Ϯ 3 Potassium (mmol/liter) 4.4 Ϯ 0.3 4.3 Ϯ 0.4 4.3 Ϯ 0.4 4.3 Ϯ 0.4 Creatinine (␮mol/liter) 99 Ϯ 18 106 Ϯ 25 98 Ϯ 15 106 Ϯ 15 Urea (mmol/liter) 6.4 Ϯ 1.8 6.8 Ϯ 3.0 6.0 Ϯ 1.7 6.2 Ϯ 1.3 Plasma neurohormones at rest* Norepinephrine (pg/ml) 338 Ϯ 179 300 Ϯ 124 329 Ϯ 113 353 Ϯ 174 Epinephrine (pg/ml) 35 Ϯ 15 37 Ϯ 17 37 Ϯ 20 30 Ϯ 20 Renin (␮U/ml) 33 Ϯ 31 25 Ϯ 24 29 Ϯ 20 25 Ϯ 12 Aldosterone (pg/ml) 104 Ϯ 50 98 Ϯ 49 75 Ϯ 49 104 Ϯ 59 Endothelin (pg/ml) 4.3 Ϯ 1.8 4.2 Ϯ 1.9 4.5 Ϯ 1.8 4.5 Ϯ 2.2 ACE (U/l) 27 Ϯ 927Ϯ 727Ϯ 627Ϯ 7 ANP (pg/ml) 115 Ϯ 49 130 Ϯ 63 181 Ϯ 96 133 Ϯ 50 BNP (pg/ml) 74 Ϯ 43 71 Ϯ 37 83 Ϯ 45 79 Ϯ 57 Concomitant drugs (number, %) Digoxin 19 23 26 24 Diuretics 25 34 25 30 Nitrates 25 21 19 20 Antiplatelet agents 34 22 27 26 Anticoagulants 13 22 18 21 ϭ ϭ ϭ CHF chronic heart failure; LV left ventricular; PWC110 physical (or pulse) working capacity at heart rate 110 beats/min. *Only available in 82 patients (see text). There were no signiﬁcant differences between the groups.

baseline, only plasma ANP and BNP were elevated; other lin were not affected, neither at rest nor during exercise. neurohormones were within the normal range (Table 2). Plasma ACE (Fig. 2C) was suppressed by Ϯ60% on all three Plasma norepinephrine concentrations at rest were not doses (all p Ͻ 0.001). Plasma ANP at rest was not signiﬁcantly affected by imidapril; during exercise they appeared to be affected by imidapril, although a trend was observed: Ϫ36 slightly attenuated (p ϭ NS) (Fig. 2A). Epinephrine was also pmol/liter on 10 mg imidapril, p ϭ 0.1 vs. placebo. During not affected. Plasma renin at rest slightly increased on all three exercise, all three imidapril doses decreased ANP, which was doses of imidapril (only signiﬁcant for 10 mg), but during dose dependent (p ϭ 0.02 for linear trend). Plasma BNP exercise this increase was more pronounced (p Ͻ 0.05 for all decreased in the 5-mg and 10-mg imidapril group at rest (both doses) (Fig. 2B). Furthermore, a dose-response relation by p Ͻ 0.05), and in the 10-mg group during exercise (p Ͻ 0.05) linear trend was observed both at rest (p ϭ 0.011) and during (Fig. 2D). This effect was dose dependent, at rest and during exercise (p Ͻ 0.001). Plasma aldosterone and plasma endothe- exercise (p Ͻ 0.05). JACC Vol. 32, No. 7 VAN VELDHUISEN ET AL. 1815 December 1998:1811–8 HIGH- VS. LOW-DOSE ACE INHIBITION IN HEART FAILURE

Table 2. Reasons for Premature Discontinuation During the Study by a reduced sympathetic drive during exercise (28), which was ϭ (n 25 of 244 patients) also observed in the present study. Imidapril Imidapril Imidapril Several other mechanisms may have played a role in the Placebo 2.5 mg 5mg 10 mg observed effects of ACE inhibition in the present study. ϭ ϭ ϭ ϭ (n 62) (n 65) (n 57) (n 60) Chronic heart failure is accompanied by a decreased skeletal Death 1 — 2 — muscle blood flow, which is an important contributing factor Progressive CHF 6* 2 1 — responsible in the exercise intolerance in these patients. The Angina pectoris 1 1 — — ACE inhibitors improve skeletal muscle blood flow (29), and Arrhythmias — — 2# — the present findings may also be related to this. Furthermore, Side effects — — 2 2 Other/“unrelated” — 2 1 2 ACE inhibitors inhibit the degradation of bradykinin and TOTAL 8 5 8 4 potentiate its actions, which may also explain part of its beneficial effects, and in animal studies this effect was also *p Ͻ 0.05 vs. active medication (3 imidapril groups together). #Including 1 patient who was successfully resuscitated for ventricular fibrillation. dose-related (30). Although it may be speculated that both mechanisms might also have been affected in a dose-related manner by imidapril, they were not investigated in the present study. Previous studies. In two earlier studies with quinapril Discussion in patients with mild CHF, a dose-dependent improvement in Because they reduce morbidity and mortality, ACE inhibi- exercise time (31) and