Patients and Methods
A total of 398 patients attending 35 outpatient centres in Austria, Germany, Italy and Spain entered a 2-week placebo washout/run-in phase, after which they were considered eligible for study enrolment if they had a de novo diagnosis of essential hypertension or a known history of essential hypertension. Patients with a sitting office systolic blood pressure (SBP) of 140–199mm Hg and a sitting office diastolic blood pressure (DBP) of 95–114mm Hg at the end of the washout/run-in phase were randomised to treatment (326 patients).
Criteria for exclusion from the study were: a history of myocardial infarction (MI) or stroke within the preceding 6 months; heart failure; haemo- dynamically relevant aortic or mitral stenosis; obstructive hypertensive cardiomyopathy; primary hyperaldosteronism; renal artery stenosis; impairment of hepatic or renal function as defined by liver function values of ≥1.5-fold the upper normal limit or serum creatinine >150 µmol/L; and a history of intolerance to the drug classes used in the study.
This was a randomised, double-blind, parallel-group, active-controlledstudy conducted in accordance with the principles of the revised Declaration of Helsinki (Somerset West, South Africa, 1996), and Good Clinical Practice. The protocol was approved by the responsible Ethics Committees (Institutional Review Boards). Prior written informed consent was obtained from all patients.
After the 2-week single-blind, placebo washout/run-in period, during which any previous antihy- pertensive drug treatment was discontinued, the 326 patients fulfilling the inclusion criteria were randomly allocated to receive amlodipine 5mg once daily or candesartan cilexetil 8mg once daily for 6 weeks in a double-blind design. Patients responding to treatment (defined as those with a sitting office DBP <90mm Hg) continued their randomised treatments at an unchanged dose for the remaining 6 weeks of the study. Patients who did not reach the target DBP of <90mm Hg had their dose uptitrated for the remaining 6 weeks: those in the amlodipine group were given amlodipine 10mg once daily, and those in the candesartan cilexetil group were given candesartan cilexetil 16mg once daily. After a total of 12 weeks of active treatment, all patients were required to take a single-blind placebo instead of the active medication, simulating a 4-day 'drug holiday'.
The treatment medications supplied for the study were amlodipine besylate 5 and 10mg tablets, candesartan cilexetil 8 and 16mg capsules, placebo matching the amlodipine tablets and placebo matching the candesartan cilexetil capsules. Treatment groups were assigned by a permutated block randomisation. The randomisation codes were sealed and were maintained blind through all procedures. In order to maintain the double-blind, all patients took one tablet and one capsule once daily (double-dummy design). Patients were instructed to take their medication every morning at breakfast time.
Patients visited the outpatient centres every 2 weeks, and had their office BP and heart rate determined in the sitting position. Office BP and heart rate were measured 22–26 hours after the previous dose using an oscillometric automated BP recorder with a printout report (Omron 705 CP, Hamburg, Germany). All patients had BP readings taken at 2-minute intervals after an initial 5-minute rest. The value of the first reading was disregarded and the mean was calculated for the second and third readings. Self-measurement of BP was performed five times a day for the whole study course using a BP wrist watch (NAIS Blood Pressure Watch EW 272 and EW 280, Matsushita Electric Works, Düsseldorf, Germany; both devices are validated by the German Society of Hypertension). After years of concern about the wrist BP monitoring devices,[14] several validated devices have in the meantime gained broader acceptance through very comfortable measurements to arm cuff devices, especially for systolic BP. All the recorded data were documented in the patient's diary. The physicians trained the patients on how to obtain valid BP readings. As a general rule, measurements were taken in the sitting position with the device at heart level. Patients measured their BP at the following times: (i) briefly before the patient got up in the morning (still in bed); (ii) briefly before the patient took the morning dose of the study drug; (iii) at noon (about 12:00h); (vi) in the afternoon (about 17:00h); and (v) upon retiring in the evening.
Standard laboratory blood tests were performed at screening and at the end of the active treatment period in all patients: creatinine, AST, ALT, uric acid, glycosylated haemoglobin (HbA1c), glucose, total cholesterol, high-density lipoprotein (HDL), low-density lipoprotein (LDL) and triglycerides. Fibrinogen was determined for the German and Austrian study sites only by a central laboratory.
The primary efficacy endpoint was the comparison of changes from baseline in trough sitting office SBP and DBP at week 12. The maximum acceptable differences for clinical equivalence in office BP were defined as 5mm Hg (SD 13mm Hg) for SBP and 3mm Hg (SD 8mm Hg) for DBP. A 10% drop-out rate from the study was assumed. Based on these prerequisites and using the two-sided t-test, a total of 274 patients were necessary to achieve a power of at least 80%.
Statistical comparisons of BP equivalence were carried out using analysis of covariance (ANCOVA) with treatment groups, baseline values and centre effects as covariates within a per-protocol population using the last-observation-carried-forward (LOCF) method. Two-sided 95% CIs were calculated for adjusted means between final and baseline values via the ANCOVA model.
The primary efficacy endpoint analysis was performed using the per-protocol population (n = 294), which included all randomised patients who received at least one dose of study medication with the data-set of the mean of the second and third SBP and DBP measurements during treatment and without any severe protocol deviation.
The differences in the frequency of adverse events were assessed using 95% Pearson Clopper CIs. The safety population (n = 326) included all patients who received medication. The mean percentage changes in laboratory measurements were calculated using the Bootstrap method.
Starting from the screening visit, patients measured their BP and heart rate five times a day. The analysis of self-measured BP was performed using the daily mean of these five measurements. The value for each visit (screening and weeks 2, 4, 6, 8, 10 and 12) was calculated as the mean of the patient's diary entries on the corresponding day as well as the four neighbouring days (± 2 days). For time-points outside the visit schedule (weeks 1, 3, 5, 7, 9 and 11) the intermediate date between two consecutive visits was identified and the same procedure (mean value over five consecutive days) was applied accordingly. The statistical analysis of self-measured BP was carried out in the same way as for office BP. In addition, an analysis using a mixed linear model was performed.
Clin Drug Invest. 2005;25(9):567-577. © 2005 Adis Data Information BV
Cite this: Amlodipine Besylate Versus Candesartan Cilexetil in Hypertensive Patients -- Office and Self-Measured Blood Pressure - Medscape - Sep 01, 2005.
