This study examined the effects of increasing the thiazide diuretic dose in a fixed-dose ARB/diuretic combination in patients with uncontrolled hypertension despite 6 weeks' open-label treatment with the ARB/diuretic combination, telmisartan 80 mg/hydrochlorothiazide 12.5 mg (T80/H12.5). 713 patients with trough seated DBP =90 mmHg were then randomized to 8 weeks' double-blind treatment with telmisartan 80 mg and an increased dose of 25 mg of hydrochlorothiazide (T80/H25) or T80/H12.5. Adjusted mean seated DBP changes from baselines of 95.3 (T80/H25) and 95.0 mmHg (T80/H12.5) were −7.1 and -−5.5 mmHg (difference: 1.6 mm Hg), respectively (P=.0012). Changes in systolic blood pressure from 147.9 mmHg (T80/H25) and 147.4 mmHg (T80/H12.5) were −9.8 and −7.1 mmHg (difference: 2.7 mm Hg) (P=.0003). Adverse events occurred in 31.5% (T80/H25) and 29.6% (T80/H12.5), with serious events in 1.4% and 0.8%, respectively. Hypokalemia was rare. These results show that higher-dose thiazide diuretic in combination with T80 in patients with hypertension uncontrolled by T80/H12.5 provides additional blood pressure reductions and is well tolerated.
Because of the detrimental long-term effects of hypertension on cardiovascular morbidity and mortality, recent guidelines advocate tight control of blood pressure.1–3 Certain patients, such as those with concomitant diabetes and/or chronic kidney disease, are at increased cardiovascular risk and require more rigorous control of blood pressure to improve prognosis.4,5 Programs to educate patients have increased their awareness of the seriousness of hypertension, and in recent years, the proportion of adults receiving treatment for hypertension has increased in both the United States and Europe.6–9 However, in many patients (most notably in Europe), elevated blood pressure is still inadequately controlled.10–12 In patients in whom target blood pressure is not reached with monotherapy, the use of ≥2 antihypertensive agents is recommended.1,2 Use of agents with complementary modes of action potentiates efficacy; adverse effects of the different agents may be counterbalanced when used in combination. One such recommended combination is an angiotensin receptor blocker (ARB) and a thiazide diuretic.1,2 Administration of the 2 drugs as a fixed-dose combination (FDC) in a single tablet also may have the advantage of encouraging patient compliance because of a simplified treatment regimen.13
Telmisartan is an ARB with a long elimination half-life.14 In clinical trials, telmisartan, as well as several other ARBs, has been shown to provide 24-hour control of blood pressure.15 The FDC of telmisartan 80 mg plus a low dose of the thiazide diuretic hydrochlorothiazide 12.5 mg (T80/H12.5) administered once daily is effective in many hypertensive patients16–18 and is approved as second-step treatment for patients whose blood pressure is not adequately controlled on other antihypertensive therapies. However, blood pressure targets still may not be achieved in some patients with this medication. In these patients, the use of telmisartan 80 mg plus a higher 25-mg dose of hydrochlorothiazide may provide additional reduction of blood pressure.
The present study (www. ClinicalTrials.gov, Identifier: NCT00239369) was conducted to determine whether a higher dose of a thiazide with the ARB resulted in further reduction in blood pressure in patients whose blood pressure was insufficiently controlled after 6 weeks' treatment with T80/H12.5. The study was also designed to compare the safety and tolerability of the 2 medications.
PATIENTS AND METHODS
The study was conducted in adults (older than 18 years) of either sex with essential hypertension who were currently taking between 1 and 3 antihypertensive medications at a stable dose for at least 4 weeks and in whom blood pressure was not sufficiently controlled on existing treatment. Inadequate control was defined as a seated diastolic blood pressure (DBP) ≥95 mm Hg on 1 current antihypertensive medication or DBP ≥90 mm Hg on 2 or 3 current antihypertensive medications; blood pressure at this stage was not necessarily measured at trough. Women who were pregnant, breast-feeding, or of childbearing potential and not using effective contraception were excluded from the study. Other exclusion criteria included known or suspected secondary hypertension, mean systolic blood pressure (SBP) ≥200 mm Hg, severe renal impairment, any clinically relevant hepatic impairment or electrolyte imbalances, any condition that would exclude treatment with ARBs or thiazide diuretics, drug or alcohol dependency within the previous 6 months, and hypertrophic obstructive cardiomyopathy or hemodynamically relevant stenosis of the aortic or mitral valve. All patients had to provide written informed consent before enrollment.
