Mean change in 24-hour diastolic blood pressure (DBP) from weeks 0 to 8. CC indicates coat core; GITS, gastrointestinal therapeutic system.
Mean office visit trough systolic (SBP) and diastolic blood pressure (DBP). CC indicates coat core; GITS, gastrointestinal therapeutic system.
Customize your JAMA Network experience by selecting one or more topics from the list below.
Hall WD, Reed JW, Flack JM, Yunis C, Preisser J, and the ISHIB Investigators Group. Comparison of the Efficacy of Dihydropyridine Calcium Channel Blockers in African American Patients With Hypertension. Arch Intern Med. 1998;158(18):2029–2034. doi:10.1001/archinte.158.18.2029
Hypertension is a prevalent disease among African Americans, and successful treatment rates are low. Since calcium channel blockers are well-tolerated and efficacious in African Americans, we undertook this study to compare the efficacy, safety, and tolerability of 3 commonly prescribed calcium channel blockers: amlodipine besylate (Norvasc), nifedipine coat core (CC) (Adalat CC), and nifedipine gastrointestinal therapeutic system (GITS) (Procardia XL).
One hundred ninety-two hypertensive patients across 10 study centers were randomly assigned to double-blind monotherapy with amlodipine besylate (5 mg/d), nifedipine CC (30 mg/d), or nifedipine GITS (30 mg/d) for 8 weeks. Patients not achieving therapeutic response after 4 weeks had their dose doubled for the next 4 weeks. The primary end point was a comparison of the average reduction (week 8 minus baseline) in 24-hour ambulatory diastolic blood pressure (DBP). Secondary end points included a comparison of average 24-hour ambulatory systolic blood pressure (SBP), office SBP or DBP reduction, responder rates, safety, and tolerability.
One hundred sixty-three patients were evaluable for efficacy after 8 weeks. There was no significant difference in the average 24-hour ambulatory DBP (−8.5, −9.0, and −6.1 mm Hg, respectively) or SBP (−14.3, −15.7, and −11.8 mm Hg, respectively) reduction. Average office SBP and DBP were reduced to a comparable degree (19-22 mm Hg [P=.50] and 12-14 mm Hg [P=.51], respectively). Responder rates (DBP <90 or reduced by ≥10 mm Hg) were similar (P=.38). Discontinuation rates and adverse event frequency were distributed similarly across the 3 treatment groups.
The efficacy, safety, and tolerability of the 3 dihydropyridine calcium channel blockers are equivalent in African Americans with stages 1 and 2 hypertension.
IN THE US adult population, hypertension is a common condition, affecting 43 million people.1,2 The age-adjusted prevalence is particularly high (ie, 32.4%) in African Americans. Awareness and treatment rates have improved during the last 2 decades, but only 27% to 29% of all hypertensive patients have attained systolic blood pressure (SBP) levels below 140 mm Hg and diastolic BP (DBP) levels below 90 mm Hg.
Long-acting calcium channel blockers (CCBs) are among the drugs recommended by the Joint National Committee for the Treatment of Hypertension.2 A recent Veterans Affairs study showed that these drugs are effective when prescribed as monotherapy for white and African American patients with hypertension.3 Among hypertensive African Americans, monotherapy with CCBs appears to be superior to that with β-blockers or angiotensin-converting enzyme inhibitors in achieving BP reduction.4-7
Unlike most classes of antihypertensives, the CCBs are relatively heterogeneous compounds represented by the phenylalkylamines (eg, verapamil hydrochloride), benzothiazepines (eg, diltiazem hydrochloride), dihydropyridines (eg, amlodipine besylate, felodipine, nicardipine hydrochloride, nifedipine, and nisoldipine).8 Comparative studies of the same or different subtypes of CCBs are few, especially with regard to large numbers of hypertensive African Americans.
The purpose of our randomized, double-blind study was to compare directly the BP-lowering efficacy of the following 3 long-acting dihydropyridine CCBs: amlodipine besylate (Norvasc), nifedipine coat core (CC) (Adalat CC), and nifedipine gastrointestinal therapeutic system (GITS) (Procardia XL) after 8 weeks of therapy in hypertensive African Americas.
The design of the study was a randomized, double-blind, parallel-group phase 4 clinical trial comparison of 3 dihydropyridine CCBs.
African American men and women (age range, 18-75 years) with average (mean of 2 standardized measurements) DBP from 95 to 110 mm Hg and SBP from 140 to 179 mm Hg were eligible for randomization. Eligibility was determined after receiving no antihypertensives for at least 3 weeks, during which time compliance to placebo was at least 80% by pill counts.
