Evaluation of Angiotensin II Receptor Blockers for 24-Hour Blood Pressure Control: Meta-Analysis of a Clinical Database


Joel M. Neutel, MD, Director, Orange County Heart Institute and Research Center, 505 South Main Street, Suite 1025, Orange, CA 92828
E-mail: jmneutel@aol.com


Angiotensin II receptor blockers are well tolerated and improve compliance in hypertensive patients. The need for 24-hour blood pressure control has focused attention on whether all agents in this class maintain smooth antihypertensive effects over a 24-hour period. Insight into this issue emerged from a meta-analysis of five large, multicenter trials in which ambulatory blood pressure monitoring was used to compare the antihypertensive effects of three angiotensin II receptor blockers: telmisartan, losartan, and valsartan. These trials used either a double-blind, placebo-controlled or a prospective, randomized, open-label, blinded-end point design. Initial analysis established the validity of combining ambulatory blood pressure monitoring data from the double-blind, placebo-controlled and prospective, randomized, open-label, blinded-end point designs. Subsequent analyses revealed that telmisartan 80 mg was significantly more effective than losartan 50 mg and valsartan 80 mg for reducing 24-hour mean blood pressure. Furthermore, telmisartan 80 mg was comparable to amlodipine 5 mg for controlling the early morning surge in blood pressure.

Epidemiologic studies have shown that people with high blood pressure (BP) are at greater risk for cardiovascular or cerebrovascular events, such as stroke or myocardial infarction.1,2 The goal in the management of hypertension is to reduce the incidence of morbidity and mortality from cardiovascular events. A meta-analysis of the effects of antihypertensive treatment on the incidence of cardiovascular complications has demonstrated that the goal of reducing such events has been achieved only partially.3 The reduction in the incidence of stroke in hypertensive patients is consistent with the observed reduction in BP. However, only about one half of the expected reduction in coronary heart disease has been attained from a comparable BP reduction. This suggests that adequate management of hypertension has not yet been achieved across a broad spectrum of hypertensive patients.

Guidelines on the management of hypertension from the World Health Organization4 or the sixth report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure5 emphasize the importance of lowering BP. However, the optimal level of BP reduction that is required to achieve improved outcomes has become apparent only recently. A series of outcome studies has compared treated hypertensive patients achieving tight BP control with those achieving less tight BP control.6,7 The UK Prospective Diabetes Study, for example, compared two groups of type 2 diabetic patients who were also hypertensive and had similar baseline Bps.7 After 9 years of follow-up, patients in the “less tight control” group had a mean BP of 154/87 mm Hg. In the “tight control” group, mean BP was 144/82 mm Hg. Comparison of the outcomes between these two groups showed a 37% reduction in the risk of microvascular complications (p=0.0092) and a 44% reduction in the risk of stroke (p=0.013) among patients with tight control of BP. Therefore, a difference in BP of 10/5 mm Hg had a dramatic impact on cardiovascular outcomes in type 2 diabetic patients who were hypertensive.

Although the importance of reducing BP is well known, only one quarter of hypertensive patients are adequately controlled to a BP of 140/90 mm Hg or less.5 This threshold suggests that approximately 45 million people within the United States are currently either untreated or have inadequately treated BP. Data from other countries also suggest that these goals are unattainable in little more than 25% of hypertensive patients.8 Therefore, despite the numbers of drugs available for the treatment of hypertension, adequate control of high BP and hence, adequate protection against the risk of cardiovascular diseases has yet to be achieved in the vast majority of hypertensive patients.