The study was approved by the independent ethics committee/institutional review board of each participating center. During the screening period of 4 to 10 days, patients continued to take their usual antihypertensive medication(s). Eligible patients stopped their usual antihypertensive medication(s) and switched to open-label T80/H12.5 once daily for a run-in period of 6 weeks. Thereafter, patients with trough seated DBP ≥90 mm Hg (ie, with inadequate control of hypertension) were randomized (1:1) to double-blind, double-dummy treatment with either T80/H25 or T80/H12.5 for 8 weeks. During the run-in and double-blind treatment periods, no additional antihypertensive medication other than the assigned study drug was permitted.
Blood pressures in the seated and standing positions were measured using standard cuff sphygmomanometry.19,20 SBP and DBP values were determined as the mean of 3 readings 2 minutes apart. Mean SBP and DBP measurements were recorded at the start of the run-in treatment, after 6 weeks' open-label T80/H12.5 run-in treatment immediately before randomization, and after 8 weeks' double-blind T80/H25 or T80/H12.5 treatment. The primary efficacy end point was the change in seated DBP at trough at the end of the study compared with the value at randomization (baseline). Secondary efficacy end points included the change from baseline in seated SBP at trough, the change from baseline in standing DBP and SBP at trough, proportion of patients in whom DBP control (trough seated DBP <90 mm Hg) was reached, proportion of patients in whom DBP response (trough seated DBP <90 mm Hg or DBP reduction ≥10 mm Hg) was achieved, proportion of patients in whom SBP response (trough seated SBP <140 mm Hg or SBP reduction ≥10 mm Hg) was achieved, and proportion of patients in whom a more rigorous SBP response (trough seated SBP <140 mm Hg or SBP reduction ≥20 mm Hg) was achieved. The nature, incidence, severity, and causal relationship of all adverse events occurring during the study were recorded. In addition, routine hematology, blood chemistry, and urinalysis were performed at a central laboratory on samples at the beginning and end of the study. Patient compliance with medication was monitored throughout the study.
A difference of 2 mm Hg in DBP between treatments was considered clinically significant.21 Previous experience has shown that the standard deviation of the change from baseline in the trough seated DBP in a study with this design would be 6.7 mm Hg (unpublished data). Using this estimate, a sample size of 240 evaluable patients per treatment group would have approximately 90% power to detect a significant difference between the different doses of the thiazide if the real difference was 2.0 mm Hg (two-sided, α=.05). To achieve this number of evaluable patients, it was estimated that approximately 1600 patients would need to be enrolled, on the assumption that effective control of trough seated DBP would not be attained in 30% of patients with T80/H12.5.
Primary efficacy analysis was performed in patients who received at least 1 dose of double-blind medication and in whom the baseline and subsequent trough seated DBP had been measured. The effects of T80/H25 and T80/H12.5 were compared using analysis of covariance, including fixed effects of treatment and country, and baseline DBP measurement a covariate. The adjusted mean treatment difference was determined with the corresponding 95% confidence interval (CI) and P value resulting from a two-sided test of statistical significance at the 5% level. The same statistical analysis was performed on the secondary efficacy end points of trough seated SBP and trough standing SBP and DBP. The proportions of patients in whom DBP control and SBP and DBP responses were achieved in the two treatment groups were compared using Mantel-haenszel statistics, with control for the effect of country. Comparisons were expressed as 95% CIs and P values.
Further analysis of changes in mean seated trough blood pressure (unadjusted) in patients in the full analysis set from the start of run-in to the end of the study was also performed.
In total, 1042 patients were enrolled at 113 centers in 16 countries in Europe (Denmark, Finland, France, Germany, Ireland, Italy, Netherlands, Norway, Spain, Sweden, Switzerland), Asia (Hong Kong, Korea, Malaysia, Taiwan), and South Africa. The patients had poorly controlled blood pressure while being treated with 1 to 3 prior antihypertensive agents, with mean seated SBP/DBP (not necessarily trough values) of 154.3±14.1/97.5±4.9 mm Hg at the start of the study. At the end of the run-in period, the mean seated SBP/DBP at trough was 145.2±14.9/92.6±7.5 mm Hg in the 918 patients with complete data. Of the 1042 patients, a target DBP of <90 mm Hg was not reached in 713 after treatment with open-label T80/H12.5 for 6 weeks; these persons were eligible for randomization. The trough seated SBP/DBP at baseline in the randomized patients was 148.0±13.0/95.3±4.7 mm Hg in the T80/H25 group compared with 147.5±13.0/94.9±4.4 mm Hg in the T80/H12.5 group. The baseline characteristics of the patients randomized to double-blind treatment with either medication were comparable (Table I). and of these randomized patients, 686 completed the study; discontinuation rates were 4.7% and 2.8%, respectively, in the lower- and higher-dose thiazide groups. Compliance with double-blind treatment was good (>97%).