hemodynamics (22) was observed. In tors have become the most important drugs in the treatment of advanced CHF, high-dose enalapril was superior to low-dose CHF. Although the large trials were conducted with high doses for symptomatology, but no difference was observed in exercise of ACE-inhibition, a large number of CHF patients are using capacity and hemodynamics (32). Recently, the NETWORK substantially lower doses (1–6). This practice of prescribing was published, in which 1,532 patients with CHF NYHA class lower doses may be related to several considerations, such as the belief that low doses are equally effective but less toxic (4,5). The present study sheds new light on this issue; it shows Figure 1. (A) Change after 12 weeks of treatment in exercise duration that within 3 months after initiating ACE inhibitor treatment, (s) in the four groups. *p Ͻ 0.05 vs. placebo; #p Ͻ 0.05 vs. 2.5-mg a higher dose is clinically more effective than a low dose, an imidapril. A significant dose-response relation was observed by the effect that is paralleled by a dose-related effect on plasma linear trend test (p ϭ 0.019). (B) Change after 12 weeks of treatment in physical (or pulse) working capacity (PWC) in watts at a heart rate neurohormones. Ͻ Potential explanations. of 110 beats/min (PWC110). *p 0.05 vs. placebo. There was also a Given the dose-related decrease of significant dose-response effect (p Ͻ 0.05). both BNP and ANP during imidapril treatment in this study, unloading of the heart appears to play an important role. Natriuretic peptides are released in response to an increase in intracardiac volume or pressure, both in the atria (ANP) and in the ventricles (BNP) (25). Furthermore, natriuretic peptides are related to the severity of CHF and prognosis, and a decrease could thus be interpreted as a favorable response (6). An important effect of ACE inhibitors in CHF is unloading of the heart, thereby preventing cardiac dilatation, and improving diastolic function (26). The latter was also found to correlate significantly with a reduction in ANP levels, which is therefore compatible with the observed (dose-related) effect in the present study. A second explanation may be the dose-related improvement in PWC110, which indicates that at a heart rate of 110 beats/min, more work is accomplished, or a lower heart rate is “required.” While both resting and peak exercise heart rates were not affected by imidapril, submaximal heart rate was lower. In CHF, cardiac output during exercise depends on chronotropic mechanisms, and the present findings would support a dose-dependent effect of imidapril. Furthermore, while resting heart rate may be similar in CHF patients of different severity, during the early stages of exercise it increases more in advanced CHF than in mild CHF (27). This blunted response of heart rate in mild CHF may be explained 1816 VAN VELDHUISEN ET AL. JACC Vol. 32, No. 7 HIGH- VS. LOW-DOSE ACE INHIBITION IN HEART FAILURE December 1998:1811–8

Figure 2. (A) Change after 12 weeks of treatment in plasma norepinephrine concentrations (in pg/ml), at rest (open bars) and peak exercise (solid bars). There are no significant differences. (B) Change after 12 weeks of treatment in plasma renin (␮U/ml), at rest (open bars) and peak exercise (solid bars). *p Ͻ 0.05 vs. placebo; **p Ͻ 0.001 vs. placebo; #p Ͻ 0.05 vs. 2.5-mg imidapril. The dose-response relation was significant both at rest and during exercise (p Ͻ 0.05 by linear trend). (C) Change after 12 weeks of treatment in plasma ACE (in % from baseline), at rest (open bars) and peak exercise (solid bars). All differences are highly significant vs. placebo (p Ͻ 0.001) but there is no dose-response relation. (D) Change after 12 weeks of treatment in plasma BNP (pg/ml), at rest (open bars) and peak exercise (solid bars). *p Ͻ 0.05 vs. placebo. The effect on BNP is dose- dependent, both at rest and during exercise (both p Ͻ 0.05).

