The Calcium Antagonist Controversy: The Emerging Importance of Drug Formulation as a Determinant of Risk

The Calcium Antagonist Controversy: The Emerging Importance of Drug Formulation as a Determinant of Risk Murray Epstein,

MD

Calcium antagonists are one of the widely available classes of antihypertensive agents. Their broad appeal is attributable to several features, including their efficacy, their beneficial characteristics such as metabolic neutrality, and the occurrence of relatively few nuisancetype side effects. Despite these attributes, a number of retrospective analyses have suggested that calcium antagonists may be detrimental and may both promote adverse cardiovascular events and increase the risk of cancer by interfering with cellular apoptosis. On the basis of this and other retrospective analyses, Furberg and Psaty (Am J Hypertens 1996;9:122–125) have proposed that the use of calcium antagonists as first-line antihypertensive agents should be discontinued. I have previously countered these allegations and have suggested that they are not relevant to the newer calcium antagonist formulations in current use. It is not widely

appreciated that different formulations of the same chemical moiety can produce markedly different hemodynamic and neurohormonal effects, due to differences in the rate of drug delivery into the systemic circulation. During chronic treatment with dihydropyridine calcium antagonists, major fluctuations in blood pressure (rapid onset and offset of antihypertensive effects) during the dosing interval may occur for drugs and formulations that are short acting. In contrast, slow-release formulations of otherwise rapidly absorbed dihydropyridines achieve a more gradual and sustained antihypertensive effect. It is probable that newer calcium antagonist formulations that are truly once daily and do not provoke intermittent sympathetic activation or a cardioacceleratory response will not promote adverse cardiovascular events. Q1997 by Excerpta Medica, Inc. (Am J Cardiol 1997;79(10A):9–19)

ince their introduction ú25 years ago, calcium S antagonists have emerged as one of the most attractive and widely used classes of antihypertensive

myocardial infarction (cases) and those who had not (controls), with respect to their prior use of various antihypertensive drug therapies. They concluded that hypertensive patients receiving calcium antagonists had a significantly greater risk of myocardial infarction compared with those receiving b blockers (relative risk [RR] Å 1.63; 95% confidence interval [CI] Å 1.23–2.16). On the basis of this and other retrospective analyses, Furberg and Psaty have proposed that the use of calcium antagonists as first-line antihypertensive agents should be discontinued.8 Undoubtedly, many readers are familiar with concerns raised about the study, including the fact that the retrospective, population-based, case control design used by Furberg and Psaty precluded ‘‘randomization of treatment allocation.’’ In her recent book, Science on Trial,9 dealing with the breast implant controversy, Marcia Angell succinctly summarizes the pitfalls in epidemiologic cohort studies:

agent. Of the 20–25 million patients receiving medication for hypertension in the United States, about one quarter are taking calcium antagonists. Their wide appeal is attributable to several features, including efficacy, beneficial characteristics such as metabolic neutrality, and the occurrence of relatively few nuisance-type side effects.1,2 In addition, recent investigations have focused on their possible protective effects on target organs, such as the heart and kidney,3,4 further enhancing their appeal. Despite these attributes, a number of retrospective analyses have suggested that calcium antagonists may be detrimental and may promote adverse cardiovascular events. More recently, Pahor et al5,6 have proposed that calcium antagonists increase the risk of cancer by interfering with cellular apoptosis. The first meta-analysis of the effects of 1,4-dihydropyridine calcium antagonists on outcome included 21 clinical trials and suggested an adverse trend on mortality and the development of subsequent myocardial infarction.7 Subsequently, Furberg and Psaty8 reported the findings of an observational, case control study of patients who had suffered a From the Nephrology Section, Veterans Affairs Medical Center, and Department of Medicine, University of Miami School of Medicine, Miami, Florida. Portions of this article are adapted with permission from an earlier review published in The American Journal of Hypertension.14 Address for reprints: Murray Epstein, MD, Nephrology Section, Veterans Affairs Medical Center, 1201 NW 16th Street, Miami, Florida 33125.

Probably the chief difficulty in epidemiologic studies is choosing groups of people who are alike in every way except for the exposure in question (in cohort studies) or the disease in question (case control studies). Yet this is essential. Otherwise, some other difference between the groups might account for the results and badly mislead everyone. Other differences between groups that may confuse the results are termed ‘‘confounding variables.’’ For example, cigarette smokers are more likely to drink alcohol than are nonsmokers. So when an epidemiologic study shows a link between cigarette smoking and disease, it is necessary to determine whether the real association is

Q1997 by Excerpta Medica, Inc.

0002-9149/97/$17.00 PII S0002-9149(97)00266-X

All rights reserved.

/ 2z0c 7203 Mp

9

Thursday May 15 03:12 PM

EL–AJC-S (suppl 7200 ’97)

7203

9

TABLE I Classification of Calcium Antagonists Calcium Antagonist Verapamil Amlodipine Felodipine Isradipine Lacidipine Nicardipine Nifedipine Nimodipine Nisoldipine Nitrendipine Diltiazem

Parent Compound

Class

Phenylalkylamine 1,4-Dihydropyridine 1,4-Dihydropyridine 1,4-Dihydropyridine 1,4-Dihydropyridine 1,4-Dihydropyridine 1,4-Dihydropyridine 1,4-Dihydropyridine 1,4-Dihydropyridine 1,4-Dihydropyridine Benzothiazepine

I II II II II II II II II II III

with smoking or whether it might possibly be with drinking (the confounding variable in this case). It could be the combination—or even some other factor—that might be different between smokers and nonsmokers. Although there are statistical methods for neutralizing confounding variables, they are not perfect, and they are of no use whatsoever unless the confounding variables are known and measured. Although meta-analyses and observational studies clearly have limits,10,11 Psaty et al12 have raised an important question that deserves consideration: whether calcium antagonists, as a group, promote adverse cardiovascular events. Furthermore, media reporting of the presentation of Drs. Furberg, Psaty, et al triggered concern among users of calcium antagonists and even among those taking other antihypertensive drug therapies. This underlines the importance and relevance of critically considering the issues raised by Psaty et al. I have previously debated Furberg on these issues and have suggested that the allegations are not relevant to the calcium antagonists in current use.13,14

In this discussion I wish to expand on these earlier discussions and counter the arguments that calcium antagonists, as a group, are dangerous. I will not limit my remarks to the merits of Furberg and Psaty’s recent retrospective analysis and the constraints of the experimental design, which have been discussed extensively of late.10 Rather, in this review, I will focus on a number of key issues that I believe are pivotal to this discussion. First, I will review the markedly disparate effects of differing calcium antagonist formulations and their clinical implications, which are clearly relevant to this controversy. Second, I will also consider the putative mechanisms proposed by Furberg and Psaty to account for the enhanced morbidity. Finally, I will present data, derived from recent prospective studies, that are inconsistent with Furberg and Psaty’s findings.