Table I. Demographics and Baseline Characteristics of the Patients Randomized to Double-Blind Treatment With a Fixed-Dose Combination of T80/H25 or T80/H12.5
Male sex, No. (%)
Mean ± SD age, y
Age <65 y, No. (%)
Race, No. (%)a
Mean ± SD body mass index (kg/m2)
Smokers, No. (%)
Mean ± SD trough seated SBP, mm Hg
Mean ± SD trough seated DBP, mm Hg
Mean ± SD duration of hypertension, y
Duration of hypertension, No. (%)
Previous antihypertensive therapy,d No. (%)
Calcium channel blocker
Abbreviations: ACE, angiotensin-converting enzyme; ARB, angiotensin receptor blocker; DBP, diastolic blood pressure; T80/H12.5, telmisartan 80 mg/hydrochlorothiazide 12.5 mg; T80/H25, telmisartan 80 mg/hydrochlorothiazide 25 mg; SBP, systolic blood pressure. aRace not recorded in 1 T80/H12.5 group patient. bn=349; cn=357. dPatients may appear in more than one row of the table.
After 8 weeks of double-blind treatment with either T80/H25 or T80/H12.5 in patients who had previously failed to respond to antihypertensive therapy, there was a reduction in the primary end point of trough seated DBP (Figure 1). The reduction was larger with T80/H25: the adjusted mean ± SE change from baseline was −5.5±0.4 mm Hg with T80/H12.5 compared with −7.1±0.5 mm Hg with T80/H25. The difference (95% CI) in the adjusted mean changes of −1.6 mm Hg (−2.5 mm Hg, −0.6 mm Hg) showed that T80/H25 provided significantly superior DBP reduction (P=.0012).
A reduction in trough seated SBP was also seen after 8 weeks' treatment (Figure 1). Again, the reduction was greater with T80/H25: adjusted mean ± SE change was −9.8±0.7 mm Hg with T80/H25 compared with −7.1±0.7 mm Hg with T80/H12.5. The difference (95% CI) in adjusted mean change between the two treatments of −2.7 mm Hg (−4.2 mm Hg, −1.2 mm Hg) in favor of T80/H25 was statistically significant (P=.0003).
The effects of the 2 treatments on trough standing SBP and DBP followed a similar pattern. The difference in adjusted mean changes in standing DBP in favor of T80/H25 was −1.5 mm Hg (–2.2 mm Hg, –0.6 mm Hg) (P=.0016). For SBP, the difference in favor of T80/H25 was –2.9 mm Hg (–4.4 mm Hg, –1.5 mm Hg) (P=.0001). Mean (unadjusted) seated blood pressure levels at the start of run-in, at baseline after 6 weeks' T80/H12.5 treatment, and after 8 weeks of randomized T80/H25 or T80/H12.5 treatment for the full analysis set are reported in Table II.
Table II. Mean (Unadjusted) Seated SBP/DBP (mm Hg) in Patients in the Full Analysis Set Treated With a Fixed-Dose Combination of T80/H25 or T80/H12.5
Start of run-ina
Baseline after 6 weeks of T80/H12.5 treatmentb
Change from start of run-in to baseline
After 8 weeks' randomized treatmentb
Change from start of run-in to end of randomized treatment
The proportions of patients in whom DBP control and SBP and DBP responses were achieved are shown in Figure 2. In the case of DBP control, defined as seated DBP at trough <90 mm Hg, this target was reached in 55.9% of T80/H25-treated patients and 49.0% of T80/H12.5-treated patients (P=.0641). DBP response, defined as trough seated DBP <90 mm Hg or a trough seated DBP reduction ≥10 mm Hg, was achieved in a significantly higher proportion of patients treated with T80/H25 (59.7% with T80/H25 vs 51.9% with T80/H12.5; P=.0336). Similarly, an SBP response, defined as a seated trough SBP <140 mm Hg or a seated trough SBP reduction ≥10 mm Hg, was attained in a significantly higher proportion of patients treated with the higher dose of the diuretic: 65.9% with T80/H25 compared with 57.3% with T80/H12.5(P=.0200). Also, using the more rigorous definition of SBP response of seated trough SBP <140 mm Hg or a seated trough reduction SBP ≥20 mm Hg, T80/H25 was superior: 57.6% compared with 48.1%(P=.0103).
