<<

A SYSTEMATIC REVIEW OF THE BLOOD PRESSURE LOWERING EFFICACY

OF BETA BLOCKERS FOR PRIMARY HYPERTENSION

by

WAN KI WONG

B.A., The Ohio State University, 2002

B.Sc., The Ohio State University, 2006

A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF

THE REQUIREMENTS FOR THE DEGREE OF

DOCTOR OF PHILOSOPHY

in

THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES

(Pharmacology and Therapeutics)

THE UNIVERSITY OF BRITISH COLUMBIA

(Vancouver)

February 2014

© Wan Ki Wong, 2014 ABSTRACT

Background: Beta blockers are a class of drug commonly prescribed for treatment of hypertension. Although the long-term goal of antihypertensive treatment is reduction of mortality and morbidity, the reduction of blood pressure (BP) is often used by clinicians to evaluate the effectiveness of treatment. Our goal was to quantify the BP lowering efficacy of all beta blockers. The magnitude of BP lowering efficacy of each provides valuable information to assist clinical decision making.

Methods: We searched the three major databases (MEDLINE, EMBASE and CENTRAL) and other sources for randomized, double blind, placebo controlled trials (DBRCT) of beta blocker mono therapy studies in primary hypertensive patients. Study duration had to be between 3 to 12 weeks. We used the RevMan program, version 5.2, to generate pooled BP, heart rate and pulse pressure estimates.

Results: Included in our review were 96 DBRCTs that examined the dose related BP lowering efficacy of 20 mono-therapy beta blockers in 9,803 primary hypertensive patients.

In patient with mild to severe hypertension, the estimates of BP lowering efficacy

(SBP/DBP) for beta blockers separated by subclasses were -10/-7 mmHg for non-selective beta blockers, -5/-4 mmHg for dual receptor blockers, -8/-4 mmHg for partial agonists and -

10/-8 mmHg for beta-1 blockers. Baseline BP and time of measurement (peak or trough) influenced the BP lowering effect. None of the beta blocker subclasses showed a convincing graded dose response BP lowering effect. Doses higher than twice the recommended starting dose provided little or no additional BP lowering effect. Beta blockers overall had little or no effect on pulse pressure. Beta blockers showed a graded dose response effect in reducing

ii heart rate and different beta blockers subclasses lowered heart rate by different amounts.

The review findings were limited by a high risk of loss of blinding bias and publication bias.

Conclusion: Different subclasses of beta blockers lowered BP and heart rate by different amounts. None of the beta blockers subclasses showed a graded dose response blood pressure lowering effect. Beta blockers overall have little or no effect on pulse pressure in comparison to other classes.

iii

PREFACE

Chapter 2 of this thesis has been published as part of methodology of four Cochrane protocols in the Cochrane library. The citations of the protocols are listed below:

 Wong GWK, Laugerotte A, Wright JM. Blood pressure lowering efficacy of non- selective beta blockers for primary hypertension (Protocol). Cochrane Database of Systematic Reviews 2008, Issue 4. Art. No.: CD007452. DOI: 10.1002/14651858.CD007452.

 Wong GWK, Laugerotte A, Wright JM. Blood pressure lowering efficacy of dual alpha and beta blockers for primary hypertension (Protocol). Cochrane Database of Systematic Reviews 2008, Issue 4. Art. No.: CD007449. DOI: 10.1002/14651858.CD007449.

 Wong GWK, Laugerotte A, Wright JM. Blood pressure lowering efficacy of partial agonist beta blockers for primary hypertension (Protocol). Cochrane Database of Systematic Reviews 2008, Issue 4. Art. No.: CD007450. DOI: 10.1002/14651858.CD007450.

 Wong GWK, Laugerotte A, Wright JM. Blood pressure lowering efficacy of beta-1 selective beta blockers for primary hypertension (Protocol). Cochrane Database of Systematic Reviews 2008, Issue 4. Art. No.: CD007451. DOI: 10.1002/14651858.CD007451.

The results in chapter 3, 4, 5 and 6 are in the process of being submitted for publication in the Cochrane library. Alexandra Laugerotte is the second reviewer for chapter

3 and 4. Dr. Heidi Boyda is the second reviewer for chapter 5 and 6. The duty of the second reviewer is confirming that the included studies meet the inclusion criteria and extracting the data independently. Dr. James M Wright formulated the idea for the thesis and developed the basis for the methodology.

No patient or animal contact occurred for the purpose of this thesis therefore ethical approval was not required.

iv

TABLE OF CONTENTS

ABSTRACT ...... ii

PREFACE ...... iv

TABLE OF CONTENTS ...... v

LIST OF TABLES ...... x

LIST OF FIGURES ...... xiii

LIST OF ABBREVIATIONS ...... xiv

ACKNOWLEDGEMENTS ...... xv

DEDICATION ...... xvi

1. INTRODUCTION ...... 1

1.1 Hypertension...... 1 1.2 Measuring the risk of cardiovascular events ...... 2 1.3 Risk reduction by antihypertensive treatments...... 3 1.4 Role of sympathetic nervous system in regulation of BP...... 4 1.5 Pharmacology of beta receptor antagonists (beta blockers) ...... 6 1.6 Clinical outcomes of beta blockers ...... 8 1.7 Importance of BP lowering information in clinical and policy decision making ..... 10 1.8 What is a systematic review? ...... 10 1.9 The Cochrane Collaboration and Cochrane library ...... 12 1.10 Importance of this review ...... 12

2. METHODS ...... 13

2.1 Objectives ...... 13 2.1.1 Primary objective ...... 13 2.1.2 Secondary objectives ...... 13 2.2 Search for studies ...... 13 2.3 Inclusion criteria ...... 14 2.3.1 Types of studies ...... 14 2.3.2 Types of participants...... 14 2.3.3 Types of interventions ...... 15 2.3.4 Types of outcome measures ...... 15 2.4 Data extraction...... 16 2.4.1 Measurement of blood pressure ...... 16 2.4.2 Handling missing data ...... 16 2.4.3 Assessment of risk of bias ...... 17 2.4.4 Data synthesis ...... 18 2.4.5 Direct and indirect comparison ...... 19

v

3. RESULTS FOR NON-SELECTIVE BETA BLOCKERS ...... 20

3.1 Search findings ...... 20 3.2 Characteristics of included studies ...... 22 3.3 Characteristics of excluded studies in non-selective beta blockers review ...... 22 3.4 Effects of on systolic blood pressure, diastolic blood pressure, heart rate and pulse pressure...... 23 3.5 Effects of on systolic blood pressure, diastolic blood pressure, heart rate and pulse pressure...... 27 3.6 Effects of on systolic blood pressure, diastolic blood pressure, heart rate and pulse pressure...... 29 3.7 Effects of on systolic blood pressure, diastolic blood pressure, heart rate and pulse pressure...... 29 3.8 Effect on heart rate of other included non-selective beta blockers ...... 30 3.9 Pooled subclass effects on systolic blood pressure, diastolic blood pressure, heart rate and pulse pressure...... 30 3.10 Subgroup analysis ...... 33 3.11 Discussion...... 33 3.11.1 Propranolol ...... 33 3.11.2 Heterogeneity in propranolol 2x starting dose subgroup ...... 34 3.11.3 Penbutolol ...... 36 3.11.4 Other included non-selective beta blockers ...... 36 3.11.5 Pooled BP lowering effects of non-selective beta blockers...... 37 3.11.6 Heart rate ...... 37 3.11.6 Pulse pressure ...... 37 3.11.7 Effect on blood pressure variability ...... 38 3.11.8 Subgroup analysis ...... 39 3.12 Risk of bias and quality of the evidence ...... 41 3.12.1 Randomization and allocation concealment (selection bias) ...... 41 3.12.2 Blinding (performance bias and detection bias) ...... 41 3.12.3 Incomplete outcome data (attrition bias) ...... 42 3.12.4 Selective reporting (reporting bias) ...... 42 3.12.5 Publication bias ...... 42 3.12.6 Quality of the evidence ...... 43

4. RESULTS FOR ALPHA AND BETA DUAL RECEPTOR BLOCKERS ...... 45

4.1 Search findings ...... 45 4.2 Characteristics of included studies ...... 47 4.3 Characteristics of excluded studies ...... 47 4.4 Effects of on systolic blood pressure, diastolic blood pressure, heart rate and pulse pressure...... 48 4.5 Effects of on systolic blood pressure, diastolic blood pressure, heart rate and pulse pressure...... 50 4.6 Pooled effects of dual receptor blockers ...... 51 4.6.1 Withdrawal due to adverse effects (WDAE) ...... 51 4.7 Subgroup and sensitivity analyses ...... 52

vi

4.7.1 Difference between published and unpublished data ...... 52 4.8 Discussion...... 52 4.8.1 Carvedilol ...... 52 4.8.2 Differences in BP lowering effect between carvedilol and labetalol ...... 53 4.8.3 Differences in published and unpublished data ...... 54 4.9 Risk of bias and quality of evidence...... 56 4.9.1 Randomization and allocation concealment (selection bias) ...... 56 4.9.2 Blinding (Performance and detection bias) ...... 57 4.9.3 Incomplete outcome data (attrition bias) ...... 57 4.9.4 Selective reporting (reporting bias) ...... 57 4.9.5 Publication bias ...... 58 4.9.6 Quality of the evidence ...... 58

5. RESULTS FOR PARTIAL AGONISTS ...... 61

5.1 Search Findings ...... 61 5.2 Characteristics of included studies ...... 63 5.3 Characteristics of excluded studies ...... 63 5.4 Effects of on systolic blood pressure, diastolic blood pressure, heart rate and pulse pressure...... 64 5.5 Effects of on systolic blood pressure, diastolic blood pressure, heart rate and pulse pressure...... 65 5.6 Effects of on systolic blood pressure, diastolic blood pressure, heart rate and pulse pressure...... 66 5.7 Effects of on systolic blood pressure, diastolic blood pressure, heart rate and pulse pressure...... 68 5.8 Effects of on systolic blood pressure, diastolic blood pressure, heart rate and pulse pressure...... 69 5.8.1 Withdrawal due to adverse effects ...... 70 5.9 Effects of on systolic blood pressure, diastolic blood pressure and pulse pressure ...... 71 5.10 Pooled effects of partial agonists on systolic blood pressure, diastolic blood pressure, heart rate and pulse pressure ...... 72 5.10.1 Blood pressure variability ...... 73 5.10.2 Subgroup and sensitivity analysis ...... 74 5.11 Discussion...... 74 5.11.1 Pooled subclass effect of partial agonists ...... 75 5.11.2 Heart rate ...... 75 5.11.3 Pulse pressure ...... 76 5.11.4 Blood pressure variability ...... 76 5.12 Risk of bias and quality of evidence...... 77 5.12.1 Allocation (selection bias) ...... 78 5.12.2 Blinding (performance bias and detection bias) ...... 78 5.12.3 Incomplete outcome data (attrition bias) ...... 78 5.12.4 Selective reporting (reporting bias) ...... 78 5.12.5 Publication bias ...... 79 5.12.6 Quality of the evidence ...... 79

vii

6. RESULTS FOR BETA-1 SELECTIVE BLOCKERS ...... 82

6.1 Search findings ...... 82 6.2 Characteristics of included studies ...... 84 6.3 Characteristics of excluded studies ...... 84 6.4 Effects of on systolic blood pressure, diastolic blood pressure, heart rate and pulse pressure...... 84 6.5 Effects of on systolic blood pressure, diastolic blood pressure, heart rate and pulse pressure...... 87 6.6 Effects of on systolic blood pressure, diastolic blood pressure, heart rate and pulse pressure...... 89 6.7 Effects of on systolic blood pressure, diastolic blood pressure, heart rate and pulse pressure...... 91 6.8 Effects of on systolic blood pressure, diastolic blood pressure, heart rate and pulse pressure...... 93 6.9 Effects of on systolic blood pressure, diastolic blood pressure, heart rate and pulse pressure...... 95 6.10 Effects of on systolic blood pressure, diastolic blood pressure, heart rate and pulse pressure...... 97 6.11 Effects of on systolic blood pressure, diastolic blood pressure, heart rate and pulse pressure...... 98 6.12 Pooled subclass effects on systolic blood pressure, diastolic blood pressure, heart rate and pulse pressure...... 98 6.12.1 Blood pressure variability ...... 100 6.12.2 Withdrawal due to adverse effects ...... 100 6.13 Discussion...... 101 6.13.1 Nebivolol ...... 101 6.13.2 Atenolol ...... 103 6.13.3 Metoprolol ...... 105 6.13.4 Bisoprolol ...... 105 6.13.5 Betaxolol, bevantolol, pafenolol and practolol ...... 106 6.13.6 Overall pooled blood pressure lowering effect of beta-1 blocker ...... 106 6.13.7 Pulse pressure ...... 108 6.14 Risk of bias assessments and quality of evidence ...... 111 6.14.1 Randomization and allocation concealment (selection bias) ...... 111 6.14.2 Blinding (performance bias and detection bias) ...... 111 6.14.3 Incomplete outcome data (attrition bias) ...... 112 6.14.4 Selective reporting (reporting bias) ...... 112 6.14.5 Publication bias ...... 112 6.14.6 Quality of evidence ...... 113

7. OVERALL DISCUSSION ...... 115

7.1 What is the effect on blood pressure and heart rate compared to placebo? ...... 115 7.1.1 Pulse pressure ...... 118 7.1.2 BP variability ...... 118 7.1.3 Peak and trough measurements ...... 119

viii

7.2 What is the mechanism by which beta blockers lower BP? ...... 120 7.3 Is there a difference in BP lowering efficacy between beta blocker subclasses? ... 122 7.4 Disagreement with other published reviews...... 124 7.5 Difference in BP lowering efficacy in beta-blockers compared to other classes? . 124 7.6 Overall quality of evidence ...... 127 7.6.1 Potential of bias in included studies ...... 128 7.6.2 Randomization (Selection bias) ...... 128 7.6.3 Blinding (performance and detection bias)...... 129 7.6.4 Reporting bias ...... 129 7.6.5 Publication bias ...... 130 7.7 Conclusion ...... 131

BIBLIOGRAPHY ...... 135

APPENDICES ...... 149

Appendix A: Search strategies ...... 149 Appendix B: Data extraction form ...... 155 Appendix C: Table of included studies in non-selective beta blocker review ...... 156 Appendix D: Table of included studies in dual receptor blocker review ...... 171 Appendix E: Table of included studies in partial agonist review ...... 176 Appendix F: Table of included studies in beta-1 selective blocker review ...... 182

ix

LIST OF TABLES

Table 1.1: Clinical outcomes of different classes of first-line antihypertensive drugs ...... 4

Table 1.2: Characteristics of adrenergic receptors ...... 5

Table 1.3: Pharmacokinetic properties of beta blockers ...... 7

Table 2.1: Classification of beta blockers ...... 15

Table 3.1: Characteristics of excluded studies in non-selective beta blockers review ...... 23

Table 3.2: Dose ranging BP, heart rate and pulse pressure lowering efficacy of propranolol ...... 24

Table 3.3: Dose ranging BP, heart rate and pulse pressure lowering efficacy of penbutolol ...... 28

Table 3.4: Dose ranging BP, heart rate and pulse pressure lowering efficacy of timolol..... 29

Table 3.5: Heart rate lowering effect of , and ...... 30

Table 3.6: Dose ranging BP, heart rate and pulse pressure lowering efficacy of non- selective beta blockers ...... 31

Table 3.7: WDAE of non-selective beta blockers ...... 32

Table 3.8: Summary of findings table for non selective beta blockers ...... 44

Table 4.1: Characteristics of excluded studies in dual receptor blockers review ...... 47

Table 4.2: Dose ranging BP, heart rate and pulse pressure lowering efficacy of carvedilol ...... 48

Table 4.3: Dose ranging BP, heart rate and pulse pressure lowering efficacy of labetalol ...... 50

Table 4.4: Summary of findings table for dual receptor blockers ...... 60

Table 5.1: Characteristics of excluded studies in partial agonists review ...... 64

Table 5.2: Dose ranging BP, heart rate and pulse pressure lowering efficacy of acebutolol ...... 65

Table 5.3: Dose ranging BP, heart rate and pulse pressure lowering efficacy of pindolol ... 66

Table 5.4: Dose ranging BP, heart rate and pulse pressure lowering efficacy of celiprolol ...... 67

x

Table 5.5: Dose ranging BP, heart rate and pulse pressure lowering efficacy of alprenolol ...... 68

Table 5.6: Dose ranging BP, heart rate and pulse pressure lowering efficacy of bopindolol ...... 70

Table 5.7: Dose ranging BP, heart rate and pulse pressure lowering efficacy of oxprenolol ...... 71

Table 5.8: Dose ranging BP, heart rate and pulse pressure lowering efficacy of partial agonists ...... 72

Table 5.9: Summary of findings table for partial agonists ...... 81

Table 6.1: Characteristics of excluded study in beta-1 selective blockers review ...... 84

Table 6.2: Dose ranging BP, heart rate and pulse pressure lowering efficacy of nebivolol ...... 85

Table 6.3: Dose ranging BP, heart rate and pulse pressure lowering efficacy of atenolol ... 88

Table 6.4: Dose ranging BP, heart rate and pulse pressure lowering efficacy of metoprolol ...... 90

Table 6.5: Dose ranging BP, heart rate and pulse pressure lowering efficacy of bisoprolol ...... 92

Table 6.6: Dose ranging BP, heart rate and pulse pressure lowering efficacy of betaxolol...... 94

Table 6.7: Dose ranging BP, heart rate and pulse pressure lowering efficacy of bevantolol...... 96

Table 6.8: Dose ranging BP, heart rate and pulse pressure lowering efficacy of pafenolol ...... 97

Table 6.9: Dose ranging BP, heart rate and pulse pressure lowering efficacy of practolol ...... 98

Table 6.10: Dose ranging BP, heart rate and pulse pressure lowering efficacy of beta-1 blockers ...... 99

Table 6.11: BP lowering efficacy of 2x starting beta-1 blockers ...... 107

Table 6.12: Summary of findings table for beta-1 blockers ...... 114

Table 7.1: The effect of varying doses of different subclasses of beta blockers on systolic blood pressure, diastolic blood pressure and heart rate ...... 121

xi

Table 7.2: Overall subclasses baseline characteristics and time of measurements ...... 122

Table 7.3: List of mean estimates of all four subclasses ...... 123

Table 7.4: BP lowering efficacy of other classes of antihypertensive drugs ...... 125

xii

LIST OF FIGURES

Figure 3.1: Non-selective beta blockers review PRISMA flow diagram ...... 21

Figure 3.2: Funnel plot: Systolic blood pressure at 2x the starting dose for propranolol...... 26

Figure 3.3: Funnel plot: Diastolic blood pressure at 2x the starting dose for propranolol ..... 27

Figure 3.4: Risk of bias summary of non-selective beta blockers ...... 40

Figure 4.1: PRISMA diagram of dual receptor blockers review ...... 46

Figure 4.2: Risk of bias summary of dual receptor blockers...... 56

Figure 5.1: PRISMA diagram of partial agonists review ...... 62

Figure 5.2: Risk of bias summary of partial agonists ...... 77

Figure 6.1: PRISMA diagram of beta-1 selective blocker ...... 83

Figure 6.2: Funnel plot: Systolic blood pressure at 1x, 2x and 4x starting dose for nebivolol ...... 102

Figure 6.3: Funnel plot: Diastolic blood pressure at 1x, 2x and 4x starting dose for nebivolol ...... 102

Figure 6.4: Funnel plot: Systolic blood pressure at starting dose for atenolol ...... 104

Figure 6.5: Funnel plot: Diastolic blood pressure at starting dose for atenolol ...... 104

Figure 6.6: Risk of bias summary of beta-1 blockers...... 110

xiii

LIST OF ABBREVIATIONS

AC Adenylate cyclase

ACE inhibitor Angiotensin converting enzyme inhibitors

ARB Angiotensin receptor blockers cAMP Cyclic adenosine monophosphate

CCB Calcium channel blockers

CHD Coronary heart disease

CVE Cardiovascular events

DBP Diastolic blood pressure

HF Heart failure

HR Heart rate

IP3 Inositol triphosphate

MAO Monoamine oxidase

MI Myocardial infarction

PP Pulse pressure

RR Relative risk

SBP Systolic blood pressure

SD Standard deviation

SEM Standard error of mean

Tmax Time to maximum serum concentration t½ Half life

WHO World Health Organization

WDAE Withdrawal due to adverse effects

xiv

ACKNOWLEDGEMENTS

It would not be possible for me to complete my Ph.D. program if I had not received help from the people of the Therapeutics initiative and Cochrane Hypertension group.

First, I would like to thank my supervisor, Dr. James Wright, for his guidance, support and encouragement. I would not be admitted to the program if Dr. Wright had not agreed to be my mentor and provide financial support. He is always available for appointment and questions. He provided inspirations and directions that helped me through this journey, as well as spent countless hours of effort in reading and editing my thesis. He taught me the skills and the integrity needed to become a scientist.

I also wish to thank Dr. Ken Bassett and Dr. Vijaya Musini for teaching me how to conduct a systematic review. The insightful skill that they taught me is greatly helpful in the research of my dissertation. Dr. Ken Bassett also provided guidance and support as a member of my supervisory committee. I also deeply appreciated my other committee members, Dr.

Barbara Mintzes and Dr. Colin Dormuth for their support over all these years.

I would like to thank Steven Adams for retrieving hundreds of citations in such a timely manner. He is one of the best librarians I have ever met and his help has greatly reduced the time required to obtain the articles.

I would also like to thank Ciprian Jauca, Chris Adlparvar and Doug Salzwedel from the Cochrane Hypertension group for technical support and Wynne Leung from the department of Pharmacology and Therapeutics for helping me navigate through all the rules and regulations.

Last but not the least, I would like to thank my parents for their unconditional love and support through the years.

xv

DEDICATION

I dedicate this thesis to my parents

Ma Li Wong and Tak Yan Wong whose love, guidance, support and unlimited patience have helped me through many obstacles in life.

xvi

1. INTRODUCTION

1.1 Hypertension

Hypertension, commonly known as elevated blood pressure, is a highly prevalent health problem in the world. It is a major risk factor for cardiovascular diseases such as stroke, coronary heart disease and peripheral vascular disease. According to the World

Health Organization’s (WHO) report in 2012, over 25% of adults in the world are reported to have hypertension [1]. The WHO report estimated that hypertension contributed to 51% of deaths by stroke and 45% of deaths by coronary heart disease worldwide [1]. In Canada, the number of people diagnosed with hypertension has continued to rise since 2001. According to Statistics Canada, 17.1% of Canadians aged 12 and older were reported to be diagnosed with hypertension in 2011 [2]. This condition puts a heavy burden on our already stretched health care resources and it is likely to increase in the foreseeable future.

According to Hypertension Canada guidelines, in order to confirm the diagnosis of hypertension, patients must have at least 3 blood pressure readings that are elevated during at least two separate office visits [3]. However, there is no clear cut off between normal and high blood pressure. At the present time hypertension is arbitrarily defined as systolic blood pressure (SBP) higher than 140 mmHg and/or diastolic blood pressure (DBP) higher than 90 mmHg. This surrogate threshold is commonly accepted as the diagnostic criterion for hypertension around the world. However, since it is an arbitrary standard, it is subject to change and has progressively decreased over the last 50 years. It is thus likely to change in the future.

1

1.2 Measuring the risk of cardiovascular events

Blood pressure (BP) measurement is commonly used in clinical practice to make the diagnosis of hypertension and to document the effect of BP lowering drugs. Blood pressure is also used as a surrogate for risk of cardiovascular events (CVE). Early observational data initially associated the risk for stroke and coronary heart disease (CHD) with elevated diastolic blood pressure [4]. More recent data, such as the Framingham Heart Study, indicates that systolic blood pressure and pulse pressure (PP) are comparable predictors for the risk of CHD in patients aged 50 to 59 while systolic blood pressure is a better predictor of the risk in older patients [5]. In addition, pulse pressure, the difference between systolic blood pressure and diastolic blood pressure, has been receiving more attention through the last decade as an independent risk factor for CHD and heart failure (HF) [6].

Blood pressure fluctuates greatly between and within individuals. Factors that can influence blood pressure include psychological state, physical activity and circadian rhythm.

Recently blood pressure variability has been receiving more attention for its predictive value of adverse cardiovascular events. The Cardiovascular Health Study, followed 1,642 elderly patients for 10 years, and found that greater systolic blood pressure variability was significantly associated with increased risk of mortality and incidence of myocardial infarction (MI) [7]. A few years ago, two case control studies, including several hundred elderly hypertensive patients, found that the increase in both systolic blood pressure and diastolic blood pressure variability was associated with an increase in cardiovascular events

[8, 9]. The association of these parameters to the risk of cardiovascular events should be carefully considered when treating hypertension because it is important to remember that the

2 goal for treating hypertension is to reduce adverse cardiovascular events and not simply the reduction of blood pressure.

1.3 Risk reduction by antihypertensive treatments

There are many different classes of antihypertensive drugs. The more commonly used classes include Angiotensin Converting Enzyme inhibitors (ACE inhibitors),

Angiotensin receptor blockers (ARB), thiazide diuretics, calcium channel blockers (CCB) and beta blockers. Treating patients age 60 and older with moderate to severe hypertension with various different antihypertensive drugs has been shown to reduce mortality and total cardiovascular events [11]. However, the mortality benefit was limited to patients between

60 to 80 years old. It is not clear whether the clinical benefits of lowering elevated blood pressure is the same for all hypertensive patients.

Furthermore, it is possible that not all classes of antihypertensive drugs produce the same clinical benefits. A Cochrane review that compared the benefits of different classes of first-line antihypertensive drugs with placebo or no treatment for mortality and morbidity, found differential effects for different drug classes (table 1.1) [10]. First-line beta blockers stand out as probably having a less beneficial effect than the other 3 classes. At the present time the reason for this differential effect is not known. It may have something to do with the fact that beta blockers lower BP by different mechanism of action than the other drug classes.

3

Table 1.1: Clinical outcomes of different classes of first-line antihypertensive drugs

All cause Risk for Stroke Risk for CHD Risk Total CVE Class mortality RR (95% CI) RR (95% CI) RR (95% CI) RR (95% CI) Low dose thiazide 0.89 (0.83, 0.96) 0.63 (0.57, 0.71) 0.84 (0.75, 0.95) 0.70 (0.66, 0.76) ACE inhibitors 0.83 (0.72-0.95) 0.65 (0.52-0.82) 0.81 (0.70-0.94) 0.76 (0.67-0.85) CCB 0.86 (0.68, 1.09) 0.58 (0.41, 0.84) 0.77 (0.55, 1.09) 0.71 (0.57, 0.87) Beta blockers 0.96 (0.86, 1.07) 0.83 (0.72, 0.97) 0.90 (0.78, 1.03) 0.89 (0.81, 0.98) Thiazide diuretics provide significant benefit in relative risk (RR) of all cause mortality and cardiovascular morbidity. The data for ACE inhibitors are less robust but show similar point estimates to thiazides. Neither calcium channel blockers (CCB) nor beta blockers provide significant benefit for all cause mortality or reduction of coronary heart disease (CHD). However, calcium channel blockers have a smaller sample size and wider 95% confidence intervals compared to beta blockers. The point estimates for calcium channel blockers are also closer to that of thiazides and ACE inhibitors than beta blockers.

1.4 Role of sympathetic nervous system in regulation of BP

The sympathetic nervous system regulates many vital physiological functions. The sympathetic nervous system acts on adrenergic receptors in the regulation of cardiovascular functions. Adrenergic receptors are present in many body systems, including the heart, blood vessels and the kidney. They play an important role in regulation of blood pressure.

The sympathetic nervous system uses chemical messengers called catecholamines.

Norepinephrine and epinephrine are the two major types of catecholamines for sympathetic regulation of human cardiovascular function. They are produced by sympathetic neurons and adrenal medulla. Sympathetic neurons and chromaffin cells in the adrenal medulla synthesize catecholamines from . In sympathetic neurons, is synthesized in the cell body and transported in vesicles to the terminal. Norepinephrine released from nerve terminals acts primarily on post-synaptic nerve terminal receptors. In the adrenal medulla, norepinephrine is mostly converted to epinephrine by phenylethanolamine-N- methyltransferase outside of the chromaffin cells before reentering the chromaffin cells for

4 storage. Eighty percent of the catecholamine stored in the adrenal medulla is epinephrine.