DIFFERENT FORMULATIONS OF THE SAME CALCIUM ANTAGONIST CAN PRODUCE MARKEDLY DIFFERENT PHARMACOKINETIC AND PHARMACODYNAMIC EFFECTS First, we must recognize that to group all of the calcium antagonists together as if they represent a uniform type of agent is erroneous. The fact that calcium antagonists are markedly heterogeneous, and are chemically dissimilar agents, is well established (Table I).15,16 Perhaps what is not so well appreciated is the fact that even different formulations of the same chemical moiety can produce markedly different hemodynamic and neurohormonal effects.17 – 19 The earliest calcium antagonists were short acting. Subsequently, the drug delivery systems for the short-acting agents were modified to provide more slowly and consistently maintained calcium antagonist activity.

FIGURE 1. Comparison of the differing pharmacokinetic profiles of a similar dose of nifedipine administered either as a capsule or as the gastrointestinal therapeutic system (GITS) formulation. Repetitive administration of the nifedipine capsule results in a rapid attainment of peak concentration (within 1 hour) with a rapid decline. In contrast, plasma drug concentration of nifedipine GITS reaches a plateau slowly with maintenance of relatively constant drug levels over an extended duration.

10

THE AMERICAN JOURNAL OF CARDIOLOGYT

/ 2z0c 7203 Mp

10

VOL. 79 (10A)

Thursday May 15 03:12 PM

MAY 22, 1997

EL–AJC-S (suppl 7200 ’97)

7203

FIGURE 2. A) Comparison of the effects of a rapid bolus dose and a slow infusion of nifedipine on blood pressure and heart rate. A rapid intravenous bolus dose (and exponential infusion) rapidly attained an effective level of nifedipine. In contrast, a progressive incremental intravenous infusion attained an identical plasma nifedipine level more gradually. Results are mean { standard deviation. B) The rapid attainment of an effective plasma concentration via a rapid intravenous bolus dose (and exponential infusion) caused no appreciable fall in blood pressure. In contrast, if an identical plasma nifedipine concentration was attained gradually, there was a significant decrease in blood pressure but no increase in heart rate. Results are mean { standard deviation. (Adapted, with permission, from J Cardiovasc Pharmcol.18)

For example, after administration of a nifedipine capsule, plasma drug concentration peaks rapidly within 1 hour to high levels and then falls rapidly (Figure 1). In contrast, plasma drug concentrations with the nifedipine gastrointestinal therapeutic system (GITS), which uses an osmotic pump system to deliver the drug in a steady infusion over the 24hour interval,20 attains a plateau slowly, peaking at approximately 6 hours, and thereafter maintains a relatively constant drug level over 24 hours. Such new slow-release formulations were developed primarily to produce a sustained 24-hour therapeutic

effect. An added benefit that is equally important, albeit not widely appreciated, is that these uniform plasma concentrations avoid provoking activation of the renin–angiotensin and the sympathetic nervous systems.17 The rate of drug delivery into the systemic circulation has profound effects on the hemodynamic and neurohormonal responses to a dihydropyridine calcium antagonist. Kleinbloesem et al18 conducted a study of intravenously administered nifedipine which showed clearly that the rate of drug delivery determined the pattern of response (Figure 2A).

A SYMPOSIUM: CALCIUM ANTAGONISTS IN HYPERTENSION

/ 2z0c 7203 Mp

11

Thursday May 15 03:12 PM

EL–AJC-S (suppl 7200 ’97)

7203

11

FIGURE 3. Introduction and growth of new prescriptions of different formulations of nifedipine since 1986. Nifedipine (Procardia) was only available in the United States as a shortacting formulation in the late 1980s. The slow-release gastrointestinal therapeutic system (GITS) formulation was not introduced until 1989, and its prescription and usage did not increase appreciably until 1991–1992. (Data adapted from IMS NPA audit, Plymouth Meeting, Pennsylvania.)

They compared the effects of a rapid bolus dose and a slow infusion on systemic hemodynamics. The rapid attainment of an effective plasma concentration via a rapid intravenous bolus dose (and exponential infusion) caused no appreciable fall in blood pressure because there was an associated increase in adrenergic activity and a marked increase in heart rate, and presumably also in cardiac output (Figure 2B). In contrast, if an identical plasma nifedipine concentration was gradually attained over several hours by means of a slow intravenous infusion, there was no adrenergic response and no increase in heart rate, but there was a significant decrease in blood pressure (Figure 2B). These studies emphasize the importance of the rate of attainment of plasma levels in determining the consequent adrenergic and cardioacceleratory response. During chronic treatment with dihydropyridines, major fluctuations in blood pressure (rapid onset and offset of antihypertensive effects) during the dosing interval may persist for short-acting drugs and formulations.19,21,22 In contrast, slow-release formulations of otherwise rapidly absorbed dihydropyridines achieved a more gradual and sustained antihypertensive effect. The studies of Frohlich et al23 are representative of several recent studies that highlight the differences between alternative formulations of calcium antagonists on intermittent increases in sympathetic activity after dosing. These investigators compared the effects of different formulations of nifedipine on sympathetic activation, assessed by measuring plasma norepinephrine at trough. They administered either nifedipine GITS once daily or an immediate release nifedipine capsule 3 times daily to 10 patients with mild-to-moderate hypertension. Both formula12

THE AMERICAN JOURNAL OF CARDIOLOGYT

/ 2z0c 7203 Mp

12

tions reduced mean arterial pressure and total peripheral resistance to a similar extent. Trough plasma norepinephrine levels of the patients receiving nifedipine GITS were virtually identical to those of the patients receiving placebo (375 { 60 vs 387 { 71 pg/mL). In contrast, trough norepinephrine levels tended to be substantially higher in patients receiving immediate release nifedipine capsules 3 times daily, although this adrenergic activation did not attain statistical significance. More recently, Rousseau et al24 compared the effects of nisoldipine coat core (CC), an extendedrelease formulation, with an immediate-release formulation of nisoldipine. They reported that no effect on heart rate or neurohormonal levels was observed following administration of nisoldipine CC and suggested this was due to the more favorable pharmacokinetic profile, which avoids sudden vasodilation and hypotension.

RELEVANCE OF DIFFERENT FORMULATIONS TO RECENT RETROSPECTIVE REVIEWS The clinical relevance of the different pharmacodynamic profiles of the rapidly absorbed agents and the more recently introduced slow-release formulations, and their disparate cardioacceleratory and neurohormonal effects, is implicit. As demonstrated in Figure 3, nifedipine was only available in the United States as a short-acting formulation in the late 1980s. The slow-release GITS formulation was not introduced until 1989, and its prescription and use did not increase appreciably until 1991–1992. The data for the benzothiazepine calcium antagonists, diltiazem, are similar. The only diltiazem formulation available in the late 1980s was the tablet.

