Journal of Human Hypertension (2009) 23, 610–619 & 2009 Macmillan Publishers Limited All rights reserved 0950-9240/09 $32.00 www.nature.com/jhh ORIGINAL ARTICLE Antihypertensive efficacy of telmisartan vs ramipril over the 24-h dosing period, including the critical early morning hours: a pooled analysis of the PRISMA I and II randomized trials

B Williams1, Y Lacourcie`re2, H Schumacher3, P Gosse4 and JM Neutel5 1Department of Cardiovascular Sciences, Leicester Royal Infirmary, Leicester, UK; 2Hypertension Research Unit, Centre Hospitalier de Que´bec, Ste-Foy, Quebec, Canada; 3Medical Data Services Department, Boehringer Ingelheim Pharma GmbH & Co. KG, Ingelheim, Germany; 4Service de Cardiologie—Hypertension Arte´rielle, Hoˆpital Saint-Andre´, Bordeaux, France and 5Orange County Research Center, Tustin, CA, USA

Cardiovascular risk is subject to circadian variation, dosing cycle. The adjusted mean treatment differences with peak morning incidence of myocardial infarction in the last 6-h mean ambulatory SBP/DBP were À5.8/– and stroke correlating with the early morning blood 4.2 mm Hg after 8 weeks and À4.1/–3.0 mm Hg after 14 pressure (BP) surge (EMBPS). Ideally, antihypertensive weeks, in favour of telmisartan (Po0.0001 for all four therapy should maintain control of BP throughout the comparisons). Secondary end point results, including 24-h dosing cycle. In two sister studies, Prospective, the mean 24-h ambulatory BP monitoring, day- and Randomized Investigation of the Safety and efficacy of night-time BP and 24-h BP load, also significantly Micardis vs Ramipril Using ABPM (ambulatory BP favoured telmisartan (Po0.0001). Both treatments monitoring) (PRISMA) I and II, BP control was compared were well tolerated; adverse events, including cough, in patients with essential hypertension (24-h mean were less common with telmisartan. These findings baseline ambulatory BPB148/93 mm Hg) randomized suggest that telmisartan is more effective than ramipril to the angiotensin receptor blocker, telmisartan throughout the 24-h period and during the EMBPS; (80 mg; n ¼ 802), or the angiotensin-converting enzyme this may be attributable to telmisartan’s long duration inhibitor, ramipril (5 or 10 mg; n ¼ 811), both dosed of effect, which is sustained throughout the 24-h in the morning. The primary end point was the change dosing period. from baseline in mean ambulatory systolic BP (SBP) Journal of Human Hypertension (2009) 23, 610–619; and diastolic BP (DBP) during the final 6 h of the 24-h doi:10.1038/jhh.2009.4; published online 19 February 2009

Keywords: ambulatory blood pressure monitoring; cardiovascular risk factors; circadian variation; ramipril; telmisartan

