Drugs 40 (Suppl. 2): 38-43, 1990 0012-6667/90/0200-0038/$3.00/0 © Adis International Limited All rights reserved. DRSUP1930.

A Comparison of Antihypertensive Drug Effects on the Progression of Extracranial Carotid Atherosclerosis The Multicenter Isradipine Diuretic Atherosclerosis Study (MIDAS)

Richard H. Grimm Jr,l John M. Flack,l Robert Byington, 2 Gene Bond 2 and Steven Brugger,3 for the MIDAS Research Groupt I Division of Epidemiology, School of Public Health, University of Minnesota, Minneapolis, USA 2 The Bowman Gray School of Medicine, Winston-Salem, North Carolina, USA 3 Sandoz Research Institute, East Hanover, New Jersey, USA

Summary

Hypertension is a major risk factor for coronary heart disease (CHD) and is the primary risk factor for stroke. Drug trials lowering blood pressure by pharmacological means have demonstrated impressive reduction in both fatal and nonfatal stroke (33 to 50%) that are virtually identical to the predicted stroke reduction, considering the observed diastolic blood pressure change (5 to 6mm Hg). On the other hand, reduction of CHD risk has been less impressive in these same trials. Although statistically significant, the reduction in CHD risk is roughly one-half (14%) of that predicted (25%) when results from these drug trials are analysed in aggregate. Most trials have used moderate to high dosages of thiazide diuretics or t3-blockers as therapies. Several factors may account for the disappointing results in CHD risk reduction. These drugs may induce metabolic disturbances in lipids, increased glucose tolerance, insulin resistance, or cause inadequate regression of left ventricular hypertrophy, thus attenuating the predicted reduction in CHD risk associated with pharmacological blood pressure lowering. Isradipine is a new dihydropyridine calcium antagonist that is highly effective in lowering blood pressure. Isradipine also has antiatherogenic properties in animal models of atherosclerosis. The effect of isradipine on atherosclerosis in humans is unknown. The Multicenter Isradipine Diuretic Atherosclerosis Study (MIDAS) is a 3-year double-blind, randomised trial in over 800 men and women with hypertension, aged 40 years or older. The primary aim of MIDAS is to compare the efficacy of isradipine 2.5 to 5.0mg twice daily vs hydrochlorothiazide 12.5 to 25mg twice daily in retarding the progression of extracranial carotid atherosclerosis. A sample of 883 men and women with hypertension and with demonstrated carotid plaque have been recruited and will be observed for at least 3 years. Participants are randomised into 2 parallel groups receiving either isradipine or hydrochlorothiazide. The primary end-point is intimal thickness and extent of plaque in the carotid arteries as measured by B-mode ultrasonography.

t

For a list of participating institutions, see p. 43.

Comparison of Antihypertensive Drug Effects

Hypertension remains a major clinical and public health problem. Hypertension is one of the major risk factors for coronary heart disease (CHD) and is the primary risk factor for stroke. In recent years, significant progress and refinements have occurred in the recognition and treatment of high blood pressure. Currently, there are several major classes of antihypertensive drugs available for use by the practising physician. In addition, effective nonpharmacological treatments have been more widely accepted and non-drug therapy is becoming more routine in medical practice. Much of this progress comes from the work in the basic sciences and clinical research. In addition, several large hypertension clinical trials carried out during this period have contributed significantly to this progress. Much has been learned; however, important questions remain. Probably the most important remaining question is 'Does pharmacological treatment of hypertension, especially 'mild' hypertension (diastolic blood pressure 90 to 104mm Hg), prevent the atherosclerotic complications of hypertension, namely CHD?' This question has been raised in recent meta-analyses in which the results of past trials have been examined to more clearly define the impact of blood pressure change on the reduction of cerebrovascular disease and CHD. Clinical trials involving the drug treatment of mild hypertension have clearly demonstrated impressive reductions in the incidence of both fatal and nonfatal stroke (33 to 50%). These same trials, when considered in aggregate, have shown a statistically significant (14%) reduction in CHD risk (Collins et al. 1990). However, this is only about one-half of the expected reduction in CHD risk (25%), given the observed change in diastolic blood pressure (5 to 6mm Hg). This disappointing outcome in the reduction in CHD risk is particularly disturbing, since CHD accounts for 3 to 4 times more deaths than does stroke. Thus, alternative therapies to the traditional moderate to high dose thiazide diuretic (50 to 100 mg/day of hydrochlorothiazide or chlor-

