Journal of the American Society of Hypertension 9(4) (2015) 257–265
Research Article
Rationale for a single-pill combination of perindopril arginine and amlodipine besylate William J. Elliott, MD, PhD Division of Pharmacology, Department of Biomedical Sciences, Pacific Northwest University of Health Sciences, Yakima, WA Manuscript received November 30, 2014 and accepted December 21, 2014
Abstract A systematic review identified 86 outcome–based clinical trials involving perindopril, amlodipine, or other antihypertensive drugs. In fixed–effects meta–analyses of 11 clinical trials (90,208 subjects), amlodipine was associated with a significant 24% increase in heart failure, but a significant decrease in death, cardiovascular death, stroke, coronary heart disease, and first major cardiovascular adverse event. In five clinical trials (52,565 subjects), perindopril was associated with a significant reduction in all six cardiovascular endpoints. Network and Bayesian meta–analyses suggested that (with the exception of amlodipine and heart failure), each agent was at least as effective as an initial diuretic to prevent these events. Short–term trials have demonstrated that the combination of perindopril and amlodipine is safe and effective, with statistically greater lowering of blood pressure than either agent alone and a potential synergistic effect on pedal edema. The single–pill combination of perindopril and amlodipine may be a useful addition to the antihypertensive armamentarium. J Am Soc Hypertens 2015;9(4):257–265. Ó 2015 American Society of Hypertension. All rights reserved. Keywords: Cardiovascular events; fixed–effects; heart failure; network and Bayesian meta–analyses; random–effects..
Introduction Hypertension continues to be a major public health concern, both in the US and worldwide.1–5 Recent data suggest that, despite an increase in prevalence attributed to an aging and more obese population,4 the prevalence of controlled hypertension has improved during the last decade.5,6 Some have attributed this, and a corresponding decrease in hypertension–related morbidity and mortality in some populations,7–9 to more widespread use of antihypertensive drug therapy, particularly single–pill combination products that improve adherence, persistence, and adverse effect profiles, compared with antihypertensive
This manuscript was developed without financial support from any entity. Conflict of interest: none. *Corresponding author: William J. Elliott, MD, PhD, Professor of Preventive Medicine, Internal Medicine and Pharmacology, Head, Division of Pharmacology, Chair, Department of Biomedical Sciences, Pacific Northwest University of Health Sciences, 200 University Parkway, Yakima, WA 98901. Tel: 509-2497726; fax: 509-249-7799. E-mail: [email protected]
drug monotherapy.10 During the last decade, the US Food and Drug Administration has approved many single–pill antihypertensive drug combinations as initial therapy, usually because of a low probability of achieving an appropriate blood pressure target with monotherapy, particularly in individuals with Stage 2 hypertension.2,3 Two of the most widely appreciated ‘‘synergies’’ of single–pill combinations for hypertension are the offsetting effects of renin–angiotensin system blockers (angiotensin–converting enzyme [ACE]–inhibitors, angiotensin receptor blockers [ARBs], or renin inhibitors) and thiazide diuretics on serum potassium concentrations, and the arteriolar dilating properties of dihydropyridine calcium antagonists and venodilating properties of renin–angiotensin system blockers on peripheral edema.11 Recent data suggest that ACE–inhibitors are more effective than other renin–angiotensin system blockers for reducing the severity and incidence of amlodipine–associated peripheral edema.12 Reduction of adverse effects attributed to single–pill combinations has resulted in a large number of marketed agents of each type, many of which are now available in generic formulations, and some of which have been studied in prospective randomized clinical trials with cardiovascular and/or renal outcomes. Our purpose is
1933-1711/$ - see front matter Ó 2015 American Society of Hypertension. All rights reserved. http://dx.doi.org/10.1016/j.jash.2014.12.012
258
W.J. Elliott / Journal of the American Society of Hypertension 9(4) (2015) 257–265
to review current data regarding a new single–pill combination of perindopril þ amlodipine.