Introduction monitoring (ABPM) are beginning to elucidate the pathogenic significance of the EMBPS, notably as a Reducing the early morning blood pressure surge risk factor for ischaemic stroke.8–10 The clinical (EMBPS) is an emerging objective of antihyperten- 1–3 application of antihypertensive therapy, however, sive therapy. The peak incidence of atherosclero- has not kept pace with the progress in circadian tic plaque rupture and acute cardiovascular events biology.11 In a recent editorial, Kario1 has described occurs in the morning, at the same time as surges in morning hypertension as the ‘blind spot’ of current BP, pulse rate, sympathetic tone and the activity 4–7 management. A key consideration is that antihyper- of the renin–angiotensin–aldosterone system. tensive medication is taken once daily, shortly after Recent prospective studies using ambulatory BP the patient wakes in the morning.12–14 After a given day’s dose, the next EMBPS occurs towards the end of the dosing cycle, as treatment efficacy approaches Correspondence: Professor B Williams, Department of its nadir. Cardiovascular Sciences, Clinical Sciences Building, Leicester Telmisartan, an angiotensin II receptor blocker Royal Infirmary, PO Box 65, Leicester LE2 7LX, UK. (ARB) with a plasma half-life (t1)ofB24 h, provides E-mail: [email protected] 2 Received 2 November 2008; revised 5 January 2009; accepted 8 sustained BP control throughout the daily dosing 15,16 January 2009; published online 19 February 2009 1 cycle. Ramipril (t2, 13–17 h), a widely used Telmisartan vs ramipril: pooled PRISMA I and II B Williams et al 611 angiotensin-converting enzyme (ACE) inhibitor, is of proven efficacy in the prevention of cardio- vascular events.17 We have pooled the data from two clinical trials comparing the antihypertensive efficacy of telmisartan and ramipril at the end of the 24-h dosing cycle (that is, in the early-morning period). This pooled analysis provides the largest- ever study of an ACE inhibitor vs an ARB with regard to 24-h BP control and the EMBPS. Figure 1 Design of the PRISMA I and II clinical trials. ABPM, ambulatory blood pressure monitoring; E, enrolment; R, rando- Methods mization. Study design This was a prospectively designed analysis of two 8 weeks were telmisartan 40/80/80 mg and ramipril studies; Prospective, Randomized Investigation of 2.5/5/10 mg. Study medication was administered the Safety and Efficacy of Micardis vs Ramipril each day at 0900±1300 hours. Using ABPM (PRISMA) I18 and PRISMA II (no. Efficacy evaluations were conducted after 8 and 502.392 and no. 502.391, respectively).19 These were 14 weeks. At each evaluation, a 90207 ambulatory prospective, randomized, open-label, blinded end BP monitor (SpaceLabs Medical Data, Redmond, point (PROBE) clinical trials that evaluated BP WA, USA) was attached to the participant’s non- control during the final 6 h of the 24-h dosing cycle dominant arm, and monitoring commenced imme- in patients randomized to a forced-titration treat- diately after the scheduled administration of the ment with either telmisartan or ramipril. PRISMA I study medication. Measurements were taken at 20- and II were conducted in 10 countries, namely min intervals throughout each monitoring period. Austria, Canada, France, Germany, The Nether- Evaluability of individual measurements and of lands, South Africa, Spain, Switzerland, the United complete ABPMs was assessed using pre-specified Kingdom and the United States. screening rules. Mean values were calculated for Adults (X18 years) were potentially eligible for each hour relative to the time of dosing. Triplicate treatment randomization if they had mild or mod- measurements of trough-seated BP were taken with erate (stage I or II) primary hypertension, with both a an office cuff sphygmomanometer, in accordance mean clinic-seated systolic BP (SBP)/diastolic BP with the guidelines of the American Society of (DBP) of o180/95–109 mm Hg at enrolment and a Hypertension.20 mean 24-h DBP 485 mm Hg at baseline. For pre- The primary end point was the change, from menopausal women, adequate contraception was an baseline to evaluation, in the mean ambulatory BP additional inclusion criterion. The principal health- during the final 6 h of the 24-h dosing schedule. related exclusion criteria were a confirmed or Secondary end points included changes from base- suspected history of coronary artery disease, con- line in the mean ambulatory BP during the 24-h gestive heart failure or a recent acute cardiovascular dosing interval, the morning (0600 to 1159 hours), event (such as stroke within 6 months or myocardial daytime (0600 to 2159 hours) and night time (2200 infarction within 3 months), poorly controlled to 0559 hours); and in the 24-h BP load. An insulin-dependent diabetes mellitus and chronic ambulatory SBP treatment response was defined as kidney disease. Employment as a night shift worker a mean 24-h SBP o130 mm Hg and/or a reduction was an additional exclusion criterion. from the mean baseline SBP X10 mm Hg, whereas ambulatory SBP control was defined as a mean 24-h SBP p130 mm Hg. An ambulatory DBP treatment Study conduct response was defined as a mean 24-h DBP Study procedures were approved by the institu- o80 mm Hg and/or a reduction from the mean tional review boards/clinical trial ethics committees baseline DBP X10 mm Hg, whereas ambulatory of the participating centres. All patients gave written DBP control was defined as a mean 24-h DBP informed consent. o80 mm Hg. Trough clinic-seated BP end points Hypertensive status was confirmed at the end of a included SBP treatment response (SBP of 2- to 4-week placebo run-in period. At baseline, o140 mm Hg and/or a reduction from baseline of patients were randomized, in a 1:1 ratio, to 14 X10 mm Hg), DBP response (DBP o90 mm Hg and/ weeks’ once-daily, open-label monotherapy with or a reduction from baseline of X10 mm Hg) and either telmisartan (MICARDISÀ, Boehringer Ingel- DBP control (DBP o90 mm Hg). heim GmbH, Ingelheim, Germany) or ramipril The safety evaluation, which included all patients (Altace, Aventis Pharma Inc., Laval, QC, Canada). who received at least one dose of study medication, Dosing in each treatment arm adhered to a forced- involved monitoring of the frequency and intensity titration regimen (Figure 1), whereby the respective of all adverse events, of those classified as serious doses at initiation (baseline), after 2 weeks and after or drug-related and of those prompting study

Journal of Human Hypertension Telmisartan vs ramipril: pooled PRISMA I and II B Williams et al 612 discontinuation. Pulse rate was monitored and Table 1 Baseline demographic and physiological profile of 1613 recorded while the patient was ambulatory, as well randomized patients in the pooled PRISMA I and II trials as between the second and third trough-seated BP Telmisartan Ramipril measurements. (n ¼ 802) (n ¼ 811)