Epinephrine is released into the blood stream and acts throughout the body. When released, norepinephrine and epinephrine target adrenergic receptors. Norepinephrine is inactivated by reuptake into the nerve terminals for storage and reuse or metabolized by monoamine oxidase (MAO) or catechol-O-methyl transferase (COMT). MAO metabolizes norepinephrine and epinephrine into vanillylmandelic acid, which is excreted in the urine

[13]. The various adrenergic receptors and their functions are listed below.

It is essential to understand the receptors that the catecholamines target in order to understand the wide variety of physiological functions which these two types of catecholamines regulate [13]. Table 1.2 lists the different subtypes of adrenergic receptors and their major physiological function [13]

Table 1.2: Characteristics of adrenergic receptors

Subtype Pathways Effect on organ or system Activate IP / Alpha-1 3 of large resistance blood vessels Ca2+ pathway  Negative feedback to sympathetic neurons Alpha-2 Inhibit cAMP  Vasoconstriction of small pre-capillary blood vessels Beta-1 Activate AC Increase heart rate and cardiac output  Relax peripheral smooth muscle in small arteries Beta-2 Activate AC  Release renin from the kidney AC - adenylate cyclase cAMP – cyclic adenosine monophosphate IP3 - inositol triphosphate

Currently two main types of adrenergic receptors (alpha and beta) and five subtypes have been identified. Norepinephrine has a greater affinity to alpha receptors, while epinephrine has a greater affinity to beta receptors [12]. All adrenergic receptors are coupled with G protein. However, the signaling pathway is different for alpha receptors than beta receptors [12, 13]. Activation of beta receptors activates adenylate cyclase (AC) and

5 increases production of cyclic adenosine monophosphate (cAMP) that lead to increased contractility and frequency in cardiac cells, through activation of L-type calcium channels.

On the other hand, beta-2 activation relaxes smooth muscle. The reason for the opposite effect is due to the inhibition of myosin light chain kinase in smooth muscle cells by cAMP

[32].

Alpha-1 receptors are predominantly present in smooth muscle cells. Activation of alpha -1 receptor activates the formation of inositol triphosphate (IP3) which then stimulates the release of calcium from sarcoplasmic reticulum and contributes to smooth muscle contraction [32]. Alpha-2 receptors predominantly act as a negative feedback mechanism of sympathetic nerve cells. Activation of alpha-2 receptors inhibits the production of AC and reduces release of norepinephrine. Alpha -2 receptors also regulate smooth muscle contraction in small pre-capillary blood vessels [32].

1.5 Pharmacology of beta antagonists (beta blockers)

Beta blockers are designed to competitively inhibit beta adrenergic receptors and modulate the sympathetic nervous system. Although they target the same set of receptors, not all beta blockers act identically to one another. Differences in pharmacokinetic properties within the class (see Table 1.3) could contribute to significant differences in clinical outcomes.

6

Table 1.3: Pharmacokinetic properties of beta blockers [13, 26, 27].

Protein Year first Bioavailability T Elimination Lipid Beta blocker binding max available in (%) (hour) t (hour) solubility (%) ½ the U.S. Non-selective beta blockers Propranolol 25 90 1-4 3-6 High 1967 Penbutolol 100 95 n/a 28 High 1987 Nadolol 30 30 3-4 20-24 Low 1978 Timolol 75 80 2 4.5 Moderate 1978 Alpha and beta dual receptor blockers Carvedilol 25 98 1 6 Moderate 1995 Labetalol 100 50 1-2 6-8 Low N/A Partial agonists Acebutolol 20-60 20 2-4 7 Low 1984 Alprenolol 10 85 n/a 2-3 n/a N/A Pindolol 100 40 1 3-4 Low 1982 Oxprenolol 25-60 92 3 2 n/a N/A Beta-1 selective blockers Atenolol 50 3 3 6-7 Low 1981 Betaxolol 89 50 3 14-22 Moderate 1985 Bisoprolol 90 30 2-4 10-12 Low 1992 Metoprolol 50 10 1-2 3.5 Moderate 1978 Nebivolol n/a 98 1.5 10 Low 2007

The mechanism by which beta blockers lower blood pressure in man is not known.

Beta blockers could lower blood pressure in several ways [22]. Firstly, blocking beta-1 receptors lowers heart rate and contractility, which would lower cardiac output potentially lowering blood pressure but also triggering compensatory processes. Secondly, blocking beta-2 receptors in the kidney inhibits renin production that would lower blood pressure by reducing the effects of angiotensin II to increase blood pressure. Thirdly beta blockers could also decrease sympathetic nervous system outflow in the brain. Finally beta blockers could

7 lower blood pressure by a presently unknown mechanism or by a combination of these effects.

1.6 Clinical outcomes of beta blockers

Beta adrenergic receptor antagonists, commonly called beta blockers, are a group of drugs that target adrenergic receptors. All of them block beta receptors, but some also block alpha receptors. The first beta blocker that was marketed, propranolol, was invented by

British pharmacologist Sir James Black in the early 1960s. He received the Nobel Prize in medicine in 1988 partly due to his invention of propranolol [14]. Propranolol was originally used to treat angina. During the treatment for angina, physicians discovered that beta blockers also lowered blood pressure [15, 16]. Since then, beta blockers have been shown to reduce mortality when given to patients with a recent myocardial infarction (MI) [17] (Odds ratio 0.81 [95% CI 0.75 to 0.87]). In addition, beta blockers have been shown to reduce mortality in heart failure patients [18] (Odds ratio 0.65 [95% CI 0.53 to 0.8]). However, a meta-analysis found that this reduction in mortality is predominantly in patients with moderate to severe heart failure [19].

In the early 1980s, beta blockers became the first-line drug to lower BP in patients with elevated BP. However, at that time the scientific evidence supporting their use for hypertension was weak. In 2000, a systematic review examining the mortality and morbidity outcomes of beta blockers in post-MI patients, found that non-selective beta blockers significantly reduced total mortality compared to placebo, whereas beta-1 selective blockers and partial agonists did not [20]. This review suggested that clinical outcomes might be different among beta blocker subclasses and thus pooling them in meta-analyses might not be appropriate.

8

A Cochrane review comparing the effectiveness and safety of first-line beta blocker therapy with placebo or no treatment for primary hypertension, concluded that beta blockers significantly reduced the incidence of stroke but not total mortality or CHD. They concluded that a beta blocker was not the best first-line drug for treatment of hypertension [21]. A more recent Cochrane review examining the effectiveness and safety of first-line antihypertensive drug classes came to a similar conclusion for beta blockers [10].

Three systematic reviews have assessed the blood pressure lowering effect of beta blockers. Magee and colleagues examined the effectiveness of beta blocker therapy in hypertension during pregnancy [23]. They found that beta blocker therapy decreased the incidence of severe hypertension and the need for additional antihypertensive therapy.

However, there was insufficient evidence to make conclusions on perinatal mortality and preterm birth.

Chen and colleagues examined the blood pressure lowering efficacy of beta blockers as second-line therapy [24]. Beta blockers as a whole class significantly lowered both systolic blood pressure and diastolic blood pressure when used as second-line therapy. The mean estimates of the effect size was -6/-4 mmHg for recommended starting doses and -8/-6 mmHg for twice the starting doses. Law and colleagues examined the blood pressure lowering efficacy of antihypertensive mono-therapy of various antihypertensive drug classes

[25]. They reported that beta blockers as a whole class significantly lowered blood pressure by an average of -8.4/-6.9 mmHg. Both of these reviews did not assess or report on the different subclasses of beta blockers separately. However, as mentioned above, different subclasses of beta blockers could have significantly different effects [20].

9

1.7 Importance of BP lowering information in clinical and policy decision making

Since mortality and morbidity data are not available for most individual beta blockers, these drugs are currently marketed and used in patients with elevated BP if they have been shown to lower BP significantly as compared to a placebo. At the present time regulators, governmental and private insurers, clinicians and patients are operating under the assumption that all beta blockers lower BP by the same amount. There is no systematic review evidence to support whether that is true. If this assumption is proven to not be true, it could lead to changes in funder and/or clinician decisions with regard to beta blockers.

Furthermore, since it is probable that beta-blockers with different mechanisms of action have different effects on morbidity and mortality, it is crucial to determine whether they lower blood pressure parameters by different amounts.

No published review or meta-analysis has provided the dose related BP lowering efficacy of individual beta blockers or subclasses of beta blockers. This review would be the first to provide this information to support decision making of the use of beta blockers to lower BP in patients with elevated BP.

1.8 What is a systematic review?

Current medical science is advancing at such a fast pace that it is difficult for medical professionals to keep up with the massive amount of information. Therefore, a scientific method is needed to analyze, summarize and disseminate the most up to date evidence for medical professionals. Systematic review is the tool designed for this purpose.

A systematic review is fundamentally different from a narrative review. A narrative review is a presentation of expert opinion on a particular topic or update on a current research field. The information presented in a narrative review is generally that which

10 supports the view of the expert. Information presented in a narrative review usually focuses on a part of the whole picture. Therefore, the summary is often incomplete and biased towards a subset of data. Narrative reviews are unscientific due to a lack of transparency of the methods and process used, making it impossible for them to be replicated.

Compared to a narrative review, a systematic review addresses a specific research question. The parameter of the research question is predefined using the PICOS format.

PICOS stands for population, intervention, comparator, outcomes and study design. These are the 5 components of the research question that determine the inclusion criteria. After that the protocol and search strategy are developed in order to search for all available articles relevant to the research objectives. During the protocol stage, the authors also specify details of the outcomes of interest and troubleshooting strategies such as methods to deal with missing data or heterogeneity.

A systematic review can be both qualitative and quantitative. Quantitative analysis employs statistical methods, such as meta-analysis, to combine large amounts of data into pooled estimates in order to increase precision and statistical power. Quality assessment employs tools, such as the risk of bias tool to critically appraise the body of data. Each category of assessment is clearly documented. These tools allow researchers to identify the strengths and weaknesses of included studies and assess overall quality of evidence [29].

A systematic review can take from a few months to few years to complete. It is a time consuming and rigorous research process. At the end, it provides the best available evidence to answer a specific research question in the most precise and unbiased way.

11

1.9 The Cochrane Collaboration and Cochrane library

The Cochrane Collaboration was founded in 1993 with a vision to provide high quality, timely research evidence for healthcare decision making [30]. With 53 review groups specialized in various clinical areas, it maintains and disseminates systematic reviews in the Cochrane library. The Cochrane library was ranked number 11 in impact factor among

151 general medical journals in 2011[31]. It is an electronic publication containing several thousands of systematic reviews. Systematic reviews published in the Cochrane library must meet the high standard for quality set by the Cochrane Collaboration. In addition to the high standard of quality, in order to maintain relevancy of the evidence, Cochrane reviews are required to be updated every two years. These requirements make Cochrane reviews and

Cochrane methodology the gold standard in conducting systematic reviews.

1.10 Importance of this review

Beta blockers have been used to treat hypertension for decades. However, evidence supporting their use is still fragmented and weak. Mortality and morbidity evidence is limited to only a small number of beta blockers. Physicians and patients have to mostly rely on the blood pressure lowering efficacy of beta blockers to make healthcare decisions.

However, BP lowering efficacy information specific to each beta blockers has not been published and is not available. Physicians are thus prescribing some beta blockers without knowing the evidence of their effect on morbidity and mortality or even knowing the average magnitude of the blood pressure lowering effects. A systematic review is needed to provide this vital information. It is also important that this review be done as a Cochrane review, employing the rigorous Cochrane methodology and being updated regularly when new evidence becomes available.

12

2. METHODS

The methodology of this thesis was developed in order to meet the scientific requirement of Cochrane systematic reviews according to the Cochrane handbook [29]. The

4 protocols for the 4 reviews were approved by the Cochrane Hypertension group and published in the Cochrane library [164, 165, 166, 167]. The methods for each protocol were the same and the common methodology is provided below.

2.1 Objectives

2.1.1 Primary objective

To quantify the dose-related effects of various doses and types of beta adrenergic receptor blockers on systolic and diastolic blood pressure compared to placebo in patients with primary hypertension.

2.1.2 Secondary objectives

1. To determine the effects of beta adrenergic receptor blockers on variability of blood

pressure.

2. To determine the effects of beta adrenergic receptor blockers on pulse pressure.

3. To quantify the dose-related effects of beta adrenergic receptor blockers on heart rate.

4. To quantify the effects of beta adrenergic receptor blockers in different doses on

withdrawals due to adverse events.

2.2 Search for studies

The trial search coordinator from the Cochrane Hypertension group developed a comprehensive search strategy to help search the databases (Appendix). We searched

MEDLINE, EMBASE and Cochrane clinical trial register (CENTRAL) up to September

2012 to identify studies that meet our inclusion criteria. We also cross referenced with

13 previously published meta-analyses for studies which had not been indexed. The initial screen of these abstracts excluded articles which were clearly irrelevant. The full text of the remaining articles were retrieved and translated into English where required. Two independent reviewers assessed the eligibility of the trials using a trial selection form

(Appendix B). If there was disagreement between the two reviewers, a third reviewer resolved discrepancies.

2.3 Inclusion criteria

2.3.1 Types of studies

Studies must meet the following criteria in order to be included:

 Placebo-controlled;

 Random allocation to beta adrenergic receptor blocker fixed dose mono-therapy and

placebo group;

 Parallel or crossover design;

 Double blinded;

 Duration of follow-up of at least three weeks;

 Blood pressure measurements at baseline (following washout) and at one or more

time points between 3 to 12 weeks after starting treatment.

2.3.2 Types of participants

Participants must be men or non-pregnant women age 18 or older with a baseline blood pressure of at least 140 mmHg systolic and/or a diastolic blood pressure of at least 90 mmHg, measured in a standard way. Participants with significant renal impairment, who have creatinine levels greater than 1.5 times the normal level, were excluded.

14

2.3.3 Types of interventions

The active treatment group must be beta blocker fixed dose mono-therapy. Studies that involve titration based on BP response were excluded. Forced titration studies with participants taking the final fixed dose for 3 to 12 weeks were eligible.

We divided beta blockers into four subclasses (Table 2.1)

Table 2.1: Classification of beta blockers

Subclasses Drugs included in the subclass , , , , , , , , , , , , , , Non-selective beta indenolol, , , , moprolol, nadolol, blockers , , penbutolol, , propranolol, , , , tertatolol, , timolol, and xibenolol. Alpha and beta dual receptor carvedilol, dilevalol, labetalol blockers Partial agonists acebutolol, celiprolol, oxprenolol, pindolol, alprenolol, bopindolol Beta-1 selective atenolol, betaxolol, bevantolol, bisoprolol, esmolol, metoprolol, nebivolol , blockers pafenolol, practolol

2.3.4 Types of outcome measures

Primary outcome

Change in trough (13 to 26 hours after the dose) and/or peak (1 to 12 hours after the dose) systolic and diastolic blood pressure compared to placebo. If blood pressure measurements were available at more than one time within the acceptable trial duration window (3-12 weeks), the weighted means of the multiple blood pressures was used.

Secondary outcomes

 Change in standard deviation compared to placebo.

 Change in pulse pressure compared to placebo.

 Change in heart rate compared to placebo.

15

 Number of patients who withdraw due to adverse effects compared to placebo.

2.4 Data extraction

Data was extracted independently by two reviewers using a standard form, and then cross-checked. All numeric calculations and graphic interpolations were confirmed by a second person.

2.4.1 Measurement of blood pressure

The position of the patient during blood pressure measurement may affect the blood pressure lowering effect. When blood pressure measurement data was available in more than one position, sitting blood pressure was the first preference. If both standing and supine were available, standing blood pressure was used. In order to not lose valuable data, if data from only one position was reported, data from that position was used.

2.4.2 Handling missing data

In case of missing information in the included studies, investigators were contacted

(using email, letter and/or fax) to obtain the missing information.

In the case of missing standard deviation of the change in blood pressure, the standard deviation was imputed based on the information in the same trial or from other trials using the same drug. The following hierarchy (listed from high to low preference) was used to impute standard deviation values:

1. standard deviation of change in blood pressure taken in a different position than that

of the blood pressure data used

2. standard deviation of blood pressure at the end of treatment

3. standard deviation of blood pressure at the end of treatment measured in a different

position than that of the blood pressure data used

16

4. standard deviation of blood pressure at baseline (except if this measure is used for

entry criteria)

5. mean standard deviation of change in blood pressure from other trials using the same

drug

The standard deviation of change was calculated when necessary from the standard deviation of end treatment values using the formula recommended by the Cochrane handbook [29].

In the equation, SDΔ is the standard deviation of change, SDe and SDc are the standard deviation of experiment and control group respectively and Corr is the correlation coefficient of the two variances. In measurement of blood pressure, the correlation coefficient is 0.5 [34, 35]. This is the correlation between the standard deviation of treatment group and the standard deviation of placebo group in cross-over trials. Because the same individual is being used in the treatment and placebo groups, the two groups are not completely independent and have a weak correlation. This formula was used for cross-over trials which reported the end treatment standard deviations of the treatment and placebo phase and did not report the standard deviation of the difference between the two phases.

2.4.3 Assessment of risk of bias

The risk of bias of the included studies was assessed by using the Cochrane risk of bias assessment tool in the Review Manager 5.2 program. The risk of bias assessment tool used four major categories of bias to assess the quality of included studies. The four categories of bias were selection bias, performance and detection bias, attrition bias and

17 reporting bias. We also used funnel plots to assess publication bias when number of included studies was sufficient.

2.4.4 Data synthesis

Data synthesis and analyses were done using the Review Manager 5.2 software. Data for changes in blood pressure and heart rate were combined using a generic inverse variance method or inverse variance for mean difference when appropriate. Withdrawals due to adverse effects were analyzed using relative risk, risk difference, and number needed to harm.

Test for heterogeneity of treatment effect between the trials were made using a standard chi-square statistic and I-square statistics for heterogeneity. A p-value of 0.05 or less of chi-square statistic would suggest significant heterogeneity in the group. The fixed effects model was applied to obtain summary statistics of pooled trials, unless significant between-study heterogeneity was present, in which case the random effects model was used.

A random effect model weights the studies more equally than the fixed effect model.

If possible, subgroup analyses would include:

1. Different regimens of the same active chemical entity.

2. Gender, Age and Race.

3. Co-morbid conditions: Ischemic heart disease, peripheral vascular disease, diabetes.

4. Baseline severity of hypertension: mild, moderate, severe.

The robustness of the results was tested using several sensitivity analyses, including:

1. Trials that are industry-sponsored versus non-industry sponsored.

2. Trials with blood pressure data measured in the sitting position vs. other measurement

positions.

18

3. Trials with reported standard deviations of blood pressure change vs. imputed

standard deviations.

2.4.5 Direct and indirect comparison

We compared the dose effects by direct or indirect comparison. When possible, the direct comparison method was preferred. In direct comparison analysis, only studies that contained different dosage groups would be meta-analyzed and the subgroups were compared directly to each other. We tested the subgroup difference using Chi2 and I2 statistics. When direct comparison was not possible, we pooled studies with the same dosage into the same subgroups and compare the difference between subgroups using the statistical method of indirect comparisons. [29]

19

3. RESULTS FOR NON-SELECTIVE BETA BLOCKERS

3.1 Search findings

All four subclasses used the same study inclusion criteria [164, 165, 166, 167]. In order to save time and effort, a comprehensive search strategy (appendix A) was developed so that all four subclasses of beta blockers were searched simultaneously. Citations were then sorted according to their respective subclasses afterward. The search was first run in May

2010, and has been updated twice since, in August 2011 and September 2012. In May 2010, the search strategy identified a total of 18,604 citations from MEDLINE, EMBASE and

CENTRAL. In Aug 2011, we identified an additional 2,462 citations in our updated search.

In Sep 2012, we identified 248 new citations in our updated search. A total of 21,314 citations were identified in all three searches since May 2010, of which 8,353 were confirmed to be duplicates. The reviewers then screened 12,961 titles and abstracts, of which 12,366 citations were excluded. Five hundred ninety five citations were judged to potentially meet the inclusion criteria based on title and abstract. These were retrieved for detailed review. Four hundred eighty full text articles did not meet our inclusion criteria and were excluded. One citation (Lepantalo 1983) was a separate publication of the same data as

Lepantalo 1985. One hundred and fourteen trials met our inclusion criteria but 18 of them were excluded for reasons listed in the characteristics of excluded studies. Ninety Six trials were included in all 4 reviews. Twenty-five studies examining BP lowering efficacy of 7 non-selective beta blockers in primary hypertensive patients were included in this review.

Please refer to Figure 3.1 for the PRISMA flow diagram.

20

Figure 3.1: Non-selective beta blockers review PRISMA flow diagram

21

3.2 Characteristics of included studies

This review comprised 25 RCTs with 1,279 participants (parallel studies, 4 and crossover studies, 21). The duration of treatment ranged between 3 to 12 weeks, with the majority of trials lasting for 4 weeks. All studies were prospective, randomized, double blinded, placebo controlled fixed dose trials. Propranolol was the most studied non-selective beta blocker with 18 trials in doses ranging from 60 to 640 mg/day. Patients enrolled into these studies were mostly middle aged with mild to moderate hypertension. The mean baseline BP of the included studies was 158/104 mmHg. Characteristics of individual studies are listed in appendix C.

3.3 Characteristics of excluded studies in non-selective beta blockers review

Ten studies that potentially meet the inclusion criteria were excluded from the non- selective beta blocker review. Some of the common reasons for exclusion were lack of useful data and titration of dose according to BP response or side effect. The reasons for excluding each trial are listed in table 3.1.

22

Table 3.1: Characteristics of excluded studies in non-selective beta blockers review

Study Reason for exclusion Data during treatment were not reported. Available data reported were 4-7 Ades 1988 [61] days after the last dose. Ades 1990 [62] Article did not provide data until after 4-7 days of last dose. Agnes 1989 [63] Article did not report drugs versus placebo data. Beilin 1972 [64] Dose was adjusted due to side effects. Patients were given active treatment for 4 weeks but placebo for only 2 Cherchi 1985 [65] weeks. Costa 1984 [66] Treatment dose was titrated to target BP Davidson 1976 [67] Treatment dose was titrated to target BP Authors claimed to have placebo control, but all patients took placebo during first 4 week of treatment. Then patients were randomly assigned to Kubik 1984 [68] three active treatment arms. The period which patients took placebo was not random. Smith 1988 [70] Article only provided ambulatory BP measurement data. Weigmann 1998 [71] Article only provided ambulatory BP measurement data.

3.4 Effects of propranolol on systolic blood pressure, diastolic blood pressure, heart

rate and pulse pressure

The recommended daily dose of propranolol according to the Canadian Pharmacists

Association product monograph is from 80 mg to 320 mg for extended release tablet once daily for treatment of hypertension [33]. Eighteen studies examining the BP lowering efficacy of propranolol at doses ranging from 60 mg/day to 640 mg/day, in durations from 4 to 12 weeks in 837 hypertensive patients were included. The majority of included studies measured peak BP using a mercury sphygmomanometer. The weighted mean baseline BP of the included studies was 163.1 mmHg of systolic blood pressure and 106.7 mmHg of diastolic blood pressure. Table 3.2 summarizes the results for propranolol.

23

Table 3.2: Dose ranging BP, heart rate and pulse pressure lowering efficacy of propranolol

Dosage (multiples of starting Total # # RCT in N in mean estimate of dose) RCT (N) subgroup subgroup difference [95% CI] Systolic blood pressure 60, 80 mg/day (1x) 3 222 -4.7 [-8.1, -1.2] 120, 160 & 180 mg/day (2x) 18 13 592 -12.2 [-14.6, -9.7] 240 & 320 mg/d (4x) (N=837) 6 99 -21.0 [-25.4, -16.6] 480 & 640 mg/d (8x) 2 205 -4.3 [-9.0, 1.0] Diastolic blood pressure 60, 80 mg/day (1x) 3 222 -3.3 [-5.6, -1.0] 120, 160 & 180 mg/day (2x) 18 13 592 -8.5 [-10.0, -7.0] 240 & 320 mg/d (4x) (N=837) 6 99 -13.9 [-16.8, -11.1] 480 & 640 mg/d (8x) 2 205 -3.0 [-6.1, 0.0] Heart rate 60, 80 mg/day (1x) 2 196 -7.3 [-9.5, -5.1] 120, 160 & 180 mg/day (2x) 12 9 497 -11.5 [-13.0, -10.1] 240 & 320 mg/d (4x) (N=694) 5 85 -18.4 [-20.9, -15.8] 480 & 640 mg/d (8x) 1 171 -10.0 [-12.8, -7.2] Pulse pressure 60, 80 mg/day (1x) 3 222 -1.5 [-4.8, 1.9] 120, 160 & 180 mg/day (2x) 18 13 592 -2.5 [-4.3, -0.8] 240 & 320 mg/d (4x) (N=837) 6 99 -3.0 [-6.5, 0.6] 480 & 640 mg/d (8x) 2 205 -2.3 [-8.4, 3.8]

Propranolol significantly lowered systolic blood pressure and diastolic blood pressure compared to placebo at doses equivalent to the recommended starting dose (80 mg), 2x the starting dose and at 4x times the starting dose (Table 3.2). There was a significant degree of heterogeneity at 2x and 4x the starting dose in both systolic blood pressure and diastolic blood pressure. This heterogeneity is explored in the discussion of this chapter.

24

Not all studies included in propranolol analyses reported the change in heart rate. All doses of propranolol significantly lowered heart rate compared to placebo. There was a significant degree of heterogeneity at the starting dose and 2 times the starting dose.

Propranolol at 2x the starting dose significantly lowered pulse pressure. The magnitude of the effect on pulse pressure for the other doses was similar but not statistically significant compared to placebo. If all 4 doses are pooled the result is a statistically significant (-2 mmHg).

We were able to use the data from Galloway 1976 [42], Shukla 1979 [57] and Sica

2004 [58] to directly assess the additional BP lowering effect of propranolol when doses were doubled. Doubling the dose of propranolol did not cause a greater BP lowering effect for systolic blood pressure or diastolic blood pressure. Doubling the dose of propranolol, however, reduced the heart rate significantly. The test for subgroup differences by direct comparison in heart rate was significant (p<0.00001). Oh 1985 [49] also directly compared the effect of propranolol when dose was doubled. However, the blood pressure lowering effect, a mean of -53 mmHg systolic and -28 mmHg diastolic, seen in this study was out of keeping with all the other data. We judged this trial an extreme outlier, as of questionable validity and therefore we did not use it in the direct comparison analyses.

Funnel plots

Funnel plots were prepared for systolic blood pressure and diastolic blood pressure at

2x the starting dose to explore the heterogeneity (Figure 3.2 and Figure 3.3). The vertical line in the middle of the inverted funnel represents the overall point estimate and the two sides of the triangle represent the 95% confidence interval. The y-axis represents the standard error and as the standard error decreases, the 95% confidence interval becomes narrower. The x-

25 axis is the BP lowering effect of the drug minus the placebo effect. Each of the points represents one single study. In the most ideal case, studies should be evenly distributed within the funnel with larger and more precise studies on top and smaller less precise studies near the bottom of the graph. The funnel plots do not show the normal pattern with a number of extreme outliers well outside the 95% confidence lines. In addition the largest studies should be near the mean effect size and that is not the case. A possible explanation for this finding is discussed in section 3.11.2.

Figure 3.2: Funnel plot: Systolic blood pressure at 2x the starting dose for propranolol.

Number standardof error

Point estimates of BP lowering effect (mmHg)

26

Figure 3.3: Funnel plot: Diastolic blood pressure at 2x the starting dose for propranolol

Number standardof error

Point estimates of BP lowering effect (mmHg)

3.5 Effects of penbutolol on systolic blood pressure, diastolic blood pressure, heart

rate and pulse pressure

The recommended dose of penbutolol for hypertension is 20 mg to 80 mg once daily

[26]. Two studies with 312 participants treated with penbutolol 20 mg/day, 40 mg/day, 80 mg/day or placebo were included. Mean baseline BP was 151.9/100 mmHg. De Plean

1981[38] was a crossover study lasting for 4 week. Schoenberger 1989 [56] was a parallel study with treatment lasting for 6 weeks. Both studies measured BP using mercury sphygmomanometer in the standing position. Most of the BP was measured at trough times

(13- 24 hours after last dose). Table 3.3 summarized the results of penbutolol.