Methods Search Strategy A formal systematic review of the literature was performed 14 times from 1999 to 2013 to identify all clinical trials enrolling hypertensive patients, assigning different treatment strategies (at least one of which involved an antihypertensive drug), and reporting cardiovascular outcomes during a median follow–up of greater than 1 year. We searched PubMed, MEDLINE, the Cochrane Library, and OVIDWeb using text words and MeSH terms: ‘‘controlled clinical trials, cerebrovascular disorders, heart disease, antihypertensive agents, angiotensin–converting enzyme inhibitors, angiotensin II type 1 blockers, adrenergic alpha– antagonists, adrenergic beta–antagonists, diuretics, sodium chloride symport inhibitors, and calcium channel blockers.’’ These searches produced 189 citations that
were examined for trials that involved perindopril, 694 for amlodipine, and another 928 citations involving agents other than placebo. From these, 182 reports of randomized clinical trials were retrieved and assessed for mortality and/ or hospitalization outcomes. Additionally, six randomized clinical trials were added from the reference lists of prior meta–analyses and other, related publications.
Meta–analyses Outcomes data from all publications were doubleentered into a Microsoft Excel (Microsoft Corporation, Redmond, WA) spreadsheet and traditional fixed–effects (Mantel–Haenszel) (and sometimes random–effects) meta–analyses were performed using StatsDirect (Cheshire, UK). A separate computerized spreadsheet collected data to compare individual agents versus their comparators, which was used as input for network and Bayesian meta–analyses. Network meta–analyses were performed using a modification of the one–line program published by Lumley,13 using R (the programming language, framework 2.51 GUI 1.20)
Figure 1. Identification and selection of outcome–based randomized clinical trials (RCTs) performed with antihypertensive drugs.
W.J. Elliott / Journal of the American Society of Hypertension 9(4) (2015) 257–265
on an Apple MacBook Pro running MacOS X.4.8 (Apple Corporation, Cupertino, CA). Bayesian meta–analyses were performed using WinBUGS, version 1.4.1,14 with non–informative priors, a burn of 1000, and 99,000 iterations.15 Convergence was achieved before the first 2000 iterations. A continuity correction factor of 0.5 was used for all entries. All reported P values are two–sided.
Results The search strategy outlined above for outcome–based clinical trials (Figure 1) identified 11 studies involving 89,768 subjects randomized to either initial16–24 or add– on25–28 amlodipine (Table 1), and five studies involving 42,947 subjects randomized to initial29,30 or add–on20,31,32 perindopril (Table 2). These trials are quite heterogeneous, not only with respect to where amlodipine or perindopril was used in the sequence of randomized drug treatments
259
(initial,16–24,29,30 second–line,20,27,28,32 or even later25,26,31), but also with respect to comparators (placebo17,25,26,29–32 or active drug therapies16,18–28), concomitant medical conditions at baseline (coronary heart disease17 or not20; diabetes25,26,30 or not27,28; chronic kidney disease25–28 or not18–20,29–32; prior stroke29 or not20); average age of subjects at baseline (>5518 or >80 years32), duration of follow-up (1.832 to 5.5 years20), and primary outcomes (serum biochemistry values16 to coronary heart disease events18,20). Despite these major differences across trials, traditional fixed–effects meta–analyses showed no significant inhomogeneity for eight of the 12 cardiovascular endpoints (six for amlodipine, six for perindopril). Figure 2 summarizes the overall results of these meta–analyses of six outcomes for amlodipine versus any and all comparators. Heart failure, which demonstrated significant inhomogeneity (P < .0001) across the 11 amlodipine trials,