Male, n (%) 509 (63.5) 513 (63.3) Statistical analyses Analysis of the primary and all secondary end Age (years) Mean±s.d. 53.3±10.4 53.0±10.0 points was based on the intention-to-treat principle, X65 years, n (%) 116 (14.5) 104 (12.8) and included data for all participants who received at least one dose of study medication and for whom BMI (kg mÀ2) valid baseline and follow-up ambulatory or manual Mean±s.d. 29.6±5.1 29.8±5.3 in-clinic BP measurements were available. Treat- Race, n (%) ment effects were compared using analysis of White 662 (82.5) 671 (82.7) covariance, adjusting for study, treatment-by-study Black 44 (5.5) 40 (4.9) interaction and baseline. Response rates were Asian 22 (2.7) 15 (1.8) evaluated using the Cochran–Mantel–Haenszel sta- Not askeda 74 (9.2) 85 (10.5) tistics, adjusting for study. A two-sided significance Hypertension duration (years) level (P) of 0.05 was assumed. Mean±s.d. 6.0±7.2 6.4±8.8

24-h Ambulatory BP (mm Hg) Mean SBP±s.d. 147.9±12.3 148.2±12.1 Results Mean DBP±s.d. 92.6±6.4 92.6±6.4

Subjects Last 6-h ambulatory BP (mm Hg) A total of 3591 patients were enrolled, but 1978 Mean SBP±s.d. 141.5±14.3 141.8±14.0 were not randomized, mainly because they did not Mean DBP±s.d. 88.4±8.1 88.3±8.0 meet the ABPM inclusion criteria. At baseline, the pooled study population comprised 1613 patients, Clinic BP (mm Hg) Mean SBP±s.d. 156.2±12.3 155.4±13.0 of whom 802 were randomized to telmisartan Mean DBP±s.d. 99.9±4.5 99.9±4.6 and 811 to ramipril. At completion of the study, 1436 (89.0%) participants remained, comprising Abbreviations: BMI, body mass index; BP, blood pressure; DBP, 716 (89.3%) in the telmisartan and 720 (88.8%) in diastolic BP; PRISMA, Prospective, Randomized Investigation of the Safety and Efficacy of Micardis vs Ramipril Using ABPM; SBP, the ramipril arm. Adverse events (telmisartan, 28; systolic BP. ramipril, 44) were the most frequent reason for study aRecording of information on race was illegal in France. discontinuation. Less frequent reasons included lack of treatment efficacy (telmisartan, 12; ramipril, 10) and withdrawal of consent (telmisartan, 18; ramipril, 10). greater with telmisartan of 80 mg than with ramipril The demographic profiles of the two treatment of 5 mg, by À5.8 mm Hg (95% confidence interval groups were closely matched (Table 1). In both (CI): À6.9, À4.7; Po0.0001) for SBP and by groups, B63% of patients were male and the mean À4.2 mm Hg (95% CI: À5.0, –3.5; Po0.0001) for age was B53 years; overall, most (B86%) partici- DBP. After 14 weeks (Figure 2b), the corresponding pants were younger than 65 years. Body mass index, reductions in SBP and DBP were greater with ethnic origin and BP profile were similarly well telmisartan of 80 mg than with ramipril of 10 mg, matched between treatment groups, with a 24-h by –4.1 mm Hg (95% CI: À5.2, À2.9; Po0.0001) mean baseline ambulatory BP of B148/93 mm Hg in for SBP and by –3.0 mm Hg (95% CI: À3.8, À2.2; both groups and comparable measurements for both Po0.0001) for DBP. Overall, the efficacy advantage the mean last 6-h ambulatory BP and the mean with telmisartan was consistent across the PRISMA I trough clinic-seated BP at baseline. The mean and II studies, with no evidence of either a study duration of diagnosed hypertension was B6 years effect or a study-by-treatment interaction; therefore, and B74% of the patients had been diagnosed as pooling of the data was appropriate. hypertensive for more than 1 year.

Secondary end points Primary end point Telmisartan of 80 mg reduced the mean 24-h, Once-daily telmisartan of 80 mg was significantly morning, daytime and night-time ambulatory SBP superior to once-daily ramipril of 5 or 10 mg during and DBP, as well as the mean 24-h BP load, to a the last 6 h of the 24-h dosing period (Figure 2), greater extent than did ramipril, at both efficacy which is consistent with its sustained duration of evaluations (Po0.0001 for all comparisons; Table 2, effect and pharmacokinetic profile. After 8 weeks Figures 3 and 4). Similarly, the rates of ambulatory (Figure 2a), the reductions in SBP and DBP were SBP and DBP treatment response and DBP control

Journal of Human Hypertension Telmisartan vs ramipril: pooled PRISMA I and II B Williams et al 613

Figure 2 Primary efficacy end points in the pooled analysis of PRISMA I and II: the effects of (a) once-daily telmisartan of 80 mg (n ¼ 739) or ramipril of 5 mg (n ¼ 751), after 8 weeks; and (b) once-daily telmisartan of 80 mg (n ¼ 706) or ramipril of 10 mg (n ¼ 712), after 14 weeks, on adjusted mean change from baseline in last 6-h SBP and DBP. CI, confidence interval; DBP, diastolic blood pressure; SBP, systolic blood pressure. were greater with telmisartan of 80 mg than with Changes in trough-seated BP were broadly con- ramipril, at both evaluations (Po0.0001 for compar- sistent with those in ambulatory BP. After 8 weeks, isons) (Table 3). the adjusted mean difference in reduction of SBP