27

Table 3.3: Dose ranging BP, heart rate and pulse pressure lowering efficacy of penbutolol

Dosage (multiples Total # RCT # RCT in Mean estimate of difference N in subgroup of starting dose) (N) subgroup [95% CI] Systolic blood pressure 10 mg/day (0.5x) 1 151 -2.4 [-8.8, 4.0] 20 mg/day (1x) 2 1 147 -6.5 [-13.0, -0.0] 40 mg/day (2x) (N=312) 1 152 -4.5 [-10.9, 1.9] 80 mg/day (4x) 1 10 -18.0 [-31.3, -4.7] Diastolic blood pressure 10 mg/day (0.5x) 1 151 -2.9 [-7.0, 1.2] 20 mg/day (1x) 2 1 147 -3.7 [-7.8, 0.4] 40 mg/day (2x) (N=312) 1 152 -3.5 [-7.5, 0.5] 80 mg/day (4x) 1 10 -11.1 [-17.9, -4.3] Heart rate 10 mg/day (0.5x) 1 151 -1.2 [-9.8, 7.4] 20 mg/day (1x) 2 1 147 -0.4 [-9.0, 8.2] 40 mg/day (2x) (N=312) 1 152 -2.8 [-11.4, 5.8] 80 mg/day (4x) 1 10 -19.7 [-27.2, -12.3] Pulse pressure 10 mg/day (0.5x) 1 151 0.5 [-7.3, 8.3] 20 mg/day (1x) 2 1 147 -2.8 [-10.7, 5.0] 40 mg/day (2x) (N=312) 1 152 -1.1 [-8.9, 6.7] 80 mg/day (4x) 1 10 -6.9 [-18.4, 4.6]

Only 80 mg/day penbutolol significantly reduced both systolic blood pressure and diastolic blood pressure compared to placebo. All other doses did not significantly lower systolic blood pressure or diastolic blood pressure compared to placebo. Direct comparison of dose effect was possible using data from Schoenberger 1989. There was no significant difference in BP lowering efficacy of systolic blood pressure (p=0.68), diastolic blood pressure (p=0.96) or heart rate (p=0.93) between 10 mg/day to 40 mg/day penbutolol.

28

3.6 Effects of timolol on systolic blood pressure, diastolic blood pressure, heart rate

and pulse pressure

The recommended daily dose of timolol for hypertension is 5 mg BID to 20 mg BID

[27, 33]. One study, Chalmers 1976 [36], examined the BP lowering efficacy of 30 mg/day timolol in 20 hypertensive patients was included. Chalmers 1976 was a crossover study with

8-week treatment period. BP was measured by mercury sphygmomanometer. Timolol 30 mg/day significantly lowered both systolic blood pressure and diastolic blood pressure compared to placebo (Table 3.4). The treatment also significantly lowered heart rate but did not significantly lower pulse pressure.

Table 3.4: Dose ranging BP, heart rate and pulse pressure lowering efficacy of timolol

Dosage (multiples Total # RCT # RCT in Mean estimate of difference N in subgroup of starting dose) (N) subgroup [95% CI] Systolic blood pressure 30 mg/day (4x) 1 (N=20) 1 20 -14.4 [-21.5, -7.3] Diastolic blood pressure 30 mg/day (4x) 1 (N=20) 1 20 -13.5 [-18.1, -9.0] Heart rate 30 mg/day (4x) 1 (N=20) 1 20 -16.0 [-18.9, -13.1] Pulse pressure 30 mg/day (4x) 1 (N=20) 1 20 -0.9 [-8.7, 6.9]

3.7 Effects of tertatolol on systolic blood pressure, diastolic blood pressure, heart

rate and pulse pressure

One study with 20 participants was included for tertatolol. Fasano 1991 [40] was a parallel study examining the BP lowering effect of 5 mg/day tertatolol for 4 weeks. The authors measured peak BP in the standing position using an automated device. The baseline

29 systolic blood pressure and diastolic blood pressure was 154.5 mmHg and 104 mmHg, respectively.

Five mg tertatolol significantly lowered systolic blood pressure and diastolic blood pressure compared to placebo (SBP -14.00 [-26.54, -1.46], DBP -15.00 [-23.33, -6.67]). 5 mg/day tertatolol also significantly lowered heart rate (-17.00 [-24.06, -9.94]) but did not lower pulse pressure compared to placebo.

3.8 Effect on heart rate of other included non-selective beta blockers

Moprolol, indenolol and nadolol studies only provided data on heart rate. Indenolol and nadolol 80mg/day significantly lowered heart rate compared to placebo but moprolol did not show significant effect (Table 3.5).

Table 3.5: Heart rate lowering effect of moprolol, indenolol and nadolol

Dosage (multiples Total # RCT # RCT in Mean estimate of difference N in subgroup of starting dose) (N) subgroup [95% CI] Moprolol 150 mg/day 1 1 23 -6.0 [-13.5, 1.5] Indenolol 120 mg/day 1 1 12 -12.0 [-15.4, -8.6] Nadolol 80 mg/day (2x) 1 1 10 -20.6 [-26.5, -14.7]

3.9 Pooled subclass effects on systolic blood pressure, diastolic blood pressure, heart

rate and pulse pressure

We pooled the BP, heart rate and pulse pressure data of all available non-selective beta blockers based on the recommended starting doses (Table 3.6). The analyses showed that 0.5 times the starting dose did not significantly lower systolic blood pressure, diastolic blood pressure, heart rate and pulse pressure. The recommended starting doses and higher doses significantly lowered systolic blood pressure, diastolic blood pressure and heart rate.

30

Pulse pressure was not significantly reduced at any dose except for 2x the starting dose subgroup.

Table 3.6: Dose ranging BP, heart rate and pulse pressure lowering efficacy of non- selective beta blockers

Dosage (multiples of starting Total # # RCT in N in Mean estimate of dose) RCT (N) subgroup subgroup difference [95% CI] Systolic blood pressure 0.5x starting dose 1 155 -2.4 [-10.8, 6.0] 1x starting dose 5 393 -5.3 [-8.4, -2.1] 22 2x starting dose 14 748 -11.5 [-13.9, -9.2] (N=1,227) 4x starting dose 8 129 -19.7 [-23.6, -15.8] 8x starting dose 2 205 -4.3 [-9.6, 1.0] Diastolic blood pressure 0.5x starting dose 1 155 -2.9 [-8.2, 2.4] 1x starting dose 5 393 -3.6 [-5.4, -1.8] 22 2x starting dose 14 748 -8.2 [-9.6, -6.7] (N=1,227) 4x starting dose 8 129 -14.1 [-16.6, -11.5] 8x starting dose 2 205 -3.0 [-6.1, 0.0] Heart rate 0.5x starting dose 1 155 -1.2 [-12.5, 10.1] 1x starting dose 17 4 367 -8.8 [-12.0, -5.5] 2x starting dose (N=1,112) 12 683 -12.6 [-14.2, -11.0] 4x starting dose 6 95 -18.5 [-21.7, -15.4] Pulse pressure 0.5x starting dose 1 155 0.5 [-9.8, 10.8] 1x starting dose 5 393 -1.5 [-4.6, 1.6] 22 2x starting dose 14 748 -2.2 [-4.0, -0.5] (N=1,227) 4x starting dose 8 129 -2.9 [-6.2, 0.3] 8x starting dose 2 205 -2.3 [-8.4, 3.8]

Test for subgroup differences by direct comparison were not significant in the recommended dose range (1x, 2x, 4x and 8x the starting dose) systolic blood pressure, diastolic blood pressure and pulse pressure. Test for subgroup differences by direct

31 comparison was significant in the recommend dose range for heart rate demonstrating a relationship between dose and effect for heart rate.

None of the studies included provided pulse pressure data. We calculated the pulse pressure by subtracting diastolic blood pressure from systolic blood pressure. The pooled analysis confirmed the findings for effect of propranolol on pulse pressure.

Withdrawal due to adverse effects

Only 2 studies, one from propranolol and the other one from penbutolol, provided data regarding WDAE (Table 3.7). WDAE were not significantly different with either propranolol or penbutolol compared to placebo. Due to the lack of information no conclusion can be made about the effect of non-selective beta blockers on WDAE.

Table 3.7: WDAE of non-selective beta blockers

Withdrawal due to adverse effects Drug Total N Relative risk [95% CI] Propranolol 427 0.9 [0.3, 2.3] Penbutolol 302 0.8 [0.2, 2.9] Overall effect 729 0.8 [0.4, 1.8]

Blood pressure variability

We assessed the blood pressure variability between treatment and placebo by comparing the end treatment standard deviation with placebo by unpaired t-test. 13 studies provided end treatment and placebo standard deviation for this analysis. The BP variability for both systolic blood pressure and diastolic blood pressure was not significantly different between treatment and placebo group. The p-value for systolic blood pressure variability was

0.2 and diastolic blood pressure variability was 0.8. The weighted mean variance ratio for systolic blood pressure/diastolic blood pressure was 1.1/1.1. The majority of the data that

32 contributed to the average standard deviation came from propranolol. More discussion on BP variability of non-selective beta blockers is located in discussion (section 3.11.7).

3.10 Subgroup analysis

We were unable to perform subgroup analysis for race or co-morbid condition due to lack of data reported separately for these parameters. There was no difference between trials in terms of age and severity and therefore we could not perform a subgroup analysis for these two parameters either. There was one study, Ravid 1985 [53], which reported data for men and women separately. This study examined the BP lowering efficacy of 160 mg/day propranolol in 66 men and 68 women. The weighted mean baseline BP was 168/112 mmHg for men and 166/110 mmHg for women. The average systolic blood pressure reduction was

(-27.8 [-35.0, -20.6]) for men and (-12.7 [-20.6, -4.8]) for women, the average diastolic blood pressure reduction was (-15.2 [-20.9, -9.5]) for men and (-13.0 [-18.8, -7.2]) for women. The reduction of systolic blood pressure but not diastolic blood pressure was significantly greater in men than women in this study. In this trial propranolol significantly lowered pulse pressure in men but not in women.

3.11 Discussion

3.11.1 Propranolol

This review provides the most up-to-date estimates of the dose related blood pressure lowering efficacy of non-selective beta blockers. Most of the data contributing to the estimates came from propranolol studies. Propranolol significantly lowered resting peak systolic blood pressure and diastolic blood pressure compared to placebo. Similarly all propranolol doses significantly lowered heart rate compared to placebo.

33

Direct comparison of doses was possible by pooling 3 studies that examined the effect of doubling the dose of propranolol. This analysis showed that doubling the dose of propranolol did not significantly further lower BP. However, this dose response analysis is limited due to the small sample size.

Significant heterogeneity and the relatively large effect size of propranolol 4x starting dose subgroup were caused by an extreme outlier (Oh 1985). The point estimate of the 4x starting dose would change from -21.0/-14.5 mmHg to -13.9/-11.5 mmHg if we exclude Oh

1985. Although this was still a large effect, the wide 95% CI indicated that this estimate was imprecise.

If there is any dose response relationship for propranolol, the pattern is that it is less with the starting dose, peaks at 4x starting dose and then decreases with higher doses. This would be an unusual effect and there are few if any proven dose responses of this pattern. If true, an explanation might be that, at higher doses, the beta-2 blocking effect leads to antagonism of the BP lowering effect by blocking vasodilation in the arteries to muscles.

More carefully conducted trials are needed to properly elucidate this dose response pattern.

Against that particular pattern being true is that it was not seen with any of the other non- selective beta blockers.

3.11.2 Heterogeneity in propranolol 2x starting dose subgroup

The 2x starting dose subgroup of systolic blood pressure contained the largest sample size in the analysis and manifested a very unusual pattern in the funnel plots (Figure 3.2 and

3.3). In order to explore the heterogeneity, we arranged the studies according to publication year and then divided them into 2 groups. We used to divide studies into two groups based on their publication year, one before 1985 and one after 1985. In the older studies, the funnel

34 plot showed that Oh 1985 [49] and Ravid 1985 [53] were two extreme outliers and the main contributors to heterogeneity. In the newer studies, using the same method, McCorvey 1993

[46] was identified as an extreme outlier.

We conducted the same procedure in order to explore the heterogeneity in diastolic blood pressure analysis of 2x starting dose subgroup. In the older studies, the funnel plot identified that Oh 1985, Ravid 1985 and Shukla 1979 [57] (120 mg/day) as extreme outliers in this subgroup. There was no significant heterogeneity among the newer studies.

The estimates for the two groups were significantly different from each other (-15.7/-

11.1 mmHg for the older studies and -5.0/-4.1 mmHg for the newer studies). The difference in effect size between these two groups is possibly due to the difference in baseline BP. The weighted mean baseline BP was higher in the older studies (170.4/112.6 mmHg) compared to the newer studies (154.1/100.6 mmHg). This reflects the fact that the definition of hypertension had changed over time. In the older studies hypertension was defined as

>160/100 mmHg and in the more recent studies as >140/90 mmHg. Higher baseline BP would be expected to be associated with greater absolute BP reductions.

Another factor that may have added to the difference is a higher risk of publication bias in the older studies and the fact that older studies were influenced by the marketing of the drug. Studies published at the time were more likely to show greater effect. In contrast in the newer studies propranolol was being studied as an active comparator to a newer drug with a possible negative bias against propranolol. The older studies could also have been subject to greater loss of blinding leading to detection bias, which would also exaggerate the effect size as well. It could be a combination of these factors that led to the difference in effect size and caused the large heterogeneity.

35

The baseline BP in the newer studies was similar to the baseline in the RCTs of the

ACE inhibitor and angiotensin receptor blocking reviews by Heran 2008 (ACEI) [72] and

Heran 2008 (ARB) [73]. Thus when comparing the BP lowering effect between non- selective beta-blockers and ACE inhibitors and ARBs, it is more appropriate to use the estimates from the newer studies.

3.11.3 Penbutolol

The larger study in penbutolol analysis, Schoenberger 1989 [56], failed to show a statistically significant BP lowering effect in dose ranging from 10 mg/day to 40 mg/day.

The relatively large effect of the smaller RCT, De Plean 1981 [38], in penbutolol 80 mg/day was out of keeping with the other data. The larger effect in 80 mg/day subgroup could be due to the fact that De Plean 1981 measured peak BP while Schoenberger 1989 measured trough

BP. At any rate there are insufficient published evidence supporting the product monograph recommendation of penbutolol 20 mg and 40 mg per day in the treatment of hypertension.

3.11.4 Other included non-selective beta blockers

There were few data available for other non-selective beta blockers. It was not possible to draw any definitive conclusion based on such small data sets. Some of these drugs, such as nadolol and timolol, are approved for clinical use in U.S., Europe and Canada since late 1970. We are therefore certain that studies were completed in order to fulfill regulatory requirements in these countries and that these studies remain hidden at this time.

Despite our effort to search for unpublished data, we were not able to find any trial data for these drugs. It is important that these trial results, be made available for scientific analysis and are available to contribute to systematic reviews. More discussion on hidden study of approved medication is located in the chapter of overall discussion (section 7.6.5).

36

3.11.5 Pooled BP lowering effects of non-selective beta blockers

We pooled the data from all available non-selective beta blockers based on the recommended starting dose in order to obtain the overall subclass effect. We recognized that this method presented certain limitations as different manufacturers might have used different methods to come up with the recommended starting doses. However, this method provided the most logical way of data standardization for the purpose of combining different drugs to be analyzed as a whole class.

The pooled data also presented similar characteristics of heterogeneity as in the propranolol analyses because propranolol was the main contributor of data in the pool. The

1x and 2x starting dose subgroups contained the largest sample size. The estimate of BP lowering efficacy for non-selective beta blockers by combining the 1x and 2x starting dose subgroup was -10/-7 mmHg.

Dose response analysis was not conclusive. Non-selective beta blockers, in the recommended dose range, did not showed a convincing dose response by direct comparison.

There was little evidence to support using dose higher than 2x starting dose of non-selective beta blockers to further lower BP.

3.11.6 Heart rate

Non-selective beta blockers starting at the 1x recommended starting doses significantly lowered heart rate. The dose response in heart rate was evident by both direct and indirect comparison.

3.11.6 Pulse pressure

Non-selective beta blockers did not significantly reduce pulse pressure in any dose subgroups except for the 2x starting dose. The point estimates in the 1x, 4x and 8x starting

37 dose subgroups were similar to the 2x starting dose subgroup. Therefore, it would appear that if non selective beta blockers do lower pulse pressure, the magnitude is likely to be around 2 mmHg.

This relative lack of effect on pulse pressure may explain why beta blockers appear to be less effective at reducing mortality than other drug classes as shown by Wisonge 2007

[21], Wright 2000 [20] and Musini 2009 [11]. However, we recognized that most of the data in people with hypertension included in Wisonge 2007 and Musini 2009 used atenolol.

Furthermore, Wright 2000 found that non-selective beta blockers demonstrated a mortality benefit in Post MI patients. This suggested that the effect of non-selective beta blockers on mortality could be different than atenolol and the lack of effect on pulse pressure might not fully explain the differences in mortality. More studies evaluating the effect on mortality and morbidity of non-selective beta blocker is needed to elucidate this question.

3.11.7 Effect on blood pressure variability

Non-selective beta blockers did not significantly change end treatment blood pressure variability compared to placebo. The systolic blood pressure variance ratio in our analysis was smaller than the one found in Webb 2011 (1.1 vs 1.3). A recently published analysis on the BP variability of beta blockers suggested that non-selective beta blockers might significant increase systolic blood pressure variability (Webb 2011 [74]). We would discuss the possible explanation for the differences in result in greater detail in section 7.1.2. In summary, differences in population and comparison group could help explain the difference in the result.

38

3.11.8 Subgroup analysis

The planned subgroup analyses were not possible due to lack of data. Ravid 1985

[53] provided some data for subgroup analysis in difference sex. We summarized the findings in this study in the results section. However, we should not draw any conclusion based on findings of one single study. In addition, the results of Ravid 1985 were extreme, raising concern about bias in this study. More large scale RCT reporting outcome separately by sex are needed before we can determine whether or not there is a significant difference in effect size between difference sexes.

39

Figure 3.4: Risk of bias summary of non-selective beta blockers

High risk of bias

Low risk of bias Unknown risk of bias

40

3.12 Risk of bias and quality of the evidence

In risk of bias summary table (figure 3.6), red bubble represented high risk of bias, yellow bubble represented unknown risk of bias and green bubble represented low risk of bias. The overall quality of evidence was summarized below.

3.12.1 Randomization and allocation concealment (selection bias)

All of the studies stated that patients were randomized and allocation of treatment was concealed. There was no obvious difference in baseline parameters between parallel groups suggesting that randomization was achieved in these trials. Therefore we judged that parallel trials have a lower risk of selection bias than crossover trials. However, this reason alone was not enough to upgrade the quality of evidence for parallel studies. Therefore, the risk of selection bias remains unknown for many of the parallel studies. The baseline data was not provided for the crossover studies so we could not be confident that randomization and allocation concealment was achieved in these studies.

3.12.2 Blinding (performance bias and detection bias)

All the studies used double blinded design. However, due to the fact that beta blockers significantly lowered heart rate and most of the studies measured BP by mercury sphygmomanometer, there was a high likelihood that the investigators could detect the assignment of intervention by the effect on the heart rate. Therefore, the risk of detection and performance bias was high in this review. The detail on the potential risk of detection bias caused by lower heart rate was described in section 7.6.3.

41

3.12.3 Incomplete outcome data (attrition bias)

Withdrawn patients were not included in the analysis for all the studies who reported drop outs. The dropout rate was low and therefore we judged that there was a low risk of attrition bias in our estimates.

3.12.4 Selective reporting (reporting bias)

Only two studies reported WDAE in this review. WDAE is an important outcome in all RCTs. Not reporting WDAE could lead to high risk of reporting bias. The reason why selective reporting of WDAE could lead to reporting bias is described in section 7.6.4.

3.12.5 Publication bias

Nadolol, indenolol and moprolol studies did not provide any useful systolic blood pressure and diastolic blood pressure data for our review. Nadolol is indicated for hypertension in both the U.S. and Canada. In order to satisfy regulatory requirement in U.S. and Canada, nadolol must have been tested in RCTs for blood pressure lowering efficacy.

The reason why lack of publicly available data for approved medication could lead to high risk of publication bias is described in section 7.6.5.

The studies published earlier before 1985 showed much greater BP lowering effect compared to studies published after 1985. They used same drug and same design for these trials. However, the patient population was different. In the older studies the patients had moderate to severe hypertension according to recent definition, whereas in the more recent studies the patients had mild to moderate hypertension. In addition, the funnel plots showed presence of extreme positive outliers. Excluding these outliers from the analysis considerably lowered the estimate in some subgroup. This suggests that the estimate could be exaggerated due to extreme positive outliers and the high risk of publication bias.

42

3.12.6 Quality of the evidence

Table 3.8 summarizes the combined effect size of the combined starting and 2x the starting dose of non-selective beta blockers. In addition it provides a judgment of the quality of evidence in this review. This review included 25 RCTs with 1,279 hypertensive patients.

The sample size should provided adequate power to draw robust conclusions. However, as for all systematic review, we were limited by the data available to us.

Most of our studies did not use automated machine to measure BP. Automated machine could mitigate the risk of clinicians or patients detecting the intervention assignment. Therefore, the risk of detection bias remains high in our review.

None of the studies reported pulse pressure as one of their outcome. We calculated pulse pressure by subtracting diastolic blood pressure from systolic blood pressure. This meant that the data on pulse pressure did not come from direct measurement in the studies.

Indirectness of this outcome is the reason the evidence was further downgraded for pulse pressure.

In addition, the risk of publication bias is high in this review. For these and other reasons outlined in the discussion, the estimates of the BP lowering effect shown in the summary of findings table are very likely an exaggeration of the true effect. This is reflected in the Table by the low to very low judgment of the quality of the evidence.

43

Table 3.8: Summary of findings table for non selective beta blockers

Summary of findings table Population: Adult patients with primary hypertension Intervention: Non-selective beta blockers Comparison: Placebo mean estimates of combining 1x Number of patients in the Quality of Outcomes and 2x starting dose subgroups (# of studies) evidence [95% CI] SBP -9.5 [-10.9, -8.1]1,2,3 949 (16) Low4,5 DBP -6.6 [-7.4, -5.8]1,2,3 949 (16) Low4,5 HR -11.8 [-12.9, -10.7]1,2 864 (13) Low4,5 PP -2.0 [-3.2, -0.9]1,2,3 949 (16) Very Low4,5,6 Acronyms and grading guide SBP: systolic blood pressure; DBP: diastolic blood pressure; HR: heart rate; PP: pulse pressure; 95%CI: 95% confident interval GRADE Working Group grades of evidence [29] High quality: Further research is very unlikely to change our confidence in the estimate of effect. Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate. Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate. Very low quality: We are very uncertain about the estimate. Footnotes 1. The recommended starting and 2x starting contained most of the data in non-selective beta blocker analysis. Combining them provided estimates that represent the overall subclass effect. 2. Most of the measurements were made at peak hours. 3. Mean baseline BP was 158/104 mmHg. 4. Quality of evidence was downgraded one level due to high risk of detection bias caused by loss of blinding. 5. Quality of evidence was downgraded one level due to high risk of publication bias and the presence of extreme outliers with exaggerated large effect. 6. Quality of evidence was downgraded one level due to indirectness, none of the studies included reported pulse pressure. Pulse pressure was calculated by subtracting DBP from SBP.

44

4. RESULTS FOR ALPHA AND BETA DUAL RECEPTOR

BLOCKERS

4.1 Search findings

The search finding and flow of study selection is the same as stated in the non- selective beta blocker (chapter 3). In total, eight studies were included in this dual receptor blockers review. Figure 4.1 described the flow of study selection.

45

Figure 4.1: PRISMA diagram of dual receptor blockers review

46

4.2 Characteristics of included studies

Included in this review were 8 RCTs examined the blood pressure lowering efficacy of dual receptor blockers in 1,493 hypertensive patients for 3 to 12 weeks. Five of the 8 included studies were parallel studies and the other 3 were crossover studies. Four RCTs studied carvedilol 12.5 mg/day to 50 mg/day in 1,381 patients and the other 4 RCTs studied labetalol 300 mg/day to 800 mg/day in 112 patients. Weighted mean baseline diastolic blood pressure was 100.9 mmHg in carvedilol studies and 105.4 mmHg in labetalol studies.

Characteristics of each individual study were listed appendix D.

4.3 Characteristics of excluded studies

Four studies that potentially meet the inclusion criteria were excluded after detailed reading. The reasons for exclusion were lack of useful data for outcomes listed in the review and misleading information on method. Please refer to characteristics of excluded studies table (Table 4.1) for the reasons of exclusion.

Table 4.1: Characteristics of excluded studies in dual receptor blockers review

Study Reason for exclusion Dupont 1987 [83] Article did not report placebo group data Horvath 1979 [84] Article reported only mean arterial blood pressure Authors claimed to have placebo controlled. But all patients took placebo during first 4 week of treatment. Then patients were randomly assigned to Kubik 1984 [85] three active treatment arms. They presented the data as if placebo was randomly assigned but the period which patients take placebo was not randomly assigned. Patients were randomized to 4 week treatment period. However, each period Wilcox 1978 [86] the drugs were taken at two doses. Each dose only lasted for two weeks.

47

4.4 Effects of carvedilol on systolic blood pressure, diastolic blood pressure, heart

rate and pulse pressure

Carvedilol is indicated for the treatment of hypertension, left ventricular dysfunction following MI and heart failure in the U.S. [26]. It is indicated only for heart failure in Canada

[33]. However, it can be used off label for hypertension. The recommended doses for hypertension in the U.S. are 6.25 mg BID to 25 mg BID for the immediate release form and

20 to 80 mg daily for the controlled release form.

Table 4.2: Dose ranging BP, heart rate and pulse pressure lowering efficacy of carvedilol

Dosage (multiples of starting Total # # RCT in N in Mean estimate of dose) RCT (N) subgroup subgroup difference [95% CI] Systolic blood pressure 6.25 mg/day (0.5x) 1 236 -0.9 [-5.2, 3.4] 12.5 mg/day (1x) 4 3 494 -3.5 [-6.4, -0.7] 20 & 25 mg/day (2x) (N=1,381) 4 644 -4.1 [-6.5, -1.7] 40 & 50 mg/day (4x) 4 527 -6.5 [-9.3, -3.7] Diastolic blood pressure 6.25 mg/day (0.5x) 1 236 -0.7 [-3.1, 1.7] 12.5 mg/day (1x) 4 3 494 -2.3 [-3.9, -0.6] 20 & 25 mg/day (2x) (N=1,381) 4 644 -3.0 [-4.4, -1.6] 40 & 50 mg/day (4x) 4 527 -5.0 [-6.6, -3.4] Heart rate 6.25 mg/day (0.5x) 1 236 -1.4 [-3.9, 1.1] 12.5 mg/day (1x) 3 2 474 -2.2 [-3.9, -0.4] 20 & 25 mg/day (2x) (N=1,337) 3 624 -4.7 [-6.1, -3.3] 40 & 50 mg/day (4x) 3 506 -6.3 [-7.9, -4.7] Pulse pressure 6.25 mg/day (0.5x) 1 236 -0.2 [-3.9, 3.5] 12.5 mg/day (1x) 4 3 494 -1.2 [-3.7, 1.2] 20 & 25 mg/day (2x) (N=1,381) 4 644 -1.2 [-3.3, 0.9] 40 & 50 mg/day (4x) 4 527 -1.6 [-4.1, 0.9]

48

Please refer to table 4.2 for the result of carvedilol. Four studies examined the blood pressure lowering efficacy of 6.25 to 50 mg/day carvedilol in 1,381 hypertensive patients were included. GSK B100 [75] and GSK B101 [76] were unpublished RCTs obtained from the manufacturer's web-site. The unpublished data contributed to 72% of the data (1000 of

1,381 subjects). Patients enrolled in these studies were middle-aged (52 to 56 years old), with no co-morbidities that could affect blood pressure. The weighted mean baseline diastolic blood pressure was 100.9 mmHg. Three of the four studies measured blood pressure at trough hours.

Carvedilol 6.25 mg/day did not significantly lower systolic blood pressure or diastolic blood pressure compared to placebo. Starting from 12.5 mg/day (the recommended starting dose), carvedilol significantly lowered systolic blood pressure and diastolic blood pressure compared to placebo. Heterogeneity was not significant in any of the subgroups. Most of the studies included have multiple dose subgroups, allowing direct comparison between doses.