Table 1 Randomized clinical trials involving amlodipine that reported cardiovascular outcomes Trial Acronym, Year (Duration)
Randomized Therapies
# of Subjects
Deaths
CV Deaths
Major Adverse CV Events
Stroke
Coronary Heart Disease
Heart Failure
FACET,16 1998 (2.5 years)
Amlodipine Fosinopril Amlodipine Placebo Amlodipine Chlorthalidone Lisinopril Amlodipine Valsartan Amlodipine Atenolol Amlodipine Candesartan Amlodipine HCTZ Amlodipine Valsartan Amlodipine Valsartan Amlodipine Irbesartan Placebo Amlodipine Ramipril Metoprolol
191 189 417 408 9048 15,255 9054 7596 7649 9639 9618 2354 2349 5721 5741 511 510 575 575 567 579 569 217 436 441
5 4 6 8 1256 2203 1314 818 841 738 820 86 73 236 262 3 2 16 22 83 87 93 22 34 49
NR NR 2 7 603 996 618 NR NR 263 342 NR NR 107 134 NR NR NR NR 37 52 46 7 12 12
23 14 23 28 1374* 2329* 1421* 621* 720* 796 937 80* 88* 288 364 11* 11* 25* 26* 128 138 144 29 61 65
10 4 5 5 377 675 457 281 322 327 422 50 61 112 133 10 10 16 13 15 28 26 9 23 23
13 10 19 20 798 1362 796 313 369 429 474 33* 28* 125 159 1 2 7* 11* 27 44 46 5 19 18
0 0 1 5 706 870 612 400 354 134 159 16 20 100 96 1 3 15 3 93 60 72 8 20 22
PREVENT,17 2000 (3.0 years) ALLHAT,18 2002 (4.9 years)
VALUE,19 2004 (4.2 years) ASCOT,20 2005 (5.5 years) CASE-J,21 2007 (3.2 years) ACCOMPLISH,22 2008 (3.3 years) VART,23 2010 (3.4 years) Nagoya Heart Study,24 2012 (3.2 years) IDNT,25,26 2001, 2003 (2.6 years)
AASK,27,28 2006 (3.6 or 4.4 years)
AASK, African American Study of Kidney Diseases and Hypertension; ACCOMPLISH, Avoiding Cardiovascular events through COMbination therapy in People Living with Systolic Hypertension; ALLHAT, Antihypertensive and Lipid Lowering (to prevent) Heart Attack Trial; ASCOT, Anglo-Scandinavian Cardiac Outcomes Trial; CASE-J, Candesartan Antihypertensive Survival Evaluation in Japan; CV, cardiovascular; FACET, Fosinopril Amlodipine Cardiac Events Trial; HCTZ, hydrochlorothiazide; IDNT, Irbesartan Diabetic Nephropathy Trial; NR, not reported; PREVENT, Prospective Randomized Evaluation of the Vascular Effects of Norvasc Trial; VALUE, Valsartan Antihypertensive Long–term Use Evaluation; VART, Valsartan Amlodipine Randomized Trial. * Estimated, using linear regression.
260
W.J. Elliott / Journal of the American Society of Hypertension 9(4) (2015) 257–265
Table 2 Randomized clinical trials involving perindopril that reported cardiovascular outcomes Trial Acronym, Year (Duration) Randomized Therapies
Stroke Coronary Heart # of Deaths CV Major Failure Heart Subjects Deaths Adverse Disease CV Events
PROGRESS,29 2001 (3.9 years) Perindopril indapamide Placebo placebo EUROPA,31 2003 (4.2 years) Perindopril Placebo ASCOT,20 2005 (5.5 years) Perindopril (after amlodipine) Bendrofluazide (after atenolol) ADVANCE,30 2007 (4.3 years) Perindopril (þindapamide) Placebo HYVET,32 2008 (1.8 years) Perindopril (after indapamide) Placebo (after placebo)
3051 3054 6110 6108 9639 9618 5569 5571 1933 1912
306 319 375 420 738 820 408 471 196 235
181 198 215 249 263 342 211 257 99 121
458 604 552 664 796 937 480 520 138 193
307 420 98 102 327 422 215 218 51 69
115* 154* 320 418 429 474 265 294 9 12
75y 93y 63 103 134 159 197y 199y 22 57
ADVANCE, Action in Diabetes and Vascular disease: preterAx and diamicroN Controlled Evaluation; ASCOT, Anglo-Scandinavian Cardiac Outcomes Trial; CV, cardiovascular; EUROPA, European Reduction Of cardiac events with Perindopril in stable coronary Artery disease; HYVET, HYpertension in the Very Elderly Trial; PROGRESS, Perindopril pROtection aGainst REcurrent Stroke Study. * Estimated, using linear regression. y Hospitalization for heart failure.