VOL. 79 (10A)

Thursday May 15 03:12 PM

MAY 22, 1997

EL–AJC-S (suppl 7200 ’97)

7203

Diltiazem sustained release (SR) was not introduced until 1988, and its usage did not peak until 1991– 1992. The more recent slow-acting formulation, diltiazem CD, was not introduced until 1991. Consequently, we must assume that studies encompassing the period of observation beginning in 1986 must have included many patients using the more rapidrelease diltiazem formulations. The study of Psaty et al12 encompassed a period of observation from 1986 to 1993 at a time when the slow-release formulations were not readily available or, in some instances, had not yet been introduced. Indeed, their investigation reported the use of 3 rapid-acting prototype calcium antagonists, nifedipine, verapamil, and diltiazem, 2 of which (nifedipine and diltiazem) are not indicated for the treatment of hypertension by the US Food and Drug Administration. Consequently, it is reasonable to assume that these rapid-acting formulations were precisely the ones previously documented to cause wide fluctuations in drug concentrations and vasodilatory effects during apparent steady-state treatment, with resultant cardioacceleration and recurring sympathetic activation.17

WHY HUMAN STUDIES ON THE REGRESSION OF ATHEROSCLEROSIS ARE NEGATIVE The previous discussion of the differences between the rapidly absorbed, rapid-acting formulations and the more recently introduced longer-acting formulations may account, at least in part, for the negative results to date of human studies on the regression of atherosclerosis. None of the calcium antagonist formulations used in regression studies to date has provided sustained blood levels over the entire 24-hour period, and all of these agents have been shown to evoke reactive cardioacceleration during apparent steady-state conditions.17 Consequently, a reasonable hypothesis to account for the increased adverse vascular events in regression studies is that medications provoking a counterregulatory response with cardioacceleration may exacerbate ischemia in patients with basal myocardial ischemia. Conversely, agents producing true steady-state levels (i.e., GITS-like preparations or amlodipine or lacidipine) do not evoke such unwanted counterregulatory responses and, therefore, may not promote increased adverse vascular events. Regression using agents that produce true steadystate levels needs to be assessed in large prospective studies to validate this hypothesis. One such study that has recently been initiated is the Prospective Randomized Evaluation of the Vascular Effects of Norvasc [amlodipine] Trial (PREVENT), a prospective 3-year, double-blind, placebo-controlled trial to evaluate regression. It will compare the effect of amlodipine with that of placebo on the development and progression of atherosclerotic lesions in coronary and carotid arteries in patients with coronary artery disease. Another regression study being conducted at present is the European Lacidipine Study on Ath-

TABLE II Potential Mechanisms for Increased Relative Risk of Cardiovascular Events with Calcium Antagonists 1. 2. 3. 4.

Negative inotropic effects Proarrhythmic effects Proischemic effects (from coronary steal) Prohemorrhagic effects

erosclerosis (ELSA).25 This 4-year European study is similar to the Multicentre Isradipine Diuretic Atherosclerosis Study (MIDAS)26 but uses lacidipine— a calcium antagonist that, like amlodipine, possesses intrinsic pharmacokinetic properties that result in relatively stable plasma drug concentrations—to assess the efficacy of calcium antagonists in retarding the progression of atherosclerotic lesions.

BLOOD PRESSURE VARIABILITY AS A DETERMINANT OF END-ORGAN DAMAGE Aside from the theoretical construct that favors a reduction in coronary risk, additional considerations suggest that the more ‘‘smoothly’’ sustained activity of the newer agents, such as amlodipine, may be of additional benefit in the treatment of hypertension. This is because the variability in blood pressure, independently of the absolute level, correlates with end-organ damage in hypertensive patients. Both cross-sectional27 and, more recently, longitudinal data28 suggest that end-organ damage in hypertension is produced by both the average daily blood pressure values and by the magnitude of daily variations in blood pressure. Consequently, drugs that are short acting and whose use is associated with major fluctuations in blood pressure may fail to attenuate end-organ damage despite ostensible ‘‘blood pressure control.’’ PROPOSED MECHANISMS TO ACCOUNT FOR THE ADVERSE CARDIOVASCULAR EVENTS ATTRIBUTED TO CALCIUM ANTAGONISTS It has been proposed that several mechanisms account for the increased incidence of adverse events that have been attributed to calcium antagonists.8,12 These are summarized in Table II. The next section will consider whether these suggested mechanisms are supported by rigorous scientific data. Negative inotropic effects: In isolated myocardial preparations, all calcium antagonists have been shown to exert negative inotropic effects.2 Nevertheless, a number of caveats are in order. Calcium antagonists vary in their effects on in vivo left ventricular function depending on a number of factors: pretreatment left ventricular function; whether or not ischemia or hypertension is alleviated concomitantly; the dose used; and the extent of arterial dilation. The negative inotropic effects of calcium antagonists are dose dependent and, with the relatively low doses of these drugs used clinically, may not produce clinically important adverse effects on car-

A SYMPOSIUM: CALCIUM ANTAGONISTS IN HYPERTENSION

/ 2z0c 7203 Mp

13

Thursday May 15 03:12 PM

EL–AJC-S (suppl 7200 ’97)

7203

13

diac performance. Finally, the intrinsic negative inotropic properties of these compounds are greatly modified by a baroreceptor-mediated reflex augmentation of b-adrenergic tone consequent to vasodilation.2 Are calcium antagonists proarrhythmic? Psaty et al12 have proposed that a proarrhythmic effect may account for the increased mortality observed in patients treated with calcium antagonists. The cellular electrophysiologic basis of arrhythmia formation in the heart has been detailed in a recent review.29 It is apparent that calcium overload in myocardial cells is an important factor in the genesis of diverse serious arrhythmias, especially rate and rhythm disorders associated with myocardial ischemia, infarction, and reperfusion. These conditions result in significant elevations in intracellular calcium levels in regions of poor or no perfusion. Calcium antagonists block the entry of calcium into heart cells, and they have been shown to be effective in preventing experimentally induced ischemic and reperfusion arrhythmias.29 Theoretically, there are circumstances when dosage with certain calcium antagonists potentially leads to myocardial calcium overload. For example, as detailed in a recent review by Chakko and Bassett,30 it is possible that a b1-adrenoceptor–mediated increase in calcium entry may disturb myocardial calcium regulation, especially in diseased hearts where cellular calcium regulation is already compromised. In this manner, dosage with a short-acting calcium antagonist may cause sufficient recurrent vasodilation several times a day to be conceivably proarrhythmic. Despite such considerations, a review of the available data fails to support the notion that calcium antagonists are proarrhythmic. Rigorous prospective data on proarrhythmic and antiarrhythmic effects of calcium antagonists are relatively sparse. The proarrhythmic mechanism proposed by Psaty et al,12 however, seems unlikely in view of the extensive experimental work demonstrating that calcium antagonists have an antiarrhythmic effect on both ischemic and reperfusion arrhythmias.29 For example, a recent Finnish study31 of 155 consecutive patients resuscitated after out-ofhospital ventricular fibrillation demonstrated a clear antiarrhythmic effect of the dihydropyridine calcium antagonist nimodipine. Ventricular fibrillation was corrected in all but 1 of 75 nimodipine-treated patients compared with 12 of 80 control subjects (p õ0.01). Furthermore, ventricular fibrillation recurred in one of the patients receiving nimodipine and 12 of those receiving placebo, a clear benefit in favor of the calcium antagonist (p Å 0.006). In summary, although excessive concentrations of various calcium antagonists may conceivably cause varying degrees of atrioventricular block, there is no evidence that the newer extended-release agents are proarrhythmic.31 Are calcium antagonists proischemic? Several reports have suggested that apart from their beneficial antianginal effects in patients with angina pectoris, dihydropyridine calcium antagonists may also pro14