39

thalidone) and !3-blocker treatment regimens have been intensively studied. In particular, therapeutic regimens which may be more effective for the primary prevention of CHD have received more emphasis of late. Such therapies include aI-antagonists, angiotensin converting enzyme (ACE) inhibitors, newer !3-blockers, calcium antagonists and nonpharmacological treatments (weight loss, dietary sodium reduction, reduction in alcohol intake, and increased physical activity). Unlike the older therapies, these newer therapies have in common the fact that they are either neutral or somewhat positive in their effects on other cardiovascular risk factors, such as lipids, glucose tolerance, insulin and insulin resistance. Among the newer therapies, calcium antagonists are intriguing antihypertensive agents in that they appear to retard the progression of atherosclerosis via mechanisms independent of their effect on blood pressure or other metabolic parameters. Animal studies using the dihydropyridine agent isradipine have observed that this agent has a potent antiatherogenic effect (Habib et al. 1986; Heider et al. 1987; Weinstein & Heider 1987), reducing the arterial uptake of cholesterol and calcium. This effect appears to be common to all of the calcium antagonists thus far studied. Isradipine produces this effect in animals with a dose range (mg/kg) similar to that prescribed for patients with hypertension. The Multicenter Diuretic Isradipine Atherosclerosis Study (MIDAS) is a study designed to examine this effect in humans. The purpose of MIDAS is to determine the efficacy of isradipine compared with the diuretic hydrochlorothiazide in preventing the progression of early carotid atherosclerosis in men and women with hypertension. MIDAS is the first large trial to employ a noninvasive end-point using an ultrasonographically derived measure of the extent of atherosclerotic arterial lesions. Isradipine is a potent new dihydropyridine calcium antagonist (Chellingsworth et al. 1988; Persson et al. 1989). It has a high affinity for vascular

40

Drugs 40 (Suppl. 2) 1990

smooth muscle and undergoes extensive first-pass hepatic metabolism. Isradipine causes a short term increase in renal plasma flow and glomerular filtration rate, but these changes are not maintained long term (2 years). Also, unlike most other direct and indirect vasodilators, except for the ACE inhibitors, isradipine has a sustained natriuretic effect. The mean terminal half-life of isradipine is approximately 8.4 hours. Cardiac output increases with isradipine treatment, and myocardial oxygen consumption decreases as a result of increased coronary blood flow. Isradipine has a high affinity for receptor-operated calcium channels in resistance vessels. Myocardial depression is infrequent, and isradipine acts preferentially on the sinoatrial node but has no effect on the atrioventricular node (Hof & Ruegg 1988). The anti-atherogenic effect of isradipine and other calcium antagonists in animal models is attributed to a variety of causes: it produces a reduction in aortic cholesterol accumulation in cholesterol-fed rabbits (Habib et al. 1986); it prevents the loss of endothelium-derived relaxation factor in the arteries of cholesterol-fed rabbits (Habib et al. 1986); it reduces the synthesis of collagen and elastin in cell cultures (Persson et al.

1987); it antagonises smooth muscle proliferation (Orekhov et al. 1987); it reduces intracellular calcium accumulation (Fleckenstein 1986); and it also inhibits neutrophil and macrophage chemotaxis (Williamson et al. 1987). Other factors may also playa role (Weinstein & Heider 1989).

1. MIDAS Design MIDAS is a multicentre clinical trial with 8 clinical centres (fig. 1), a coordinating centre, and a central B-mode ultrasonography reading centre. Funding is provided by Sandoz Research, Hanover, New Jersey. The study is administered by a Steering Committee made up of the principal investigators from each clinical center, the coordinating center, the reading center, and selected representatives from Sandoz. A Policy Advisory Committee made up of individuals outside the study is responsible for examining the unblinded data, and for ethical and policy issues that arise during the trial. MIDAS is a double-blind parallel randomised trial. Participants who qualify after the initial eligibility visits are randomised to receive either isradipine 2.5 to 5.0mg twice daily or hydrochloro-

University of Minnesota. Minneapolis

Sandoz Research Institute. East Hanover Bowman Gray ~,..--+- School of Medicine. Winston-Salem

University of California. Davis University of Tennessee. Memphis

Fig. 1. Institutions participating in the MIDAS study.