was significantly increased in the groups receiving amlodipine (by 24%). Conversely, all five other cardiovascular endpoints were significantly reduced in the groups receiving amlodipine, with little evidence of inhomogeneity across trials (note that the numbers of cardiovascular
deaths were not reported for Fosinopril Amlodipine Cardiac Events Trial,16 Valsartan Antihypertensive Long– term Use Evaluation,19 CASE–J,21 Valsartan Amlodipine Randomized Trial,23 or the Nagoya Heart Study,24 which limits the statistical power of any meta–analysis of this
Figure 2. Results of fixed–effects meta–analysis of amlodipine versus other randomized therapies. Results of meta–analyses of clinical trials in hypertension comparing amlodipine as either initial16–24 or add-on25–28 therapy versus any and all other comparators are presented. All P–values for homogeneity [P (homo)] were >0.05 (except for heart failure [HF]) for the trials that randomized subjects to initial amlodipine therapy, but those for cardiovascular (CV) death, coronary heart disease (CHD), and major cardiovascular disease (CVD) events were all
Research Article
Rationale for a single-pill combination of perindopril arginine and amlodipine besylate William J. Elliott, MD, PhD Division of Pharmacology, Department of Biomedical Sciences, Pacific Northwest University of Health Sciences, Yakima, WA Manuscript received November 30, 2014 and accepted December 21, 2014
Abstract A systematic review identified 86 outcome–based clinical trials involving perindopril, amlodipine, or other antihypertensive drugs. In fixed–effects meta–analyses of 11 clinical trials (90,208 subjects), amlodipine was associated with a significant 24% increase in heart failure, but a significant decrease in death, cardiovascular death, stroke, coronary heart disease, and first major cardiovascular adverse event. In five clinical trials (52,565 subjects), perindopril was associated with a significant reduction in all six cardiovascular endpoints. Network and Bayesian meta–analyses suggested that (with the exception of amlodipine and heart failure), each agent was at least as effective as an initial diuretic to prevent these events. Short–term trials have demonstrated that the combination of perindopril and amlodipine is safe and effective, with statistically greater lowering of blood pressure than either agent alone and a potential synergistic effect on pedal edema. The single–pill combination of perindopril and amlodipine may be a useful addition to the antihypertensive armamentarium. J Am Soc Hypertens 2015;9(4):257–265. Ó 2015 American Society of Hypertension. All rights reserved. Keywords: Cardiovascular events; fixed–effects; heart failure; network and Bayesian meta–analyses; random–effects..
Introduction Hypertension continues to be a major public health concern, both in the US and worldwide.1–5 Recent data suggest that, despite an increase in prevalence attributed to an aging and more obese population,4 the prevalence of controlled hypertension has improved during the last decade.5,6 Some have attributed this, and a corresponding decrease in hypertension–related morbidity and mortality in some populations,7–9 to more widespread use of antihypertensive drug therapy, particularly single–pill combination products that improve adherence, persistence, and adverse effect profiles, compared with antihypertensive
This manuscript was developed without financial support from any entity. Conflict of interest: none. *Corresponding author: William J. Elliott, MD, PhD, Professor of Preventive Medicine, Internal Medicine and Pharmacology, Head, Division of Pharmacology, Chair, Department of Biomedical Sciences, Pacific Northwest University of Health Sciences, 200 University Parkway, Yakima, WA 98901. Tel: 509-2497726; fax: 509-249-7799. E-mail: [email protected]
drug monotherapy.10 During the last decade, the US Food and Drug Administration has approved many single–pill antihypertensive drug combinations as initial therapy, usually because of a low probability of achieving an appropriate blood pressure target with monotherapy, particularly in individuals with Stage 2 hypertension.2,3 Two of the most widely appreciated ‘‘synergies’’ of single–pill combinations for hypertension are the offsetting effects of renin–angiotensin system blockers (angiotensin–converting enzyme [ACE]–inhibitors, angiotensin receptor blockers [ARBs], or renin inhibitors) and thiazide diuretics on serum potassium concentrations, and the arteriolar dilating properties of dihydropyridine calcium antagonists and venodilating properties of renin–angiotensin system blockers on peripheral edema.11 Recent data suggest that ACE–inhibitors are more effective than other renin–angiotensin system blockers for reducing the severity and incidence of amlodipine–associated peripheral edema.12 Reduction of adverse effects attributed to single–pill combinations has resulted in a large number of marketed agents of each type, many of which are now available in generic formulations, and some of which have been studied in prospective randomized clinical trials with cardiovascular and/or renal outcomes. Our purpose is
1933-1711/$ - see front matter Ó 2015 American Society of Hypertension. All rights reserved. http://dx.doi.org/10.1016/j.jash.2014.12.012
258
W.J. Elliott / Journal of the American Society of Hypertension 9(4) (2015) 257–265
to review current data regarding a new single–pill combination of perindopril þ amlodipine.
Methods Search Strategy A formal systematic review of the literature was performed 14 times from 1999 to 2013 to identify all clinical trials enrolling hypertensive patients, assigning different treatment strategies (at least one of which involved an antihypertensive drug), and reporting cardiovascular outcomes during a median follow–up of greater than 1 year. We searched PubMed, MEDLINE, the Cochrane Library, and OVIDWeb using text words and MeSH terms: ‘‘controlled clinical trials, cerebrovascular disorders, heart disease, antihypertensive agents, angiotensin–converting enzyme inhibitors, angiotensin II type 1 blockers, adrenergic alpha– antagonists, adrenergic beta–antagonists, diuretics, sodium chloride symport inhibitors, and calcium channel blockers.’’ These searches produced 189 citations that
were examined for trials that involved perindopril, 694 for amlodipine, and another 928 citations involving agents other than placebo. From these, 182 reports of randomized clinical trials were retrieved and assessed for mortality and/ or hospitalization outcomes. Additionally, six randomized clinical trials were added from the reference lists of prior meta–analyses and other, related publications.