Journal of Human Hypertension ora fHmnHypertension Human of Journal 614

Table 2 Adjusted mean changes, from baseline to evaluation, in selected secondary efficacy end points II and I PRISMA pooled ramipril: vs Telmisartan

After 8 weeks After 14 weeks

Telmisartan 80 mg Ramipril 5 mg Telmisartan 80 mg Ramipril 10 mg Adjusted mean change±s.e. Difference (telmisartan–ramipril) Adjusted mean change±s.e. Difference (telmisartan–ramipril)

N 739 751 Adjusted mean (95% CI) P-value 706 712 Adjusted mean (95% CI) P-value Williams B Mean ambulatory BP 24-h

SBP (mm Hg) –13.2±0.35 –9.2±0.35 –4.0 (–4.9, –3.0) o0.0001 –14.1±0.38 –11.0±0.38 –3.1 (–4.2, –2.1) o0.0001 al et DBP (mm Hg) –9.1±0.24 –5.9±0.23 –3.2 (–3.8, –2.5) o0.0001 –9.6±0.25 –7.2±0.25 –2.4 (–3.1, –1.7) o0.0001

Morning SBP –13.5±0.39 –8.4±0.39 –5.1 (–6.2, –4.1) o0.0001 –14.3±0.41 –10.0±0.41 –4.3 (–5.4, –3.1) o0.0001 DBP –9.2±0.27 –5.2±0.26 –4.1 (–4.8, –3.3) o0.0001 –9.8±0.27 –6.4±0.27 –3.3 (–4.1, –2.6) o0.0001

Daytime SBP –13.9±0.38 –10.4±0.37 –3.5 (–4.6, –2.5) o0.0001 –14.9±0.40 –11.9±0.40 –3.0 (–4.1, –1.9) o0.0001 DBP –9.5±0.25 –6.6±0.25 –2.9 (–3.6, –2.2) o0.0001 –10.1±0.27 –7.6±0.26 –2.5 (–3.2, –1.8) o0.0001

Night time SBP –11.7±0.39 –6.7±0.39 –5.0 (–6.1, –3.9) o0.0001 –12.4±0.42 –9.2±0.42 –3.2 (–4.4, –2.1) o0.0001 DBP –8.2±0.28 –4.4±0.27 –3.8 (–4.5, –3.0) o0.0001 –8.5±0.29 –6.3±0.29 –2.3 (–3.1, –1.5) o0.0001

24-h BP load (%) SBP –29.9±0.85 –21.2±0.84 –8.7 (–11.1, –6.4) o0.0001 –32.1±0.90 –25.6±0.90 –6.5 (–9.0, –4.0) o0.0001 DBP –28.0±0.73 –19.2±0.73 –8.8 (–10.8, –6.7) o0.0001 –30.0±0.75 –23.3±0.75 –6.7 (–8.8, –4.6) o0.0001

Abbreviations: BP, blood pressure; CI, confidence interval; DBP, diastolic BP; SBP systolic BP. Telmisartan vs ramipril: pooled PRISMA I and II B Williams et al 615

Figure 3 Profiles of the mean (a) SBP and (b) DBP at hourly intervals relative to the time of dosing, at baseline and after 8 weeks of once- daily treatment with either telmisartan of 40/80 mg or ramipril of 2.5/5 mg (forced titration of doses after 2 weeks). DBP, diastolic blood pressure; SBP, systolic blood pressure. between telmisartan of 80 mg and ramipril of telmisartan and 364 (44.9%) patients with ramipril. 5mg was À6.4 mm Hg (95% CI: À7.6, À5.1; The only significant difference in adverse events Po0.0001). For DBP, the corresponding difference between the two groups was in the incidence of was À3.8 mm Hg (95% CI: À4.5, À3.0; Po0.0001). cough, which affected 11 (1.4%) patients in the After 14 weeks, the adjusted mean difference in telmisartan group, as opposed to 68 (8.4%) patients reduction of SBP between telmisartan of 80 mg and in the ramipril group (Po0.0001, Fisher’s exact test). ramipril of 10 mg was À4.7 mm Hg (95% CI: À3.4, Telmisartan was also associated with a lower À6.1; Po0.0001). The corresponding difference for incidence of adverse events that was considered by DBP was À2.6 mm Hg (95% CI: À1.7, À3.4; Po0.0001). the investigators to be related to treatment (7.4%) Telmisartan also provided superior response rates than was ramipril (11.6%) (P ¼ 0.0038). This was and control of trough-seated BP (Table 4). primarily because of the lower incidence of treat- ment-related cough (0.4 vs 6.9%). By medication and dose, the numbers of patients Tolerability and safety who reported severe adverse events were telmisar- Both antihypertensive agents were well tolerated, tan of 40 mg, 4 (0.5%); telmisartan of 80 mg, 24 and the overall incidence of adverse events in the (3.1%); ramipril of 2.5 mg, 15 (1.8%); ramipril of two treatment arms was low. The number of patients 5 mg, 18 (2.3%); and ramipril of 10 mg, 14 (1.9%) reporting one or more adverse events, typically of patients. Serious adverse events were experienced mild or moderate intensity, was 323 (40.3%) with by 8 telmisartan-treated and 11 ramipril-treated