Test for subgroup differences by direct comparison was not significant for systolic blood pressure and diastolic blood pressure within the recommended dose range (1x, 2x and 4x starting dose). The 1x and 2x starting dose subgroups contained the largest patient sample in carvedilol. The estimate for combining the average trough BP lowering effect of 1x and 2x starting dose carvedilol was -3.9/-2.7 mmHg.

Carvedilol 6.25 mg/day did not significantly lower heart rate compared to placebo.

Test for subgroup differences in carvedilol 12.5 mg/day and higher doses was significant for heart rate (p<0.0005). Carvedilol did not significantly change pulse pressure compared to placebo in any of the dose subgroups. If we combined the pulse pressure data of 1x and 2x

49 starting dose subgroups, the result becomes statistically significant. However, the effect of carvedilol on pulse pressure was small (1 mmHg).

BP variability was assessed by comparing the end treatment standard deviation of carvedilol and placebo groups with t-test. Twelve standard deviations from carvedilol subgroups and 4 standard deviations from placebo subgroups were included. Carvedilol did not significantly change blood pressure variability in systolic blood pressure (p=0.37) or diastolic blood pressure (p=0.65). The weighted mean standard deviation for carvedilol group (SBP/DBP) was 15.9/9.0 and for placebo group was 16.4/9.2.

4.5 Effects of labetalol on systolic blood pressure, diastolic blood pressure, heart

rate and pulse pressure

Labetalol is indicated for the treatment of hypertension in both the U.S. and Canada

[26, 33]. The recommended doses for hypertension in U.S. and Canada are 100 mg BID to

400 mg BID.

Table 4.3: Dose ranging BP, heart rate and pulse pressure lowering efficacy of labetalol

Dosage (multiples of starting Total # # RCT in N in Mean estimate of dose) RCT (N) subgroup subgroup difference [95% CI] Systolic blood pressure 300 & 400 mg/day (2x) 4 110 -10.4 [-14.0, -6.8] 4 (N=112) 600 & 800 mg/day (4x) 2 44 -19.7 [-25.4, -14.0] Diastolic blood pressure 300 & 400 mg/day (2x) 4 110 -6.6 [-8.6, -4.6] 4 (N=112) 600 & 800 mg/day (4x) 2 44 -14.6 [-18.0, -11.2] Heart rate 300 & 400 mg/day (2x) 4 110 -7.1 [-9.1, -5.0] 4 (N=112) 600 & 800 mg/day (4x) 2 44 -7.7 [-10.4, -4.9] Pulse pressure 300 & 400 mg/day (2x) 4 110 -3.6 [-6.7, -0.5] 4 (N=112) 600 & 800 mg/day (4x) 2 44 -4.6 [-9.7, 0.4]

50

Please refer to table 4.3 for result of labetalol. Four studies examined the BP lowering efficacy of 400 to 800 mg/day labetalol in 112 hypertensive patients were included in the analysis. The baseline characteristics of patients were not well described in the articles. The mean baseline BP of patients enrolled in the studies was 165/105 mmHg. 60% of the patients were middle age men with no other co-morbidity.

We did not find any data for the recommended starting dose (200 mg/day ) of labetalol. Both labetalol 400 and 800 mg/day subgroups significantly lowered systolic blood pressure and diastolic blood pressure. Test for subgroup differences was significant by direct comparison for systolic blood pressure (p= 0.09) and diastolic blood pressure (p= 0.0006).

Heterogeneity was not significant in any of the subgroups. Both 400 and 800 mg/day labetalol significantly lowered heart rate. The effect on heart rate was not significantly different between 400 and 800 mg/day labetalol.

Labetalol 400 mg/day significantly reduced pulse pressure but 800 mg/day subgroup did not significantly change pulse pressure.

4.6 Pooled effects of dual receptor blockers

Our analysis showed that the BP lowering effect of carvedilol was much smaller than labetalol. Since the BP lowering effects of carvedilol and labetalol appeared to be different we did not pool the data. Possible reasons for the difference between the drugs are explained in the discussion.

4.6.1 Withdrawal due to adverse effects (WDAE)

We pooled the WDAE data of carvedilol and labetalol together in order to include all valuable data. Five studies provided data for WDAE in 1,412 patients. Four of the studies

51 used carvedilol as active treatment. There was no significant difference in relative risk of

WDAE between dual receptor blocker treatment and placebo (0.9 [0.5, 1.4], p=0.6).

4.7 Subgroup and sensitivity analyses

With the exception of baseline BP, the baseline characteristics of patients were similar in all studies. None of the studies reported data separated by subgroups. Therefore the subgroup analysis for sex, co-morbid conditions, race, age or severity of disease was not possible.

4.7.1 Difference between published and unpublished data

We performed a sensitivity analysis in order to examine the differences between estimated mean effect size of published and unpublished trials. The estimated means

(SBP/DBP) for the 2x starting dose subgroup from the published data were -3.2/-4.2 mmHg as compared to -4.5/-2.4 mmHg from the unpublished data. The estimate of SBP/DBP in 4x starting dose subgroup from published data was -8.3/-6.5 mmHg as compared to -5.4/-3.8 mmHg from the unpublished data. For the 4x starting dose, the published estimate exceeded the unpublished estimate by -2.9/-2.7 mmHg.

4.8 Discussion

4.8.1 Carvedilol

Carvedilol 12.5 mg/day and higher doses significantly lowered blood pressure. This finding coincided with the product monograph recommendation. However, it was inconclusive whether there was a dose response effect. The BP lowering efficacy point estimates were higher for higher doses, but the test for subgroup differences was not significant. The large number of subjects in carvedilol studies, particularly the data coming from the unpublished studies, provided good estimates of the blood pressure lowering effect

52 of carvedilol. The average trough BP lowering effect of 1 time and 2 times the recommended starting dose is only -4/-3 mmHg. This suggested that carvedilol was relatively ineffective at lowering BP compared to other classes of antihypertensive drugs or other beta blockers.

4.8.2 Differences in BP lowering effect between carvedilol and labetalol

The number of patients included in trials for labetalol was much smaller than carvedilol. In addition, we did not find any data for 200 mg/day labetalol which was the recommended starting dose. This suggested that some studies of the BP lowering effect of labetalol have not been published in any form. From the studies available, labetalol appeared to significantly lower BP to a greater extent than carvedilol. The average peak BP lowering effect of labetalol at 2 times the recommended starting dose was -10/-7 mmHg.

The pharmacology of both dual receptor blockers was examined in order to explore the possible explanation for the differences. In basic pharmacology studies, carvedilol and labetalol showed similar properties in time to maximum serum concentration and elimination half life for pharmacokinetic parameters and similar receptor affinity in pharmacodynamic parameters [13, 26, 33, 87]. This suggested that the difference in BP lowering effect that we observed was unlikely to be the result of differences in pharmacological properties.

Next we examined the difference in baseline BP. The mean baseline blood pressure of labetalol studies (105.4 mmHg) were higher compared to the carvedilol studies (100.8 mmHg). Because of this, the absolute blood pressure lowering effect could be bigger for labetalol due to the higher baseline. However, if the BP lowering was expressed as the diastolic blood pressure percent change from baseline, it was still larger for labetalol (-6%) as compared to carvedilol (-4%). This suggested that differences in baseline BP could only partly explain the differences in effect size.

53

Three of four carvedilol studies stated that the blood pressure was measured at trough hours (13-24 hours after last dose). None of the labetalol studies specified whether blood pressure was measured at peak or trough hours. However, since labetalol was taken two to three times a day, the blood pressure must have been measured at peak hours (1-12 hours after the dose) in these studies. Measurements during the peak effect time could result in greater blood pressure lowering effect. Therefore the difference in time of measurements could also contribute to the difference in effect size.

The paucity of large trials assessing the BP lowering effect of labetalol indicated that the pivotal trials that would have been required to achieve a license in different countries had not been published and the evidence supporting the use of labetalol for hypertension was weak. Our search for this unpublished data for labetalol was unsuccessful.

The pharmacological properties, basic trial design and study population were similar in carvedilol and labetalol studies. Compared to the smaller effect size for carvedilol, the labetalol estimate could likely be exaggerated due to lack of large pivotal trials. If the difference in baseline BP and time of measurements were taken into account, it is possible that the effect sizes of the two drugs are actually not different.

In addition, labetalol needed to be taken two to three times a day compared to carvedilol control release tablet which could be taken once daily for hypertension. Physicians and patients making decision to use dual receptor blockers need to be made aware of the small BP lowering effect of this class of drugs.

4.8.3 Differences in published and unpublished data

From the sensitivity analysis with carvedilol, we found that published trials could overestimated the BP lowering effect by as much as -2.9 mmHg systolic and -2.7 mmHg

54 diastolic compared to unpublished trials. This was one of the first examples where the effect of publication bias could be quantified. It showed the amount of over-estimation caused by deliberate decision to publish studies that showed greater effect while not publishing studies that showed less effect. This provides good support for the extra effort to look for unpublished data. However, despite our effort to search, many trials remain completely hidden. The Cochrane collaboration had been advocating for the publication of all data, positive and negative, for many years. In recent years, the pressure from Cochrane and

European regulatory agency had resulted in the release of some unpublished data from the industry [88].

In addition, the public are becoming more aware of the impact publication bias may have on medical decision making. This is a positive change in the right direction. It is likely that the unpublished data we found in GSK web site is the result of the effort by many

Cochrane researchers pushing the company to release all RCT data.

55

Figure 4.2: Risk of bias summary of dual receptor blockers

High risk of bias

Low risk of bias Unknown risk of bias

4.9 Risk of bias and quality of evidence

The risk of bias summary is presented in figure 4.2. In the figure, a red bubble represented high risk of bias, yellow bubble represented unknown risk of bias and green bubble represented low risk of bias. The overall quality of evidence was summarized below.

4.9.1 Randomization and allocation concealment (selection bias)

Kane 1976 [78] and Lechi 1982 [79] described the detail on method of randomization and allocation concealment. Majority of studies did not report any information on this matter.

56

Therefore it was difficult to assess the risk of bias in this category. The baseline characteristics of parallel groups were similar, thus we did not find any evidence that randomization was not properly done. It was impossible to assess the integrity of randomization in crossover studies without additional information. Most of the studies included in this review were parallel studies therefore the risk of selection bias was low in this review. Even though we did not find any evidence to suspect any bias in randomization and allocation concealment, better reporting would allow us to better assess the quality of evidence.

4.9.2 Blinding (Performance and detection bias)

Beta blockers generally have a high risk of detection bias as they lower heart rate. It was possible for assessors to detect the treatment group based on the change in heart rate.

This risk could be mitigated by using automated BP machine. However, none of the studies in this review reported that they used automated BP machine. Therefore, the risk for detection and performance bias was high in this review. The detail on the potential risk of detection bias caused by lower heart rate was described in section 7.6.3.

4.9.3 Incomplete outcome data (attrition bias)

All but one included study, Frick 1976 [77], reported the information on dropouts.

The dropout rate was low in the included studies. Therefore the risk of attrition bias was low in this review.

4.9.4 Selective reporting (reporting bias)

McPhillips 1988 [80] did not report the data for heart rate according to intervention groups and Frick 1976 did not report WDAE. Other than that, blood pressure and heart rate

57 were reported in all studies. In addition, five of the eight studies reported useful WDAE. We judged that the risk of selective reporting bias was low in this review.

4.9.5 Publication bias

As discussed in previous sections, it was possible that large pivotal trials that would have been required to achieve a license for labetalol were not published. As the result, magnitude of the blood pressure lowering effect was likely over-estimated for labetalol. It is important that both positive and negative data are made available to provide the best estimate of the true effect. The reason for why selective publishing of trials for approved medication could lead to high risk of publication bias was explained in section 7.6.5.

4.9.6 Quality of the evidence

The summary of findings table (table 4.4) summarizes the effect size of the starting dose and 2 x the starting dose and provides a judgment of the quality of evidence in this review. This review included 8 studies examining two dual receptor blockers in 1,493 hypertensive patients. The sample size was sufficient to allow a robust conclusion with regard to carvedilol primarily because we were able to obtain data on 2 large unpublished

RCTs. However, like all systematic reviews, we were limited by the data that were available to us.

The data was insufficient for labetalol and we found no RCTs for dilevelol. As discussed above, data from large pivotal trials is missing in the labetalol analysis. As the result, magnitude of the blood pressure lowering effect was likely over-estimated for labetalol. For this reason, the risk of publication bias for labetalol is high. The risk of publication bias for carvedilol is much lower because of the two unpublished trials. The

58 quality of evidence would be upgraded to moderate if we only consider estimates of carvedilol.

Similar to non-selective beta blockers, most of the dual receptor blocker studies did not use automated machine which could have mitigated the risk of detection bias caused by loss of blinding. For this reason, the risk of detection bias remains high in this review and the quality of evidence was downgraded by one level.

Similar to the included studies of other three subclasses, none of the studies reported pulse pressure as one of their outcomes. Data for pulse pressure was not a direct measurement from the trials. We calculated pulse pressure by subtracting diastolic blood pressure from systolic blood pressure. Indirectness of this outcome is the reason quality of evidence was further downgraded for pulse pressure to very low.

59

Table 4.4: Summary of findings table for dual receptor blockers

Summary of findings table Population: Adult patients with primary hypertension Intervention: Alpha and beta dual receptor blockers Comparison: Placebo mean estimates of combining 1x Number of patients in the Quality of Outcomes and 2x starting dose subgroup (# of studies) evidence [95% CI] SBP -5.2 [-6.9, -3.6]1,2,3 997 (8) Low4,5 DBP -3.5 [-4.5, -2.6]1,2,3 997 (8) Low4,5 HR -4.4 [-5.4, -3.5]1,2 977 (7) Low4,5 PP -2.1 [-4.0, -0.2]1,2 997 (8) Very low4,5,6 Acronyms and grading guide SBP: systolic blood pressure; DBP: diastolic blood pressure; HR: heart rate; PP: pulse pressure; 95%CI: 95% confident interval GRADE Working Group grades of evidence [29] High quality: Further research is very unlikely to change our confidence in the estimate of effect. Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate. Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate. Very low quality: We are very uncertain about the estimate. Footnotes 1. The recommended starting and 2x starting contained most of the data in dual receptor blockers analysis. Combining them provided estimates that represent the overall subclass effect. 2. Most of the measurements were made at trough hours. 3. Weighted mean baseline DBP was 101 mmHg. 4. Quality of evidence was downgraded one level due to high risk of detection bias caused by breaking of blinding. 5. Quality of evidence was downgraded one level due to high risk of publication bias in labetalol studies. 6. Quality of evidence was downgraded one level due to indirectness, none of the studies included reported pulse pressure. Pulse pressure was calculated by subtracting DBP from SBP.

60

5. RESULTS FOR PARTIAL AGONISTS

5.1 Search Findings

The search finding and flow of study selection is the same as stated in the non- selective beta blocker chapter. In total, 13 studies were included in this partial agonists review. Please refer to figure 5.1 for the flow of study inclusion.

61

Figure 5.1: PRISMA diagram of partial agonists review

62

5.2 Characteristics of included studies

Please refer to appendix E for detail descriptions for each included study. We included 13 studies examining the blood pressure lowering efficacy of 6 partial agonists in

612 hypertensive patients. Six hundred and five of 612 patients randomized to placebo or partial agonist mono therapy, completed the studies. Five of the 13 included studies were parallel studies and the other 8 were crossover studies. Most included studies had duration lasting for 4 weeks. The average age of the participants was 52 and weighted mean baseline

BP was 174.9/106.7 mmHg. Seven of 9 included studies that provided baseline BP had baseline BP over 160/100 mmHg, therefore most of the patients included in these studies had moderate to severe hypertension. The baseline characteristic of partial agonist studies was different from the patient populations included in most trials in the other beta blocker subclasses, which were mild hypertensive patients.

Celiprolol was the most studied drug in this class with the largest sample size. Only two of the 13 included studies specified the time (peak or trough) when BP was measured.

Vandongen 1986 [100] measured BP at peak (2 hour after dose) hour and Watson 1980 [101] measured at trough hour (12-20 hour after dose). The medications were either taken multiple times a day or in the morning, therefore it should be reasonable to assume that most of the studies measured BP at peak hours.

5.3 Characteristics of excluded studies

Three studies that potentially meet the inclusion criteria were excluded from this review. The reasons for exclusion were lack of useful information, adjustment of treatment according to BP and lack of randomization in placebo group. Please see table 5.1 for the reasons for exclusion.

63

Table 5.1: Characteristics of excluded studies in partial agonists review

Study Reason for exclusion Erley 1997 [102] Only reported MAP data. If BP was not adequately controlled at any stage, prozosin was added in Fraser 1986 [103] standard dosage. The placebo period was place between treatment periods. Therefore, the Lewis 1984 [104] time to give placebo was not in random order.

5.4 Effects of acebutolol on systolic blood pressure, diastolic blood pressure, heart

rate and pulse pressure

Acebutolol is indicated for treatment of hypertension and ventricular arrhythmia in the US [26] and the EU [27]. The recommended dose for acebutolol in Canada to treat hypertension is 200 mg twice daily to 400 mg twice daily [33]. Three studies examining the blood pressure lowering efficacy of acebutolol in 85 hypertensive participants were included.

Forette 1979 [91] and Watson 1980 [101] were crossover studies. Hansson 1977 [92] was a parallel study. Only 400 mg/day acebutolol was studied for treatment duration ranged from 4 to 8 weeks. The mean baseline BP of acebutolol studies was 176/113 mmHg.

Please refer to table 5.2 for the results of acebutolol. 400 mg/day acebutolol significantly lowered systolic blood pressure but did not significantly lower diastolic blood pressure compared to placebo. 400 mg/day acebutolol significantly lowered heart rate compared to placebo but did not significantly change pulse pressure.

64

Table 5.2: Dose ranging BP, heart rate and pulse pressure lowering efficacy of acebutolol

Dosage (multiples of starting Total # # RCT in N in Mean estimate of dose) RCT (N) subgroup subgroup difference [95% CI] Systolic blood pressure 400 mg/day (1x) 3 (n=85) 3 85 -5.1 [-8.9, -1.4] Diastolic blood pressure 400 mg/day (1x) 3 (n=85) 3 85 -1.8 [-3.9, 0.3] Heart rate 400 mg/day (1x) 3 (n=85) 3 85 -8.7 [-10.9, -6.5] Pulse pressure 400 mg/day (1x) 3 (n=85) 3 85 -2.4 [-5.4, 0.6]

5.5 Effects of pindolol on systolic blood pressure, diastolic blood pressure, heart rate

and pulse pressure

Pindolol is indicated for treatment of hypertension and angina pectoris in Canada and the EU [27, 33], and only for hypertension in the US [26].The recommended dose for pindolol in Canada to treat hypertension is 5 mg twice daily to 15 mg three times daily [33].

Three studies examining blood pressure lowering efficacy of 10 and 30 mg/day pindolol in

50 hypertensive patients were included. Forty five of 50 randomized patients completed the studies. All of pindolol studies were crossover studies with duration ranging from 3 to 8 weeks. Only one study provided baseline diastolic blood pressure information which was 93 mmHg. Please refer to table 5.3 for pindolol results.

65

Table 5.3: Dose ranging BP, heart rate and pulse pressure lowering efficacy of pindolol

Dosage (multiples of starting Total # # RCT in N in Mean estimate of dose) RCT (N) subgroup subgroup difference [95% CI] Systolic blood pressure 10 mg/day (1x) 2 29 -14.5 [-19.8, -9.1] 3 (n=45) 30 mg/day (4x) 1 16 -12.5 [-20.2, -4.8] Diastolic blood pressure 10 mg/day (1x) 2 29 -6.9 [-10.2, -3.6] 3 (n=45) 30 mg/day (4x) 1 16 -8.8 [-14.8, -2.8] Heart rate 10 mg/day (1x) 2 (n=29) 2 29 -5.9 [-9.5, -2.3] Pulse pressure 10 mg/day (1x) 2 29 -8.0 [-12.7, -3.3] 3 (n=45) 30 mg/day (4x) 1 16 -3.7 [-10.3, 2.9]

Both 10 mg/day and 30 mg/day pindolol significantly lowered systolic blood pressure and diastolic blood pressure compared to placebo. Heart rate data was not reported for 30 mg/day. Indirect comparison did not find significant difference in systolic blood pressure and diastolic blood pressure between 10 mg/day and 30 mg/day pindolol. Pindolol

10 mg/day significantly lowered pulse pressure compared to placebo. However, 30 mg/day pindolol did not significantly change pulse pressure.

5.6 Effects of celiprolol on systolic blood pressure, diastolic blood pressure, heart

rate and pulse pressure

Celiprolol is indicated for treatment of hypertension in the EU. The recommended dose for celiprolol for treatment of hypertension is 200 to 400 mg once daily [27]. Three studies examining the blood pressure lowering efficacy of 200 to 600 mg/day celiprolol in

267 hypertensive patients were included in this analysis. Two of the RCTs were parallel studies and the other one was a crossover study. These studies had duration ranging from 4 to

66

12 weeks. Mean baseline BP for celiprolol studies was 163/100 mmHg. Please refer to table

5.4 for the results of celiprolol.

Table 5.4: Dose ranging BP, heart rate and pulse pressure lowering efficacy of celiprolol

Dosage (multiples of starting Total # # RCT in N in Mean estimate of dose) RCT (N) subgroup subgroup difference [95% CI] Systolic blood pressure 200 mg/day (1x) 3 171 -7.9 [-11.9, -3.8] 400 mg/day (2x) 3 (n=267) 2 120 -8.6 [-13.4, -3.8] 600 mg/day (4x) 1 90 -10.0 [-16.5, -3.5] Diastolic blood pressure 200 mg/day (1x) 3 171 -4.6 [-6.6, -2.5] 400 mg/day (2x) 3 (n=267) 2 120 -4.9 [-7.3, -2.5] 600 mg/day (4x) 1 90 -6.0 [-9.5, -2.5] Heart rate 200 mg/day (1x) 1 26 -2.4 [-4.9, 0.1] 1 (n=26) 400 mg/day (2x) 1 26 -0.6 [-3.4, 2.2] Pulse pressure 200 mg/day (1x) 3 171 -3.8 [-7.3, -0.3] 400 mg/day (2x) 3 (n=267) 2 120 -3.9 [-8.1, 0.3] 600 mg/day (4x) 1 90 -4.0 [-9.6, 1.6]

All three doses of celiprolol significantly lowered both systolic blood pressure and diastolic blood pressure compared to placebo. The direct comparisons of dose effect did not find significant difference between 200 mg/day and 400 mg/day in both systolic blood pressure and diastolic blood pressure (test for subgroup differences, I2 = 0%, p=0.9 for systolic blood pressure and diastolic blood pressure). The overlapped 95% CIs agreed with such findings. Trafford 1989 was the only study that report information on heart rate in celiprolol. 200 mg/day and 400 mg/day celiprolol did not significantly lowered heart rate

67 compared to placebo. Celiprolol 200 mg/day was the only subgroup that significantly lowered pulse pressure compared to placebo.

5.7 Effects of alprenolol on systolic blood pressure, diastolic blood pressure, heart

rate and pulse pressure

Alprenolol is not available in Canada, the U.S. or the European Union. We did not find the product monograph from these government agencies, therefore we were unable to provide the indication and recommended dosage for alprenolol. Two studies examining the blood pressure lowering efficacy of 400 mg/day alprenolol in 27 hypertensive patients were included in this analysis. Both of the studies were crossover studies with duration ranging from 8 to 10 weeks. Please refer to table 5.5 for the results of alprenolol. 400 mg/day alprenolol significantly lowered systolic blood pressure, diastolic blood pressure, heart rate and pulse pressure compared to placebo.

Table 5.5: Dose ranging BP, heart rate and pulse pressure lowering efficacy of alprenolol

Dosage (multiples of starting Total # # RCT in N in Mean estimate of dose) RCT (N) subgroup subgroup difference [95% CI] Systolic blood pressure 400 mg/day 2 (n=27) 2 27 -16.9 [-23.9, -10.0] Diastolic blood pressure 400 mg/day 2 (n=27) 2 27 -6.7 [-10.1, -3.4] Heart rate 400 mg/day 1 (n=16) 1 16 -8.0 [-15.5, -0.6] Pulse pressure 400 mg/day 2 (n=27) 2 27 -8.7 [-14.8, -2.7]

68

5.8 Effects of bopindolol on systolic blood pressure, diastolic blood pressure, heart

rate and pulse pressure

Bopindolol is not available in Canada, the U.S. or the European Union. We did not find the product monograph of bopindolol from any government agencies therefore we cannot provide the indication and recommended dosage for bopindolol. Bopindolol is a prodrug of pindolol with an ester. One parallel study examining blood pressure lowering efficacy of 0.5 to 2 mg once daily bopindolol for 4 weeks in 117 hypertensive patients was included in this review. A hundred and fifteen out of 117 randomized patients completed the studies.

69

Table 5.6: Dose ranging BP, heart rate and pulse pressure lowering efficacy of bopindolol

Dosage (multiples of starting Total # # RCT in N in Mean estimate of dose) RCT (N) subgroup subgroup difference [95% CI] Systolic blood pressure 0.5 mg/day 1 55 -5.9 [-15.8, 4.0] 1 mg/day 1 (n=115) 1 60 -6.1 [-15.6, 3.4] 2 mg/day 1 60 -5.8 [-14.6, 3.0] Diastolic blood pressure 0.5 mg/day 1 55 -1.5 [-6.8, 3.8] 1 mg/day 1 (n=115) 1 60 -4.0 [-9.1, 1.1] 2 mg/day 1 60 -2.4 [-7.2, 2.4] Heart rate 0.5 mg/day 1 55 -2.2 [-7.3, 2.9] 1 mg/day 1 (n=115) 1 60 -5.6 [-10.1, -1.0] 2 mg/day 1 60 -6.0 [-10.2, -1.9] Pulse pressure 0.5 mg/day 1 55 -4.8 [-13.4, 3.8] 1 mg/day 1 (n=115) 1 60 -2.1 [-10.4, 6.2] 2 mg/day 1 60 -3.4 [-11.1, 4.3] WDAE All dose combined 1 117 RR 0.7 [0.1, 7.7]

Bopindolol, in all doses, did not significantly lowered systolic blood pressure or diastolic blood pressure compared to placebo. Bopindolol 1 mg/day and 2 mg/day significantly lowered heart rate compared to placebo. Bopindolol did not significantly changed pulse pressure compared to placebo in all doses tested.

5.8.1 Withdrawal due to adverse effects

Moleur 1988 was the only study that reported usable WDAE data in this review.

Bopindolol was not significantly different from placebo in terms of WDAE (RR 0.72 [0.07,

7.67]).

70

5.9 Effects of oxprenolol on systolic blood pressure, diastolic blood pressure and

pulse pressure

Oxprenolol was discontinued in Canada and the EU. It was indicated for treatment of hypertension in the EU. The recommended dose of oxprenolol for treatment of hypertension is 80 to 320 mg once daily [27]. One parallel study examining blood pressure lowering efficacy of 20 to 80 mg once daily oxprenolol in 66 hypertensive patients for 4 weeks was included in this review. Please refer to table 5.7 for the results of oxprenolol.

Table 5.7: Dose ranging BP, heart rate and pulse pressure lowering efficacy of oxprenolol

Dosage (multiples of starting Total # # RCT in N in Mean estimate of dose) RCT (N) subgroup subgroup difference [95% CI] Systolic blood pressure 20 mg/day (0.25x) 1 26 3.4 [-8.6, 15.4] 40 mg/day (0.5x) 1 (n=66) 1 27 -6.2 [-17.9, 5.5] 80 mg/day (1x) 1 39 -11.5 [-21.6, -1.4] Diastolic blood pressure 20 mg/day (0.25x) 1 26 4.2 [-2.3, 10.7] 40 mg/day (0.5x) 1 (n=66) 1 27 0.4 [-6.0, 6.8] 80 mg/day (1x) 1 39 -3.3 [-8.9, 2.3] Pulse pressure 20 mg/day (0.25x) 1 26 5.3 [-5.1, 15.7] 40 mg/day (0.5x) 1 (n=66) 1 27 0.7 [-9.5, 10.9] 80 mg/day (1x) 1 39 2.7 [-7.7, 13.1]

Oxprenolol at 80 mg/day (the starting dose) significantly lowered systolic blood pressure compared to placebo. No other doses of the oxprenolol significantly lowered systolic blood pressure or diastolic blood pressure compared to placebo. Motolese 1975 did not report any data on heart rate. Oxprenolol did not significantly changed pulse pressure in any dose subgroup.