Major study exclusion criteria included a serum creatinine level above 212 µmol/L (2.4 mg/dL), 4+ proteinuria on routine urinalysis, uncontrolled diabetes mellitus (fasting glucose level, >11 mmol/L [198 mg/dL]), aspartate aminotransferase levels of greater than 1.5 times upper normal limit, 1+ or more pretibial edema on results of clinical examination, known secondary causes of hypertension, comorbid cardiovascular conditions (previous stroke, coronary artery disease, arrhythmias, or heart failure), pregnancy or planned pregnancy during the study, average alcohol intake of 3 or more drinks per day, and (because of the ambulatory BP monitoring) working the night shift or having a mid–upper arm circumference of greater than 37.5 cm such that technically valid computerized BP recordings were less likely.
All centers had to receive institutional review board approval before receipt of antihypertensives, and all patients gave written informed consent before enrollment.
All protocol visits occurred in the mornings, 22 to 26 hours after the last dose of blinded medication. Following a 3-week, single-blind placebo lead-in period (during which participants were seen at least weekly), monotherapy was begun with amlodipine besylate, 5 mg/d; nifedipine CC, 30 mg/d; or nifedipine GITS, 30 mg/d, with follow-up visits at 2-week intervals. If DBP remained at least 90 mm Hg after 4 weeks, the dosage was titrated up to 10 mg/d of amlodipine besylate or 60 mg/d of either form of nifedipine. Body weight was measured (to the nearest 0.2 kg) in light clothing, without shoes.
Participant safety was improved by enrollment of only patients with stage 1 or 2 hypertension, exclusion of patients with underlying cardiovascular disease, frequent clinic visits with monitoring for adverse events, and discontinuation of any patients whose SBP or DBP was confirmed to be at least 180 or 110 mm Hg, respectively.
Blood pressure was measured in the seated position, using the same arm (right). Observers were pretrained using an explicit measurement protocol that followed the most recent recommendations of the American Heart Association.9 Cuff sizes were selected according to midarm circumference (adult regular for 22.6-30.0 cm, large arm for 30.1-37.5 cm, and pediatric for 16.0-22.5 cm). The clinic visit BP was defined as the average of duplicate readings separated by 2 minutes. Heart rate values represent 30-second counts of the pulse rate, multiplied by 2.
Twenty-four–hour ambulatory BP monitoring was performed the day preceding randomization and at the end of the 8-week treatment using the ambulatory BP monitor (Spacelabs model 90207, Medifacts Ltd, Rockville, Md) with measurements every 15 minutes during the day and every 20 minutes at night. A technically valid baseline recording with a minimum of 24 hours following the last placebo dose was a prerequisite to randomization.
A standardized baseline questionnaire was administered to ascertain age, race, educational status (total years), smoking (yes or no), history of antihypertensive use, and history of comorbid conditions (eg, diabetes). Ankle edema was evaluated at each protocol visit using results of clinical examination and self-reported history, both of which were used in the analyses.
Adverse effects were elicited at randomization and each follow-up protocol visit. This included a designated list of 21 possible symptoms potentially associated with hypertension or antihypertensives and objective clinical evaluation by the investigators.
The primary end point of the study was the change in the average 24-hour ambulatory DBP after 8 weeks of monotherapy. The hypothesis was that the 3 treatment regimens would produce equivalent lowering of 24-hour ambulatory DBP. We designed the study to detect a 3-mm Hg or greater difference in 24-hour ambulatory DBP change with 90% power (ie, 10% chance of a type II error) in pairwise comparisons between nifedipine CC and amlodipine or nifedipine GITS. Sample size estimates were based on an intraperson SD of ambulatory DBP of 5.5 mm Hg and a 1-tail type I error of 5%. A total of 174 evaluable participants was necessary to satisfy the statistical assumptions. However, a total of 210 participants were to be randomized (ie, 70 in each group), given the a priori assumption that 17% of participants would drop out or be nonevaluable.
Secondary end points included the change in average 24-hour ambulatory SBP, office visit SBP and DBP reduction, responder rate (DBP <90 or reduced by ≥10 mm Hg), and safety and tolerability of CCBs after 8 weeks of monotherapy.
Randomization was accomplished by using envelopes numbered consecutively and containing the randomization assignment for each patient. For each clinical site, randomization was stratified according to previous antihypertensive use (yes or no), such that patients were assigned predetermined randomization numbers from opposite ends of their clinical site's random code. Stratified randomization was performed to balance treatment assignment depending on whether participants had used antihypertensives previously, a variable that could confound the relations of treatment exposure and BP response. Logic checks were developed for data entry fields, and a 5% random sample was rekeyed for quality control purposes.
An analysis of covariance with effects for clinic and previous medications was used to assess equivalence. Responder rates were compared using Mantel-Haenszel χ2 statistics, stratified by clinic. For continuous variables (eg, height, weight, heart rate), a multivariate analysis of variance was used (with effects for treatment, clinic, and previous medications) as the primary method of analysis.