Patient compliance with the dosing regimen is the most important reason that treatment goals are not achieved with antihypertensive therapy.5 Compliance is largely dependent on drug side effects and convenience of the dosing regimen.9 If compliance can be improved, the number of patients with controlled hypertension will increase. The Merck-Medco database, which contains information on all antihypertensive agents used in the United States, shows greater drug compliance at 1 year among patients on angiotensin II receptor blockers (ARBs), compared with those on any other class of antihypertensive drug.10

ARBs are well tolerated, and all agents in this class exhibit a side effect profile comparable to that of placebo.11 For example, results from clinical trials indicate that the incidence of side effects with telmisartan and placebo is similar, although patients treated with telmisartan report a significantly lower incidence of headaches than those treated with placebo.11 These data, together with evidence for other ARBs, suggest that headache, nocturia, and fatigue are important symptoms of hypertension that many patients discount as normal, everyday occurrences. As the entire class of ARBs appears to be better tolerated than placebo, this would suggest that hypertension is not truly an asymptomatic disease. Many hypertensive patients are non-compliant because they do not consider the benefits of treatment to justify the inconvenience of taking medication.12 The potential of ARBs to reduce the incidence of subtle symptoms of hypertension to a degree perceptible to the patient may have important consequences for promoting compliance.

Patients tend to be more compliant with once-daily dosing, as they prefer the convenience of taking their medication only once during the day—usually in the morning, when it is more discreet and easier to maintain a routine.9 With the newer generation of ARBs, such as telmisartan, we now have drugs that have an extremely long duration of action, which provide smooth BP control over a 24-hour dosing period.13–15 We can be confident that once-daily dosing with these agents will not compromise the patient at the end of the dosing period, when antihypertensive protection is often most needed.


The primary reason for selecting an antihypertensive drug is efficacy. Concern over the efficacy of ARBs, notably losartan, may explain why fewer than 10% of patients around the world are prescribed ARBs, despite the advantages provided by superior compliance.16 Several small studies, however, have compared ARBs with commonly used antihypertensive agents, such as atenolol, lisinopril, and amlodipine.15,17,18 In each of these studies, the antihypertensive efficacy of the ARBs was at least comparable or superior to that of other antihypertensive agents. The concerns raised about the efficacy of ARBs prompted us to conduct a meta-analysis of telmisartan clinical trials.


This meta-analysis was limited to studies using ambulatory BP monitoring (ABPM). ABPM provides more clinical data and is a better predictor of target-organ damage than other BP monitoring methods.19–21 Considerable focus was placed on the use of ABPM in the clinical program of telmisartan, providing the largest ABPM database for a single ARB. Hence, although the selected studies were performed in a number of geographic regions (United States, Canada, and Europe), five studies were selected for inclusion in the meta-analysis because of consistency among inclusion/exclusion criteria, criteria for ABPM, and the type of BP monitor used for assessments (SpaceLabs 90207 device, Redmond, WA).

Comparison of PROBE and DBPC Trial Designs. The only major difference between the studies included in the meta-analysis was that two used a double-blind, placebo-controlled (DBPC) design and three used a prospective, randomized, open-label, blinded-end point (PROBE) design. Before combining the data, we needed to investigate whether the results from these two study designs were consistent. If data obtained from the DBPC and PROBE studies were consistent, we could legitimately examine the comparative efficacies of telmisartan, losartan, valsartan, and amlodipine for reduction of 24-hour ambulatory BP and pulse pressure.

There is an increasing trend toward PROBE-design studies, as they are easier to conduct in clinical practice. In recent times, many of the phase 4 studies investigating antihypertensive drugs have used the PROBE design. The main advantages of PROBE trials are that there is no need for a matched placebo group and that the open-label design allows patients to receive a marketed form of the drug.22 Hence, patients are more amenable to and more comfortable with this trial design. In addition, PROBE studies offer a considerable savings in time and cost over DBPC studies, as no additional packaging of drugs is required.22 The major disadvantage of PROBE studies, however, is the possibility of patient and/or investigator bias. The blinded end point (in this case ABPM) is used to attenuate this disadvantage. However, many researchers remain concerned that PROBE studies are inferior to DBPC studies with respect to the reliability of the data.

Comparison of the demographics of the DBPC and PROBE studies selected for the meta-analysis showed that age and gender distribution was similar between the two study groups. Similarly, baseline systolic BP (SBP) and diastolic Bps (DBP) between the two groups were comparable, whether measured in the office or as the 24-hour mean of ABPM. Indeed, baseline mean ambulatory Bps for both groups were 150–152/93–95 mm Hg. The number of subjects treated with telmisartan 80 mg within the PROBE design study group was about 5.8-fold greater than in the DBPC group, which was factored into the statistical analysis.