71

5.10 Pooled effects of partial agonists on systolic blood pressure, diastolic blood

pressure, heart rate and pulse pressure

We pooled the results of partial agonists based on the manufacturer's recommended starting doses (table 5.8). This allowed us to analyze the effect of partial agonists as a whole subclass.

Table 5.8: Dose ranging BP, heart rate and pulse pressure lowering efficacy of partial agonists

Dosage (multiples of starting Total # # RCT in N in Mean estimate of dose) RCT (N) subgroup subgroup difference [95% CI] Systolic blood pressure 0.25x starting dose 1 26 3.4 [-8.6, 15.4] 0.5x starting dose 2 82 -6.2 [-12.4, 0.0] 1x starting dose 11 (n=578) 10 384 -8.1 [-10.4, -5.8] 2x starting dose 3 180 -8.6 [-13.4, -3.8] 4x starting dose 2 106 -11.0 [-16.0, -6.1] Diastolic blood pressure 0.25x starting dose 1 26 4.2 [-2.3, 10.7] 0.5x starting dose 2 82 0.4 [-6.0, 6.8] 1x starting dose 11 (n=578) 10 384 -3.8 [-5.0, -2.6] 2x starting dose 3 180 -4.9 [-7.3, -2.5] 4x starting dose 2 106 -6.7 [-9.7, -3.7] Heart rate 0.5x starting dose 1 55 -2.2 [-7.3, 2.9] 1x starting dose 7 (n=255) 7 200 -6.0 [-7.4, -4.5] 2x starting dose 2 86 -2.3 [-4.6, 0.0] Pulse pressure 0.25x starting dose 1 26 5.3 [-5.1, 15.7] 0.5x starting dose 2 82 -2.5 [-9.1, 4.0] 1x starting dose 11 (n=578) 10 384 -3.6 [-5.5, -1.6] 2x starting dose 3 180 -3.8 [-7.4, -0.2] 4x starting dose 2 106 -3.9 [-8.2, 0.0]

72

Partial agonists in 0.25x and 0.5x starting contained little data and generally did not significantly lower systolic blood pressure or diastolic blood pressure compared to placebo although the 0.5x starting does subgroup showed marginal significant result. Starting from the recommended starting dose, partial agonists significantly lowered systolic blood pressure and diastolic blood pressure compared to placebo. Test for subgroup differences within the recommended dose range (1x, 2x and 4x the starting dose) by direct comparison was not significant for systolic blood pressure (p= 0.84) or diastolic blood pressure (p=0.74). This suggested that higher dose partial agonists did not offer greater BP lowering effect compared to the recommended starting dose of partial agonists.

The recommended starting dose of partial agonists was the only dose that significantly lowered heart rate compared to placebo. Test for subgroup differences for heart rate using direct comparison was not significant. We calculated the pulse pressure by subtracting diastolic blood pressure from systolic blood pressure. Partial agonists at starting and 2x starting doses significantly lowered pulse pressure. 4x starting dose subgroup showed similar point estimate but the 95% CI was wide and not statistically significant.

Only one study (bopindolol) reported WDAE. Bopindolol was not significantly different compared to placebo in terms of WDAE. It is a small study with 117 patients taking beta blockers for 4 weeks. We did not find any data on WDAE for other partial agonists.

5.10.1 Blood pressure variability

We tested the BP variability by paired t-test comparing the standard deviations of the partial agonist group and the placebo group. The average end treatment standard deviation of systolic blood pressure in partial agonist group was 16.3 and in placebo group was 14.8. The average standard deviation of diastolic blood pressure for the treatment group was 8.4 and

73 placebo group was also 8.4. The p-value for the paired t-test for systolic blood pressure was

0.02 and for diastolic blood pressure was 0.64. The result suggested that partial agonists might increase variability of systolic blood pressure but not diastolic blood pressure. We only included a small sample of 9 end-treatment standard deviations for each group.

Therefore, more data is needed in order to answer this question.

5.10.2 Subgroup and sensitivity analysis

Due to lack of data and small sample size, we were not able to perform any of the subgroup or sensitivity analyses.

5.11 Discussion

Celiprolol was the most studied partial agonist in this review. The BP and heart rate lowering profile of celiprolol reflected the effects that were expected from a partial agonist.

Celiprolol significantly lowered systolic blood pressure and diastolic blood pressure compared to placebo while having no significant effect on heart rate. The test for subgroup differences by direct comparison was not significant. This suggested that higher dose celiprolol did not lower blood pressure more than the lower dose.

Motolese 1975 was the only study included in oxprenolol analyses. The baseline of the 60 and 80 mg/day subgroups (188/112 mmHg) were much higher than placebo (175/109 mmHg). The unbalance baseline BP raised concerns for the adequacy of randomization.

Since no other oxprenolol subgroup significantly lowered BP and the risk of selection bias was high in this subgroup, the effect in the 60 & 80 mg/day subgroup could be an exaggeration. This subgroup only had a small sample size and did not significantly affect the overall pooled results.

74

The sample sizes of other partial agonists were too small to draw any definitive conclusion. It is troubling that acebutolol and pindolol, which are approved Canada or the

U.S. to treat hypertension, provided little publicly available data.

5.11.1 Pooled subclass effect of partial agonists

The patients enrolled in the studies generally had moderate to severe hypertension.

Most of the studies were published back in the 70s and 80s. The definition of hypertension at that time was >160/100 mmHg. Therefore, the mean baseline BP of the included studies

(175/107 mmHg) was higher compared to other subclasses.

Only two of the 13 included studies specified whether BP was measured at peak (1-

12 hours after dose) or trough hours (13 to 24 hours after dose). However, the fact that many of the partial agonists used in the studies were taken two to three time a day, it is reasonable to assume that BP was measured at peak hours in these studies.

In the pooled data, 0.25 and 0.5 times the recommended starting dose did not significantly lowered systolic blood pressure or diastolic blood pressure. The recommended starting dose and higher doses significantly lowered both systolic blood pressure and diastolic blood pressure. Direct comparison showed no significant different in BP lowering effect between the starting dose and higher doses. The 1x and 2x starting dose subgroups contained the largest sample size. The estimate of BP lowering efficacy for partial agonists by combining the 1x and 2x starting dose subgroup was -8/-4 mmHg.

5.11.2 Heart rate

The combined effect of 1x and 2x starting dose subgroups on heart rate for partial agonist (-5 beats per minute) was smaller than non-selective beta blockers (-9 beats per

75 minute). This result suggests that partial agonists might be less likely to cause bradycardia than other non-selective beta blockers.

The rationale of developing partial agonists was that having agonistic properties, partial agonists could produce fewer side effects of beta antagonism, such as bradycardia, fatigue, and rebound hypertension after withdrawal. The available data did not answer whether this was the case as many trials did not report the effect on heart rate or WDAE.

5.11.3 Pulse pressure

Partial agonists significantly lowered pulse pressure compared to placebo. The relatively small effect on diastolic blood pressure could help explain why partial agonist, with smaller sample size, was the only beta blocker subclass that showed an effect on pulse pressure of this magnitude. The agonistic property of partial agonists on peripheral beta-2 receptors causing vasodilation and improving vascular compliance which would lower pulse pressure and possibly explain this finding.

5.11.4 Blood pressure variability

In this review, we found that the systolic blood pressure end treatment standard deviation of partial agonists was significantly higher compared to placebo. We had a small sample size and thus it was difficult to draw a definitive conclusion. Only 6 of 13 studies provided the end treatment standard deviations in the partial agonist and placebo groups.

Attempts to contact the authors in order to obtain this information were not successful. In addition, measuring BP at peak hours might increase variability. It was possible that the higher variability in systolic blood pressure might disappear when BP was measured at trough hours. Large, well conducted studies comparing the BP variability of partial agonists and placebo are needed in order to shed more light on this matter.

76

Figure 5.2: Risk of bias summary of partial agonists

High risk of bias

Low risk of bias Unknown risk of bias

5.12 Risk of bias and quality of evidence

The risk of bias summary was presented in figure 5.2. In the figure, a red bubble represented high risk of bias, yellow bubble represented unknown risk of bias and green bubble represented low risk of bias. The overall quality of evidence was summarized below.

77

5.12.1 Allocation (selection bias)

The procedure of randomization was poorly reported in most studies. We examined the baseline characteristics of each groups in parallel studies. The baseline BP was significantly difference in Motolese 1975 between the groups. This raised the concern about selection bias. This issue was mentioned in the discussion. We did not find any reason to suspect that high risk of selection bias in any other studies. In crossover studies, information was not sufficient to make any judgment on selection bias.

5.12.2 Blinding (performance bias and detection bias)

Beta blockers generally lowered heart rate. For this reason, blinding could be compromised if the investigators had used mercury sphygmomanometer to measure blood pressure. Only one study used automated machine which would mitigate the risk of detection bias. This risk of detection bias remained high in this review. The detail on the potential risk of detection bias caused by lower heart rate was described in section 7.9.3.

5.12.3 Incomplete outcome data (attrition bias)

The blood pressure and heart rate were reported by all the included studies. Due to the short duration of the studies, the dropout rate was low. Most of the patients randomized were included into the analyses. We judged that the risk of attrition bias was low.

5.12.4 Selective reporting (reporting bias)

Systolic blood pressure and diastolic blood pressure were reported in all the included studies. They were reported as the end treatment values for each intervention group. Some of the studies did not report heart rate. The difference in the effect on heart rate compared to non-selective beta blockers could potentially distinguish partial agonists from other classes.

Failure to report heart rate could be an indication of potential bias in reporting.

78

Although the dropout rate was low, only one study reported WDAE. The overall subclass assessment of adverse effects serious enough to cause withdrawal could not be done due to lack of data. WDAE is an important outcome to assess drug tolerability, particularly in short term studies. The selective reporting of WDAE raises concern about the risk of reporting bias. The reason why selective reporting of WDAE could lead to reporting bias is described in section 7.6.4.

5.12.5 Publication bias

The sample size of this review was the smallest among the four beta blockers subclasses. Acebutolol and pindolol are approved to treat hypertension in Canada and the

United States. However, we only found small studies examining these partial agonists. The lack of data on drugs that are marketed in North America meant that many trials conducted to achieve licensing had not been published. The lack of incentive to publish these studies might suggest that the real effect of partial agonists could be less than the current estimate.

These two drugs showed similar problem of selective publishing as labetalol. The reason why lack of publicly available data for approved medication could lead to high risk of publication bias and the possible impact on evidence is described in section 7.6.5.

5.12.6 Quality of the evidence

Table 5.9 summarizes the combined effect size of the combined starting and 2x the starting dose of partial agonists. In addition it provides a judgment of the quality of evidence in this review. Included in this review were 13 studies that examined the BP lowering effect of partial agonists in 605 middle aged hypertensive patients. The average baseline blood pressures were mostly in the severely elevated range.

79

Only one study used automated machine to measure blood pressure. This suggests that the risk of detection bias is high due to loss of blinding in this review. Because of the high risk of detection bias, the quality of evidence was downgraded by one level.

Due to the small sample size of this review, the estimates are likely to change when additional data are made available in the future. This is the reason that the quality of evidence was downgraded by another level.

In addition, search for acebutolol and pindolol trials, which are approved for treatment of hypertension in North America, had yielded only a small amount of data. This raises suspicion that large trials required to achieve licensing was not published. Therefore the risk of publication bias in this review is high which downgraded the quality of evidence by another level. For these reasons, the estimates of the BP lowering effect shown in the summary of findings table are very likely an exaggeration of the true effect. This is reflected in the Table by the very low judgment of the quality of the evidence.

80

Table 5.9: Summary of findings table for partial agonists

Summary of findings table Population: Adult patients with primary hypertension Intervention: Partial agonists Comparison: Placebo mean estimates of combining 1x Number of patients in the Quality of Outcomes and 2x starting dose subgroup (# of studies) evidence [95% CI] SBP -8.1 [-10.1, -6.1]1,2,3,4 490 (10) Very low5,6,7 DBP -4.0 [-5.1, -2.9]1,2,3,4 490 (10) Very low5,6,7 HR -4.9 [-6.2, -3.7]2,3 256 (7) Very low5,6,7 PP -3.6 [-5.3, -1.9]1,2 490 (10) Very low5,6,7,8 Acronyms and grading guide SBP: systolic blood pressure; DBP: diastolic blood pressure; HR: heart rate; PP: pulse pressure; 95%CI: 95% confident interval GRADE Working Group grades of evidence [29] High quality: Further research is very unlikely to change our confidence in the estimate of effect. Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate. Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate. Very low quality: We are very uncertain about the estimate. Footnotes 1. Test for subgroup differences showed that 1x and 2x starting dose subgroups were not significantly different from each other. 2. The recommended starting and 2x starting contained most of the data in dual receptor blockers analysis. Combining them provided estimates that represent the overall subclass effect. 3. BP was measured at peak hours. 4. Weighted mean baseline BP of included studies was 175/107 mmHg. 5. Quality of evidence was downgraded by one level due to small sample size and less precise estimates. 6. Quality of evidence was downgraded by one level due to high or unknown risk of detection bias caused by loss of blinding. 7. Quality of evidence was downgraded by one level due to high risk of publication bias. 8. Quality of evidence should be even lower due to indirectness. None of the studies included reported pulse pressure. Pulse pressure was calculated by subtracting DBP from SBP.

81

6. RESULTS FOR BETA-1 SELECTIVE BLOCKERS

6.1 Search findings

The search finding and flow of study selection is the same as stated in the non- selective beta blocker chapter. In total, 55 studies were included in this beta-1 selective blocker review. Figure 6.1 summarized the flow of study selection.

82

Figure 6.1: PRISMA diagram of beta-1 selective blocker

83

6.2 Characteristics of included studies

Fifty five RCTs examining the BP lowering efficacy of 8 beta-1 blockers in 6,426 primary hypertensive patients were included in this review. Twenty five RCTs were parallel studies and the other 30 RCTs were crossover studies with duration of treatment ranging from 3 to 16 weeks. When study duration was longer than 12 weeks, only the data obtained between 3 to 12 weeks was used. The mean baseline BP of the patients randomized in the studies was 155.6/101.1 mmHg. Please refer to appendix F for detail of included studies.

6.3 Characteristics of excluded studies

One study that potentially meets the inclusion criteria was excluded for reason listed in table 6.1.

Table 6.1: Characteristics of excluded study in beta-1 selective blockers review

Study Reason for exclusion Non-hypertensive and hypertensive patients were mixed and randomized. Wald 2008 [156] The data of hypertensive patients were not reported separately.

6.4 Effects of nebivolol on systolic blood pressure, diastolic blood pressure, heart

rate and pulse pressure

Nebivolol is a beta-1 selective blocker which was recently marketed in Canada in early 2013. Nebivolol is indicated for treatment of hypertension in Canada and the United

States [26, 33]. The recommended doses of nebivolol are 5 mg to 20 mg daily [33]. 12 RCTs examining the blood pressure lowering efficacy of 1 mg to 40 mg per day nebivolol in 3,209 hypertensive patients were included in this review. Eight of them were parallel studies and the other 4 were crossover studies. In addition, NEB-305 [137] was an unpublished RCT from an FDA report. The mean baseline BP of the patients in the included studies was

154.6/100.4 mmHg. Please refer to table 6.2 for nebivolol results.

84

Table 6.2: Dose ranging BP, heart rate and pulse pressure lowering efficacy of nebivolol

Dosage (multiples of starting Total # # RCT in N in Mean estimate of dose) RCT (N) subgroup subgroup difference [95% CI] Systolic blood pressure 1.0, 1.25 mg/day (0.25x 3 366 -4.5 [-7.3, -1.8] starting) 2.5 mg/day (0.5x starting) 3 415 -5.0 [-7.7, -2.3] 5 mg/day (starting dose) 12 12 1280 -8.8 [-10.2, -7.4] (N=2,941) 10 mg/day (2x starting) 5 855 -5.3 [-7.2, -3.3] 20 mg/day (4x starting) 3 652 -7.0 [-9.3, -4.6] 30, 40 mg/day (8x starting) 2 331 -8.3 [-11.2, -5.4] Diastolic blood pressure 1.0, 1.25 mg/day (0.25x 3 366 -3.6 [-5.2, -1.9] starting) 2.5 mg/day (0.5x starting) 3 415 -4.2 [-5.8, -2.7] 5 mg/day (starting dose) 12 12 1280 -6.7 [-7.5, -5.8] (N=2,941) 10 mg/day (2x starting) 5 855 -5.8 [-7.0, -4.7] 20 mg/day (4x starting) 3 652 -5.8 [-7.2, -4.4] 30, 40 mg/day (8x starting) 2 331 -7.2 [-9.0, -5.4] Heart rate 1.0, 1.25 mg/day (0.25x 1 136 -2.9 [-5.86, 0.1] starting) 2.5 mg/day (0.5x starting) 2 218 -3.9 [-6.2, -1.6] 5 mg/day (starting dose) 7 7 487 -8.2 [-9.7, -6.8] (N=1,039) 10 mg/day (2x starting) 2 291 -5.9 [-8.0, -3.9] 20 mg/day (4x starting) 2 298 -8.4 [-10.5, -6.4] 30, 40 mg/day (8x starting) 2 331 -9.2 [-11.2, -7.2] Pulse pressure 1.0, 1.25 mg/day (0.25x 3 366 -0.9 [-3.2, 1.5] starting) 2.5 mg/day (0.5x starting) 3 415 -1.0 [-3.3, 1.4] 5 mg/day (starting dose) 12 12 1280 -1.6 [-2.8, -0.4] (N=2,941) 10 mg/day (2x starting) 5 855 0.5 [-1.2, 2.2] 20 mg/day (4x starting) 3 652 -1.3 [-3.4, 0.9] 30, 40 mg/day (8x starting) 2 331 -1.1 [-3.7, 1.6] All nebivolol doses significantly lowered trough systolic blood pressure and diastolic blood pressure compared to placebo. Test for subgroup differences by direct comparison included five large RCTs with multiple dosage subgroups. There was no dose related effect

85 within the recommended dose range (1x, 2x, 4x starting dose) for systolic blood pressure

(p=0.47) and diastolic blood pressure (p=0.52). The maximum BP lowering effect was showed at 5 mg/day (1x starting). Since there was no dose response within the recommended doses, we pooled the three dosages subgroups (5 mg, 10 mg and 20 mg) to calculate the estimates for blood pressure lowering efficacy. The estimate for blood pressure lowering efficacy for nebivolol is -7.5/-6.3 mmHg. Heterogeneity was significant in these subgroups and the validity of this estimate is discussed further in the discussion of this chapter.

Nebivolol 1.25 mg/day did not significantly lower heart rate. Starting from 2.5 mg/day, nebivolol significantly lowered heart rate compared to placebo. Heterogeneity in the

5 mg/day subgroup was significant. Test of subgroup difference by direct comparison was significant (p=0.0004) for heart rate. Only nebivolol 5 mg/day significantly lowered pulse pressure. However, the 5 mg/day subgroup was also the only subgroup in which heterogeneity was significant.

Four large studies provided both peak and trough data in the same patients. We compared the difference in peak and trough effect for systolic blood pressure and diastolic blood pressure. Peak measurements were not significantly different from trough measurements for systolic blood pressure. The estimated mean differences varied between -

1.2 to 1.4 mmHg. But the BP lowering effect at peak was significantly greater than trough measurements for diastolic blood pressure, averaging 2 mmHg.

The blood pressure variability was not significantly different between nebivolol and placebo for systolic blood pressure (p=0.61) and diastolic blood pressure (p=0.52).

86

6.5 Effects of atenolol on systolic blood pressure, diastolic blood pressure, heart rate

and pulse pressure

Atenolol is indicated for the treatment of hypertension and angina in Canada and the

United States [26, 33]. The recommended doses for hypertension are 50 mg to 100 mg daily

[26]. 23 RCTs examining the blood pressure lowering efficacy of 25 to 200 mg per day atenolol in 1,119 hypertensive patients were included in this review. Seven of the included studies were parallel studies and the other 16 were crossover studies. The mean baseline BP for the atenolol studies was 162.3/104.2 mmHg. Please refer to table 6.3 for atenolol results.

87

Table 6.3: Dose ranging BP, heart rate and pulse pressure lowering efficacy of atenolol

Dosage (multiples of starting Total # # RCT in N in Mean estimate of dose) RCT (N) subgroup subgroup difference [95% CI] Systolic blood pressure 25 mg/day (0.5x starting) 1 87 -7.0 [-12.5, -1.5] 50 mg/day (1x starting) 22 11 669 -10.3 [-11.9, -8.6] 100 mg/day (2x starting) (N=1,119) 12 495 -15.3 [-16.9, -13.8] 150, 200 mg/day (4x starting) 3 121 -11.3 [-14.0, -8.5] Diastolic blood pressure 25 mg/day (0.5x starting) 1 87 -4.0 [-6.7, -1.2] 50 mg/day (1x starting) 22 11 669 -7.8 [-8.8, -6.7] 100 mg/day (2x starting) (N=1,119) 12 495 -12.9 [-13.9, -11.9] 150, 200 mg/day (4x starting) 3 121 -8.8 [-10.6, -6.9] Heart rate 25 mg/day (0.5x starting) 1 87 -4.0 [-9.6, 1.6] 50 mg/day (1x starting) 17 7 470 -12.0 [-13.4, -10.7] 100 mg/day (2x starting) (N=846) 10 413 -13.7 [-14.8, -12.6] 150, 200 mg/day (4x starting) 4 129 -18.3 [-20.3, -16.4] Pulse pressure 25 mg/day (0.5x starting) 1 87 -3.0 [-7.8, 1.8] 50 mg/day (1x starting) 22 11 669 -2.3 [-3.7, -0.9] 100 mg/day (2x starting) (N=1,119) 12 495 -2.0 [-3.2, -0.8] 150, 200 mg/day (4x starting) 3 121 -2.6 [-4.8, -0.4]

All atenolol doses significantly lowered systolic blood pressure and diastolic blood pressure compared to placebo. The maximum BP lowering effect was shown at 100 mg/day

(2x starting). Test for subgroup differences within the recommended dose range (1x, 2x and

4x starting dose) by direct comparison was not significant for systolic blood pressure

(p=0.56) and diastolic blood pressure (p=0.22). However, only two small studies provided information for direct comparison. Test for subgroup differences by indirect comparison was not significant for systolic blood pressure (p=0.31) but significant for diastolic blood

88 pressure (p=0.04). Given this inconsistency, it is inconclusive whether atenolol shows a dose response effect.

Significant heterogeneity was present for both the 50 mg/day and the 100 mg/day subgroups for systolic blood pressure and diastolic blood pressure. The 50 mg/day (starting dose) and 100 mg/day (2x starting) subgroups contained the largest sample size. The estimate of blood pressure lowering efficacy combining 1x and 2x starting dose was -13/-11 mmHg.

Atenolol 25 mg/day did not significantly lower heart rate. Starting from 50 mg/day and higher, atenolol significantly lowered heart rate compared to placebo. Test for subgroup differences in the recommended dose range by direct comparison was significant for heart rate (p=0.008).

Atenolol 50 mg/day and higher significantly lowered pulse pressure compared to placebo. There were no differences between 50 mg/day, 100 mg/day and 200 mg/day by direct comparison for pulse pressure.

Blood pressure variability was not significantly different between the treatment and placebo groups for systolic blood pressure (p=0.3) and diastolic blood pressure (p=0.13).

6.6 Effects of metoprolol on systolic blood pressure, diastolic blood pressure, heart

rate and pulse pressure

Metoprolol is indicated for the treatment of hypertension, angina and stable acute MI in Canada and the United States. The recommended doses for treatment of hypertension are

100 to 200 mg daily in Canada and 100 to 450 mg daily in the United States [26, 33]. Nine

RCTs examining the blood pressure lowering efficacy of 25 to 400 mg per day metoprolol in

1,004 hypertensive patients were included in this review. Four of the included studies were

89 parallel studies and the other 5 were crossover studies. The mean baseline BP of the metoprolol studies was 154.4/100.3 mmHg. Please refer to table 6.4 for metoprolol results.

Table 6.4: Dose ranging BP, heart rate and pulse pressure lowering efficacy of metoprolol

Dosage (multiples of starting Total # # RCT in N in Mean estimate of dose) RCT (N) subgroup subgroup difference [95% CI] Systolic blood pressure 25 mg/day (0.25x starting) 2 424 -6.3 [-9.2, -3.4] 50 mg/day (0.5x starting) 2 339 -6.3 [-9.5, -3.1] 9 100 mg/day (starting dose) 3 410 -5.4 [-8.1, -2.6] (N=1,004) 200 mg/day (2x starting) 5 284 -11.5 [-14.4, -8.7] 400 mg/day (4x starting) 2 193 -10.9 [-15.4, -6.4] Diastolic blood pressure 25 mg/day (0.25x starting) 2 424 -3.6 [-5.4, -1.9] 50 mg/day (0.5x starting) 2 339 -4.6 [-6.4, -2.7] 9 100 mg/day (starting dose) 3 410 -4.7 [-6.5, -2.9] (N=1,004) 200 mg/day (2x starting) 5 284 -10.3 [-11.9, -8.5] 400 mg/day (4x starting) 2 193 -7.7 [-10.7, -4.7] Heart rate 50 mg/day (0.5x starting) 1 94 -5.6 [-10.0, -1.2] 100 mg/day (starting dose) 1 24 -15.0 [-18.9, -11.1] 5 (N=166) 200 mg/day (2x starting) 2 40 -13.3 [-17.0, -9.5] 400 mg/day (4x starting) 1 8 -20.0 [-25.9, -14.1] Pulse pressure 25 mg/day (0.25x starting) 2 424 -2.7 [-5.2, -0.2] 50 mg/day (0.5x starting) 2 339 -1.7 [-4.5, 1.0] 9 100 mg/day (starting dose) 3 410 -0.6 [-3.0, 1.9] (N=1,004) 200 mg/day (2x starting) 5 284 -1.2 [-3.6, 1.2] 400 mg/day (4x starting) 2 193 -3.3 [-7.3, 0.7]

All metoprolol doses significantly lowered systolic blood pressure and diastolic blood pressure compared to placebo. The maximum BP lowering effect was seen at 200 mg/day

(2x starting). Test for subgroup differences in the recommended dose range by direct

90 comparison was not significant for both systolic blood pressure (p=0.12) and diastolic blood pressure (p=0.12). Significant heterogeneity was present in the systolic blood pressure 400 mg/day and diastolic blood pressure 200 mg/day subgroups. Since there was no definitive dose response within the recommended range, we pooled all the recommended dose, 100 mg,

200 mg/day and 400 mg/day, subgroups to estimate the blood pressure lowering effects. The estimate of blood pressure lowering effect of metoprolol was -8.8/-7.6 mmHg.

The sample size for heart rate was fairly small for metoprolol, as few studies reported heart rate. However, all metoprolol doses significantly lowered heart rate compared to placebo. Test for subgroup differences by indirect comparison was significant in heart rate

(p=0.0007). Metoprolol did not significantly change pulse pressure.

Metoprolol did not significantly change blood pressure variability compared to placebo for systolic blood pressure (p=0.56) or diastolic blood pressure (p=0.86).

6.7 Effects of bisoprolol on systolic blood pressure, diastolic blood pressure, heart

rate and pulse pressure

Bisoprolol is indicated for the treatment of hypertension in Canada and the United

States. The recommended doses for the treatment of hypertension are 5 mg to 20 mg/day in

Canada and the U.S. [26, 33]. Seven RCTs examining the blood pressure lowering efficacy of 5 to 20 mg/day bisoprolol in 622 hypertension patients were included in this review. Three of the included studies were parallel design and the other 4 were crossover design. The mean baseline BP was 151.2/100.1 mmHg. Please refer to table 6.5 for bisoprolol results.