During the 6-week T80/H12.5 FDC run-in phase, 219 patients (22.6%) experienced adverse events. In 51 of these patients, the events were considered related to study drug. Discontinuation due to an adverse event occurred in 27 patients. In the 8-week double-blind randomized phase of the study, the proportion of patients reporting ≥1 adverse event was 30.6% (Table III). In the T80/H25 and T80/H12.5 groups, the incidence of adverse events (31.5% vs 29.6%, respectively), serious adverse events (1.4% vs 0.8%, respectively), and discontinuations due to adverse events (1.7% vs 3.0%) were comparable. The majority of reported adverse events were mild (17.9% of T80/H25 patients and 19.1% of T80/H12.5 patients) or moderate (11.6% of T80/H25 patients and 9.1% of T80/H12.5 patients) in intensity. Adverse events of severe intensity occurred in 7 patients receiving T80/H25 and 5 receiving T80/H12.5. Drug-related adverse events were reported in 5.7% and 5.0%, respectively, of patients in the T80/H25 and T80/H12.5 groups. The most frequent drug-related events were erectile dysfunction (2 vs 3 patients), headache (1 vs 4 patients), vertigo (3 vs 1 patient), and dizziness (3 T80/H25 patients). Severe events considered by the investigators to be related to the study drug occurred in 1 patient receiving T80/H25 (hypotension) and in 2 patients receiving T80/H12.5 (third-degree atrioventricular block, back pain). Serious adverse events occurred in 5 T80/H25 and 3 T80/H12.5 patients. Two of the serious adverse events were considered related to the study drug: atrial flutter in a patient receiving T80/H25 and third-degree atrioventricular block in a patient receiving T80/H12.5. Death from a head injury after falling from a bicycle occurred in a patient treated with T80/H12.5.
Table III. Comparison of the Safety of a Fixed-Dose Combination of T80/H25 or T80/H12.5 During the Randomized Phase
Incidence (% of Patients)a
Any adverse event
Adverse events of severe intensity
Discontinuations due to adverse events
Serious adverse events
Abbreviations: T80/H12.5, telmisartan 80 mg/hydrochlorothiazide 12.5 mg; T80/H25, telmisartan 80 mg/hydrochlorothiazide 25 mg. aNo. of patients with event/No. in group.
No treatment-emergent laboratory abnormalities were reported as adverse events. Hypokalemia (serum potassium <3.5 mmol/L) was detected in 2 patients treated with T80/H12.5 and in 1 patient treated with T80/H25. No case was considered clinically significant, and none required treatment with a potassium supplement. Potassium levels <3.0 mmol/L were never detected during the study in any patient. In addition, there were no instances of hyperkalemia (serum potassium >5.8 mmol/L) during randomized treatment.