THE AMERICAN JOURNAL OF CARDIOLOGYT

/ 2z0c 7203 Mp

14

mote proischemic complications.17,32 In their metaanalysis, Furberg and Psaty8 related the proischemic effect to ‘‘coronary steal.’’ The steal phenomenon presumably induces myocardial ischemia by reducing coronary resistance and directing flow from poststenotic underperfused areas to normally perfused myocardium.33,34 Highlighted in a recent editorial, the evidence to date supporting the concept that nifedipine causes coronary steal in humans is not rigorous.35 A number of examples are inconsistent with this notion. Malacoff et al36 measured regional myocardial blood flow in patients with coronary artery disease using xenon-133 with positron emission tomography and observed that nifedipine improved regional myocardial blood flow to ischemic segments. An important study of myocardial perfusion in patients with variant angina during exercise was reported by Kugiyama et al.37 They observed that nifedipine improved exercise duration and reduced the size of the perfusion defect on thallium-201 scintigraphic imaging; conversely the b blocker propranolol shortened exercise duration and actually increased the size of the perfusion defect compared with placebo. Other studies have also failed to show that nifedipine causes a coronary steal phenomenon when coronary perfusion was assessed38; most suggest that nifedipine has a favorable effect on coronary perfusion in humans.39 In addition, studies have shown that nifedipine improves left ventricular dysfunction associated with ischemia in humans, a finding that argues against coronary steal.40 Thus, the concept that nifedipine is proischemic and promotes coronary steal in humans remains speculative. A subsequent study in patients with chronic stable angina by Parmley et al41 disclosed that nifedipine GITS reduced the weekly number of anginal episodes from 5.7 to 1.8 (p Å 0.0001) and the number of ischemic events (assessed by ambulatory ECG monitoring) from 7.3 to 4.0 (p Å 0.0001). The drug, either alone or in combination with a b blocker, reduced ischemia over a 48-hour period. Specifically, Parmley et al did not observe a proischemic effect (worsening of ischemia on ambulatory ECG monitoring) with nifedipine GITS. Such studies clearly call into question generalizations that calcium antagonists, as a class, are proischemic.

DATA FROM RECENT PROSPECTIVE STUDIES To date, much of the controversy has centered on the limitations of meta-analyses and retrospective studies. Although such analyses are helpful and suggestive, they are no substitute for randomized, prospective studies with hard endpoints. Within the past year, four prospective studies in well-defined patient populations have become available: the Prospective Randomized Amlodipine Survival Evaluation (PRAISE) study,42 the Shanghai Trial of Nifedipine in the Elderly (STONE) study,43 the VasodilatorHeart Failure Trial (V-HeFT III),44 and the Doppler flow and Echocardiography in Functional cardiac In-

VOL. 79 (10A)

Thursday May 15 03:12 PM

MAY 22, 1997

EL–AJC-S (suppl 7200 ’97)

7203

sufficiency: Assessment of Nisoldipine Therapy (DEFIANT) II.45 All four studies provide data that are relevant to the present controversy. Both PRAISE and V-HeFT III provided mortality data on patients with severe congestive heart failure. DEFIANT II provided data in the setting of postmyocardial infarction. All studies failed to reveal an increased mortality in the patients treated with calcium antagonists. The STONE study was a single-blind, placebocontrolled trial of 3 years’ duration that compared nifedipine with placebo in 1,632 subjects aged 60– 79 years.43 Essential hypertensive patients were allocated to either nifedipine or placebo treatment. The terminating events for the survival analysis were stroke, heart failure, uremia, myocardial infarction, angina pectoris, severe arrhythmia, death, and hospitalization for severe illness. An intention-to-treat analysis was used. Nifedipine produced a highly significant (p õ0.001) decrease in the total number of terminating events. The decrease in events among patients in the nifedipine group was significant whether or not concomitant treatment was included, or whether ‘‘all’’ events were analyzed or the analysis was restricted to those related to hypertension. Finally, in a preliminary analysis of two of the largest hypertension trials of nifedipine GITS and amlodipine,46 there were no deaths, and the incidence of myocardial infarction was similar to that reported for diuretics and b blockers by Psaty et al.12 A recent assessment47 of the association of calcium antagonist use with mortality risk in hypertensive subjects from the Framingham Heart Study fails to support the allegations of Psaty et al. Separate analyses were performed for patients with and without a history of coronary heart disease at baseline. A total of 3,547 subjects (52% female) were eligible for analysis. The mean follow-up was 7 years (maximum 12.7 years). Mortality as a function of baseline coronary heart disease status and calcium antagonist use is summarized in Table III. Proportional hazard analyses, adjusting for age, sex, blood pressure, and other covariables, revealed no evidence of increased mortality risk in association with calcium antagonist use. Results were similar when short-term mortality (4 years) was examined. The authors47 concluded that in this substantive population-based sample, they found no evidence of increased mortality risk in hypertensive calcium antagonist users (Table III).

DO CALCIUM ANTAGONISTS INCREASE THE RISK OF CANCER? The indictments against first-generation calcium antagonists seem to lengthen steadily. The controversy regarding calcium antagonists has been extended recently beyond the relatively narrow limits of cardiovascular safety. Last June, the lead article in the American Journal of Hypertension comprised a report by Pahor and colleagues of an apparent increase in the risk of cancer in patients taking calcium antagonists for hypertension.5 Subsequently, they extended these findings to a larger cohort of elderly