II–IV were randomized to 2.5, 5 or 10 mg enalapril twice daily conversion (36). Furthermore, in a clinical study there was no (33). After 6 months, no differences were observed with regard correlation between the clinical effects and the degree of to mortality or hospitalizations for CHF. Even more recently, plasma ACE inhibition (37). Therefore, plasma renin (partic- results of the Assessment of Treatment with Lisinopril and ularly during exercise) may be a better parameter to judge Survival (ATLAS) study have also become available (34). In dose-related suppression of the renin-angiotensin system. that study, 3,164 patients with NYHA class II–IV CHF were Plasma renin increases in response to ACE-inhibition both at randomized to either low-dose (2.5 or 5 mg) or high-dose (32.5 rest and during exercise; the latter effect is probably due to an or 35 mg) lisinopril, and they were followed for almost 4 years. acute release from the juxtaglomerular cells in the kidney, Although results of the ATLAS study have not yet been fully which is susceptible to sympathetic blockade (38). Thus, the published, the data show that, compared to low-dose ACE higher exercise-related increase in renin with higher doses of inhibition, high-dose lisinopril does not cause a statistically imidapril may reflect larger stores of renin in the kidney, which significant reduction in all-cause mortality, but it is associated suggests a more effective decrease of local angiotensin II, and with a significant reduction (Ϫ12%) in the combined end point confirms the importance of tissue inhibition (39). of all-cause mortality and hospitalization. A third large-scale Study limitations. Given the rather small patient popula- trial of comparable size (ACHIEVE) is currently still under- tion and the short follow-up, the present results are clearly not way (6). definitive, for they do not provide data on mortality and Regarding the neurohormonal effects, Nussberger et al. (long-term) morbidity. However, this study was designed to (22) showed a dose-dependent decrease in angiotensin II after examine whether high-dose ACE inhibition would be superior quinapril, while renin increased. A similar effect was also found to low-dose, with regard to exercise capacity and neurohor- after short-term lisinopril treatment in patients with mild CHF mones, and in this respect the data may be complementary to (24), but in severe CHF (32), neurohormones were not af- larger survival studies such as ATLAS. A study period of 3 fected in a dose-dependent manner by enalapril. In the present months was chosen because it was deemed unethical to refrain study, only ANP and BNP were increased at baseline, which patients from ACE inhibition any longer at that time. Given supports their value in mild CHF (35), and confirms that these this short follow-up, however, it can also not be excluded that neurohormones may also be very useful in the assessment of low-dose imidapril might also have been effective during drug-induced changes in mild CHF (6). prolonged treatment, and in more advanced CHF. In addition, Inhibition of the renin-angiotensin system. Plasma renin these data obtained with imidapril should not automatically be increased during imidapril treatment in a dose-related manner, extrapolated to other ACE-inhibitors, as these compounds but plasma ACE was suppressed to a similar extent, which may differ—for instance, with regard to tissue penetration (40). suggests that plasma ACE activity poorly reflects the degree to Furthermore, the observed superiority of high-dose ACE which the renin-angiotensin system is suppressed. This is inhibition in mild CHF may not be similar in advanced CHF supported by a recent experimental study in which the degree (32), which might be due to differences in sensitivity to ACE of plasma ACE inhibition underestimated local angiotensin I inhibition. JACC Vol. 32, No. 7 VAN VELDHUISEN ET AL. 1817 December 1998:1811–8 HIGH- VS. LOW-DOSE ACE INHIBITION IN HEART FAILURE