Meta–analyses Outcomes data from all publications were doubleentered into a Microsoft Excel (Microsoft Corporation, Redmond, WA) spreadsheet and traditional fixed–effects (Mantel–Haenszel) (and sometimes random–effects) meta–analyses were performed using StatsDirect (Cheshire, UK). A separate computerized spreadsheet collected data to compare individual agents versus their comparators, which was used as input for network and Bayesian meta–analyses. Network meta–analyses were performed using a modification of the one–line program published by Lumley,13 using R (the programming language, framework 2.51 GUI 1.20)
Figure 1. Identification and selection of outcome–based randomized clinical trials (RCTs) performed with antihypertensive drugs.
W.J. Elliott / Journal of the American Society of Hypertension 9(4) (2015) 257–265
on an Apple MacBook Pro running MacOS X.4.8 (Apple Corporation, Cupertino, CA). Bayesian meta–analyses were performed using WinBUGS, version 1.4.1,14 with non–informative priors, a burn of 1000, and 99,000 iterations.15 Convergence was achieved before the first 2000 iterations. A continuity correction factor of 0.5 was used for all entries. All reported P values are two–sided.
Results The search strategy outlined above for outcome–based clinical trials (Figure 1) identified 11 studies involving 89,768 subjects randomized to either initial16–24 or add– on25–28 amlodipine (Table 1), and five studies involving 42,947 subjects randomized to initial29,30 or add–on20,31,32 perindopril (Table 2). These trials are quite heterogeneous, not only with respect to where amlodipine or perindopril was used in the sequence of randomized drug treatments
259
(initial,16–24,29,30 second–line,20,27,28,32 or even later25,26,31), but also with respect to comparators (placebo17,25,26,29–32 or active drug therapies16,18–28), concomitant medical conditions at baseline (coronary heart disease17 or not20; diabetes25,26,30 or not27,28; chronic kidney disease25–28 or not18–20,29–32; prior stroke29 or not20); average age of subjects at baseline (>5518 or >80 years32), duration of follow-up (1.832 to 5.5 years20), and primary outcomes (serum biochemistry values16 to coronary heart disease events18,20). Despite these major differences across trials, traditional fixed–effects meta–analyses showed no significant inhomogeneity for eight of the 12 cardiovascular endpoints (six for amlodipine, six for perindopril). Figure 2 summarizes the overall results of these meta–analyses of six outcomes for amlodipine versus any and all comparators. Heart failure, which demonstrated significant inhomogeneity (P < .0001) across the 11 amlodipine trials,
Table 1 Randomized clinical trials involving amlodipine that reported cardiovascular outcomes Trial Acronym, Year (Duration)
Randomized Therapies
# of Subjects
Deaths
CV Deaths
Major Adverse CV Events
Stroke
Coronary Heart Disease
Heart Failure
FACET,16 1998 (2.5 years)
Amlodipine Fosinopril Amlodipine Placebo Amlodipine Chlorthalidone Lisinopril Amlodipine Valsartan Amlodipine Atenolol Amlodipine Candesartan Amlodipine HCTZ Amlodipine Valsartan Amlodipine Valsartan Amlodipine Irbesartan Placebo Amlodipine Ramipril Metoprolol
191 189 417 408 9048 15,255 9054 7596 7649 9639 9618 2354 2349 5721 5741 511 510 575 575 567 579 569 217 436 441
5 4 6 8 1256 2203 1314 818 841 738 820 86 73 236 262 3 2 16 22 83 87 93 22 34 49
NR NR 2 7 603 996 618 NR NR 263 342 NR NR 107 134 NR NR NR NR 37 52 46 7 12 12
23 14 23 28 1374* 2329* 1421* 621* 720* 796 937 80* 88* 288 364 11* 11* 25* 26* 128 138 144 29 61 65
10 4 5 5 377 675 457 281 322 327 422 50 61 112 133 10 10 16 13 15 28 26 9 23 23
13 10 19 20 798 1362 796 313 369 429 474 33* 28* 125 159 1 2 7* 11* 27 44 46 5 19 18