patients.6 The validity of the association with cancer necessitates careful scrutiny. Pahor et al5,6 have hypothesized that calcium antagonists increase the risk of cancer by interfering with the physiologic mechanisms that regulate cancer cell growth. Evidence is emerging that calcium antagonists can block apoptosis,48,49 an efficient mechanism for limiting cancer growth.50 – 53 Calcium antagonists might affect cancer risk generally or be limited to specific sites where calcium mechanisms predominate. For example, colon cancer has been related to reduced calcium ingestion.54 Recently, in the lead article in The Lancet,6 Pahor et al have extended the initial analyses performed in individuals receiving treatment for hypertension5 to the general older population.6 The aim was to assess whether individuals taking calcium antagonists, for any indication, were at higher risk of developing cancer than those not taking such drugs. They observed that the hazard ratio for cancer associated with calcium antagonists (1,549 personyears; 47 events), compared with those not taking calcium antagonists (17,225 person-years; 373 events), was 1.72 (95% CI Å 1.27 – 2.34; p Å 0.0005), after adjustment for confounding factors. A significant dose–response gradient was found. Hazard ratios associated with verapamil, diltiazem, and nifedipine did not differ significantly from each other. The association between calcium antagonists and cancer was found with most of the common cancers. As pointed out by Dargie in an accompanying editorial,55 the validity of the association of calcium antagonists with cancer is questionable. The ‘‘evidence’’ presented is based entirely on observational data in elderly patients who would have first received calcium antagonists late in life. The previous drug history of these individuals is unknown and, in the current paper, a group of patients being treated with calcium antagonists for various reasons is compared with a heterogeneous control group, thereby rendering adjustment for possible confounders very difficult. Exposure information: A major limitation of the analyses carried out by Pahor et al5,6 continues to be the weak exposure information. Information on drug use was collected only at the baseline visit in 1988 using a 2-week window, and was not collected after that time. For the current analysis, patients were followed from the baseline 1988 time period for a mean of 3.7 years until a cancer event, death, or study completion in 1992. For patients receiving calcium antagonists, exposure was assumed to be continuous from the baseline period until the outcomes described above. For those not exposed, it was assumed that calcium antagonists were not used for the entire time period. This assumption of continuous exposure or nonexposure is the basis of the proportional hazards model and the use of person-time as a denominator. This use of person-time would be appropriate if the hypothesis being tested were that a one-time (e.g., atomic bomb) exposure conferred a long-term continuing risk of cancer. It is not appropriate when

A SYMPOSIUM: CALCIUM ANTAGONISTS IN HYPERTENSION

/ 2z0c 7203 Mp

15

Thursday May 15 03:12 PM

EL–AJC-S (suppl 7200 ’97)

7203

15

TABLE III Mortality as a Function of Baseline Coronary Heart Disease (CHD) Status and Calcium Antagonist Use* CHD Present

No CHD Calcium Antagonist Users

Calcium Antagonist Nonusers

Calcium Antagonist Users

Calcium Antagonist Nonusers

182 65.8 22.5

2,751 62.4 21.9

219 66.9 37.4

395 70.2 49.9

Subjects at risk Age (years) Deaths (%) *Adapted from Circulation.47

continuous exposure or exposure close in time to the event is the basis of the hypothesis (i.e., apoptosis). Therefore, the use of person-time is the denominator for the incidence calculations on which the reported relative risks are based is inappropriate, and attributing subsequent follow-up time to calcium antagonist use is not valid. If person-time were eliminated and the crude rates based on the number of actual individuals known to be exposed or not exposed at baseline is analyzed instead, the result is a smaller relative risk that is not statistically significant. Pahor et al5,6 have suggested that inhibition of apoptosis is a possible explanation for the development of cancer on the basis that calcium is an intracellular messenger in this process. Calcium, however, is such a ubiquitous component of cell signaling systems that almost any adverse drug reaction could be said to be explained on the basis of this mechanism. This explanation must, therefore, be regarded as highly speculative. Indeed, the converse may be possible—as calcium antagonists have shown favorable results in the treatment of cancer.56

LESSONS LEARNT FROM HISTORY It is worthwhile remembering that we have encountered previous cancer scares with respect to antihypertensive agents, and history has taught us to

be cautious. Two examples come readily to mind. In the atenolol study of ú2000 elderly hypertensive patients, the relative risk of cancer in men was increased two- or three-fold.57 Subsequently, a large retrospective group of patients was studied, linking data from the Glasgow Blood Pressure Clinic to the West of Scotland Cancer registry, and these results were shown to refute this claim.58 In 6,528 patients, there was no overall association between cancer and the use of atenolol. A likely explanation is that an increased risk of cancer of up to three times is often a nonspecific finding of questionable significance.59 Another cancer scare centered on reserpine. In 1974, The Lancet published 3 separate retrospective studies from Oxford (UK), Boston (USA), and Helsinki (Finland), linking reserpine to breast cancer, together with an accompanying editorial that commented on a three- to four-fold increase of breast cancer in women exposed to reserpine compared with those not exposed.60 The authors of the 3 papers included internationally recognized authorities such as Sir Richard Doll, who is credited with establishing the links between smoking and lung cancer. Yet, further work subsequently disproved these claims. The same group that had provided evidence for the increased risk with reserpine in 1974,61 disproved the hypothesis a full 10 years later.62 These discordant

FIGURE 4. Percentage of participants still taking the initially assigned treatment at 48 months in the Treatment of Mild Hypertension Study (TOMHS). Amlodipine was the best tolerated drug treatment (82.5% for amlodipine and 67.5–77.8% for the other drug treatments; p Å 0.1 for differences). Data adapted from Hypertension.64

16

THE AMERICAN JOURNAL OF CARDIOLOGYT

/ 2z0c 7203 Mp

16

VOL. 79 (10A)

Thursday May 15 03:12 PM

MAY 22, 1997

EL–AJC-S (suppl 7200 ’97)

7203

findings were attributable to differences in sample size. In the 1974 study, there were 150 women with breast cancer, of whom only 11 had used reserpine; but, in the 1984 study, there were 1,881 women with breast cancer of whom 65 had used reserpine. Thus, even the apparently markedly increased relative cancer risk of ú3, coming from 3 centers in 3 different countries and involving very respected investigators, was subsequently discounted. These lessons from the past highlight the danger of focusing on small numbers of cases, such as the 11 fatal cancers with calcium antagonists in the recent Pahor study.5

SUMMARY AND FUTURE PERSPECTIVES What, if anything, should physicians make of these apparently disparate hazards in a group of drugs that they prescribe so widely? I, and many of my colleagues, still believe that many patients do benefit from the judicious use of calcium antagonists. I believe that there is no reason to suspend their use on the basis of current evidence. The only evidence on which that action would be justified is a randomized, controlled trial. Ironically, if the use of calcium antagonists were suspended, the large randomized, controlled trials that we have espoused and that are now under way might flounder through lack of interest or concern. In this case, the current debate on many of the adverse effects, particularly cancer, would never be resolved. As I have noted in a recent editorial, the evidence that exists for the safety of ‘‘third-generation’’ longer-acting calcium antagonists currently under investigation in these large outcome studies gives us no real cause for concern. WHY SHOULD WE CARE WHETHER CALCIUM ANTAGONISTS CONTINUE TO SERVE AS FIRST-LINE ANTIHYPERTENSIVE AGENTS? A final consideration that has received relatively little attention to date is whether it is inconsequential if calcium antagonists remain a therapeutic option for antihypertensive treatment. Clearly, an extensive antihypertensive armamentarium with many efficacious agents is now available. If concerns have arisen regarding the safety of calcium antagonists, why not merely discard them? I believe that such a simplistic approach is patently inappropriate, and there are compelling reasons for retaining calcium antagonists as part of our antihypertensive armamentarium. First, we should remember that, despite a massive publicity effort and promulgation of management guidelines by many national advisory bodies, we have not achieved great success in our attempts to control blood pressure.63,64 The recent National Health and Nutrition Evaluation (NHANES III) survey63 has clearly demonstrated that during the 20-year period from 1971 to 1991, patient awareness of hypertension increased markedly in the United States, from 51% to 84% of the population. Concomitantly, the percentage of patients being treated for hypertension has increased from 36% to 83%.63 Nevertheless, pa-