Journal of Human Hypertension Telmisartan vs ramipril: pooled PRISMA I and II B Williams et al 616

Figure 4 Profiles of the mean (a) SBP and (b) DBP at hourly intervals relative to the time of dosing, at baseline and after 14 weeks of once-daily treatment with either telmisartan of 40/80/80 mg or ramipril of 2.5/5/10 mg (forced titration of doses after 2 weeks and 6 weeks). DBP, diastolic blood pressure; SBP, systolic blood pressure.

Table 3 Rates of ambulatory SBP and DBP response and of DBP control, with once-daily telmisartan of 80 mg, or ramipril of 5 or 10 mg

8 weeks 14 weeks

Telmisartan 80 mg Ramipril 5 mg P-value Telmisartan 80 mg Ramipril 10 mg P-valuea (n ¼ 739) (n ¼ 751) (n ¼ 706) (n ¼ 712)

SBP responseb 68.9 53.5 o0.0001 72.5 62.8 o0.0001 (%) DBP responsec 50.6 29.2 o0.0001 53.4 38.9 o0.0001 (%) DBP controld (%) 36.7 18.1 o0.0001 38.4 24.7 o0.0001

Abbreviations: DBP, diastolic blood pressure; SBP, systolic blood pressure. aThe Cochran–Mantel–Haenszel test adjusted for study. b24-h Mean ambulatory SBP o130 mm Hg and/or reduction from baseline of at least 10 mm Hg. c24-h Mean ambulatory DBP o80 mm Hg and/or reduction from baseline of at least 10 mm Hg. d24-h Mean ambulatory DBP o80 mm Hg.

Journal of Human Hypertension Telmisartan vs ramipril: pooled PRISMA I and II B Williams et al 617 Table 4 Rates of trough-seated clinic SBP and DBP response and of DBP control, with once-daily telmisartan of 80 mg, or ramipril of 5or10mg

8 weeks 14 weeks

Telmisartan 80 mg Ramipril 5 mg P-value Telmisartan 80 mg Ramipril 10 mg P-valuea (n ¼ 765) (n ¼ 781) (n ¼ 722) (n ¼ 742)

SBP responseb 74.4 59.1 o0.0001 76.2 66.9 o0.0057 (%) DBP responsec 63.2 46.3 o0.0001 61.9 54.8 o0.0532 (%) DBP controld (%) 52.4 32.7 o0.0001 54.2 40.4 o0.0001

Abbreviations: DBP, diastolic blood pressure; SBP, systolic blood pressure. aThe Cochran–Mantel–Haenszel test adjusted for study. bSeated SBP o140 mm Hg and/or reduction from baseline of at least 10 mm Hg. cSeated DBP o90 mm Hg and/or reduction from baseline of at least 10 mm Hg. dSeated DBP o90 mm Hg.