91

Table 6.5: Dose ranging BP, heart rate and pulse pressure lowering efficacy of bisoprolol

Dosage (multiples of starting Total # # RCT in N in Mean estimate of dose) RCT (N) subgroup subgroup difference [95% CI] Systolic blood pressure 5 mg/day (1x starting) 5 422 -11.4 [-13.7, -9.1] 10 mg/day (2x starting) 6 (N=557) 2 134 -7.0 [-11.1, -2.9] 20 mg/day (4x starting) 1 121 -7.6 [-12.6, -2.6] Diastolic blood pressure 5 mg/day (1x starting) 6 462 -8.2 [-9.5, -6.8] 10 mg/day (2x starting) 7 (N=622) 3 173 -7.4 [-9.9, -5.0] 20 mg/day (4x starting) 1 121 -8.5 [-11.6, -5.4] Heart rate 5 mg/day (1x starting) 5 236 -6.9 [-8.8, -4.9] 10 mg/day (2x starting) 6 (N=432) 3 173 -10.2 [-12.9, -7.5] 20 mg/day (4x starting) 1 121 -11.1 [-14.9, -7.2] Pulse pressure 5 mg/day (1x starting) 5 422 -3.4 [-5.3, -1.4] 10 mg/day (2x starting) 6 (N=557) 2 134 1.7 [-1.9, 5.3] 20 mg/day (4x starting) 1 121 0.9 [-3.5, 5.3]

All doses of bisoprolol significantly lowered systolic blood pressure and diastolic blood pressure compared to placebo. There was significant heterogeneity for 5 mg/day

(starting dose) subgroup for both outcomes. Test for subgroup differences in the recommended dose range by direct comparison was not significant in systolic blood pressure

(p=0.76) and diastolic blood pressure (p=0.32). Since there was no significant difference between the subgroups, we combined the three subgroups to obtain the estimate of BP lowering of bisoprolol, -10/-8.1 mmHg.

All doses of bisoprolol significantly lowered heart rate compared to placebo. Test for subgroup differences by direct comparison was not significant in heart rate (p=0.12). Only 5

92 mg/day (starting dose) bisoprolol significantly lowered pulse pressure. This effect was not seen in other subgroups.

There was no significant difference in blood pressure variability between treatment and placebo for systolic blood pressure (p=0.66) or diastolic blood pressure (p=0.96).

6.8 Effects of betaxolol on systolic blood pressure, diastolic blood pressure, heart

rate and pulse pressure

Please refer to the table 6.6 for the betaxolol results. Betaxolol is indicated for treatment of hypertension in the United States. The recommended daily doses of betaxolol are 10 mg/day to 20 mg/day. Two studies examined the BP lowering efficacy of betaxolol at dosage of 5 mg/day to 20 mg/day in hypertensive 627 participants were included in this review. The duration of both studies was 4 weeks. Ameling 1991 [106] was a crossover study and Williams 1992 [158] was a parallel study. Both studies measured BP using mercury sphygmomanometer. Williams 1992 reported that they measured BP 24 hours after the last dose (trough). Ameling 1991 did not report the time of the measurement. The mean baseline BP of the included studies was 158.9/102.8 mmHg.

93

Table 6.6: Dose ranging BP, heart rate and pulse pressure lowering efficacy of betaxolol

Dosage (multiples of starting Total # # RCT in N in Mean estimate of dose) RCT (N) subgroup subgroup difference [95% CI] Systolic blood pressure 5 mg/day (1x starting) 1 100 -5.8 [-15.0, 3.4] 10 mg/day (2x starting) 2 (N=627) 1 98 -8.2 [-17.4, 1.0] 20 mg/day (4x starting) 2 429 -11.6 [-13.7, -9.5] Diastolic blood pressure 5 mg/day (1x starting) 1 100 -3.8 [-7.5, -0.1] 10 mg/day (2x starting) 2 (N=627) 1 98 -6.5 [-10.8, -2.2] 20 mg/day (4x starting) 2 429 -8.1 [-9.4, -6.8] Heart rate 5 mg/day (1x starting) 1 100 -7.4 [-17.0, 2.2] 10 mg/day (2x starting) 2 (N=627) 1 98 -8.2 [-17.4, 1.0] 20 mg/day (4x starting) 2 429 -14.4 [-16.4, -12.5] Pulse pressure 5 mg/day (1x starting) 1 100 -1.1 [-9.1, 6.9] 10 mg/day (2x starting) 2 (N=627) 1 98 -1.7 [-9.7, 6.3] 20 mg/day (4x starting) 2 429 -3.5 [-5.4, -1.6]

5 mg/day and 10 mg/day betaxolol did not significantly lowered systolic blood pressure compared to placebo. However, they significantly lowered diastolic blood pressure compared to placebo. 20 mg/day betaxolol significantly lowered both systolic blood pressure and diastolic blood pressure. Williams 1992 provided data for direct comparison between the dose subgroup. There was no significant difference between in BP lowering effect between the dose subgroups. Since there was no significant difference between the dosages, we pooled all three doses to estimate the mean effect. The estimated BP lowering effect of betaxolol is -11.2/-7.5 mmHg.

Only 20 mg/day betaxolol significantly lowered heart rate. In additional, 20 mg/day betaxolol also significantly lowered pulse pressure.

94

We were not able to perform analysis on BP variability because neither of the two included studies reported standard deviation.

6.9 Effects of bevantolol on systolic blood pressure, diastolic blood pressure, heart

rate and pulse pressure

Please refer to table 6.7 for the bevantolol results. Bevantolol is not available in

Canada, the U.S. or the E.U. We did not find the product monograph or recommended starting dose for bevantolol from these government agencies. One parallel study (Okawa

1986) examining the BP lowering efficacy of 100 mg/day to 400 mg/day bevantolol in 139 hypertensive patients for 6 weeks was included in this review.

95

Table 6.7: Dose ranging BP, heart rate and pulse pressure lowering efficacy of bevantolol

Dosage (multiples of starting Total # # RCT in N in Mean estimate of dose) RCT (N) subgroup subgroup difference [95% CI] Systolic blood pressure 100 mg/day 1 35 -13.0 [-29.6, 3.6] 200 mg/day 1 35 -9.0 [-25.6, 7.6] 1 (N=35) 300 mg/day 1 35 -7.0 [-23.6, 9.6] 400 mg/day 1 35 -8.0 [-24.6, 8.6] Diastolic blood pressure 100 mg/day 1 35 -5.00 [-15.5, 5.5] 200 mg/day 1 35 -9.00 [-19.5, 1.5] 1 (N=35) 300 mg/day 1 35 -9.00 [-19.5, 1.5] 400 mg/day 1 35 -10.0 [-20.5, 0.5] Heart rate 100 mg/day 1 35 -13.0 [-29.6, 3.6] 200 mg/day 1 35 -9.0 [-25.6, 7.6] 1 (N=35) 300 mg/day 1 35 -7.0 [-23.6, 9.6] 400 mg/day 1 35 -8.0 [-24.6, 8.6] Pulse pressure 100 mg/day 1 35 -8.0 [-22.5, 6.5] 200 mg/day 1 35 0.0 [-14.5, 14.5] 1 (N=35) 300 mg/day 1 35 2.0 [-12.6, 16.6] 400 mg/day 1 35 2.0 [-12.5, 16.5]

Bevantolol did not significantly lower systolic blood pressure, diastolic blood pressure, heart rate or pulse pressure compared to placebo. Bevantolol did not significantly change BP variability. The end treatment standard deviation for systolic blood pressure was

20.8 for bevantolol group and 19.8 for placebo group. The end treatment standard deviation for diastolic blood pressure is 13.1 for bevantolol group and 12.6 for placebo group.

96

6.10 Effects of pafenolol on systolic blood pressure, diastolic blood pressure, heart

rate and pulse pressure

Please refer to table 6.8 for pafenolol results. Pafenolol is not available in Canada, the

U.S. or the E.U. We did not find the product monograph or recommended starting dose for pafenolol in the website of these government agencies. Two studies examining the blood pressure lowering efficacy of 25 to 100 mg/day pafenolol in 161 hypertensive patients were included. Both studies were parallel studies with treatment periods of 4 weeks. The mean baseline BP was 161.1/109.6 mmHg.

Table 6.8: Dose ranging BP, heart rate and pulse pressure lowering efficacy of pafenolol

Dosage (multiples of starting Total # # RCT in N in Mean estimate of dose) RCT (N) subgroup subgroup difference [95% CI] Systolic blood pressure 25 mg/day 1 21 3.0 [-11.8, 17.8] 50 mg/day 2 (N=54) 2 33 -5.9 [-17.6, 5.7] 100 mg/day 1 14 -6.0 [-22.9, 10.9] Diastolic blood pressure 25 mg/day 1 21 -8.0 [-16.7, 0.7] 50 mg/day 2 (N=54) 2 33 -4.4 [-11.2, 2.4] 100 mg/day 1 14 -2.0 [-12.8, 8.8] Heart rate 25 mg/day 1 21 -3.0 [-12.9, 6.9] 50 mg/day 2 (N=54) 2 33 -8.1 [-15.9, -0.4] 100 mg/day 1 14 -20.0 [-32.6, -7.4] Pulse pressure 25 mg/day 1 21 11.0 [-1.8, 23.8] 50 mg/day 2 (N=54) 2 33 -1.8 [-11.9, 8.4] 100 mg/day 1 14 -4.0 [-18.9, 10.9] Pafenolol did not significantly lower systolic blood pressure, diastolic blood pressure or pulse pressure. Pafenolol 50 mg/day and 100 mg/day significantly lowered heart rate.

97

6.11 Effects of practolol on systolic blood pressure, diastolic blood pressure, heart

rate and pulse pressure

Please refer to table 6.9 for practolol results. Practolol is not available in Canada, the

U.S. or the E.U. We did not find the product monograph or recommended starting dose for pafenolol in the website of these government agencies. One crossover study examining the blood pressure lowering effect of 600 mg/day practolol in 24 hypertensive patients for 4 weeks was included in this review. The baseline BP was 182.9/123.3 mmHg.

Practolol 600 mg/day significantly lowered systolic blood pressure, diastolic blood pressure, heart rate and pulse pressure compared to placebo.

Table 6.9: Dose ranging BP, heart rate and pulse pressure lowering efficacy of practolol

Dosage (multiples of starting Total # # RCT in N in Mean estimate of dose) RCT (N) subgroup subgroup difference [95% CI] systolic blood pressure 600 mg/day 1 (N=24) 1 24 -21.2 [-29.3, -13.1] diastolic blood pressure 600 mg/day 1 (N=24) 1 24 -13.9 [-19.0, -8.8] Heart rate 600 mg/day 1 (N=24) 1 24 -14.0 [-21.5, -6.6] Pulse pressure 600 mg/day 1 (N=24) 1 24 -7.3 [-14.4, -0.2]

6.12 Pooled subclass effects on systolic blood pressure, diastolic blood pressure, heart

rate and pulse pressure

We pooled the data for all available beta-1 selective blocker together based on the recommended starting doses. This allowed us to estimate the blood pressure lowering effect of beta-1 selective blockers as a whole subclass, as well as to compare it to other classes of antihypertensive drugs. Please refer to table 6.10 for the overall beta-1 blockers results.

98

Table 6.10: Dose ranging BP, heart rate and pulse pressure lowering efficacy of beta-1 blockers

Dosage (multiples of starting Total # # RCT in N in Mean estimate of dose) RCT (N) subgroup subgroup difference [95% CI] Systolic blood pressure 0.25x starting 5 790 -5.3 [-7.3, -3.4] 0.5x starting 7 941 -5.7 [-7.6, -3.8] 1x starting 49 30 2879 -9.3 [-10.2, -8.4] 2x starting (N=6,248) 26 2197 -11.3 [-12.2, -10.3] 4x starting 9 1087 -8.9 [-10.5, -7.3] 8x starting 2 331 -8.3 [-11.2, -5.4] Diastolic blood pressure 0.25x starting 5 790 -3.6 [-4.8, -2.4] 0.5x starting 7 941 -4.3 [-5.3, -3.2] 1x starting 50 31 2907 -7.1 [-7.6, -6.5] 2x starting (N=6,313) 27 2236 -9.5 [-10.1, -8.9] 4x starting 9 1087 -7.2 [-8.2, -6.2] 8x starting 2 331 -7.2 [-8.9, -5.4] Heart rate 0.25x starting 1 136 -2.9 [-5.9, 0.1] 0.5x starting 5 499 -4.3 [-6.2, -2.5] 1x starting 36 20 1315 -9.8 [-10.5, -8.9] 2x starting (N=3357) 19 1346 -12.2 [-13.1, -11.4] 4x starting 8 556 -13.7 [-15.0, -12.4] 8x starting 2 331 -9.2 [-11.2, -7.2] Pulse pressure 0.25x starting 5 790 -1.7 [-3.4, 0.0] 0.5x starting 7 941 -1.5 [-3.1, 0.2] 1x starting 49 30 2913 -2.0 [-2.8, -1.2] 2x starting (N=6,248) 26 2197 -1.5 [-2.4, -0.7] 4x starting 9 1087 -1.8 [-3.2, -0.4] 8x starting 2 331 -1.0 [-3.7, 1.6]

All pooled beta-1 blockers doses significantly lowered systolic blood pressure and diastolic blood pressure compared to placebo. The starting dose and 2x starting subgroups

99 each contained over 2,000 patients which provided good estimates to represent this class of beta blockers. Heterogeneity was significant in these two subgroups. The source of heterogeneity is explored in discussion of this chapter. Test for subgroup differences in the

1x, 2x, 4x and 8x starting dose subgroups by direct comparison was not significant for systolic blood pressure (p=0.23) and diastolic blood pressure (p=0.11). The BP lowering estimate (SBP/DBP) combining the 1x and 2x starting dose subgroup was -10/-8 mmHg.

All doses of beta-1 blockers significantly lowered heart rate. Beta-1 selective blockers significantly lowered pulse pressure at the starting, 2x starting and 4x starting doses compared to placebo. The test for subgroup differences by direct comparison in pulse pressure was not significant.

6.12.1 Blood pressure variability

We tested the overall effect of beta-1 blockers on BP variability using an unpaired t- test. End treatment standard deviation from parallel studies was extracted for beta-1 blocker group and placebo group. Beta-1 blocker did not significantly change BP variability in systolic blood pressure (p=0.83) or diastolic blood pressure (p=0.52). The overall weighted mean end treatment standard deviation for beta-1 blocker (SBP/DBP) was 14.5/8.6 and placebo group was 14.9/8.5.

6.12.2 Withdrawal due to adverse effects

We pooled the data for WDAE. Out of the 55 included studies, only two studies reported WDAE data that could be used for the analysis. In these 2 RCTs there was no significant difference in WDAE between treatment and placebo (RR 0.85 [0.37, 1.96]).

100

6.13 Discussion

6.13.1 Nebivolol

Nebivolol 5 mg/day showed the maximum blood pressure lowering effect. Doses higher than 5 mg/day did not provide additional BP lowering effect. The data from 1x, 2x and 4x the starting dose were pooled to estimate the average BP lowering effect of nebivolol,

-7.5/-6.3 mmHg. The funnel plots for nebivolol showed a paucity of small negative studies

(Figure 6.2, Figure 6.3). This suggested that small negative studies were not published. The estimate above is likely an over estimation of the true effect. The funnel plots also showed several outliers on the left hand side of the graph. These positive outliers could also exaggerate the effect.

101

Figure 6.2: Funnel plot: Systolic blood pressure at 1x, 2x and 4x starting dose for

nebivolol

Number standardof error

Point estimates of BP lowering effect (mmHg)

Figure 6.3: Funnel plot: Diastolic blood pressure at 1x, 2x and 4x starting dose for

nebivolol

Number standardof error

Point estimates of BP lowering effect (mmHg)

102

Nebivolol lowered systolic and diastolic BP to a similar degree therefore it had only a small effect on pulse pressure. Heterogeneity in the 5 mg/day subgroup suggested that the statistically significant effect on pulse pressure could be caused by variation as no other dose of nebivolol showed a similar effect.

Peak to trough difference

Only 4 nebivolol studies reported both peak and trough measurements. If there was a greater effect at peak (seen with some diastolic BP measurements) the effect was small

(2mmHg).

6.13.2 Atenolol

Both the recommended starting dose and 2x the starting dose contained a large number of subjects and provided a good estimate of the blood pressure lowering efficacy of atenolol. Funnel plots were used to identify extreme outliers and assessment of bias. The funnel plots of the recommended starting dose did not show any extreme outliers or provide evidence of asymmetry which would suggest publication bias (Figure 6.4, Figure 6.5).

103

Figure 6.4: Funnel plot: Systolic blood pressure at starting dose for atenolol

Number standardof error

Point estimates of BP lowering effect (mmHg)

Figure 6.5: Funnel plot: Diastolic blood pressure at starting dose for atenolol

Number standardof error

Point estimates of BP lowering effect (mmHg)

In the 2x starting dose subgroup, the funnel plot identified Lischner 1987 [134] as an extreme outlier for both systolic blood pressure and diastolic blood pressure. Other studies from the same laboratory were also identified as extreme outlier in the non-selective beta blocker review (Ravid 1985 [53]). These data are therefore of questionable validity. This represented a common problem in performing systematic review and meta-analysis as such

104 studies lead to an overestimation of the magnitude of blood pressure lowering. If we removed the extreme outlier, the estimate of 2x starting dose decreased from -15/-13 mmHg to -12/-10 mmHg. Removing it would also decrease the overall estimate of the pooled atenolol 1x and 2x starting dose from -13/-11 mmHg to -11/-9 mmHg.

Atenolol was the only beta-1 blockers that significantly lowered pulse pressure across several doses. The combined estimate of the pulse pressure effect of 1x, 2x and 4x starting dose atenolol is small (-2.2 [1.4-3.1] mmHg). It did not show dose response effect and is unlikely to be clinically significant.

6.13.3 Metoprolol

Most of the data for metoprolol came from 2 studies which tested the dose response effect of the drug at several doses (Frishman 2006 [123], Papademetriou 2006 [139]). These two studies showed a greater BP lowering in the recommended dose range compared to lower doses (0.25x and 0.5x starting dose). However, the dosage within the recommended dose range did not show significant dose response effect. The mean BP lowering effect of the starting dose, 2 times and 4 times the starting dose was -9/-8 mmHg.

6.13.4 Bisoprolol

No additional BP lowering effect was seen for doses higher than the recommended starting dose of bisoprolol. The mean BP lowering effect of the starting dose, twice the starting dose and 4x starting dose was -11/-8 mmHg. This was likely to be exaggerated because of the presence of two extreme positive outliers (Deary 2001[119], Deary 2002

[120]) in the data.

105

6.13.5 Betaxolol, bevantolol, pafenolol and practolol

The sample sizes of these four beta-1 blockers were small. They contributed little weight to the overall pooled estimates. Their estimates are reported in the results.

6.13.6 Overall pooled blood pressure lowering effect of beta-1 blocker

Beta-1 selective blockers comprised the largest sample size of the four subclasses of beta blockers. The pooled data included 6,313 patients and multiple dosages. The data set provided the best opportunity to explore whether there is a graded dose response effect. The findings showed a similar and smaller BP lowering effect at 0.25x and 0.5x the starting dose, but then a flat and similar BP lowering for the starting dose, 2x, 4x and 8x the starting dose

(see Table 6.10). Twice the starting dose subgroups had the most data and exhibited considerable heterogeneity, which is explored below. The lack of a dose response suggested that high dose beta-1 blockers are not more effective in lowering BP.

In table 6.11, we demonstrate that the primary source of heterogeneity in 2x starting dose subgroup was the difference in BP lowering effect between the individual beta-1 blockers.

106

Table 6.11: BP lowering efficacy of 2x starting beta-1 blockers

Dosage (multiples of starting Total # # RCT in N in Mean estimate of dose) RCT (N) subgroup subgroup difference [95% CI] systolic blood pressure Atenolol 12 495 -15.3 [-16.9, -13.8] Metoprolol 5 284 -11.5 [-14.4, -8.7] 26 Betaxolol 2 429 -11.6 [-13.7, -9.5] (N=2197) Nebivolol 5 855 -5.3 [-7.2, -3.3] Bisoprolol 2 134 -7.0 [-11.1, -2.9] diastolic blood pressure Atenolol 12 495 -12.9 [-13.9, -11.9] Metoprolol 5 284 -10.3 [-11.9, -8.5] 27 Betaxolol 2 429 -8.1 [-9.4, -6.8] (N=2236) Nebivolol 5 855 -5.8 [-6.9, -4.7] Bisoprolol 3 173 -7.4 [-9.9, -5.0]

Atenolol showed the largest effect size among the five beta-1 blockers. In the discussion for atenolol, we explained that the estimate of 2x starting dose subgroup of atenolol could be exaggerated due to extreme outlier. The effect size of this atenolol subgroup would change from -15/-13 mmHg to -12/-10 mmHg if we removed the extreme outlier from the analysis. We also considered the fact that mean baseline BP of atenolol studies (162/104 mmHg) was higher than the other beta-1 blockers. In addition, most studies for atenolol, metoprolol and betaxolol measured BP at peak hours. Peak measurements could also contribute to a greater effect size.

Difference in pharmacodynamic properties could also have contributed to the difference in BP lowering efficacy. Heart rate reduction during exercise was used for many beta-1 blockers to test the potency of beta-1 blockade. Bisoprolol 10 mg (2x starting dose) had equivalent exercise heart rate percentage change from baseline compared to 50 mg atenolol (1x starting dose) and 100 mg metoprolol (1x starting dose) [157]. 5 mg nebivolol

107

(1x starting dose) had an equivalent effect compared to 100 mg atenolol (2x starting dose)

[158]. If beta-1 blockade was the dominant mechanism by which blood pressure is lowered by beta blockers, nebivolol should have lowered BP the most and bisoprolol the least. As this did not fit with the data potency of beta-1 blocking ability does not explain the difference in blood pressure lowering effect.

Beta-1 selectivity might explain the differences. Both nebivolol and bisoprolol are highly beta-1 selective. The beta-1/beta-2 selectivity ratios are 321 fold for nebivolol and 100 fold for bisoprolol [157, 158]. The selectivity ratios for atenolol, metoprolol and betaxolol are much less, at 35 fold, 40 fold and 20 fold respectively [157]. In this case, it appears that beta blockers with lower beta-1 selectivity lower BP by a greater magnitude.

The explanation for the difference in BP lowering effect likely lies within the mechanism by which beta blockers lowered blood pressure. However, studying the pharmacodynamic properties of beta blockers is notoriously difficult. The methods to test pharmacodynamic properties vary between research groups. The outcomes are most often not comparable to each other [28]. At this time, it is difficult to fully explain the observed differences in BP lowering effect between different beta-1 blockers.

The 1x and 2x starting dose subgroups contained the largest sample size. The estimate of BP lowering efficacy for beta-1 blockers by combining the 1x and 2x starting dose subgroup was -10/-8 mmHg.

6.13.7 Pulse pressure

Atenolol was the only beta-1 blocker that consistently lowered pulse pressure at different doses. This effect was also seen in the pooled analysis. Since, no other beta-1

108 blocker exhibited a similar effect on pulse pressure we are not convinced that there is an overall subclass effect on pulse pressure.

109

Figure 6.6: Risk of bias summary of beta-1 blockers

High risk of bias

Low risk of bias

Unknown risk of bias

110

6.14 Risk of bias assessments and quality of evidence

The risk of bias summary is presented in figure 6.6. In the figure, a red bubble represents high risk of bias, yellow bubble represents unclear risk of bias and green bubble represents a low risk of bias. The overall quality of evidence is summarized below.

6.14.1 Randomization and allocation concealment (selection bias)

Although many of the studies were published recently, the reporting of detail important to bias assessment was still poor in many cases. The detail regarding randomization procedure and allocation concealment was not well reported.

It was difficult to judge the potential for bias in the category. In the parallel studies, we examined the baseline characteristics of patients in each intervention groups. We did not find any significant difference in baseline characteristics between the intervention groups.

Hence, we did not find any evidence that suggested a high selection bias in the parallel studies. It was impossible to assess whether randomization was conducted properly in crossover studies, thus there is an unclear risk of selection bias in cross-over trials. Without more details, it is difficult to judge the overall potential for selection bias in this review.

6.14.2 Blinding (performance bias and detection bias)

Beta blockers generally lowered heart rate. Beta blockers are well known for their ability to lower heart rate. The assessor can detect the difference in heart rate when measuring blood pressure. Using an automated machine to measure BP would mitigate this risk of detection bias. However this was not done in most of the included trials in this review.

Therefore the risk of detection and performance bias is high in this review. The detail on the potential risk of detection bias caused by lower heart rate is described in section 7.9.3.

111

6.14.3 Incomplete outcome data (attrition bias)

Most of the studies reported the method by which they handled dropouts. The dropout rates were low and most of studies used an ITT analysis. Therefore we judged that the risk of attrition bias in this review was low.

6.14.4 Selective reporting (reporting bias)

All the studies included reported the systolic blood pressure, diastolic blood pressure and heart rate outcomes of the patients. WDAE is an important outcome in clinical trials.

Only two studies reported useful data on WDAE. The risk of reporting bias was high because most of the studies failed to report WDAE. The reason why selective reporting of WDAE could lead to reporting bias is described in section 7.6.4.

6.14.5 Publication bias

Funnel plots of the pooled data showed a paucity of small and less effective studies and positive outliers. The asymmetry of funnel plots indicated the high potential for publication bias. Some of the extreme outliers were large studies with a large effect that was questionable. It was not possible to prove that these trials were flawed, therefore we kept them in the review, however, they add to the suspicion of a high risk of bias. As a result BP lowering estimates of particularly twice the starting dose are likely to be overestimated.

112

6.14.6 Quality of evidence

The summary of findings table (table 6.12) summarizes the important findings of starting plus 2x starting doses of beta-1 selective blockers as well as the quality of evidence in this review. Fifty five RCTs examining the BP lowering efficacy of 8 beta-1 blockers in

6,426 hypertensive patients were included in this review. The large sample size provided adequate power to draw robust conclusions.

The risk of detection bias is of concern in beta blocker studies. Most of the included studies did not use automated machine when measuring BP which would mitigate the risk of detection bias. Therefore, the risk of detection bias by loss of blinding is high. Nebivolol and atenolol contributed a large portion of data to our pooled analysis. Both of the beta-1 blockers showed high risk of publication bias either due to skewed funnel plots or the presence of extreme outliers.

Pulse pressure data was reported in the trial. The data on pulse pressure did not come from direct measurement in the studies. We calculated pulse pressure by subtracting diastolic blood pressure from systolic blood pressure. Indirectness of this outcome is the reason the evidence was further downgraded for pulse pressure. For these reasons, the quality of evidence is judged to be low to very low in the summary of findings table.

113

Table 6.12: Summary of findings table for beta-1 blockers

Summary of findings table Population: Adult patients with primary hypertension Intervention: Beta-1 selective blockers Comparison: Placebo mean estimates of combining 1x Number of patients in the Quality of Outcomes and 2x starting dose subgroup (# of studies) evidence [95% CI] SBP -10.3 [-10.9, -9.6]1,2,3 4,301 (46) Low4,5 DBP -8.2 [-8.6, -7.8]1,2,3 4,336 (47) Low4,5 HR -11.1 [-11.7, -10.5]1,2 2,330 (32) Low4,5 PP -1.7 [-2.3, -1.2]1,2 4,301 (46) Very low4,5,6 Acronyms and grading guide SBP: systolic blood pressure; DBP: diastolic blood pressure; HR: heart rate; PP: pulse pressure; 95%CI: 95% confident interval GRADE Working Group grades of evidence [29] High quality: Further research is very unlikely to change our confidence in the estimate of effect. Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate. Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate. Very low quality: We are very uncertain about the estimate. Footnotes 1. The recommended starting and 2x starting contained most of the data in dual receptor blockers analysis. Combining them provided estimates that represent the overall subclass effect. 2. About half of the data of BP was measured at peak hours and the other half was measured at trough hours. 3. The mean baseline BP was 156/101 mmHg. 4. The quality of evidence was downgraded by one level due to publication bias, extreme outliers and significant heterogeneity within some subgroups. 5. The quality of evidence was downgraded by one level due to high risk of detection bias caused by breaking of blinding 6. Quality of evidence was downgraded one level due to indirectness, none of the studies included reported pulse pressure. Pulse pressure was calculated by subtracting DBP from SBP.