Clinical implications. The present study shows that al- 9. Saruta T, Omae T, Kuramochi M, et al. Imidapril hydrochloride in essential hypertension: a double-blind comparative study using enalapril maleate as a ready within the first 3 months of initiating ACE inhibitor control. J Hypertens 1995;13 Suppl 3:S23–30. treatment with imidapril, high-dose treatment is more effective 10. Narita H, Kaburaki M, Doi H, et al. Prolonging action of imidapril on the than low dose treatment, which is reflected by a preservation of lifespan expectancy of cardiomyopathic hamsters. J Cardiovasc Pharmacol exercise capacity, and a more pronounced decrease in the 1996;27:861–71. atrial peptides (ANP and BNP). Although the present study 11. Mizino Y, Yashue H, Oshima S, et al. Effects of angiotensin-converting enzyme inhibitor on plasma B-type natriuretic peptide levels in patients with provides insight into underlying mechanisms of dose-related acute myocardial infarction. J Card Fail 1997;3:287–93. effects, further data from larger survival trials will be needed to 12. Pinto YM, Van Veldhuisen DJ, Tjon-Ka-Jie RT, Rooks G, Netzer T, Lie KI. confirm a dose-dependent effect on prognosis. Dose-finding study with imidapril, a novel angiotensin-converting enzyme inhibitor, in patients with stable chronic heart failure. Eur J Clin Pharmacol 1996;50:265–8. 13. Rost R, Hollmann W. Belastungsuntersuchungen in der Praxis. Grundlagen, Appendix Technik und Interpretation ergometrischer Untersuchungsverfahren. Stutt- gart and New York: Georg Thieme-Verlag, 1982:82–5. Participating centers and local investigators: 14. Eugene M, Vandewalle H, Bertholon JF, Teillac A. Arterial elasticity and The Netherlands physical working capacity in young men. J Appl Physiol 1986;61:1720–3. Maria Ziekenhuis, Tilburg, H.A.M. van Kesteren; St. Jans Gasthuis Weert, 15. Predel HG, Rohden C, Heine O, Prinz U, Rost RE. ACE inhibition and H.J.A.M. Penn; University Hospital Groningen, D.J. van Veldhuisen; St. Elisa- physical exercise: studies on physical work capacity, energy metabolism, and beth Ziekenhuis, Tilburg, W.H. Pasteuning; Ziekenhuis Hilversum, K.L. Liem; maximal oxygen uptake in well-trained, healthy subjects. J Cardiovasc Refaja Ziekenhuis, Stadskanaal, L.M. van Wijk; Reinier de Graaf Ziekenhuis, Pharmacol 1994;23 Suppl 1:S25–8. Delft, A.J.A.M. Withagen; St. Gemini Ziekenhuis, Den Helder, J.G.M. Tans; 16. Van den Berg MP, Crijns HJGM, Gosselink ATM, et al. Chronotropic Twenteborg Ziekenhuis Almelo, J.L. Darmanata; Het Nieuwe Spittaal, Zutphen, response to exercise in patients with atrial fibrillation: relation to functional A.C. Tans. state. Br Heart J 1993;70:150–3. Belgium 17. Hjemdal P, Daleskog M, Kana T. Determination of plasma catecholamines Ziekenhuis Genk, W. Van Mieghem. by high performance liquid chromatography with electrochemical detection: comparison with a radioenzymatic method. Life Sci 1979;25:131–8. Germany 18. Derkx FHM, De Bruin RJA, Van