patients, and none was classed by the investigators Control of morning BP is often poor, even in as treatment-related. In the course of the study, no patients with controlled office BP. In the analysis of patient died and there were no clinically significant the control of BP using ABPM (ACAMPA study), changes in vital signs. 62% of patients with controlled office BP were found to have ambulatory morning BP 4138/ 85 mm Hg.29 In the Jichi Morning Hypertension Discussion Research (J-MORE) study, which used home BP monitoring, the equivalent figure was 61%.8 This The pooled PRISMA I and II ambulatory BP database study shows that more widespread use of longer- is among the largest ever compiled from clinical acting agents could improve these figures. trials prospectively comparing an ARB with an ACE The treatment difference favouring telmisartan inhibitor. Analysis of this database has shown that was not confined to the last quarter of the daily once-daily telmisartan of 80 mg was significantly dosing schedule, but was evident in each hour superior to once-daily ramipril of 5 or 10 mg in the during 24-h ABPM, as well as in the mean 24-h last 6 h of the 24-h dosing interval. Although at the measurements. The clinical relevance of this result 8-week time point the ramipril dose of 5 mg once is suggested by the fact that 24-h ABPM gives a high daily might be considered sub-optimal when com- degree of correlation with the risk of cardiovascular pared with the telmisartan dose of 80 mg once daily, disease (superior, indeed, to that obtained with the significant differences in BP control persisted in office BP measurements).30 The difference between the last 6 h of the 24-h dosing interval, even after the telmisartan and ramipril in mean 24-h SBP, amount- ramipril dose had been force-titrated to 1 mg once ing to À3.1 mm Hg, is clinically significant, given daily from the 8th to the 14th week of the study. The that a À2 mm Hg reduction in SBP has been shown superiority of telmisartan over ramipril in this time to be associated with a reduction in mortality of 7% period is consistent with the established pharmaco- for ischaemic heart disease and 10% for stroke.31 kinetic profiles of these agents.21–23 It is also The recent large outcome study, the Ongoing consistent with the results of other trials using 24- Telmisartan Alone and in Combination with Rami- h ABPM that have shown superior efficacy of pril Global Endpoint Trial (ONTARGET), investi- telmisartan compared with antihypertensives from gated the cardiovascular protective benefits of a range of classes, including losartan,24,25 valsar- telmisartan and ramipril in a broad range of high- tan,15,16 amlodipine26 and perindopril.27 cardiovascular-risk patients.32 In the ONTARGET Evidence that reduction of morning BP provides study, telmisartan was non-inferior to ramipril in clinical benefit comes from a study of patients with preventing a composite of cardiovascular events that type II diabetes, hypertension and nephropathy, in formed the primary end point, that is a composite of whom a reduction of morning SBP in response to death from cardiovascular causes, myocardial in- antihypertensive treatment was shown to correlate farction, stroke and hospitalization for heart failure. significantly with a slowing of the progression of Mindful of the fact that our study shows that renal damage.28 In the analysis reported here, telmisartan is more effective than ramipril at con- telmisartan produced reductions in BP in the last trolling BP, the result of ONTARGET might seem 6 h of the dosing interval that were greater than surprising. However, it is important to recognize that those produced by ramipril of 10 mg by À4.1 mm Hg the ONTARGET study was not a study of hyperten- for SBP and À3.0 mm Hg for DBP, a difference that is sion treatment. Approximately 30% of patients were large and more likely to be clinically meaningful. considered normotensive at randomization and the

Journal of Human Hypertension Telmisartan vs ramipril: pooled PRISMA I and II B Williams et al 618 remaining patients had their BP treated with a variety maintained throughout the 24-h dosing cycle, it was of concomitant medications and these increased especially marked during the last 6 h, when as the study progressed. The average BP of the the EMBPS is typically associated with a peak in ONTARGET population at randomization was B142/ cardiovascular risk. This reflects the longer duration of 82 mm Hg. This is important because the ability to action of the ARB. Moreover, the superior efficacy show differences in BP-lowering efficacy between of telmisartan over ramipril was comprehensively treatments diminishes at a lower baseline BP and is shown, across the range of pre-specified end points, further diminished by the addition of concomitant with treatment differences persisting despite the forced medications. Thus, the smaller difference in