tients achieving blood pressure control to below 140/ 90 mm Hg currently remain at only 24%.63 A recent survey indicates that the experience in Europe is similar. The extent of the problem is revealed in data from the Cardiomonitor study, an independent survey of approximately 1,500 physicians performed biannually in 10 European countries and the United States by an independent healthcare market research company.65 Patients were treated with diuretics, calcium antagonists, b blockers, and angiotensin-converting enzyme (ACE) inhibitors (singly and in combination). The survey, conducted between April and June 1992, demonstrated that of the 11,613 hypertensive patients receiving treatment included in the survey, only 37% had reached the target diastolic blood pressure set by the doctor. Although the reasons for this striking failure are complex and multifactorial, one reasonable explanation must relate to compliance and the need for patients to remain on prescribed therapy. The Treatment of Mild Hypertension Study (TOMHS)66 disclosed that the percentage of patients remaining on initial therapy at 4 years was highest in the calcium antagonist group, who were receiving amlodipine (Figure 4). Elliot67 has also reported on drug discontinuation in a large group of hypertensive patients followed for 7 years. He observed that the relative likelihood of discontinuing hypertensive therapy was greatest with b blockers and least with calcium antagonists.

CONCLUSIONS In summary, based on a compelling body of evidence, it seems likely that reports of an increase in myocardial infarction in hypertensive patients attributed to calcium antagonists may in reality be restricted to the more rapidly absorbed, short-acting calcium antagonists. The period of observation encompassed by most of the earlier observational studies precedes the introduction of the newer, slow-release formulations of both dihydropyridines and benzothiazepines. Short-acting calcium antagonists can induce recurrent sympathetic neurohormonal activation and evoke a reactive cardioacceleration that may be detrimental to patients with underlying myocardial disease. Recent studies suggest that such changes are not encountered with slow-release formulations, which are capable of achieving a more gradual and sustained antihypertensive effect. Moreover, studies are emerging that suggest the slower onset and longer duration of action of the calcium antagonist formulations used in clinical practice today are more efficacious and produce fewer side effects than the older formulations. Large prospective, randomized outcome studies using such slow-release agents are urgently needed to validate these formulations in patients with chronic stable angina, Prinzmetal’s angina, syndrome X, and hypertension. Hopefully, such studies will allay the apprehension engendered by the recent retrospective observational studies.

A SYMPOSIUM: CALCIUM ANTAGONISTS IN HYPERTENSION

/ 2z0c 7203 Mp

17

Thursday May 15 03:12 PM

EL–AJC-S (suppl 7200 ’97)

7203

17

Fortunately, a number of such prospective studies have recently been initiated, including the Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack (ALLHAT)68 and International Nifedipine Study, Intervention as a Goal in Hypertension Treatment (INSIGHT)69 studies. The ALLHAT study is an $85 million, 9-year study supported by the National Heart, Lung, and Blood Institute that will examine the ability of several antihypertensive agents, including the long-acting calcium antagonist amlodipine, to reduce coronary heart disease among hypertensive patients.67 INSIGHT is a randomized, 3-year study of 6,600 high-risk hypertensive patients in Europe, comparing the effects of nifedipine GITS with diuretic therapy on cardiovascular and cerebrovascular morbidity and mortality.69 A third study is Swedish Trial in Old Patients (STOP) Hypertension 2.70 The study is being conducted in Sweden to evaluate b blockers and/or diuretics against ACE inhibitors (enalapril and lisinopril) and calcium antagonists (isradipine and felodipine) in 6,600 elderly hypertensive patients. The primary aim is to assess the effects of the different interventions on cardiovascular mortality; the primary endpoints are stroke, myocardial infarction, and other cardiovascular events. While we await the completion of such studies, it is reasonable and prudent to continue using calcium antagonists as antihypertensive agents. History is replete with examples of instances where concerns regarding adverse events were raised that were ultimately not substantiated. One example is the fears that were engendered regarding excess mortality resulting from the use of diuretics in the management of patients with diabetes mellitus.71 Subsequently, these allegations were refuted. Indeed, the most recent consensus guidelines for managing hypertension associated with diabetes mellitus released by the National High Blood Pressure Education Program now include a recommendation for the use of thiazide diuretics.72 Despite the absence of prospective, randomized outcome studies of slow-release formulations, a prudent interim approach would be to restrict the use of calcium antagonists to the newer slow-release formulations that, by virtue of their ability to attain more gradual and sustained plasma levels, do not evoke reactive sympathetic activation. Concomitantly, such formulations should promote increased patient compliance and thereby favorably influence hypertension-related morbidity and mortality. Acknowledgment: The author thanks Elsa V. Reina for her secretarial assistance.

1. Epstein M. Calcium antagonists in the management of hypertension. In: Ep-

stein M, ed. Calcium Antagonists in Clinical Medicine. Philadelphia: Hanley and Belfus, 1992:213–230. 2. Frishman WH. Current status of calcium channel blockers. Curr Prob Cardiol 1994;19:637–688. 3. Epstein M. Calcium antagonists and renal protection: current status and future perspectives. Arch Intern Med 1992;152:1573–1584.