University of Alabama. Birmingham

Miami Heart Institute. Miami Beach

41

Comparison of Antihypertensive Drug Effects

Isradipine (2.5mg twice daily)

Initial visit8

Hydrochlorothiazide (12.5mg twice daily) isation C

I

I 3 to 6 36 r-weeks - + - - - - - - monthsd ------I Fig. 2. Design of the MIDAS study.

a Initial B-mode carotid ultrasound to determine study eligibility is performed. b Three- to 6-week placebo run-in phase; placebo is continued until the average of 2 sitting diastolic blood pressure readings is > 90mm Hg but < 115mm Hg on 3 successive weekly visits. c Randomisation to double-blind parallel group active drug therapy; second carotid ultrasound is obtained to assess reproducibility. d Minimum follow-up on active drug therapy; quantitative B-mode carotid ultrasonography is performed semiannually.

30% compared with placebo. A 2-tailed a-level was set at 0.05 with a power of 90% ({3 = 0.l0). The resulting estimated sample size was 800. Additional patients were recruited to account for losses to follow-up and to study drop-out, and drop-ins to treatment.

2. End-Points The B-mode system of ultrasound was selected (Biosound 2000) as the major end-point measurement in MIDAS. Extensive technician training was undertaken and a central reading centre established. Details of the B-mode ultrasonic measurement and methods are provided elsewhere (Furberg et a1. 1989). The primary objective of MIDAS is to compare the effectiveness of 2 hypertensive treatment regimens in reducing the rate of progression of early extracranial carotid atherosclerTable I. Inclusion and exclusion criteria for the MIDAS trial

thiazide 12.5 to 25mg twice daily. The design of the study is provided in figure 2. The following patients were included: men and women 40 years and over with hypertension (diastolic blood pressure ~ 90 to 114mm Hg) at the end of each of the last 3 weeks of an initial washout period. Table I provides the inclusion and major exclusion criteria for MIDAS. Participants were recruited from a variety of sources, including referrals, media, brochure mailings, and field screening. After randomisation, patients are seen every 2 weeks until blood pressure is controlled. They are then followed up every 2 months for 1 year, and at 3-month intervals thereafter. Participants whose blood pressures are not controlled after at least 4 weeks on the maximum dose are treated ina non-blinded fashion with the ACE inhibitor enalapri1. Follow-up continues until all participants have reached their third annual visit. Sample size is based on the assumption that the rate of progression of atherosclerotic lesions in the comparison group (hydrochlorothiazide) will be between 0.15 and 0.20 mm/year. The treatment effect of the isradipine group is hypothesised to be

Inclusion criteria Age", 40 years (males and females) Hypertension: average of 2 sitting diastolic pressures must be > 90 and < 115mm Hg at each of 3 clinical visits during placebo washout Presence of carotid plaque: at least one lesion in extracranial carotid arteries with a measured plaque thickness of 1.3 to 3.5mm, as measured by quantitative 8-mode ultrasonography Exclusion criteria Accelerated hypertension, secondary hypertension, or symptomatic orthostatic hypotension Total cholesterol';; 260 mg/dl, triglycerides .;; 300 mg/dl Thyroid abnormality Serum creatinine> 1.8 mg/dl Liver enzymes > 3 x upper limits of normal History of myocardial infarction within 3 months History of stroke or transient ischaemic attack Previous carotid endarterectomy Unstable angina or angina pectoris not controlled with nitrates Pregnancy, lactation, or lack of effective birth control in premenopausal women Drugs (estrogens, androgens, progesterones, other steroid hormone, regular aspirin use, hydantoin anticonvulsants) Antihypertensive drugs (if participants refused to stop treatment by at least 4 weeks before randomisation) History of allergies to thiazides, calcium antagonists or ACE inhibitors