0 0 1 5 706 870 612 400 354 134 159 16 20 100 96 1 3 15 3 93 60 72 8 20 22
PREVENT,17 2000 (3.0 years) ALLHAT,18 2002 (4.9 years)
VALUE,19 2004 (4.2 years) ASCOT,20 2005 (5.5 years) CASE-J,21 2007 (3.2 years) ACCOMPLISH,22 2008 (3.3 years) VART,23 2010 (3.4 years) Nagoya Heart Study,24 2012 (3.2 years) IDNT,25,26 2001, 2003 (2.6 years)
AASK,27,28 2006 (3.6 or 4.4 years)
AASK, African American Study of Kidney Diseases and Hypertension; ACCOMPLISH, Avoiding Cardiovascular events through COMbination therapy in People Living with Systolic Hypertension; ALLHAT, Antihypertensive and Lipid Lowering (to prevent) Heart Attack Trial; ASCOT, Anglo-Scandinavian Cardiac Outcomes Trial; CASE-J, Candesartan Antihypertensive Survival Evaluation in Japan; CV, cardiovascular; FACET, Fosinopril Amlodipine Cardiac Events Trial; HCTZ, hydrochlorothiazide; IDNT, Irbesartan Diabetic Nephropathy Trial; NR, not reported; PREVENT, Prospective Randomized Evaluation of the Vascular Effects of Norvasc Trial; VALUE, Valsartan Antihypertensive Long–term Use Evaluation; VART, Valsartan Amlodipine Randomized Trial. * Estimated, using linear regression.
260
W.J. Elliott / Journal of the American Society of Hypertension 9(4) (2015) 257–265
Table 2 Randomized clinical trials involving perindopril that reported cardiovascular outcomes Trial Acronym, Year (Duration) Randomized Therapies
Stroke Coronary Heart # of Deaths CV Major Failure Heart Subjects Deaths Adverse Disease CV Events
PROGRESS,29 2001 (3.9 years) Perindopril indapamide Placebo placebo EUROPA,31 2003 (4.2 years) Perindopril Placebo ASCOT,20 2005 (5.5 years) Perindopril (after amlodipine) Bendrofluazide (after atenolol) ADVANCE,30 2007 (4.3 years) Perindopril (þindapamide) Placebo HYVET,32 2008 (1.8 years) Perindopril (after indapamide) Placebo (after placebo)
3051 3054 6110 6108 9639 9618 5569 5571 1933 1912
306 319 375 420 738 820 408 471 196 235
181 198 215 249 263 342 211 257 99 121
458 604 552 664 796 937 480 520 138 193
307 420 98 102 327 422 215 218 51 69
115* 154* 320 418 429 474 265 294 9 12
75y 93y 63 103 134 159 197y 199y 22 57
ADVANCE, Action in Diabetes and Vascular disease: preterAx and diamicroN Controlled Evaluation; ASCOT, Anglo-Scandinavian Cardiac Outcomes Trial; CV, cardiovascular; EUROPA, European Reduction Of cardiac events with Perindopril in stable coronary Artery disease; HYVET, HYpertension in the Very Elderly Trial; PROGRESS, Perindopril pROtection aGainst REcurrent Stroke Study. * Estimated, using linear regression. y Hospitalization for heart failure.
was significantly increased in the groups receiving amlodipine (by 24%). Conversely, all five other cardiovascular endpoints were significantly reduced in the groups receiving amlodipine, with little evidence of inhomogeneity across trials (note that the numbers of cardiovascular
deaths were not reported for Fosinopril Amlodipine Cardiac Events Trial,16 Valsartan Antihypertensive Long– term Use Evaluation,19 CASE–J,21 Valsartan Amlodipine Randomized Trial,23 or the Nagoya Heart Study,24 which limits the statistical power of any meta–analysis of this
Figure 2. Results of fixed–effects meta–analysis of amlodipine versus other randomized therapies. Results of meta–analyses of clinical trials in hypertension comparing amlodipine as either initial16–24 or add-on25–28 therapy versus any and all other comparators are presented. All P–values for homogeneity [P (homo)] were >0.05 (except for heart failure [HF]) for the trials that randomized subjects to initial amlodipine therapy, but those for cardiovascular (CV) death, coronary heart disease (CHD), and major cardiovascular disease (CVD) events were all