Gool JMG, et al. Clinical validation of Sta¨dtischesKrankenhaus, Bietigheim-Bissingen, D. Hey; Klinikum Su¨dstadt, renin monoclonal-antibody based sandwich assays of renin and prorenin, and Rostock, G. Naumann; Johannes-Gutenberg-Universita¨tMedizinische Klinik, use of renin inhibitor to enhance prorenin immunoreactivity. Clin Chem Mainz, H. Darius; Krankenhaus Pfungstadt, K.H. Baumgartl; Stadtkrankenhaus 1996;42:1051–63. Neuwied, W. Mansury; Stadtkrankenhaus Worms, P. Limbourg; Krankenhaus 19. Van Veldhuisen DJ, Man in ’t Veld AJ, Dunselman PHJM, et al. Double- Rosenheim, R. Rackwitz; Medizinische Akademie, Erfurt, I. Assmann; Kran- kenhaus Kaarst, D. Maring; Krankenhaus Bellheim, R. Philipp; Krankenhaus blind, placebo-controlled study of ibopamine and digoxin in patients with Berlin, R. Bohm; Krankenhaus Darmstadt, D. Gengnagel; Katharinen- mild to moderate heart failure: results of the Dutch Ibopamine Multicenter Krankenhaus, Frankfurt/Main, H.J. Gilfrich; Medizinische Universita¨tsklinik, Trial (DIMT). J Am Coll Cardiol 1993;22:1564–73. Tu¨bingen, R. Haasis. 20. Rodeheffer RJ, Lerman A, Heublein DM, Burnett JC Jr. Increased plasma concentrations of endothelin in congestive heart failure. Mayo Clin Proc Core Laboratories for Plasma Neurohumoral Determinations 1988;67:719–24. Plasma (nor)epinephrine, renin, aldosterone, and ACE activities were 21. Cushman DW, Cheung HS. Concentrations of angiotensin-converting en- determined at Bioscientia, Ingelheim/Rhein, Germany (Dr. A. Schmitt). Plasma zyme in tissues of the rat. Biochem Biophys Acta 1971;250:261–5. endothelin was determined at Merck KGaA Clinical Research, Darmstadt, 22. Nussberger J, Fleck E, Bahrmann H, Delius W, Schultheiss HP, Brunner Germany (Dr. A. Klieber), and ANP and BNP were determined at the University HR. Dose-related effects of ACE inhibition in man: quinapril in patients Hospital Dijkzigt, Rotterdam, The Netherlands (Dr. F. Boomsma). with moderate congestive heart failure. Eur Heart J 1994;15 Suppl D:113–22. 23. Bhaggoe UM, Boomsma F, Admiraal PJJ, Man in ’t Veld AJ, Schalekamp MADH. Stability of human plasma atrial natriuretic peptide during storage References at Ϫ80°C. Clin Chim Acta 1993;223:179–84. 24. Davidson NC, Coutie WJ, Webb DJ, Struthers AD. Hormonal and renal 1. Cohn JN. The management of chronic heart failure. N Engl J Med differences between low- and high-dose angiotensin-converting enzyme 1996;335:490–8. inhibitor treatment in patients with chronic heart failure. Heart 1996;75: 2. Pitt B. ACE inhibitor co-therapy in patients with heart failure: rationale for 576–81. the randomized aldactone evaluation study (RALES). Eur Heart J 1995;16 25. Struthers AD. Ten years of natriuretic peptide research: a new dawn for their Suppl N:107–10. diagnostic and therapeutic use? BMJ 1994;308:1615–19. 3. Van Veldhuisen DJ, Charlesworth A, Hampton JR, Crijns HJGM. Large 26. Greenberg B, Quinones MA, Koilpillai C, et al, for the SOLVD Investiga- differences in drug treatment of chronic heart failure between the European tors. Effects of long-term enalapril therapy on cardiac structure and function countries [abstract]. J Am Coll Cardiol 1998;31 Suppl A:115A. in patients with left ventricular dysfunction: results of the SOLVD Echocar- 4. Cleland JGF, Poole-Wilson PA. ACE inhibitors for heart failure: a question diographic Substudy. Circulation 1995;91:2573–81. of dose. Br Heart J 1994;72 Suppl:S106–10. 