42

Drugs 40 (Suppl. 2) 1990

osis. The primary end-point for the trial is the comparison of the progression of the mean maximum intimal plus medial thickness over 12 carotid segments (inner and far walls of the common, the bifurcation, and internal segments, right and left sides of the neck) over a 3-year period. For this end-point, the greatest intimal plus medial thickness in each of the 12 segments is recorded at baseline and every 6 months thereafter for 36 months. Then for each time point, the mean of these 12 maximum segmental thicknesses is calculated. At the end of the trial, a least-squares regression line will be estimated for each individual subject using the means from 7 time periods (I baseline and 6 follow-up visits). The mean slopes for the 2 treatment groups will also be compared. Other end-points for the trial include the following: comparison of the progression of the mean maximum intimal plus medial thickness of the 4 common, the 4 bifurcation and the 4 internal carotid segments each examined separately; comparison of those single segments with the greatest progression over the 3-year period (identified retrospectively); and comparison of the progression of the single thickest segment identified at baseline. Secondary end-points include evaluation of safety, blood lipids, echocardiographic left ventricular mass, and quality of life.

3. Baseline Characteristics A total of 883 participants were recruited over a 19-month period in 8 US clinical centres. Selected baseline characteristics of these participants overall and by study group are provided in table II. The average age was 58.5 years, 79% were men and 22% were Black. The average resting systolic and diastolic blood pressures were 149.8mm Hg and 96.5mm Hg, respectively. Cigarette smokers made up 19.9% of the group at baseline. Another 4.3% and 5.5% were pipe and cigar smokers, respectively. Lipid measurements and special tests (table III) revealed the mean level of cholesterol to be 215.8 mg/dl, the LDL-cholesterol 145.6 mg/dl, HDL-cholesteroI47.3 mg/dl, and triglycerides 143.5 mg/dl. Apo A-I and Apo B were 147.0 and 130.4

Table II. Baseline characteristics of study participants (n = 883) Factor

Mean

Age (years)

58.5 78.6 72.4 21.5 6.1 184.7 149.8 96.5 71.5 10.3 19.9 4.3 5.5 4.3

Male (%) White(%) Black (%) Other (%) Weight (Ibs) Systolic blood pressure (mm Hg) Diastolic blood pressure (mm Hg) Heart rate (beats/min) Duration of hypertension (years) Cigarette smokers (current, %) Pipe smokers (%) Cigar smokers (%) Alcohol intake (drinks/week)

mg/dl, respectively. Other special tests showed a baseline fibrinogen level of 340.0 mg/dl, haemoglobin Ale of 6.7% and a fasting serum insulin of 15.5 JLIU /L. There were no significant differences in these characteristics at baseline between the 2 study groups.

4. Discussion There is an increasing need to examine issues concerning the optimal treatment of hypertension. The class as well as the specific drug effects are probably important considerations over and above Table III. Mean baseline lipoprotein values and other haematological variables in the MIDAS trial (n = 791) Mean Lipid (mg/dl)

Total cholesterol LDL-cholesterol HDL-cholesterol Triglycerides Apo A-I Apo B

215.8 145.6 47.3 143.8 147.0 130.4

Other Fibrinogen (mg/dl) Haemoglobin A1c (%) Serum insulin ("IU/L)

340.0 6.7 15.5

Comparison of Antihypertensive Drug Effects

the blood pressure lowering attributed to treatment. Results from animal studies have been encouraging because, in addition to being an efficacious blood pressure-lowering agent, isradipine also appears to have a potent antiatherogenic effect. The MIDAS study is a large scale trial that is being conducted to test the relative efficacy of 2 commonly used classes of antihypertensive agents in retarding the progression of proliferative arterial disease in the extracranial carotid arteries. The plan is to observe each participant for at least 3 years. The study aims to achieve similar blood pressure lowering in the 2 groups, allowing for an assessment of the antiatherogenic effects of these agents that is independent of their ability to lower blood pressure. The results of MIDAS may be relevant to the progression and development of atherosclerotic plaque in other vascular beds such as the coronary arteries. This trial is unique in having as its major endpoint a sensitive, noninvasive measurement of early carotid atherosclerosis. If these noninvasive endpoints prove to be valid and useful in studies like MIDAS, this methodology will be a major advance for clinical trials in the future. Current studies utilising fatal and nonfatal clinical events require much larger sample sizes and are becoming prohibitively expensive. Since there is a high correlation between the presence of carotid and coronary atherosclerosis, and since the mechanisms of plaque deposition in these cases may be similar, the results of MIDAS are potentially relevant to the unresolved issue of optimal antihypertensive therapy.