Our hypothesis suggested that data from DBPC and PROBE designs were equivalent if the decrease in BP from baseline to the end of the treatment period differed between the two groups by less than 3 mm Hg for SBP and less than 2 mm Hg for DBP. The actual differences between the decrease in 24-hour mean ambulatory BP was rounded to 0.2 mm Hg (95% confidence interval [CI], −1.8, 2.1) for SBP and 0.2 mm Hg (95% CI, −1.1, 1.5) for DBP. Hence, the PROBE and DBPC designs provided comparable data, despite the difference in study design, and we concluded that it was valid to pool data from these two types of studies. The added confidence in PROBE studies provided by these data should make PROBE-design studies of antihypertensive drugs the preferred choice, since double-blind, placebo-controlled studies appear to be an unnecessary inconvenience at an additional cost for phase IV studies.


Data from 1566 patients were included in the meta-analysis (Table I). The baseline demographics showed patients treated with placebo, telmisartan 40 or 80 mg, or comparators (losartan 50 mg, valsartan 80 mg, or amlodipine 5 mg) to be similar with respect to age, race, and gender. Baseline office or mean 24-hour ambulatory BP did not differ among the six groups.

Table I.  Number of Patients in the ABPM Meta-Analysis
TreatmentNumber of Patients With Analyzable ABPM
Telmisartan 40 mg124
Telmisartan 80 mg860
Losartan 50 mg50
Valsartan 80 mg197
Amlodipine 5 mg206
Total (all treatments)1566
ABPM=ambulatory blood pressure monitoring

Comparison of BP Reduction

A comparison of the mean change from baseline in either SBP or DBP over 24 hours showed all ARBs included in this analysis to be comparable to amlodipine 5 mg (Figure 1). Most physicians regard amlodipine as the “gold standard” in terms of efficacy in reducing BP. By inference, therefore, ARBs are also highly efficacious in reducing BP. Based on the current metaanalysis, telmisartan 80 mg exhibited the greatest efficacy, producing significantly greater reductions in mean 24-hour SBP than losartan 50 mg, valsartan 80 mg, or telmisartan 40 mg, and in mean 24-hour DBP than losartan 50 mg or valsartan 80 mg (p<0.0125). In light of the many treatment comparisons made, all between-agent analyses were performed at an α=0.0125 (=0.05/4) (two-sided) level of significance. An α=0.05 (two-sided) was used to compare the 40 mg and 80 mg doses of telmisartan, since a monotonically increasing relationship for these two doses has been established previously for SBP.11

Figure 1.

Figure 1.

Mean change from baseline in 0–24-hour ambulatory systolic (SBP) and diastolic blood pressure (DBP) in patients treated with placebo, losartan 50 mg (L50), valsartan 80 mg (V80), telmisartan 40 mg (T40), amlodipine 5 mg (A5), or telmisartan 80 mg (T80). All treatments significantly reduced mean SBP and DBP compared with placebo (p<0.001).

Comparison of the SBP profiles over 24 hours indicated that ARBs are not equal in terms of duration of action (Figure 2). Toward the end of the dosing interval, a decrease in BP control was observed with valsartan 80 mg and with losartan 50 mg. In contrast, both doses of telmisartan maintained efficacy comparable to that of amlodipine throughout the 24-hour dosing period.

Figure 2.

Figure 2.

Comparison of systolic blood pressure (SBP) over 24 hours in hypertensive patients treated with angiotensin receptor blockers, amlodipine, or placebo

Much has been written regarding the importance of controlling the morning surge in BP.23–25 Morning BP is a good assessment of whether the antihypertensive drug is providing smooth 24-hour BP control. Morning was defined as the period between 6–11:59 a.m. Examining the reduction in BP from baseline during this period shows that telmisartan 40 mg and 80 mg continued to provide good BP control (Figure 3). Hence, telmisartan provides sustained reduction in BP throughout the entire dosing interval.

Figure 3.

Figure 3.