There were 192 patients randomized by 10 sites; 13 had average BP out of range at the randomization visit and 7 dropped out before any follow-up visits. Of the 172 evaluable patients, 163 completed the trial. The remaining 9 patients were discontinued because of adverse events (n=5; 1 patient from the amlodipine group and 2 patients from each nifedipine group), being unavailable for follow-up (n=1), or other reasons (n=3). In addition, 3 patients (1 patient from each treatment group) who completed all office visits did not provide ambulatory BP data at follow-up. The 160 patients used for the analysis of the primary end point provided 88% power to detect a between-group (pairwise contrast) change of at least 3 mm Hg.
Table 1 provides an outline of selected baseline characteristics of the randomized participants, according to their treatment assignment. Overall, the 3 groups were comparable except for a significantly higher baseline heart rate (P=.03) in the nifedipine CC group compared with both other groups. Previous use of antihypertensives (62.7% to 66.7%) was similar across all 3 treatment groups.
The average reductions in 24-hour ambulatory BP (SBP/DBP) at the 8-week visit were −14.3/−8.5 for amlodipine, −15.7/−9.0 for nifedipine CC, and −11.8/−6.1 for nifedipine GITS (Table 2). Figure 1 depicts the mean change in 24-hour DBP, commencing 1 hour after the drug dose, about 9 AM (time 0). The graph compares the 3 drugs and shows that, on average, amlodipine and nifedipine CC reduced DBP similarly in 24 hours.
Table 2 also provides data on the 8-week changes in ambulatory and office visit BP, according to the treatment assignment for 163 patients. The mean 8-week reductions in office visit DBP were −13.6, −13.1, and −12.2 mm Hg in the amlodipine, nifedipine CC, and nifedipine GITS groups, respectively (P=.51). The mean reductions in SBP were −21.8, −19.1, and −19.6 mm Hg, respectively. These mean changes did not differ significantly among the 3 treatment groups (P=.50). Figure 2 shows the time course of reduction in office visit SBP (top) and DBP (bottom) during the 8 weeks of monotherapy.
Responder analysis revealed that there was no significant difference among treatment groups in responder rates, defined as DBP of less than 90 mm Hg or reduced by at least 10 mm Hg. The overall responder rates for amlodipine, nifedipine CC, and nifedipine GITS after 8 weeks of antihypertensive therapy were 79.6%, 64.9%, and 74.5%, respectively (P=.38). Ninety-five patients (56.2%) were titrated to the second-level dosage (60.0%, 57.9%, and 50.9% for the amlodipine, nifedipine CC, and nifedipine GITS groups, respectively). At the end of the 8 weeks, the average prescribed daily doses for amlodipine, nifedipine CC, and nifedipine GITS were 8.0, 47.3, and 45.6 mg, respectively.
The mean 8-week changes in office visit heart rate were +2.1, −1.4, and +1.2 in the amlodipine, nifedipine CC, and nifedipine GITS groups, respectively. These changes were not significantly different from the baseline heart rate within or among treatment groups (P=.70).
The average 8-week changes in body weight were −0.7, +0.2, and +0.7 kg in the amlodipine, nifedipine CC, and nifedipine GITS groups, respectively. These modest changes also were not statistically significantly different (P=.70 among groups).
In the amlodipine group, 6 adverse events (1 serious; 1 drug-related) were reported by 6 of 56 patients. The serious event was a transient ischemic attack, not attributed to the study drug. In the nifedipine CC group, 5 adverse events (1 probably drug-related; none serious) were reported by 3 of 57 patients. In the nifedipine GITS group, 3 adverse events (2 probably drug-related; 1 serious) were reported by 2 of 59 patients. The serious event was a supraventricular tachycardia, which reverted to normal sinus rhythm following cardioversion. No deaths were reported in any treatment group. A description of the adverse events experienced by patients in each of these study groups is summarized in Table 3. When compared with baseline symptoms, 10 of the 14 adverse events were not related to study drug. Self-reported adverse effects judged to be related to active treatment were uniformly distributed across each treatment group (P>.69).
Our randomized, double-blind study is the first of its kind to provide a direct comparison of 3 CCBs in a large number of African Americans. The primary end point, change in the average ambulatory DBP, did not differ for amlodipine, nifedipine CC, or nifedipine GITS (−8.5, −9.0, and −6.1, respectively (P=.10). The results indicate that any differences in the 8-week efficacy of 3 broadly used CCBs in African Americans are relatively small at the prescribed doses. Each produced an overall reduction of 19 to 22 mm Hg in office visit SBP and 12 to 14 mm Hg in office visit DBP. Any overall differences in hemodynamic responses were small, and the 3 drugs were equally effective in reducing BP.