27. De Vries RJM, Van Veldhuisen DJ, Dunselman PHJM, Van de Veer N, 5. Packer M. Do angiotensin-converting enzyme inhibitors prolong life in Crijns HJGM. Physiologic parameters during initial stages of cardiopulmo- patients with heart failure treated in clinical practice? J Am Coll Cardiol nary exercise testing in patients with chronic heart failure: their value in the 1996;28:1323–7. assessment of clinical severity and prognosis. Eur Heart J 1997;18:1921–30. 6. Pitt B, Nicklas JM. Target doses of ACE inhibitors in heart failure: where 28. Francis GS, Goldsmith SR, Ziesche S, Nakajima H, Cohn JN. Relative should we aim? Eur Heart J 1998;19:370–1. attenuation of sympathetic drive during exercise in patients with congestive 7. Kubo M, Ochiai T, Kato J, Ishida R. Pharmacological studies on TA-6366, heart failure. J Am Coll Cardiol 1985;5:832–9. a new ACE-inhibitor: II. effect of long-term administration from the 29. Coats AJS. Interruption of the progression of heart failure: are ACE pre-hypertensive stage on blood pressure, relative heart weight and ACE- inhibitors the solution? Cardiology 1996;87 Suppl 1:11–5. activity of various tissues in spontaneously hypertensive rats (SHRs). Jpn 30. Minshall RD, Tan F, Nakamura F, et al. Potentiation of the actions of J Pharmacol 1991;57:517–26. bradykinin by angiotensin I-converting enzyme inhibitors. Circ Res 1997;81: 8. Vandenburg MJ, MacKay EM, Dews I, Pullan T, Brugier S. Dose-finding 848–56. studies with imidapril—a new ACE inhibitor. Br J Clin Pharmacol 1994;37: 31. Riegger GAJ. Effects of quinapril on exercise tolerance in patients with mild 265–72. to moderate heart failure. Eur Heart J 1991;12:703–11. 1818 VAN VELDHUISEN ET AL. JACC Vol. 32, No. 7 HIGH- VS. LOW-DOSE ACE INHIBITION IN HEART FAILURE December 1998:1811–8

32. Pacher R, Stanek B, Globits S, et al. Effects of two different enalapril doses plasma versus tissue ACE by utibapril: biochemical and functional on clinical, haemodynamic, and neurohumoral response of patients with evidence for inhibition of vascular ACE activity. J Cardiovasc Pharmacol severe congestive heart failure. Eur Heart J 1996;17:1223–32. 1997;29:684–91. 33. The NETWORK Investigators. Clinical outcome with enalapril in symp- 37. Van den Broek SAJ, De Graeff PA, Van Veldhuisen DJ, et al. Clinical and tomatic chronic heart failure: a dose comparison. Eur Heart J 1998;19: neurohumoral differences between spirapril and captopril in mild to mod- 481–9. erate chronic congestive heart failure. J Card Fail 1997;3:165–71. 34. The ATLAS Investigators. Comparative effects of low-dose versus high-dose 38. Taher MS, McLain LG, McDonald KM, et al. Effect of beta-adrenergic lisinopril on survival and major events in chronic heart failure: the Assess- blockade on renin response to renal nerve stimulation. J Clin Invest ment of Treatment with Lisinopril and Survival (ATLAS) Study [abstract]. 1976;57:459–65. Eur Heart J 1998;19:142. 39. Dzau VJ, Re R. Tissue angiotensin system in cardiovascular medicine: a 35. Cowie MR, Struthers AD, Wood DA, et al. Value of natriuretic peptides in paradigm shift? [editorial]. Circulation 1994;89:493–8. assessment of patients with possible new heart failure in primary care. 40. Hirsch AT, Talsness CE, Smith AD, Schunkert H, Ingelﬁnger JR, Dzau VJ. Lancet 1997;350:1349–53. Differential effects of captopril and enalapril on tissue renin-angiotensin 36. Buikema H, Pinto YM, Van Geel PP, et al. Differential inhibition of systems in experimental heart failure. Circulation 1992;86:1566–74.