BP titration of the ramipril dose, from 5 to 10 mg, between between the ramipril vs telmisartan arms of the evaluations. The findings of our study are important ONTARGET study (difference 0.9/0.6 mm Hg in fa- because they suggest that the better quality of 24-h BP vour of telmisartan) can be accounted for by the control and improved tolerability of the ARB could design of the ONTARGET study, which was not ultimately turn out to be the most important difference focused on showing differences in BP-lowering between ARBs and ACE inhibition. efficacy between treatment arms. Ongoing analyses of the ABPM substudy of the ONTARGET study patients will provide more data in this area. The ambulatory BP response rates to telmisartan What is known about this topic: K Peak incidence of atherosclerotic plaque rupture and acute in the PRISMA studies were high, with a DBP cardiovascular events occurs at the same time as the response in 450% of patients and an SBP response EMBPS and surges in pulse rate, sympathetic tone and the in B70% of patients, which is consistent with activity of the renin–angiotensin–aldosterone system.4–7 earlier studies.33 Control of trough DBP was K Recent prospective studies using ABPM are beginning to achieved in 54.2% of patients treated with telmi- elucidate the pathogenic significance of the EMBPS, notably as a risk factor for ischaemic stroke.8–10 Therefore, sartan of 80 mg, as opposed to only 40.4% with reducing EMBPS is an emerging objective of ramipril of 10 mg, a potentially important finding antihypertensive therapy.1–3 given that office BP currently remains uncontrolled in more than half of treated hypertensive patients in What this study adds: 34 K The pooled PRISMA I and II ambulatory BP database is Europe and the United States. among the largest ever compiled from clinical trials Although the PRISMA studies used a PROBE prospectively comparing an ARB with an ACE inhibitor. design, rather than the more accepted double-blind K The antihypertensive efficacy of once-daily telmisartan of design, this is not a major limitation. The PROBE 80 mg was significantly superior to that of once-daily design was used because it closely reflects clinical ramipril of 5 or 10 mg in the last 6 h of the 24-h dosing interval when the EMBPS is typically associated with a practice, whereas the end point assessment is peak in cardiovascular risk. The superiority of telmisartan clearly independent from knowledge of the patients’ over ramipril in this time period is consistent with the treatment allocation. Furthermore, PROBE studies results of other trials using 24-h ABPM that have shown have been shown earlier to provide results equiva- superior efficacy of telmisartan compared with 35 antihypertensives from a range of classes, including lent to those achieved by double-blind studies, and losartan,24,25 valsartan,15,16 amlodipine26 and perindopril.27 are also acceptable for use with ABPM. As discussed K These findings reflect the longer duration of action of the above, an additional potential limitation of the ARBs and are important because they suggest that the better study design could be that the ramipril dose used quality of 24-h BP control and improved tolerability of the in the present studies is submaximal (that is, at a ARB could ultimately turn out to be the most important difference between ARBs and ACE inhibition. minimum in the United States); however, the doses of drugs selected for initial and maximal therapy were those typically used in clinical practice to treat hypertension. The PRISMA studies were conducted Acknowledgements in both Europe (PRISMA I) and North America (PRISMA II). Telmisartan of 80 mg was chosen The PRISMA study programme was investigator-led, because it is the maximum approved dose in both and was supported by a research grant from Europe and the United States. The maximum Boehringer Ingelheim. approved dose of ramipril varies in Europe (10 mg) and the United States (20 mg). For consistency, ramipril of 10 mg was used in both PRISMA studies, References and the patients were force-titrated to this dose for 6 weeks before the final efficacy comparison. More- 1 Kario K. Morning surge and variability in blood over, this dose of ramipril has shown earlier BP- pressure: a new therapeutic target? Hypertension lowering efficacy and cardiovascular protection.17,36 2005; 45: 485–486. 2 Kario K. Time for focus on morning hypertension: In conclusion, pooled analysis of PRISMA I and II pitfall of current antihypertensive medication. Am J shows that the antihypertensive efficacy of once-daily Hypertens 2005; 18: 149–151. telmisartan of 80 mg was consistently superior to 3 Marfella R, Siniscalchi M, Nappo F, Gualdiero P, that of once-daily ramipril of 5 or 10 mg. Although Esposito K, Sasso FC et al. Regression of carotid the treatment difference favouring telmisartan was atherosclerosis by control of morning blood pressure