114

7. OVERALL DISCUSSION

The primary goal of this thesis was to quantify the best evidence of the blood pressure lowering effect of beta blockers. Included in our review was a large amount of data from 96

RCTs involving 9,803 hypertensive participants. The thesis is composed of the review of 4 beta blocker subclasses looking at the BP lowering effect of four beta blocker subclasses. In order to minimize bias, only double blinded randomized placebo controlled trials were included. The duration of treatment was limited to 3 to 12 weeks in order to minimize dropouts. Beta blocker treatments in the studies had to be fixed dose mono-therapy so that the effect of dose could be assessed.

These inclusion criteria were similar to the systematic review of diuretics [159], ACE inhibitors [72] and ARBs [73]. Thus it is possible to compare the results of these 4 subclasses with the other three classes.

7.1 What is the effect on blood pressure and heart rate compared to placebo?

The review provide the most up-to-date clinical evidence on dose related blood pressure lowering effects of beta blockers in hypertensive patients. It is the first publication to provide dose related estimates of blood pressure lowering efficacy for each beta-blocker subclass as well as for each individual beta blocker.

Most beta blockers significantly lowered BP compared to placebo. The BP lowering effects for the subclasses was estimated by combining the recommended starting dose and twice the starting dose subgroups of the pooled subclass data.

In patient with mild to moderate hypertension, non-selective beta blockers lowered peak BP by an average of -10/-7 mmHg and reduced heart rate by 12 beats per minute.

Propranolol and penbutolol were the two drugs that contributed to most of the data for non-

115 selective beta blockers. This estimate is likely exaggerated due to the presence of extreme outliers and other sources of bias. Mean estimate from studies published before 1985 were much higher compared to recent studies. Baseline BP, time of measurement and publication bias help to explain the differences in effect size.

Non-selective beta blockers did not show a convincing dose response for systolic blood pressure and diastolic blood pressure. Test for subgroup differences was not significant. The evidence for an increasing magnitude of effect at higher doses was weak.

However, it was evident that non selective beta blockers showed a dose response on heart rate. This suggests that non selective beta blockers cause more bradycardia at higher doses compared to lower doses. Non-selective beta blockers showed a small statistical significant effect on pulse pressure that was small (2 mmHg) and the same for all doses.

In patient with mild to moderate hypertension, dual receptor blockers lowered trough

BP by an average of -5/-4 mmHg and reduced heart rate by 4 beats per minute. Two of the carvedilol studies were unpublished data. The risk of publication bias was lower for carvedilol because we found these two large unpublished trials. Therefore, carvedilol provided a better estimate of BP lowering effect than the other dual receptor blocker, labetalol, and other beta blockers. The BP lowering estimate from combining carvedilol 1x and 2x starting doses is -4/-3 mmHg. Doses higher than the recommended starting dose did not provide significant additional BP reduction.

Test for subgroup differences by direct comparison was significant for heart rate. This showed that higher dose dual receptor blockers caused more bradycardia than lower doses.

Carvedilol did not significantly change pulse pressure.

116

In patients with moderate to severe hypertension, partial agonists lowered peak BP by an average of -8/-4 mmHg and reduced heart rate by 5 beats per minute. There was no evidence of significant additional BP lowering effect at doses higher than the starting doses.

The BP lowering effect appeared to be less than non-selective and beta-1 selective blockers in comparable trials and less than ACE inhibitors and angiotensin receptor blockers in patients with mild to moderate elevations in BP.

Partial agonists did not show a convincing dose response effect on heart rate. Higher dose partial agonists decreased heart rate to a lesser degree. Partial agonists were the only subclass of beta blockers that did not show a convincing dose response on heart rate.

Starting from the recommended dose, partial agonists consistently lowered pulse pressure.

The magnitude of pulse pressure reduction was -3 to -4 mmHg and similar for all doses.

In patient with mild to moderate hypertension, beta-1 selective blockers lowered BP by an average of -10/-8 mmHg and reduce heart rate by 11 beats per minute. Half of the beta-

1 blocker data measured BP at peak and the other half measured at trough hours. The beta-1 blockers that measured at peak hours showed a greater BP reduction compared to beta-1 blockers that measured at trough hours (-12/-9 mmHg vs -8/-7 mmHg). The maximum BP reduction of beta-1 blockers was shown at twice the starting doses. Beta-1 blockers did not show a graded dose response effect on systolic blood pressure and diastolic blood pressure over the recommended dose range.

A graded dose response of beta-1 blockers on heart rate was evident. Higher dose beta-1 blockers significant lower heart rate more than lower doses. Only atenolol consistently showed a significant effect on pulse pressure compared to placebo. The effect of atenolol on pulse pressure, -2 mmHg, is small.

117

7.1.1 Pulse pressure

None of the studies included in the review measured pulse pressure as an outcome.

We had to calculate the pulse pressure by subtracting diastolic blood pressure from systolic blood pressure. Beta blockers individually did not significantly change pulse pressure, with the exceptions of partial agonists and atenolol. It is important to note that BP was measured at peak hours for both partial agonists and atenolol. The trough hour pulse pressure effect could be smaller than that seen in this review. For both partial agonists and atenolol, the effect on pulse pressure was similar for all doses.

7.1.2 BP variability

We compared the end treatment BP variability between beta blocker and placebo groups in all available parallel studies included in the review. There were 10 large parallel studies with 4,755 participants across all 4 subclasses that provided end treatment standard deviation for analysis. From these 10 studies, we collected 35 standard deviation of treatment from different dose subgroups in the RCT and 10 standard deviation of placebo for both systolic blood pressure and diastolic blood pressure. The end treatment standard deviation of treatment groups was not significantly different from placebo groups for both systolic blood pressure (p=0.66) and diastolic blood pressure (p=0.81). The average standard deviation

(SBP±SD/DBP±SD) was 15.5±2.1/8.8±1.2 for treatment group and 15.1±1.9/8.9±1.3 for placebo group. The variance ratio (SBP/DBP) was 1.05/0.98.

We did not use standard deviation from crossover studies because crossover studies usually report the standard deviation of change. In addition, the quality of the data was not as good for parallel studies because the value of standard deviation in crossover studies varied greatly from study to study.

118

Webb 2011 [74] reported that beta blockers increased the variability of systolic blood pressure. There were several important differences between our analysis and Webb 2011 that could have caused the differences in findings. First, Webb 2011 included patients with hypertension, post myocardial infarction (MI) and congestive heart failure (CHF) in their analysis. Our analysis only included patients with primary hypertension. The effect of beta blockers on BP variability could be different in primary hypertension compared to post MI and CHF patients.

Secondly, Webb 2011 also compared beta blockers to other antihypertensive treatments while we compared beta blockers to placebo alone. Beta blockers could tend to increase BP variability compared to placebo, while other antihypertensive drugs could tend to lower BP variability compared to placebo. By comparing beta blockers to other antihypertensive drugs directly, the larger difference could have led to a statistically significant result. A systematic review of BP variability of all antihypertensive drugs classes compared to placebo in hypertensive patients is needed to determine the truth.

None of the studies included in the analysis was designed to examine the difference in BP variability. Future RCT designed to evaluate the BP variability of beta blockers are needed in order to clarify this issue.

7.1.3 Peak and trough measurements

In chapter 6, from nebivolol data, we showed that the BP lowering effect measured at peak and trough hours could be different. In fact, beta blockers that measured BP at trough hours, such as carvedilol and nebivolol, generally showed smaller BP lowering effects.

Therefore, in each study, the time of BP measurement must be reported. Most of our included studies did not report the time of measurements. Reporting of this parameter was

119 better in new studies as most of them reported trough hour measurements. If the difference between peak and trough measurements is true this difference may contribute to differences in BP effect estimates and heterogeneity.

This review was not able to conclusively prove that BP lowering was greater at peak than at trough though most of the findings suggested that was true. Clinic BP measurements were used in this review. This has the disadvantage of being a snap shot of the blood pressure at one single moment. Ambulatory BP measurement would be a much better way to assess whether peak and trough measurements are different. Several protocols published in the Cochrane library which examine the ABPM effect of beta blockers will help answer this question [160, 161, 162].

7.2 What is the mechanism by which beta blockers lower BP?

In table 7.1, the close correlation between systolic blood pressure and heart rate was evident in all four subclasses except for partial agonists. The beta blockers that showed a greater lowering of systolic blood pressure were also associated with a greater lowering of heart rate. It appeared that the systolic blood pressure lowering efficacy of beta blockers was related to their ability to reduce heart rate. This would suggest that the main BP lowering effect of beta blockers was mediated through the blocking the effect on beta-1 receptors in the heart affecting heart rate and contractility. However, beta-1 antagonism was not likely to be the sole pathway that beta blockers lower BP. If it was truly the sole pathway, nebivolol, the most beta-1 selective blocker, would be expected to have the greatest BP lowering effect.

120

Table 7.1: The effect of varying doses of different subclasses of beta blockers on systolic blood pressure, diastolic blood pressure and heart rate

Outcome 0.5x starting dose 1x starting dose 2x starting dose 4x starting dose [95% CI] Non-selective beta blockers SBP -2.4 [-10.8, 6.0] -5.3 [-8.4, -2.1] -11.7 [-14.0, -9.4] -20.8 [-25.1, -16.5] DBP -2.9 [-8.2, 2.4] -3.6 [-5.4, -1.8] -8.4 [-9.8, -7.1] -14.1 [-17.0, -11.4] HR -1.2[-12.5, 10.1] -8.8 [-12, -5.5] -12.6 [-14.2, -11] -18.5 [-21.7, -15.4] Dual blockers SBP -0.9 [-5.2, 3.4] -3.5 [-6.4, -0.7] -6.1 [-8.1, -4.1] -9.2 [-11.7, -6.6] DBP -0.7 [-3.1, 1.7] -2.3 [-3.9, -0.6] -4.2 [-5.3, -3.0] -6.7 [-8.2, -5.3] HR -1.4 [-3.9, 1.1] -2.1 [-3.9, -0.4] -5.4 [-6.6, -4.3] -6.8 [-8.1, -5.3] Partial agonists SBP -6.1 [-11.4, -0.9] -8.1 [-10.4, -5.8] -8.0 [-12.2, -3.7] -11.0 [-16.0, -6.1] DBP -0.7 [-4.8, 3.4] -3.8 [-5.1, -2.6] -4.4 [-6.5, -2.3] -6.7 [-9.7, -3.7] HR -2.2 [-7.3, 2.87] -6.0 [-7.4, -4.5] -2.3 [-4.6, 0] No data Beta-1 selective blocker SBP -5.7 [-7.6, -3.8] -9.3 [-10.2, -8.4] -11.3 [-12.2, -10.3] -8.9 [-10.5, -7.4] DBP -4.3 [-5.3, -3.2] -7.1 [-7.6, -6.5] -9.5 [-10.1, -8.9] -7.2 [-8.2, -6.2] HR -4.4 [-6.2, -2.5] -9.8 [-10.7, -8.9] -12.2 [-13.1, -11.4] -13.7 [-15, -12.4]

The mechanism by which beta blockers lower blood pressure is likely to be multi- factorial. The effect of beta antagonism on various physiological systems contributes and results in the blood pressure lowering efficacy of beta blockers.

The sympathetic nervous system is one of the major cardiovascular regulatory mechanisms in humans. Sympathetic activity has many functions in the body. In terms of

BP regulation, increase sympathetic activity would increase heart rate and cardiac contractility, constrict peripheral blood vessels and release epinephrine from the adrenal medulla. The circulating epinephrine then would release renin from the adrenal cortex, decrease peripheral vascular resistance and also increase heart rate and cardiac contractility

121

[22]. Beta antagonism would block the physiological effects mediated by beta-1 and beta-2 receptors. In addition, counter mechanisms opposed to the sympathetic activity, such as parasympathetic activity also played a role in regulation of BP [22]. It is likely to be the combination of response by these physiological pathways to beta antagonism that produces the overall BP lowering effect of beta blockers. However, the exact BP lowering mechanism remains unknown.

7.3 Is there a difference in BP lowering efficacy between beta blocker subclasses?

Before we compare the BP lowering efficacy between beta blockers, we must note that baseline BP and time of measurement varied between the different trials.

Table 7.2: Overall subclasses baseline characteristics and time of measurements

Beta blocker subclasses Baseline BP Time of measurements Non-selective 158/104 mmHg Peak Dual blockers 151/101 mmHg Trough Partial agonists 175/107 mmHg Peak Beta-1 selective 156/101 mmHg Peak/Trough

Table 7.2 shows the difference in baseline BP and time of measurements among the subclasses. When different subclasses of beta blockers are compared to each other, it is essential to appreciate that higher baseline BP would be associated with a greater BP lowering effect. In addition, peak BP lowering effects are likely greater than trough BP lowering effects. The summary of findings tables at the end of each chapter showed the estimates and quality of evidence of the BP lowering efficacy, heart rate and pulse pressure reduction for each of the subclasses of beta blockers. Table 7.3 shows the estimates from the summary of findings table side by side for comparison.

122

Table 7.3: List of mean estimates of all four subclasses

mean estimates of mean estimates of mean estimates of mean estimates combining 1x and combining 1x and combining 1x and of combining 1x Beta blocker 2x starting dose 2x starting dose 2x starting dose for and 2x starting subclass for SBP for DBP HR dose for PP [95% CI] [95% CI] [95% CI] [95% CI] Non- selective -9.5 [-10.9, -8.1] -6.6 [-7.4, -5.8] -11.8 [-12.9, -10.7] -2.0 [-3.2, -0.9] beta blockers Dual blockers -5.2 [-6.9, -3.6] -3.5 [-4.5, -2.6] -4.4 [-5.4, -3.5] -2.1 [-4.0, -0.2] Partial agonists -8.1 [-10.1, -6.1] -4.0 [-5.1, -2.9] -4.9 [-6.2, -3.7] -3.6 [-5.3, -1.9] Beta-1 blockers -10.3 [-10.9, -9.6] -8.2 [-8.6, -7.8] -11.1 [-11.7, -10.5] -1.7 [-2.3, -1.2]

Different subclasses of beta blockers appeared to lower BP by different magnitude.

Measuring BP at peak hours, non-selective beta blockers had greater BP lowering effect compared to partial agonists. With similar baseline BP, dual receptor blockers lowered BP by a smaller magnitude compared to non-selective beta blockers and beta-1 blockers. These are indirect comparisons, but represent fairly strong evidence that the BP lowering efficacy of different subclasses of beta blockers are not the same. Head to head trial at equivalent dose will allow direct comparison to assess whether or not there are differences in BP lowering efficacy between different beta blockers.

All previous publications assumed that all beta blockers reduced BP to a similar degree and combined all of the data together. However, pooling the data of beta blockers is misleading, as the average effect for the whole class does not apply to individual beta blockers or subclass of beta blockers. Since our review had shown that this assumption is untrue, future systematic reviews should not combine the data from different subclasses of beta blockers.

123

7.4 Disagreement with other published reviews

Two previous systematic reviews [24, 25] examined the BP lowering efficacy of beta blockers. These two publications estimated that beta blockers lowered BP by -8/-6 mmHg

[24] and -8/-7 mmHg [25]. However, neither of them separated beta blockers into subclasses, considered the baseline BP or time of BP measurement. Their estimates were the result of pooling all beta blockers in all settings and doses together.

This method of analysis presents a problem in interpretation. From a pharmacologist point of view, it is not logical to assume that drugs with different pharmacokinetic and pharmacodynamic properties would lower BP to the same extent. The advantage of a systematic review is that it allows one to be able to compare different individual beta blockers, subclasses and doses of beta blockers and examine whether the assumption that they are all the same is true.

As discussed in previous sections, the blood pressure lowering efficacy of beta blockers was different in different subclasses. Pooling the data of beta blockers resulted in a misrepresentation of the effect. The average effect for the class did not apply to all products or subgroups of beta blockers. Because of this, it was not appropriate to compare our results to these two publications.

7.5 Difference in BP lowering efficacy in beta-blockers compared to other classes?

Before we compare the BP lowering effect of beta blockers to other classes, we must ascertain whether it is appropriate to compare them. The summary of findings table listed the baseline BP and time of measurement for each subclass. We must consider that baseline and time of measurements when comparing to other classes. Table 7.4 shows the average BP lowering efficacy of other classes of antihypertensive drugs at peak and trough times.

124

Table 7.4: BP lowering efficacy of other classes of antihypertensive drugs

Class of antihypertensive Peak BP lowering effect Trough BP lowering effect ACE inhibitors [72] -11/-6 mmHg -8/-5 mmHg ARBs [73] -12/-7 mmHg -9/-6 mmHg Thiazide diuretics [159] -10/-4 mmHg

Since beta-blockers had a different effect on pulse pressure than other classes of drugs it is best to consider systolic and diastolic separately when comparing the magnitude of

BP lowering.

Non selective beta blockers lowered peak systolic blood pressure by a smaller degree

(-10/-7 mmHg) than the peak measurement of ACE inhibitors and ARBs. However, they lowered peak diastolic blood pressure by a similar degree compared to ARBs and greater degree compared to ACE inhibitors. The effect of non selective beta blockers is likely exaggerated due to the presence of extreme outliers and other biases. If we remove the extreme outliers from the analysis, the estimate for non selective beta blockers is lower (-8/-5 mmHg). The most striking difference between non selective beta blockers and the other classes is the smaller effect on pulse pressure.

Dual blockers (-5/-4 mmHg) lowered trough BP by a smaller magnitude compared to the trough effect of thiazide diuretic (-10/-4 mmHg), ACE inhibitors (-8/-5 mmHg) and

ARBs (-9/-6 mmHg). Like other beta blockers, the fact that diastolic blood pressure was reduced to a similar amount to systolic BP resulted in dual receptor blockers having little effect on pulse pressure. If we would consider carvedilol alone, the estimate was even smaller (-4/-3 mmHg). Thus dual receptor blockers lower BP less than other beta blocker subclasses as well as other antihypertensive drug classes.

125

Compared to the peak effect of ACE inhibitors and ARBs, partial agonists (-8/-4 mmHg) lowered peak BP by a smaller magnitude. The estimate of the effect of partial agonists on pulse pressure was larger at 4 mmHg than other beta blockers, but because of the very low quality of the evidence more studies are needed to know whether partial agonists have a different effect on pulse pressure than other subclasses of beta blockers.

The baseline BP of partial agonists (175/107 mmHg) was much higher than ACE inhibitors (157/101 mmHg) and ARBs (156/101 mmHg). The absolute amount of reduction of BP is expected to be higher when the baseline BP was higher. It is likely that the trough

BP lowering effect of partial agonists in a mild hypertensive population would be smaller than the current estimate.

Beta-1 blockers lowered peak systolic blood pressure (-10 mmHg) by smaller degree while the magnitude in diastolic blood pressure (-8 mmHg) was greater compared to ACE inhibitors and ARBs. If we considered only the beta-1 blockers that measured BP at trough hour, such as nebivolol and bisoprolol, the same pattern was found (-8/-7 mmHg). Beta-1 blockers lowered systolic blood pressure by a similar degree and lowered diastolic blood pressure by a greater magnitude compared to diuretics, ACE inhibitors and ARBs. Lowering diastolic blood pressure by a similar extent as systolic blood pressure explains the relative little effect on pulse pressure by beta-1 blockers.

Thus most beta blocker subclasses had a smaller systolic blood pressure lowering effect and greater reduction in diastolic blood pressure compared to thiazide diuretics, ACE inhibitors and ARBs. This could be part of the reason why beta blockers were associated with less reduction in adverse cardiovascular outcomes [21]. The greater reduction in diastolic blood pressure can potentially be harmful as the heart is perfused during diastole.

126

Lowering diastolic pressure might cause under-perfusion of the heart. In addition, over the years, systolic blood pressure continues to rise but diastolic blood pressure plateaus in older people which result in increasing pulse pressure [5]. Greater reduction of diastolic blood pressure could also lead to an increase in pulse pressure. The lack of effect on pulse pressure by beta blockers could help explain why beta blockers are less effective in preventing coronary heart disease. However this did not take into account that most of the clinical data came from studies that used first-line atenolol [21], which in this review caused a similar reduction in systolic blood pressure and a greater reduction in diastolic blood pressure as compared to thiazides, ACE inhibitors and ARBs.

Furthermore beta blockers, such as non-selective beta blockers, carvedilol and acebutolol, have been shown to provide a mortality benefit in other cardiovascular diseases

[18, 19, 20]. Therefore it is not clear whether the differing BP lowering efficacy of beta blockers is the reason for the different mortality and morbidity outcomes associated with the beta blockers. These findings contradict the common belief that the morbidity and mortality benefit of antihypertensive drugs can be explained based on how much they lower BP. It is obviously much more complicated than that.

7.6 Overall quality of evidence

The rigorous application of explicit inclusion and exclusion criteria as well as the comprehensive search strategies minimized the potential for bias during the selection of included studies. The search of citations in the three databases was performed by the trial search coordinator of the Cochrane Hypertension group. Once we had the set of citations, we sorted through them following the inclusion criteria. This process minimized the possibility of missing trials that could have introduced bias.

127

When data were only presented in figures, two reviewers extracted data independently and then the average of the two values was used. This minimized the potential of human error during the data extraction process. We are confident that the potential for bias during the review process is low.

7.6.1 Potential of bias in included studies

Like all systematic reviews, our review was limited by the quality and quantity of data available to us. The quantity of data we found was large enough to provide adequate power for examining the outcomes in most of our beta blockers. The quality was assessed according to the headings below.

7.6.2 Randomization (Selection bias)

Randomization is vital to all RCTs. Most of our included studies did not report detail on the procedure to ensure proper randomization. We examined the integrity of randomization in parallel studies by assessing the baseline characteristics between intervention groups. We did not find any significant difference between the intervention groups in any relevant baseline characteristics in the parallel studies. On the other hand, there was no way to assess the integrity of randomization in the crossover studies. Most of them only mentioned that participants were randomly assigned without giving any other information. Such vague reporting did not provide sufficient information for judging whether randomization and allocation concealment was properly done. Therefore, we did not have enough information to adequately judge whether there was significant risk of selection bias in crossover studies.

128

7.6.3 Blinding (performance and detection bias)

All the included studies were double blind trials. Some studies described using

“double dummy” technique. Blinding presents a challenge for beta blocker studies. Since it is well known that beta blockers lower heart rate, physicians and patients can guess the intervention assignment by measuring heart rate. The risk of detection bias was greater in crossover studies as patients could also detect a lower heart rate once they start on beta blockers, as compared with the period in which placebo was taken. In addition, lower heart rate could lead to underestimating systolic blood pressure and overestimating diastolic blood pressure when BP is measured by a mercury sphygmomanometer. This problem could be mitigated with BP measurement using an automated machine. However, only a few included studies used automated machines to measure blood pressure. In many of our included studies, this issue remains a problem.

7.6.4 Reporting bias

Systolic blood pressure and diastolic blood pressure were the most frequently reported outcomes in our included studies. The time which these two outcomes were measured was often not reported in the publications. Some of our newer studies would report that instruction was given to patients in order to ensure that it was 13-24 hours after last dose when their BP was taken. However, this important piece of information was often not mentioned in older studies. If the medications were taken at multiple times daily, we assumed that BP was taken at peak hours (1 to 12 hours after the dose) in those studies. On the other hand, when medications were only taken once daily, it was difficult to determine if the BP measurements were made during peak or trough times. As peak and trough

129 measurements may be significantly different, it is important for the studies to provide this information.

The selective reporting of withdrawal due to adverse effects (WDAE) was prevalent in all 4 subclasses. It was unlikely that any trial protocol would be approved without mentioning the procedure for handling dropouts. The investigators probably collected data for WDAE but chose not to report it in the article. The decision to not report this particular outcome leads to a high risk of reporting bias. The majority of our included studies, particularly the older studies, did not even mention WDAE. WDAE is an important measure of drug tolerability, particularly in short term trials. Selective reporting of WDAE leads to a serious under estimation of the harms of the medication. All RCT should report WDAE as one of their outcomes.

Standard deviation (SD) or standard error of mean (SEM) was another outcome that the RCTs often failed to report. We had to impute standard deviation when the information was missing. These are the most basic of statistical parameters. Any study with a sound statistic methodology should not omit the reporting of standard deviation or SEM and editors should not allow publication of trials that do not report this parameter.

7.6.5 Publication bias

In the absence of publication bias, a funnel plot should resemble a symmetrical inverted funnel with large studies concentrate at the top and smaller studies scatter more widely at the bottom of the graph. Some of our funnel plots are asymmetric suggesting that small studies, particularly those with negative results, have not been published. Selective publishing of small negative studies could exaggerate the overall effect estimates and also lead to higher risk of publication bias.

130

Furthermore, some of the beta blockers, such as labetalol, acebutolol and pindolol, have been marketed with few or no available RCTs to support its use. We were only able to find data in 100 participants or less for these drugs. These beta blockers are available in

Canada, the U.S. and the European Union since late 70s and early 80s. They were approved after the stricter drug approval process had been instituted in most developed countries. We estimated that in order to fulfill regulatory requirements, data on the blood pressure lowering effect from at least 1,000 participants must be submitted to regulatory agencies. This meant that more than 90% of the data for these drugs has not been published or made available in some other way. A recent analysis found that unpublished trial data submitted to regulatory agencies provided considerably more information on patient-relevant outcomes than publicly available sources [163]. The result of this analysis corroborated our suspicion that study data was hidden. The regulatory agencies could improve the situation by making the unpublished data submitted to them publicly available.

In the conclusion of systematic reviews, authors usually mention that more RCTs would be needed in order to answer some of the knowledge gaps discovered during the review. However, in this case, we are confident that data already exist. Thus the first step is to ensure that this data is made available so that it can be included in an update of this review. The Cochrane collaboration is at the forefront in pushing for open access to all clinical trial data [88]. All scientists who are searching for the truth should join this cause.

7.7 Conclusion

1. The review provides the most up-to-date dose related blood pressure lowering effect

estimates for each beta blocker subclass as well as for each individual beta blocker.

The BP lowering estimates of each subclass were -10/-7 mmHg for non-selective beta

131

blocker, -5/-4 mmHg for dual receptor blockers, -8/-4 mmHg for partial agonists and

-10/-8 mmHg for beta-1 selective blockers. The data shows that different subclasses

of beta blockers lower BP by different magnitudes.

2. One of the most important findings of this review is the demonstration that beta

blockers lower systolic and diastolic blood pressure to a similar degree and thus have

little or no effect on pulse pressure. This pattern is clearly different than the other

classes of drugs that have been studied, such as ACE inhibitors, angiotensin receptor

blockers and diuretics. Even in the case where a reduction in pulse pressure was

shown it was small, 2 mmHg for atenolol and 4 mmHg for partial agonists. There

was no dose response effect for pulse pressure.

3. The data did not convincingly show a graded dose response effect on systolic blood

pressure or diastolic blood pressure for the most the beta blockers. In most cases, the

maximum blood pressure lowering effect was shown at either the recommended

starting dose or twice the recommended starting dose. There was little or no further

reduction in blood pressure for doses higher than twice the recommended starting

dose.

4. Most beta blockers significantly lowered heart rate and showed a dose related effect.

This suggests that higher dose beta blockers are more likely to cause bradycardia

without providing additional BP reduction compare to lower doses of beta blockers.

5. The blood pressure lowering estimates of beta blocker subclass that mostly measured

blood pressure at peak hours were generally greater than the estimates of the subclass

that measured at trough hours. This suggests blood pressure lowering of beta

blockers at peak hours is greater than at trough hours. More study is needed to

132

confirm this observation and quantify the magnitude of the difference between peak

and trough measurements for beta blockers.

6. End of treatment standard deviation of beta blocker and placebo groups for both

systolic and diastolic blood pressure were not different thus not suggesting that beta

blockers increased blood pressure variability in a primary hypertensive population.

7. High risk of bias was prevalent in all four subclasses of this review, including

detection bias due to loss of blinding caused by lower heart rate and publication bias

(ample evidence that many studies done to fulfill regulatory requirements were not

published).

8. Many beta blockers, such as dual receptor blockers and partial agonists, do not lower

blood pressure as much as other classes of drugs. There is little to no further

reduction in blood pressure for doses of beta blockers greater than the twice the

recommended starting dose.

9. Research implications:

i. Future blood pressure lowering trials must report: resting blood pressure and

heart rate measured in a standard sitting position; the amount of time after

administration of the drug that the BP is measured: the standard deviation or

standard error of mean of all measurements; and withdrawals due to adverse

effects in the treatment and placebo arms.

ii. RCTs are needed to directly compare the blood pressure lowering efficacy of

beta blockers at standard doses with other classes of drugs at standard doses.