In hypertensive patients, blood pressure is frequently poorly controlled despite pharmacologic intervention; target values are not achieved in many patients. In our study, patients had previously received up to 3 different antihypertensive agents without adequate reduction in blood pressure. This inadequacy of blood pressure control is consistent with the findings of recent large-scale surveys in different European countries.10–12 poor blood pressure control may be attributed to the use of low doses of medication or not using multiple agents.22 Poor blood pressure control, however, is not always the fault of the prescribing physician; part of the problem may be due to poor patient compliance or lack of appreciation of the clinical significance of hypertension.22
Prior to entering this study, patients had received a variety of antihypertensive agents. The most common class of antihypertensive medication given was diuretics, either alone or together with other classes of agents. Hydrochlorothiazide 12.5 mg administered once daily has been shown to reduce blood pressure in approximately one-half to two-thirds of patients who are responsive to this class of drugs.23 The use of higher doses may provide additional control of blood pressure but may be associated with increased incidence of adverse effects. These adverse effects may provide an explanation of why some patients are more likely to discontinue thiazide diuretics than other classes of antihypertensives, such as ARBs.24 ARBs were the next most commonly prescribed antihypertensive. The ineffective lowering of blood pressure among patients receiving such treatment may have been due to some ARBs not providing 24-hour control. A recent meta analysis reported that telmisartan was statistically more effective than valsartan or losartan in reducing 24-hour mean blood pressure.25
After the end of the run-in phase, which involved switching from previous antihypertensive therapy to an FDC of T80/H12.5, mean blood pressure was substantially reduced, even in patients in whom the guideline-defined blood pressure goal of seated DBP < 90 mm Hg at trough was not reached.1–3 Our study showed that in those patients in whom the target DBP was not attained with T80/H12.5, there was a further significant reduction in seated trough DBP (the primary efficacy end point) if patients were treated with the higher dose of a thiazide with the ARB. T80/H25 also proved superior to T80/H12.5 in reducing trough seated SBP, resulting in a further reduction in SBP of 2.7 mm Hg. This blood pressure reduction could be of long-term clinical significance, as a meta-analysis of data from more than 1 million patients participating in 61 outcome studies suggested that a reduction in SBP of 2 mm Hg could be associated with a 10% reduction in the risk of fatal stroke and a 7% reduction in the risk of death from other vascular causes.21
Although the start of run-in blood pressure values cannot be directly compared with the end point values (since not all were taken at trough), the changes from the start of the run-in to end point are of interest since they illustrate blood pressure reductions in patients whose blood pressure was uncontrolled despite receipt of as many as 3 antihypertensive agents and who were then switched to an ARB plus hydrochlorothiazide. The SBP/DBP reductions (14.0/9.2 and 12.3/7.9 mm Hg for T80/25 and T80/12.5, respectively) were statistically significant and likely to be of clinical significance.
The secondary efficacy end points also demonstrate the superiority of the higher-thiazide dose formulation in providing blood pressure control in a higher proportion of patients. The findings of our study are consistent with a large-scale, placebo-controlled study conducted in the United States in patients with stage 1 or 2 hypertension that demonstrated the antihypertensive efficacy of T80/H25 after 8 weeks' treatment and showed its superiority to another ARB combination of valsartan 160 mg plus hydrochlorothiazide 25 mg.26
The large proportion (68%) who did not respond adequately to T80/H12.5 may be due to the restrictive inclusion criteria, which only enrolled patients with documented uncontrolled hypertension. Patients enrolled in this study had elevated blood pressure (154.1/97.6 mm Hg) despite prior treatment with 1 to 3 antihypertensive agents. Analysis of previous trials (which did not deliberately include patients resistant to prior therapy) of T80/H12.5 indicated that 35% of patients did not respond adequately to this regimen (unpublished data). It should be noted that even in patients who failed to respond to 6 weeks of open-label T80/H12.5 therapy, further reductions in blood pressure were seen in patients who were randomized to further T80/H12.5 therapy during the 8-week double-blind stage of the study. There are a number of possible explanations for this observation. One is that the 6-week open-label treatment was not sufficiently long to achieve a stable reduction in blood pressure in patients with uncontrolled hypertension. Alternatively, it may be that patients were expecting to receive a more effective treatment after randomization, and hence there was a placebo-like effect. Regression to the mean may also have contributed to the additional reduction in blood pressure after 8 weeks' double-blind treatment with the lower-dose formulation.
The additional blood pressure lowering provided by T80/H25 did not appear to have any detrimental effect on adverse events, and both the low- and high-dose hydrochlorothiazide formulations were equally well tolerated. A previous dose-ranging study has shown that tolerability of various dose combinations of telmisartan and hydrochlorothiazide was comparable to that of telmisartan monotherapy.27 Most notably, the combination protected against potassium depletion, which is widely recognized as a relatively common side effect of higher-dose hydrochlorothiazide, especially dosages >50 mg/d. In our study, hypokalemia was a rare event and was detected in only 3 patients, only 1 of whom was treated with T80/H25; none of these patients required a potassium supplement.
In conclusion, treatment with T80/H25 provides a further reduction of blood pressure in patients with difficult-to-treat hypertension that does not respond adequately to treatment with T80/H12.5. The additional blood pressure lowering, which may have a positive benefit on long-term cardiovascular prognosis, was achieved without compromising tolerability and safety.
Acknowledgments and disclosure:
Statistics and data management were performed by D. Massey and A. Monckton of Boehringer Ingelheim Ltd, Bracknell, UK. This was a Boehringer Ingelheim-sponsored study. Dr Edwards is an employee of Boehringer Ingelheim.