Journal of Human Hypertension Telmisartan vs ramipril: pooled PRISMA I and II B Williams et al 619 peak in newly diagnosed hypertensive patients. Am J excretion of intravenously and orally administered Hypertens 2005; 18: 308–318. [14C]telmisartan in healthy volunteers. J Clin Pharma- 4 Millar-Craig MW, Bishop CN, Raftery EB. Circadian col 2000; 40: 1312–1322. variation of blood-pressure. Lancet 1978; i: 795–797. 22 Stangier J, Su CAPF, Roth W. Pharmacokinetics of 5 Muller JE, Stone PH, Turi ZG, Rutherford JD, Czeisler orally and intravenously administered telmisartan in CA, Parker C et al. Circadian variation in the frequency healthy young and elderly volunteers and in hyper- of onset of acute myocardial infarction. N Engl J Med tensive patients. J Int Med Res 2000; 28: 149–167. 1985; 313: 1315–1322. 23 Meisel S, Shamiss A, Rosenthal T. Clinical pharmaco- 6 Marler JR, Price TR, Clark GL, Muller JE, Robertson T, kinetics of ramipril. Clin Pharmacokinet 1994; 26:7–15. Mohr JP et al. Morning increase in onset of ischemic 24 Ding PY, Chu KM, Chiang HT, Shu KH. A double-blind stroke. Stroke 1989; 20: 473–476. ambulatory blood pressure monitoring study of the 7 Tanaka A, Kawarabayashi T, Fukuda D, Nishibori Y, efficacy and tolerability of once-daily telmisartan Sakamoto T, Nishida Y et al. Circadian variation of 40 mg in comparison with losartan 50 mg in the plaque rupture in acute myocardial infarction. Am J treatment of mild-to-moderate hypertension in Taiwa- Cardiol 2004; 93: 1–5. nese patients. Int J Clin Pract Suppl 2004; 58: 16–22. 8 Kario K, Pickering TG, Umeda Y, Hoshide S, Hoshide 25 Smith DH, Cramer MJ, Neutel JM, Hettiarachchi R, Y, Morinari M et al. Morning surge in blood pressure as Koval S. Comparison of telmisartan versus losartan: a predictor of silent and clinical cerebrovascular meta-analysis of titration-to-response studies. Blood disease in elderly hypertensives: a prospective study. Press Monit 2003; 8: 111–117. Circulation 2003; 107: 1401–1406. 26 Lacourcie`re Y, Asmar R. A comparison of the efficacy 9 Kario K, Shimada K, Pickering TG. Clinical implica- and duration of action of the angiotensin II receptor tion of morning blood pressure surge in hypertension. blocker telmisartan to amlodipine. Am J Hypertens J Cardiovasc Pharmacol 2003; 42(Suppl 1): S87–S91. 1999; 12: 1181–1187. 10 Gosse P, Lasserre R, Minifie´ C, Lemetayer P, Clementy 27 Ragot S, Ezzaher A, Meunier A, Poterre M, Bourkaib R, J. Blood pressure surge on rising. J Hypertens 2004; 22: Herpin D. Comparison of trough effect of telmisartan 1113–1118. vs perindopril using self blood pressure measurement: 11 Prisant LM. Hypertension and chronotherapy: shifting EVERESTE study. J Hum Hypertens 2002; 16: 865–873. the treatment paradigm. Am J Hypertens 2001; 14: 28 Suzuki H, Kanno Y, Nakamoto H, Okada H, Sugahara S. 277S–279S. Decline of renal function is associated with proteinuria 12 Mallion JM, Baguet JP, Siche JP, Tremel F, de Gaudemaris and systolic blood pressure in the morning in diabetic R. Compliance, electronic monitoring and antihyperten- nephropathy. Clin Exp Hypertens 2005; 27: 129–138. sive drugs. JHypertens1998; 16(Suppl 1): S75–S79. 29 Redo´n J, Roca-Cusachs A, Mora-Macia J. Uncontrolled 13 Fujii J, Seki A. Compliance and compliance-improving early morning blood pressure in medicated patients: strategies in hypertension: the Japanese experience. the ACAMPA study. Analysis of the control of blood J Hypertens Suppl 1985; 3(1): S19–S22. pressure using ambulatory blood pressure monitoring. 14 Morgan T, Anderson A, Jones E. The effect on 24 h Blood Press Monit 2002; 7: 111–116. blood pressure control of an angiotensin converting 30 Clement DL, De Buyzere ML, De Bacquer DA, de enzyme inhibitor (perindopril) administered in the Leeuw PW, Duprez DA, Fagard RH, Gheeraert PJ et al. morning or at night. J Hypertens 1997; 15: 205–221. Prognostic value of ambulatory blood-pressure record- 15 Lacourcie`re Y, Krzesinski JM, White WB, Davidai G, ings in patients with treated hypertension. N Engl J Schumacher H. Sustained antihypertensive activity of Med 2003; 348: 2407–2415. telmisartan compared with valsartan. Blood Press 31 Lewington S, Clarke R, Qizilbash N, Peto R, Collins R, Monit 2004; 9: 203–210. Prospective Studies Collaboration. Age-specific rele- 16 White WB, Lacourcie`re Y, Davidai G. Effects of the vance of usual blood pressure to vascular mortality: a angiotensin II receptor blockers telmisartan versus meta-analysis of individual data for one million adults valsartan on the circadian variation of blood pressure: in 61 prospective studies. Lancet 2002; 360: 1903–1913. impact on the early morning period. Am J Hypertens 32 The ONTARGET Investigators. Telmisartan, ramipril, 2004; 17: 347–353. or both in patients at high risk for vascular events. N 17 Mann JF, Gerstein HC, Pogue J, Bosch J, Yusuf S. Renal Engl J Med 2008; 358: 1547–1559. insufficiency as a predictor of cardiovascular outcomes 33 Neutel J, Smith DH. Evaluation of angiotensin II and the impact of ramipril: the HOPE randomized trial. receptor blockers for 24-hour blood pressure control: Ann Intern Med 2001; 134: 629–636. meta-analysis of a clinical database. J Clin Hypertens 18 Williams B, Gosse P, Lowe L, Harper R, PRISMA I 2003; 5: 58–63. Study Group. The prospective, randomized investiga- 34 Wolf-Maier K, Cooper RS, Kramer H, Banegas JR, tion of the safety and efficacy of telmisartan versus Giampaoli S, Joffres MR et al. Hypertension treatment ramipril using ambulatory blood pressure monitoring and control in five European countries, Canada, and (PRISMA I). J Hypertens 2006; 24: 193–200. the United States. Hypertension 2004; 43: 10–17. 19 Lacourcie`re Y, Neutel JM, Davidai G, Koval S. A 35 Smith DH, Neutel JM, Lacourcie`re Y, Kempthorne- multicenter, 14-week study of telmisartan and ramipril Rawson J. Prospective, randomized, open-label, in patients with mild-to-moderate hypertension using blinded-endpoint (PROBE) designed trials yield the ambulatory blood pressure monitoring. Am J Hyper- same results as double-blind, placebo-controlled trials tens 2006; 19: 104–112. with respect to ABPM measurements. J Hypertens 20 American Society of Hypertension. Recommendations 2003; 21: 1291–1298. for routine blood pressure measurement by indirect cuff 36 The Heart Outcomes Prevention Evaluation Study sphygmomanometry. Am J Hypertens 1992; 5: 207–209. Investigators. Effects of an angiotensin-converting- 21 Stangier J, Schmid J, Tu¨ rck D, Switek H, Verhagen A, enzyme inhibitor, ramipril, on cardiovascular events Peeters PA et al. Absorption, metabolism, and in high-risk patients. N Engl J Med 2000; 342: 145–153.

Journal of Human Hypertension