18

THE AMERICAN JOURNAL OF CARDIOLOGYT

/ 2z0c 7203 Mp

18

4. Messerli FH, McLoughlin M. Cardiac effects of calcium antagonists in hypertension. In: Epstein M, ed. Calcium Antagonists in Clinical Medicine. Philadelphia: Hanley and Belfus, 1992:137–150. 5. Pahor M, Guralnik JM, Salive ME, Corti MC, Carbonin P, Havlik RJ. Do calcium channel blockers increase the risk of cancer? Am J Hypertens 1996;9:695–699. 6. Pahor M, Guralnik JM, Ferrucci L, Corti MC, Salive ME, Cerhan JR. Calcium channel blockade and incidence of cancer in aged populations. Lancet 1996;348:493–497. 7. Held PH, Yusuf S, Furberg CD. Calcium channel blockers in acute myocardial infarction and unstable angina: an overview of the result from the randomized trials. BMJ 1989;299:1187–1192. 8. Furberg CD, Psaty BM. Calcium antagonists: not appropriate as first line antihypertensive agents. Am J Hypertens 1996;9:122–125. 9. Angell M. Science on Trial. New York: WW Norton and Company, 1996. 10. Buring JE, Glynn RJ, Hennekens CH. Calcium channel blockers and myocardial infarction: a hypothesis formulated but not yet tested. JAMA 1995;274:654–655. 11. Opie LH, Messerli FH. Nifedipine and mortality. Grave defects in the dossier. Circulation 1995;92:1068–1073. 12. Psaty BM, Heckbert SR, Koepsell TD, Siscovick DS, Ragunathan TE, Weiss NS, Rosendaal FR, Lemaitre RN, Smith NL, Wahl PW, et al. The risk of myocardial infarction associated with antihypertensive drug therapies. JAMA 1995;274:620–625. 13. Epstein M. Calcium antagonists should continue to be used for first-line treatment of hypertension. Arch Intern Med 1995;155:2150–2156. 14. Epstein M. Calcium antagonists: still appropriate as first-line antihypertensive agents. Am J Hypertens 1996;9:110–121. 15. Morris AD, Meredith PA, Reid JL. Pharmacokinetics of calcium antagonists: implications for therapy. In: Epstein M, ed. Calcium Antagonists in Clinical Medicine. Philadelphia: Hanley and Belfus, 1992:49–67. 16. Triggle DJ. Biochemical and pharmacologic differences among calcium channel antagonists: clinical implications. In: Epstein M, ed. Calcium Antagonists in Clinical Medicine. Philadelphia: Hanley and Belfus, 1992:1–27. 17. Ruzicka M, Leenen FHH. Relevance of intermittent increases in sympathetic activity for adverse outcome on short-acting calcium antagonists. In: Laragh JH, Brenner BM, eds. Hypertension: Pathophysiology, Diagnosis, and Management, 2nd Ed. New York: Raven Press, 1995:2815–2825. 18. Kleinbloesem CH, van Brummelen P, Danlo¨f M, Faber H, Urquhart J, Breimer DD. Rate of increase in the plasma concentration of nifedipine as a major determinant of its hemodynamic effects in humans. Clin Pharmacol Ther 1987;41:26–30. 19. Myers MG, Raemsch KD. Comparative pharmacokinetics and antihypertensive effects of the nifedipine tablet and capsule. J Cardiovasc Pharmacol 1987;10(suppl 10):S76–S78. 20. Swanson DR, Barclay BL, Wong PSL, Theeuwes F. Nifedipine gastrointestinal therapeutic system. Am J Med 1987;83(suppl 6B):3–9. 21. Chung M, Reitberg DP, Gafney M, Singleton W. Clinical pharmacokinetics of nifedipine gastrointestinal therapeutic system. A controlled release formulation of nifedipine. Am J Med 1987;83(suppl 6B):10–14. 22. Stokes GS, Shenfield GM, Johnston HJ, Marwood JF, Dunagan F, Boutagy J. Timing of blood pressure measurements in determining anomalies in duration of effect of an antihypertensive drug: assessment of isradipine. J Cardiovasc Pharmacol 1990;15(suppl 1):S65–S69. 23. Frohlich ED, McLoughlin MJ, Losem CJ, Ketelhut R, Messerli FH. Hemodynamic comparison of two nifedipine formulations in patients with essential hypertension. Am J Cardiol 1991;68:1346–1350. 24. Rousseau MF, Melin J, Benedict CR, Ahn S, Raphael D, Bornemann M, Pouler H. Effects of nisoldipine therapy on myocardial perfusion and neurohormonal status in patients with severe ischaemic left ventricular dysfunction. Eur Heart J 1994;15:957–964. 25. Bond G, Del Palu C, Hansson L, Magnani B, Giuseppe M, Neiss A, Rahn KH, Reid JL, Rodicio JL, Safar M, Zanchetti A. ELSA: European Lacidipine Study on Atherosclerosis. (Abstr.) J Hypertens 1993;11(suppl 5):S405. 26. Borhani NO, Mercuri M, Borhani PA, Buckalew VM, Canossa-Terris M, Carr AA, Kappagoda T, Rocco MV, Schnaper HW, Sowers JR, Bond G. Final outcome results of the multicenter isradipine diuretic atherosclerosis study (MIDAS). JAMA 1996:276:785–791. 27. Parati G, Pomidossi G, Albini F, Malaspina D, Mancia G. Relationship of 24-hour blood pressure mean and variability to severity of target organ damage in hypertension. J Hypertens 1987;5:93–98. 28. Frattola A, Parati G, Cuspidi C, Albini F, Mancia G. Prognostic value of 24-hour blood pressure variability. J Hypertens 1993,11:1133–1137. 29. Billman GE. The antiarrhythmic effects of calcium antagonists. In: Epstein M, ed. Calcium Antagonists in Clinical Medicine. Philadelphia; Hanley and Belfus, 1992:183–212. 30. Chakko S, Bassett AL. Are calcium antagonists proarrhythmic? In: Epstein M, ed. Calcium Antagonists in Clinical Medicine, 2nd Ed. Philadelphia; Hanley and Belfus, 1997. 31. Roine RO, Kaste M, Kinnunen A, Nikki P, Sarna S, Kajaste S. Nimodipine after resuscitation from out of hospital ventricular fibrillation. JAMA 1990;264:3171–3177. 32. Waters D. Proischemic complications of dihydropyridine calcium channel blockers. Circulation 1991;84:2598–2600.

VOL. 79 (10A)

Thursday May 15 03:12 PM

MAY 22, 1997

EL–AJC-S (suppl 7200 ’97)