Acknowledgement The authors would like to acknowledge Kimberly Kaye Lindstrom and Jeanne Clark, PhD, for their assistance with the manuscript.

The Midas Research Group The following is a list of participating institutions, with principal investigators in parentheses. Clinical Centers: Bowman Gray School of Medicine, Winston-Salem, North Carolina (V. Buckalow); Medical College of Georgia, Augusta, Georgia (A. Carr); Miami Heart Institute, Miami Beach, Florida (J. Raines); University of Alabama, Birmingham, Alabama (H. Schnaper); University of Cali-

43

fornia at Davis, Davis, California (N. Borhani); University of Minnesota, Minneapolis, Minnesota (R. Grimm); University of Tennessee, Memphis, Tennessee (w. Applegate); Wayne State University, Detroit, Michigan (J. Sowers). Operations/Analysis Center: Bowman Gray School of Medicine, Winston-Salem, North Carolina (R. Byington). Ultrasound Core Laboratory: Bowman Gray School of Medicine, Winston-Salem, North Carolina (G. Bond). Central Laboratory: MetPath, Teterboro, New Jersey. Sandoz Research Institute, East Hanover, New Jersey (H. Miller. S. Baumgardner. S. Brugger). The Chairman of the Investigators' Committee is C. Furberg; the Chairman of the Policy and Data Monitoring Committee is J. Cohn.

References Chellingsworth MC, Willis lV, lack DB, Kendall Ml. Pharmacokinetics and pharmacology of isradipine (PN 200-110) in young and elderly patients. American 10urnal of Medicine 84 (Supp!. 3B): 72-79, 1988 Collins R, Peto R, MacMahon S, et a!. Blood pressure, stroke, and coronary heart disease. Part 2. Short-term reductions in blood pressure: overview of randomized drug trials in their epidemiological context. Lancet 335: 827-838, 1990 Fleckenstein A. Model experiments on anticalcinotic and antiatherosclerotic arterial protection with calcium antagonists. American 10urnal of Cardiology 19 (Supp!. II): 109-121, 1986 Furberg CD, Byington RP, Borhani NA. For the MIDAS Research Group. Multicenter Isradipine Diuretic Atherosclerosis Study (MIDAS). American 10urnal of Medicine 86 (Supp!. 4A): 37-39, 1987 Habib lB, Bossaller C, Wells S, Williams C, Morrisett 10, et a!. Preservation of endothelium-dependent vascular relaxation in cholesterol-fed rabbit by treatment with the calcium channel blocker, PN-200-11 O. Circulation Research 58: 305-309, 1986 Heider lG, Weinstein DB, Pickens CE, Lan S, Su CM. Antiatherogenic activity of calcium channel blocker isradipine (PN200-110): a novel effect on matrix synthesis independent of calcium channel blockade. Transplantation Proceedings 19 (Supp!. 5): 96-101, 1987 Hof RP, Ruegg U. Pharmacology of the new calcium antagonist isradipine and its metabolites. American 10urnal of Medicine 84 (Supp!. 3B): 13-17, 1988 Orekhov AN, Tertov VV, Kashimov KA, Kudryashov SS, Smirnov VN. Evidence of antiatherosclerotic action of verapamil from direct effects in arterial cells. American 10urnal of Cardiology 59: 495-496, 1987 Persson B, Andersson OK, Wysocki M, Hedner T, Aurell M. Renal and hemodynamic effects of isradipine in essential hypertension. American 10urnal of Medicine 86 (Supp!. 4A): 60-64, 1989 Weinstein DB, Heider lG. Antiatherogenic properties of calcium antagonists. American 10urnal of Cardiology 59: 163B-I72B, 1987 Weinstein DB, Heider lG. Antiatherogenic properties of calcium antagonists. American 10urnal of Medicine 86 (Supp!. 4A): 2732, 1989 Williamson KC, Tauber AL, Navarro 1. Amlodipine inhibits formyl-methionyl-Ieucylphenylalanine receptor-coupled calcium transport in human neutrophils. 10urnal of Leukocyte Biology 42: 239-244, 1987

Correspondence and reprints: Dr Richard H. Grimm Jr, Division of Epidemiology, School of Public Health, University of Minnesota, Stadium Gate 27, 611 Beacon Street, MN 55455, USA.