Safety analysis revealed no substantial differences in adverse events among amlodipine, nifedipine CC, and nifedipine GITS. A similar percentage of patients exhibited adverse events as defined by a predesignated list of 21 possible related symptoms. Nevertheless, the absence of a placebo control group makes interpretation of the absolute rates difficult, given that many of these symptoms are also related to hypertension per se. Furthermore, the frequency, type, and distribution of adverse events were relatively low and similar among treatment groups, indicating that these drugs were well tolerated in African Americans. The occurrence of edema on clinical examination was uncommon in the doses used in this particular population, although it was self-reported by 14.3%, 10.5%, and 8.5% of participants receiving amlodipine, nifedipine CC, and nifedipine GITS, respectively. However, patients were seen on morning visits. Moreover, potential study participants were excluded if they had 1+ or more edema at baseline.
The results support the hypothesis that these 3 dihydropyridine CCBs can be used interchangeably without regard for potential differences in efficacy or safety in African American patients. This confirms the clinical results demonstrated in previous studies conducted in mixed patient populations.10-13 Thus, the pattern of use of each of these drugs will be determined by which drug is associated with the best compliance and the least overall cost.
Comparable efficacy and safety of slow-release nifedipine and amlodipine have been reported, although the incidence of edema was greater in the amlodipine-treated group.14,15 However, other studies have suggested that amlodipine besylate (5 mg/d) was superior to slow-release nifedipine (20 mg twice daily) in therapeutic response, compliance, and incidence of adverse events.16,17 Two other studies indicated that nifedipine CC and nifedipine GITS lowered BP to a similar degree.18,19 Previous observations have demonstrated that patients covered by Medicaid insurance could switch from nifedipine GITS to nifedipine CC with no apparent effects on BP control, incidence of adverse drug reactions, or nonprescription health care costs.19
Limitations of our study include the relatively short duration (ie, 8 weeks), the inclusion of patients with stages 1 and 2 hypertension only, and a limited power to assess differences in categorical events. Our results, however, support and extend the results of smaller studies that compared nifedipine CC with amlodipine in predominantly white populations.
Accepted for publication May 12, 1998.
This work was funded by a grant from Bayer Pharmaceuticals, West Haven, Conn. Ambulatory BP-monitoring equipment and training were provided by Medifacts, Ltd, Rockville, Md.
Presented at the 71st annual session of the American Heart Association, Orlando, Fla, November 9, 1997.
This study was coordinated by the International Society on Hypertension in Blacks (ISHIB) by Kermit G. Payne (project coordinator), Christopher T. Fitzpatrick (associate project coordinator), and Carrie T. Hamilton (assistant project coordinator). The data coordinating center included Mizpah Johnson (assistant project manager), Kathy Lane (programmer), John M. Flack, MD, MPH (director), and Carla Yunis, MD, MPH (project manager). Central laboratory was SmithKline Beecham, Van Nuys, Calif. Senior collaborators from Bayer Pharmaceuticals were Tim M. Shannon, Joe Carafano, Deborah Deakin, Steven Ripa, Lawrence Schwartz, and Kenneth Pomerantz. The ISHIB investigators and sites are as follows (asterisk indicates principal investigator): DeAnna Cheek, MD,* Vernette Brown, RN (Medical University of South Carolina, Charleston); Mandeep Dhadley, MD,* Johnnie Bloom, LPN (Roxbury Heart Center, Roxbury, Mass); Keith C. Ferdinand, MD,* Daphne Ferdinand, RN (Heartbeats Life Center, New Orleans, La); Charles K. Francis, MD,* Leroy Herbert, MD, Velvie Pogue, MD, Hope Bannister (Harlem Hospital Center, New York, NY); Henry Gift, MD,* Louis Gottlieb, MD, Ahang Zafari, MD (St Mary's Family Health Center, Waterbury, Conn); Ami Patel, MD,* C. Andrew Brown, Ardell Hinton (Veterans Affairs Medical Center, Jackson, Miss); Emanuel O. Quaye, MD,* Tim Briscoe, PharmD (Morehouse School of Medicine, Atlanta, Ga); Malcolm Taylor, MD,* Gwen Taylor (Jackson Cardiology Associates, Jackson, Miss); Harry Ward, MD,* Bella Forster, RN (King Drew Medical Center, Los Angeles, Calif); and Matthew R. Weir, MD,* Wallace Johnson, MD, Elijah Saunders, MD, B. J. Shaneman, LPN (University of Maryland, Baltimore).
Reprints: W. Dallas Hall, MD, International Society on Hypertension in Blacks, 2045 Manchester St NE, Atlanta, GA 30324-4110.