7203

33. Egstrup K, Andersen PE. Transient myocardial ischemia during nifedipine therapy in stable angina pectoris, and its relation to coronary collateral flow and comparison with metoprolol. Am J Cardiol 1993;71:177–183. 34. Schulz W, Jost S, Kober G, Kaltenbach M. Relation of antianginal efficacy of nifedipine to degree of coronary arterial narrowing and to presence of coronary collateral vessels. Am J Cardiol 1985;55:26–32. 35. Kloner RA. Nifedipine in ischemic heart disease. Circulation 1995;92: 1074–1078. 36. Malacoff RF, Lorell BH, Mudge GH, Holman BL, Idoine J, Bifolck L, Cohn PF. Beneficial effects of nifedipine on regional myocardial blood flow in patients with coronary artery disease. Circulation 1982;65(suppl I):I32 – I-37. 37. Kugiyama K, Yasue H, Horio Y, Morikami Y, Fujii H, Koga Y, Kojima A, Takahashi M. Effects of propranolol and nifedipine on exercise thallium-201 myocardial scintigraphy with quantitative rotational tomography. Circulation 1986;2:374–380. 38. Bonaduce D, Muto P, Morgano G, Cononico V, Breglio R, Salvatore M, Condorelli M. Effect of nifedipine on dipyridamole thallium-201 myocardial scintigraphy. Clin Cardiol 1986;9:285–288. 39. Hayward R, Hunter GJS, Swanton H. Effects of nifedipine on coronary perfusion: recent assessment by parametric digital subtraction. Eur Heart J 1987;8(suppl K):1–7. 40. Nesto RW, White HD, Ganz P, Koslowski J, Wynne J, Holman BL, Antman E. Addition of nifedipine to maximal beta-blocker nitrate therapy: effects on exercise capacity and global left ventricular performance at rest and during exercise. Am J Cardiol 1985;55:3E–8E. 41. Parmley WW, Nesto RW, Singh BN, Deanfield J, Gottlieb SO, and the N-CAP Study Group. Attenuation of the circadian patterns of myocardial ischemia with nifedipine GITS in patients with chronic stable angina. J Am Coll Cardiol 1992;19:1380–1389. 42. Packer M, O’Connor CM, Ghali JK, Pressler ML, Carson PE, et al. for the Prospective Randomized Amlodipine Survival Evaluation Study Group (PRAISE). Effect of amlodipine on morbidity and mortality in severe chronic heart failure. N Engl J Med 1996;335:1107–1114. 43. Gong L, Zhang W, Zhu Y, Zhu J, 11 collaborating centres in the Shanghai area, Kong D, Page V, Ghadirian P, LeLorier J, Hamet P. Shanghai trial of nifedipine in the elderly (STONE). J Hypertens 1996;14:1237–1245. 44. Cohn JN, Ziesche SM, Loss LE, Anderson GF, and the VHeFT Study Group. Effect of felodipine on short term exercise and neurohormone and long term mortality in heart failure: results of VHeFT III. Circulation 1995;92:I– 143. 45. DEFIANT-II Research Group, Poole-Wilson PA. Doppler flow and echocardiography in functional cardiac insufficiency: assessment of nisoldipine therapy. Eur Heart J 1997;18:31–40. 46. Kloner RA, Vetrovec G. Long-acting calcium channel blockers nifedipine GITS and amlodipine in hypertension. Rate of myocardial infarction. (Abstr.) J Am Coll Cardiol 1996;27(suppl A):177A. 47. Abascal VM, Larson MG, Evans JC, Poll KA, Levy D, and the Framingham Study Group. Calcium channel blocker use is not associated with increased mortality. (Abstr.) Circulation 1996;94 (suppl 1):3383. 48. Connor J, Sawczuk IS, Benson MC, Tomashefsky P, O’Toole KM, Olsson CA, Buttyan R. Calcium channel antagonists delay regression of androgen dependent tissues and suppress gene activity associated with cell death. Prostate 1988;13:119–130. 49. Ray SD, Kamendulis LM, Gurule MW, Yorkin RD, Corcoran GB. Ca2/ antagonists inhibit DNA fragmentation and toxic cell death induced by acetaminophen. FASEB J 1993;7:453–463. 50. Carson DA, Ribeiro JM. Apoptosis and disease. Lancet 1993;341:1251– 1254.

51. Trump BF, Berezesky IK. Calcium mediated cell injury and cell death. FASEB J 1995;9:219–228. 52. Martin SJ, Green DR. Apoptosis and cancer: the failure of controls on cell death and cell survival. Crit Rev Oncol Hematol 1995;18:137–153. 53. Whitfield JF. Calcium signals and cancer. Crit Rev Oncol 1992;3:55–90. 54. Garland C, Shekelle RB, Barrett Connor E, Criqui MH, Rossof AH, Paul O. Dietary vitamin D and calcium and risk of colorectal cancer: a 19-year prospective study in men. Lancet 1985;i:307–309. 55. Dargie HJ. Calcium channel blockers and the clinician. Lancet 1996;348: 488–489. 56. Zacharski LR, Moritz TE, Haakenson CM, O’Donnell JF, Ballard HS, Johnson GJ, Ringenberg QS, Schilsky RL, Spaulding MB, Tornyosk K, Williams CC. Chronic calcium antagonist use in carcinoma of the lung and colon: a retrospective cohort observational study. Cancer Invest 1990;8:451 – 459. 57. MRC Working Party. Medical Research Council trial of treatment of hypertension in older adults: principal results. BMJ 1992;304:405–412. 58. Hole DJ, Hawthorne VM, McGhee SM, Robertson JW, Gillis CR, Wapshaw JA, Lever AF. Incidence of mortality from cancer in hypertensive patients. BMJ 1993;306:609–611. 59. Taubes G. Epidemiology faces its limits. Science 1995;269:164–169. 60. Anonymous. Rauwolfia derivatives and cancer. Lancet 1974;ii:701–702. 61. Boston Group, Program Boston Collaborative Drug Surveillance. Reserpine and breast cancer. Lancet 1974;ii:669–671. 62. Shapiro S, Parsells JL, Rosenberg L, Kaufman DW, Stolley PD, Schottenfeld D. Risk of breast cancer in relation to the use of rauwolfia alkaloids. Eur J Clin Pharmacol 1984;26:143–146. 63. Burt VL, Whelton P, Roccella EJ, Brown C, Cutler JA, Higgins M, Horan MJ, Labarthe D. Prevalence of hypertension in the US adult population. Results from the Third National Health and Nutrition Examination Survey, 1988–1991. Hypertension 1995;25:305–313. 64. Frohlich ED. There is good news and not so good news. (Editorial). Hypertension 1995;25:303–304. 65. Taylor Nelson Healthcare. Cardiomonitor 1992. In: Hosie J, Wiklund I. Managing hypertension in general practice: can we do better? J Hum Hypertens 1995;9:S15–S18. 66. Neaton JD, Grimm RH, Jr, Prineas RJ, Stamler J, Grandits GA, Elmer PJ, Cutler JA, Flack JM, Schoenberger JA, McDonald R, Lewis CE, Liebson PR. Treatment of Mild Hypertension Study. JAMA 1993;270:713–724. 67. Elliot WJ. Drug therapy and ‘‘dropouts’’ in a tertiary hypertension clinic. (Abstr.) Am J Hypertens 1994;7:26A. 68. Cushman WC for the ALLHAT Research Group. The antihypertensive and lipid lowering treatment to prevent heart attack trial (ALLHAT). Design and initial characteristics. (Abstr.) Hypertension 1995;25:1356. 69. Brown MJ, Castaigne A, Ruilope LM, Mancia G, Rosenthal T, de Leeuw PW, Ebner F. INSIGHT: International nifedipine GITS study intervention as a goal in hypertension treatment. J Hum Hypertens 1996;10 (suppl 3):S157– S160. 70. Dahlo¨f B, Hansson L, Lindholm LH, Schersten B, Webster PO, Ekbom T, Hedner T, De Faire U. STOP Hypertension 2: a prospective intervention trial of ‘‘newer’’ versus ‘‘older’’ treatment alternatives in old patients with hypertension. Blood Press 1993;2:136–141. 71. Warram JH, Laffel LMB, Valsania P, Christlieb AR, Krolewski AS. Excess mortality associated with diuretic therapy in diabetes mellitus. Arch Intern Med 1991;151:1350–1356. 72. The National High Blood Pressure Education Program Working Group. National high blood pressure education program working group report on hypertension in diabetes. Hypertension 1994;23:145–158.

A SYMPOSIUM: CALCIUM ANTAGONISTS IN HYPERTENSION

/ 2z0c 7203 Mp

19

Thursday May 15 03:12 PM

EL–AJC-S (suppl 7200 ’97)

7203

19