Mean reduction in systolic blood pressure (A) and diastolic blood pressure (B) during the morning period (6–11:59 a.m.) by angiotensin receptor blockers and amlodipine 5 mg p<0.0125 vs. losartan 50 mg or valsartan 80 mg; p<0.0125 vs. losartan 50 mg or valsartan 80 mg; p<0.05 vs. telmisartan 40 mg.

Comparative Pulse Pressure Reductions

Some concern has been raised about the decrease in pulse pressure with the various antihypertensive agents.26 Verdecchia et al.27 recently suggested that reducing pulse pressure to <53 mm Hg seemed to be clinically important. In this meta-analysis, the decrease in mean 24-hour pulse pressure ranged between 3.1 and 4.4 mm Hg for all ARBs studied, with telmisartan 80 mg inducing the greatest reduction in 24-hour mean pulse pressure (4.4 mm Hg), comparable to that of amlodipine 5 mg (4.3 mm Hg).

Comparative Response Rate

The response rate for any single class of antihypertensive agent is often within the range of 50%–70%.28 A response was defined as the reduction of 24-hour mean BP to <85 mm Hg diastolic or a reduction from baseline of geqslant R: gt-or-equal, slanted10 mm Hg for DBP, or a reduction to <130 mm Hg systolic or reduction from baseline of geqslant R: gt-or-equal, slanted15 mm Hg for SBP. Using these stringent definitions, the DBP response rates for the various ARBs in this meta-analysis ranged between 38% and 52%, with telmisartan 80 mg inducing the highest response rate (Table II). This compares well with the range of response rates observed for other antihypertensive agents using less stringent definitions of response.28 Response rates for SBP were lower than for DBP for each of the ARBs as well as for amlodipine, ranging from 17.5% for amlodipine 5 mg to 30.6% for telmisartan 80 mg (Table II). However, this is consistent with studies using other antihypertensive agents, such as the Hypertension Optimal Treatment [HOT] study.6

Table II.  Comparative Efficacy-Response Rates
  Response Rates
TreatmentNSBP (%)DBP (%)
Losartan 50 mg502446
Valsartan 80 mg19723.437.6
Telmisartan 40 mg12421.844.4
Amlodipine 5 mg20617.540.8
Telmisartan 80 mg86030.6*52.1
Response is defined as 24-hour mean systolic blood pressure (SBP) <130 mm Hg or reduction from baseline of ≥15 mm Hg and diastolic blood pressure (DBP) <85 mm Hg or reduction from baseline of =10 mm Hg.
*p<0.0125 vs. amlodipine 5 mg; p<0.05 vs. valsartan 80 mg or telmisartan 40 mg; p<0.0125 vs. valsartan 80 mg or amlodipine 5 mg.

Although these response rates for achieving stringent goals with ARBs are typical for monotherapy, it is clear that we currently are unable to manage hypertension successfully in a large proportion of patients with a single drug. Only with combination therapy have response rates for DBP goals within the range of 85%–90% been achieved, even with the less stringent goal for DBP of 95 mm Hg used in many studies.29 Treatment strategies using combination therapy are more likely to achieve adequate management of hypertension in a greater majority of patients. Currently, there is a dearth of data available on combination therapy with ARBs, except for the addition of small doses of hydrochlorothiazide.29–34 Hydrochlorothiazide also augments the therapeutic response to angiotensin-converting enzyme inhibitors.35,36


ARBs represent an important therapeutic advance in the treatment of hypertension. They are extremely well tolerated, which has significant advantages in terms of improving the poor compliance often encountered with hypertensive patients. Some of these agents, particularly telmisartan, have a duration of action of 24 hours or greater and can be prescribed for once-daily dosing without compromising BP control at the end of the dosing period. ARBs are highly efficacious antihypertensive agents, and are at least as effective as other antihypertensive drugs. However, differences in efficacy and duration of action are observed within the ARB class. Of those evaluated in the current meta-analysis, telmisartan was the most efficacious and exhibited the longest duration of action. When applied to daily clinical practice, these qualities of ARBs may optimize 24-hour BP control, including the critical early morning period.