133 iii. Future systematic reviews should focus on comparing the blood pressure

lowering efficacy of different beta blocker subclasses in head to head trials

and on the blood pressure lowering effect of beta blockers using 24-hour

ambulatory blood pressure measurements.

134

BIBLIOGRAPHY

1. World Health Organization. World health statistics. [Accessed on June 2013]. Available online at www.who.int 2. Statistics Canada. Health fact sheet (82-625-X). [Accessed on June 2013]. Available online at http://www.statcan.gc.ca/pub/82-625-x/2012001/article/11663-eng.htm. 3. Hypertension Canada. 2013 CHEP recommendation. [Accessed on June 2013. Available online at http://www.hypertension.ca/diagnosis. 4. MacMahon S, Peto R, Collins R et al. Blood pressure, stroke, and coronary heart disease. Lancet 1990; 335:765-774. 5. Franklin SS, Larson MG, Khan SA et al. Does the relation of blood pressure to coronary heart disease risk change with aging? The Framingham Heart Study. Circulation. 2001; 103:1245-49. 6. Haider AW, Larson MG, Franklin SS, Levy D. Systolic Blood Pressure, Diastolic Blood Pressure, and Pulse Pressure as Predictors of Risk for Congestive Heart Failure in the Framingham Heart Study. Ann Intern Med. 2003;138(1):10-16. 7. Suchy-Dicey AM, Wallace ER, S V Elkind M et al. Blood pressure variability and the risk of all cause mortality, incident myocardial infarction, and incident stroke in the cardiovascular health study. Am J Hypertens. 2013; e-publication. 8. Eto M, Toba K, Akishita M et al. Impact of blood pressure variability on cardiovascular events in elderly patients with hypertension. Hypertens Res 2005;28:1- 7. 9. Hata Y, Muratani H, Kimura Y et al. Office blood pressure variability as a predictor of acute myocardial infarction in elderly patients receiving antihypertensive therapy. J Human Hypertens 2002;16:141-146. 10. Wright JM, Musini VM. First-line drugs for hypertension. Cochrane Database of Systematic Reviews 2009, Issue 3. Art No.: CD001841. DOI:10.1002/14651858.CD001841.pub2. 11. Musini VM, Tejani AM, Bassett K, Wright JM. Pharmacotherapy for hypertension in the elderly. Cochrane Database of Systematic Reviews 2009, Issue 4. Art. No.: CD000028. DOI: 10.1002/14651858.CD000028.pub2. 12. Sherwood L. Human Physiology: From cells to systems. 7th ed. Belmont, CA: Brooks/Cole; 2010. 13. Brunton LL, Chabner BA, Knollman BC (editors). Goodman and Gilman’s the pharmacological basic of therapeutics. 12th ed. New York: McGraw-Hill; 2011.

135

14. Nobelprize.org. Stockholm, Sweden: Nobel Foundation; c1988 [Accessed 13 June, 2013] Available: http://www.nobelprize.org/nobel_prizes/medicine/laureates/1988/black.html 15. Prichard BNC, Gillam PMS. Propranolol in hypertension. American Journal of Cardiology 1966;18(3):387-391. 16. Prichard BNC, Gillam PMS. Treatment of Hypertension with Propranolol. BMJ 1969;1:7-16. 17. Teo KK, Yusuf S, Furberg CD. Effects of Prophylactic AntiarrhythmicDrug Therapy in Acute Myocardial Infarction. JAMA 1993;270:1589-1595. 18. Brophy JM, Joseph L, Rouleau JL. Beta Blockers in Congestive Heart Failure. A Bayesian Meta-Analysis. Annals of Internal Medicine 2001;134:550-560. 19. Whorlow SL, Krum H. Meta-analysis of effect of beta blocker therapy on mortality in patients with New York Heart Association class IV congestive heart failure. Am J Cardiol 2000;86:886-889. 20. Wright JM.. Choosing a first-line drug in the management of elevated blood pressure: What is the evidence? 2: β-Blockers. CMAJ 2000;163(2):188-192. 21. Wisonge CS, Bradley H, Mayosi BM, Maroney R, Mbewu A, Opie LH, Volmink J. Beta blockers for hypertension. Cochrane Database of Systematic Reviews 2007, Issue 1. Art. No.: CD002003. DOI: 10.1002/14651858.CD002003.pub2. 22. Barrett KE, Boitano S, Barman SM, Brooks HL. Ganong’s review of medical physiology. 24th ed. New York, NY: the McGraw-Hill Companies; 2012. 23. Magee LA, Duley L. Oral beta-blockers for mild to moderate hypertension during pregnancy. Cochrane Database of Systematic Reviews 2003, Issue 1. Art. No.: CD002863. DOI: 10.1002/14651858.CD002863. 24. Chen JMH, Heran BS, Perez MI, Wright JM.. Blood pressure lowering efficacy of beta-blockers as second-line therapy for primary hypertension.. Cochrane Database of Systematic Reviews 2010, Issue 1. Art. No.: CD007185. DOI: 10.1002/14651858.CD007185.pub2. 25. Law M, Morris JK, Jordan R, Wald N. Headaches and the treatment of blood pressure results from a meta-analysis of 94 randomized placebo-controlled trials with 24 000 participants. Circulation 2005;112:2301-2306. 26. FDA label search. Silver spring, MD: FDA; c2013 [Accessed Jun 20, 2013] Available: http://labels.fda.gov/. 27. Electronic Medicines Compendium. Surrey, UK: EMC; c2013 [Accessed Jun 19, 2013] Available: http://www.medicines.org.uk/emc/default.aspx.

136

28. Fitzgerald JD. The applied pharmacology of beta-adrenoceptor antagonists (beta blockers) in relation to clinical outcomes. Cardiovascular drugs and therapy 1991;5:561-576. 29. Higgins JPT, Green S (editors). Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 [updated March 2011]. The Cochrane Collaboration, 2011. Available from www.cochrane-handbook.org. 30. Cochrane.org. Freiburg, Germany: The Cochrane Collaboration; c2013 [Accessed Jun 20, 2013] Available www.cochrane.org. 31. Thecochranelibrary.com. West Sussex, UK: Wiley & Son Ltd; c2013 [Accessed Jun 20, 2013] Available www.thecochranelibrary.com. 32. Klabunde RE. Cardiovascular physiology concept. 2nd ed. Baltimore, MD: Lippincott Williams and Wilkins; 2012. 33. CPS product monograph. Ottawa, ON: Canadian Pharmacists Association; c2013 [Accessed Sep, 2011] Available https://www.e-therapeutics.ca/home.whatsnew.action 34. Abrams KR, Gillies CL, Lambert PC. Meta-analysis of heterogeneously reported trials assessing change from baseline. Statist. Med. 2005; 24:3823–3844. 35. Follmann D, Elliott P, Suh I, Cutler J. Variance imputation for overviews of clinical trials with continuous response. J Clin Epidemiol. 1992; 45(7):769-73. 36. Chalmers J, Tiller D, Horvath J et al. Effects of timolol and hydrochlorothiazide on blood-pressure and plasma renin activity. Double-blind factorial trial. Lancet 1976; 2(7981):328-331. 37. Dargie H, Cleland J, Findlay I, Murray G, McInnes G. Combination of verapamil and beta blockers in systemic hypertension. Am J Cardiol 1986;80d-82d. 38. de-Plaen JF, Vander EE, Van-Ypersele-de SC. Penbutolol or hydrochlorothiazide once a day in hypertension. A controlled study with home measurements. Br J Clin Pharmacol 1981; 12:215-221. 39. Dupont AG, Bossuyt AM, Jonckheer MH, Six RO. Effect of nadolol on ambulatory blood pressure, renal haemodynamics and cardiac function in hypertension. 19[1], 129p. 1984. 40. Fasano ML, Gaeta I, De SG et al. Cardiovascular response to adrenergic stimulation during treatment with tertatolol. A new non-cardioselective beta-blocking agent in primary hypertensive patients. Jpn Heart J 1991; 32(4):435-444. 41. Ferrara LA, Capaldo B, Rivellese A et al. Adrenergic system and carbohydrate metabolism. Effects of beta-receptor blockade on insulin secretion and peripheral insulin sensitivity in normoglycaemic patients. Eur J Clin Pharmacol 1987; 33(3):273- 277.

137

42. Galloway DB, Glover SC, Hendry WG et al. Propranolol in hypertension: a dose- response study. Br Med J 1976; 2(6028):140-142. 43. Kelemen MH, Effron MB, Valenti SA et al. Exercise training combined with antihypertensive drug therapy. Effects on lipids, blood pressure, and left ventricular mass.[see comment]. JAMA 1990; 263(20):2766-2771. 44. Lepantalo MJ, Totterman KJ, Lepantalo MJ, Totterman KJ. Lower limb haemodynamics during antihypertensive treatment with metoprolol and propranolol. Int Angiol 1985; 4(2):225-228. 45. Malacco E, Mailland F, Bevilacqua M, et a. Long-term hemodynamic effects of indenolol in hypertension at rest and during isometric exercise. CURR THER RES , CLIN EXP 1985; 37:537-544. 46. McCorvey E Jr, Wright JT, Jr., Culbert JP et al. Effect of hydrochlorothiazide, enalapril, and propranolol on quality of life and cognitive and motor function in hypertensive patients. Clin Pharm 1993; 12(4):300-305. 47. McInnes GT, Findlay IN, Murray G et al. Cardiovascular responses to verapamil and propranolol in hypertensive patients. J Hypertens Suppl 1985; 3(3):S219-S221. 48. Nicaise J, Desmedt J, Kunstler M, et a. (-)-Moprolol as antihypertensive therapy in the elderly. CURR THER RES , CLIN EXP 1981; 30:93-97. 49. Oh VM, Chia BL, Taylor EA, Oh VM, Chia BL, Taylor EA. Effects of long-acting propranolol on blood pressure and heart rate in hypertensive Chinese. Br J Clin Pharmacol 1985; 20(2):144-147. 50. Pearson RM, Bulpitt CJ, Havard CW, Pearson RM, Bulpitt CJ, Havard CW. Biochemical and haematological changes induced by tienilic acid combined with propranolol in essential hypertension. Lancet 1979; 1(8118):697-699. 51. Petrie JC, Galloway DB, Jeffers TA et al. and propranolol or practolol in moderate hypertension. Br Med J 1976; 2(6028):137-139. 52. Petrie JC, Jeffers TA, Robb OJ et al. Atenolol, sustained-release oxprenolol, and long- acting propranolol in hypertension. Br Med J 1980; 280(6231):1573-1574. 53. Ravid M, Lang R, Jutrin I, Ravid M, Lang R, Jutrin I. The relative antihypertensive potency of propranolol, oxprenolol, atenolol, and metoprolol given once daily. A double-blind, crossover, placebo-controlled study in ambulatory patients. Arch Intern Med 1985; 145(7):1321-1323. 54. Richardson DW, Freund J, Gear AS et al. Effect of propranolol on elevated arterial blood pressure. Circ Cardiovasc Genet 1968; 37(4):534-542. 55. Rosen RC, Kostis JB, Jekelis A et al. Sexual sequelae of antihypertensive drugs: treatment effects on self-report and physiological measures in middle-aged male hypertensives. Arch Sex Behav 1994; 23(2):135-152.

138

56. Schoenberger JA, Schoenberger JA. Usefulness of penbutolol for systemic hypertension. Penbutolol Research Group. Am J Cardiol 1989; 63(18):1339-1342. 57. Shukla SK, Bhatnagar MK, Agarwal RK. Beta-blocker (propranolol) in mild hypertension. A dose response study. Indian Heart J 1979; 31(6):330-332. 58. Sica DA, Neutel JM, Weber MA et al. The antihypertensive efficacy and safety of a chronotherapeutic formulation of propranolol in patients with hypertension. J Clin Hypertens (Greenwich ) 2004; 6(5):231-241. 59. van Herwaarden CL, Fennis JF, Binkhorst RA et al. Haemodynamic effects of during treatment of hypertensive patients with propranolol and metoprolol. Eur J Clin Pharmacol 1977; 12(6):397-402. 60. West MJ, Wing LM, Mulligan J et al. Comparison of labetalol, hydrallazine, and propranolol in the therapy of moderate hypertension. MED J AUST 1980; 1(5):224- 225. 61. Ades PA, Gunther PGS, Meacham CP, Handy MA. Hypertension, exercise and beta adrenergic blockade. Annals of Internal Medicine 1988;109:629-634. 62. Ades PA, Gunther PGS, Meyer WL, Gibson TC, Maddalena J, Orfeo T. Cardiac and skeletal muscle adaptation to training in systemic hypertension and effect of beta blockade (metoprolol or propranolol). American Journal of Cardiology 1990;66:591- 596. 63. Agnes E, Vandermecken C, Prost JF, Safar ME. The tertatolol international multicenter study. American Journal of Hypertension 1989;2(11):296S-302S. 64. Beilin LJ, Juel-Jensen BE. Alpha and beta adrenergic blockade in hypertension.. Lancet 1972;1:979-982. 65. Cherchi A.. Controlled study of the antihypertnsive efficacy of the combination of nadolol and bendroflumethiazide with repect to nadolol alone and placebo. Clin Ter 1985;113:301-306. 66. Costa FV, Caldari R, Borghi C, Bassein L, Ambrosioni E. Acebutalol, atenolol and nadolol: comparison of their antihypertensive efficacy at rest and during exercise.. Current Therapeutics Research 1984;35(6):961-973. 67. Davidson C, Thadani U, Singleton W, Taylor SH. Comparison of antihypertensive active of beta blocking drugs during chronic treatment.. British Medical Journal 1976;2:7-9. 68. Kubik MM, Coote JH. Propranolol vs metoprolol vs labetalol: a comparative study in essential hypertension.. European Journal of Clinical Pharmacology 1984;26:1-6. 69. Lepantalo M, Totterman KJ. Effect of long term beta adrenergic blockade on calf blood flow in hypertensive patients. Clinical Physiology 1983;3:35-42.

139

70. Smith SA, Littler WA. The effect of epanolol on intra-arterial ambulatory blood pressure and baroreceptor heart rate reflex in essential hypertension. Clinical and Experimental Pharmacology and Physiology 1988;15:327-331. 71. Weigmann I, Terhaag B, Wierich W, Herrmann WM. Antihypertensive effects of different talinolol dosages after four weeks of treatment in comparison with placebo. Drug Research 1998;48(3):240-244. 72. Heran BS, Wong MMY, Heran IK, Wright JM. Blood pressure lowering efficacy of angiotensin converting enzyme(ACE) inhibitors for primary hypertension. Cochrane Database of Systematic Reviews 2008, Issue 4. Art. No.: CD003823. DOI: 10.1002/14651858.CD003823.pub2. 73. Heran BS,WongMMY, Heran IK,Wright JM. Blood pressure lowering efficacy of angiotensin receptor blockers for primary hypertension. Cochrane Database of Systematic Reviews 2008, Issue 4. Art. No.: CD003822. DOI: 10.1002/14651858.CD003822.pub2. 74. Webb AJS, Fischer U, Rothwell PM. Effect of beta-blocker selectivity on blood pressure variability and stroke. Neurology 2011;77:731-737. 75. GSK study registry. Brentford, UK: result summary for B100; c 2001-2013 [assessed Sep 3, 2011]. Available from: http://www.gsk- clinicalstudyregister.com/result_detail.jsp?protocolId=105517%2F021&studyId=80A0 B2AA-C084-4ACF-8FA7-23C564527CF2&compound=carvedilol. 76. GSK study registry. Brentford, UK: result summary for B101; c 2001-2013 [assessed Sep 5, 2011]. Available from: http://www.gsk- clinicalstudyregister.com/result_detail.jsp?protocolId=105517%2F022&studyId=CD32 5C8A-3EBC-4621-91DE-C403E8D28CFD&compound=carvedilol 2005. 77. Frick MH, Porsti P. Combined alpha- and beta-adrenoceptor blockade with labetalol in hypertension. BMJ 1976;1:1046-1048. 78. Kane J, Gregg I, Richards DA. A double blind trial of labetalol. Br J Clin Pharmaco. 1976;Suppl:737-741. 79. Lechi A, Pomari S, Berto R, Buniotto P, Parrinello A, Marini F, Cogo L, Tomasi A, Baretta G. Clinical evaluation of labetalol alone and combined with chlorthalidone in essential hypertension: a double blind multicenter controlled study. Eur J Clin Pharmacol 1982;22:289-293. 80. McPhillips JJ, Schwemer GT, Scott DI, Zinny M, Patterson D.. Effect of Carvedilol on Blood pressure in patients with mild to moderate hypertension. A dose response study. Drugs 1988;36(Suppl 6):82-91. 81. Rouffy J, Bakir R, Chanu B, Djian F, Goy-Loeper J, Henry Amar M, Renon D, Pappo M. Effect of 400mg/day labetalol on lipids, lipoproteins and apoproteins in hypertensive patients. Path Biol 1986;34(4):253-258.

140

82. Weber MA, Sica DA, Tarka EA, Iyengar M, Fleck R, Bakris GL.. Controlled Release Carvedilol in the treatment of essential hypertension.. Am J Cardiol 2006;98 (suppl):32L-38L. 83. Dupont AG, Van der Niepen P, Taeymans Y, Ingels M, Piepsz A, Bossuyt AM, Block P, Six RO, Jonckheer MH, Vanhaelst L. Effect of carvedilol on ambulatory blood pressure, renal hemodynamics, and cardiac function in essential hypertension. J Cardiovascular Pharmacol 1987;10 (Suppl 11):S130-S136. 84. Horvath JS, Caterson RJ, Collett P, Duggin GG, Kelly DH, Tiller DJ. Labetalol and bendrofluazide: comparison of their antihypertensive effects. Med J Australia 1979;1:626-628. 85. Kubik MM, Coote JH. Propranolol vs metoprolol vs labetalol: A comparative study in essential hypertension. Eur J Clin Pharmacol 1984;26:1-6. 86. Wilcox RG. Randomised study of six beta-blockers and a thiazide diuretic in essential hypertension.. BMJ 1978;2(6134):383-385. 87. McTavish D, Campoli-Richards D, Sorkin EM. Carvedilol: A review of its pharmacodynamic, pharmacokinetic properties and therapeutic efficacy. Drugs 1993;45(2):232-258. 88. Thomas K. Breaking the Seal on Drug Research. New York Times. 2013 Jun 29; page BU1. 89. Chalmer JP, Korner PI, Tiller DJ, Bune AJ, Steiner JD, West MJ, Wing LM, Uther JF.. Double blind factorial trial of pindolol and hydrochlorothiazide in hypertension. Med J Australia 1976;1:650-653. 90. Chalmers JP, Wing LMH, Grygiel JJ, West MJ, Graham JR, Bune AJ. Effect of once daily indapamide and pindolol on blood pressure, plasma aldosterone concentration and plasma renin activity in a general practice setting.. Eur J Clin Pharmacol 1982;22:191-196. 91. Forette MF, Henry JF, Hervy MP, Forette B, Berthaux P. The treatment of hypertension in elderly using a beta-blocker: acebutolol. [Traitement de l'hyertension artérielle du sujet âgé par un bêta-bloquant: l'acébutolol]. Nouv Presse Méd 1979;8:2881-2884. 92. Hansson L, Berglund G, Andersson O, Holm M. Controlled trial of acebutolol in hypertension. Eur J Clin Pharmacol. 1977;12:89-92. 93. Moleur P, Peyrieux JC, Luciani J, David D, Boissel JP.. Bopindolol in the treatment of moderate hypertension: a dose response study. Fundam Clin Pharmacol 1988;2:431- 440. 94. Motolese M, Muisan G, Colombi A. Hypotensive effect of oxprenolol in mild to moderate hypertension: a multicentre controlled study. Eur J Clin Pharmacol 1975;8:21-31.

141

95. Olkinuora JT, Wiikari J, Vanhanen H, Niina M, Kalliomaki T. Effect of celiprolol and simvastatin on the calculated risk of coronary heart disease. Scand. Cardiovascul J 2006;40:160-166. 96. Salako LA, Falase AO, Fadeke Aderounmu A. Placebo-controlled, double blind clinical trial of alprenolol in Afraican hypertensive patients. Curr Med Res Opin 1979;6:358-363. 97. Stump K, Kolloch R, Mathieu M, Capone P. A comparison of celiprolol and atenolol in the treatment of hypertension: A placebo controlled double blind study. British Journal of Clinical Practice 1985;Spp 40:73-75. 98. Tibblin G, Ablad B. Antihypertensive therapy with alprenolol, a beta-adrenergic .. Acta med scand 1969;186:451-457. 99. Trafford JAP, Latta D, Little PS, Parsley J, Ankier SI, Quail D. A multi-centre, placebo controlled comparative study between 200 mg and 400 mg celiprolol in patients with mild to moderate essential hypertension.. Curr Med Res Opin 1989;11:550-556. 100. Vandongen R, Margetts B, Deklerk N, Beilin LJ, Rogers P.. Plasma catecholamines following exercise in hypertensives treated with pindolol: comparison with placebo and metoprolol. Br. J. Clin. Pharmac. 1986;21:627-632. 101. Watson RDS, Stallard TJ, Littler WA. Comparison of once and twice daily administration of acebutolol in hypertension. Br J Clin Pharmacol 1980;9:209-212. 102. Erley CM, Berger ED, Heyne N, Klass M, Kramer D, Braun N, Wolf S, Risler T. A double-blind placebo-controlled comparison of the effects of beta receptor blockers and ACE inhibitor on renal haemodynamics and proteinuria in chronic glomerulonephritis. DMW 1997;122:953-958. 103. Fraser DM, Nimmo GR, Poloniecki JD.. Acebutolol in the treatment of diabetic patients with hypertension. Curr. Med. Res. Opin. 1986;10:122-127. 104. Lewis MJ, Jones DM, Dart AM, Henderson AH.. The psychological side effects of acebutolol and atenolol. Br J Clin Pharmac 1984;17:364-366. 105. Ades PA, Gunther PG, Meyer WL et al. Cardiac and skeletal muscle adaptations to training in systemic hypertension and effect of beta blockade (metoprolol or propranolol). Am J Cardiol 1990; 66(5):591-596.

106. Ameling EH, de Korte DF, Man i', V, Ameling EH, de Korte DF, Man in 't Veld A. Impact of diagnosis and treatment of hypertension on quality of life: a double-blind, randomized, placebo-controlled, cross-over study of betaxolol. J Cardiovasc Pharmacol 1991; 18(5):752-760.

107. Asmar RG, Kerihuel JC, Girerd XJ et al. Effect of bisoprolol on blood pressure and arterial hemodynamics in systemic hypertension. Am J Cardiol 1991; 68(1):61-64.

142

108. Baez MA, Garg DC, Jallad NS, Weidler DJ. Antihypertensive effect of in hypertensive patients: comparison with atenolol. Br J Clin Pharmacol 1986;63s-67s.

109. Bateman DN, Dean CR, Mucklow JC et al. Atenolol and chlorthalidone in combination for hypertension. Br J Clin Pharmacol 1979; 7(4):357-363.

110. Bengtsson C, Bengtsson C. The effect of metoprolol --a new selective adrenergic beta1-receptor blocking agent-- in mild hypertension. Acta Med Scand 1976; 199(1- 2):65-70.

111. Berglund G, de FU, Castenfors J et al. Monitoring 24-hour blood pressure in a drug trial. Evaluation of a noninvasive device. Hypertension 1985; 7:688-694.

112. Broekman CP, Haensel SM, van d, V et al. Bisoprolol and hypertension: effects on sexual functioning in men. J Sex Marital Ther 1992; 18(4):325-331.

113. Chan TY, Woo KS, Nicholls MG, Chan TY, Woo KS, Nicholls MG. The application of nebivolol in essential hypertension: a double-blind, randomized, placebo-controlled study. INT J CARDIOL 1992; 35(3):387-395.

114. Chan TYK, Woo KS, Critchley JAJH, Swaminathan R, Nicholls MG. Effects of nebivolol on blood pressure, hormones and cardiac function in Chinese subjects with essential hypertension. DRUG INVEST 1991; 3:98-104.

115. Chrysant SG, Chappel C, Farnham DJ et al. Antihypertensive and metabolic effects of single and combined atenolol regimens. J Clin Pharmacol 1992; 32(1):61-65.

116. Cilliers AJ, Cilliers AJ. Atenolol as primary therapy in previously untreated hypertensives and as an adjuvant to other therapy. A South African Multicentre Study. SAMJ , S 1979; Suid-Afrikaanse Tydskrif Vir Geneeskunde. 55(9):321-324.

117. Dahlof B, Sjoberg KH, Flygt C et al. Pafenolol in hypertension: a double-blind randomized trial of a new beta 1-selective adrenoceptor blocker. J Cardiovasc Pharmacol 1986; 8(1):55-59.

118. Davidov ME, Singh SP, Vlachakis ND et al. Bisoprolol, a once-a-day beta-blocking agent for patients with mild to moderate hypertension. CLIN CARDIOL 1994; 17(5):263-268.

119. Deary AJ, Schumann AL, Murfet H et al. Double-blind, placebo-controlled crossover comparison of five classes of antihypertensive drugs.[see comment]. J Hypertens 2002; 20(4):771-777.

120. Deary AJ, Schumann AL, Murfet H et al. Influence of drugs and gender on the arterial pulse wave and natriuretic peptide secretion in untreated patients with essential hypertension.[see comment]. CLIN SCI 2002; 103(5):493-499.

143

121. Dhakam Z, Yasmin-, McEniery CM, Burton T, Brown MJ, Wilkinson IB. A comparison of atenolol and nebivolol in isolated systolic hypertension. J Hypertens 2008; 26:351-356.

122. Frishman WH, Burris JF, Mroczek WJ et al. First-line therapy option with low-dose bisoprolol fumarate and low-dose hydrochlorothiazide in patients with stage I and stage II systemic hypertension. J Clin Pharmacol 1995; 35(2):182-188.

123. Frishman WH, Hainer JW, Sugg J et al. A factorial study of combination hypertension treatment with metoprolol succinate extended release and felodipine extended release results of the Metoprolol Succinate-Felodipine Antihypertension Combination Trial (M-FACT). Am J Hypertens 2006; 19(4):388-395.

124. Frisk-Holmberg M, Juhlin-Dannfelt A, Astrom H, Frisk-Holmberg M, Juhlin-Dannfelt A, Astrom H. Haemodynamic and metabolic responses to prolonged exercise after chronic beta 1-adrenoceptor blockade in hypertensive man. Clin Physiol 1985; 5(3):231-242.

125. Gostick NK, Mayhew SR, Million R et al. A dose-response study of atenolol in mild to moderate hypertension in general practice. Curr Med Res Opin 1977; 5(2):179-184.

126. Gudbjornsdottir S, Fowelin J, Elam M et al. The effect of metoprolol treatment on insulin sensitivity and diurnal plasma hormone levels in hypertensive subjects. Eur J Clin Invest 1997; 27(1):29-35.

127. Hansson L, Aberg H, Karlberg BE et al. Controlled study of atenolol in treatment of hypertension. Br Med J 1975; 2(5967):367-370.

128. Himmelmann A, Hedner T, Snoeck E et al. Haemodynamic effects and pharmacokinetics of oral d- and l-nebivolol in hypertensive patients. Eur J Clin Pharmacol 1996; 51(3-4):259-264.

129. Houston MC, Burger C, Hays JT et al. The effects of hydrochloride versus atenolol monotherapy on serum lipids, lipid subfractions, and apolipoproteins in mild hypertension. American Heart Journal 1990; 120(1):172-179.

130. Jaattela A, Baandrup S, Houtzagers J et al. The efficacy of low dose metoprolol CR/ZOK in mild hypertension and in elderly patients with mild to moderate hypertension. J Clin Pharmacol 1990; 30(2 Suppl):S66-S71.

131. Jeffers TA, Webster J, Petrie JC et al. Atenolol once-daily in hypertension. Br J Clin Pharmacol 1977; 4(5):523-527.

132. Lacourciere Y, Arnott W, Lacourciere Y, Arnott W. Placebo-controlled comparison of the effects of nebivolol and low-dose hydrochlorothiazide as monotherapies and in combination on blood pressure and lipid profile in hypertensive patients. J Hum Hypertens 1994; 8(4):283-288.

144

133. Lepantalo M, Totterman KJ, Lepantalo M, Totterman KJ. Effect of long-term beta- adrenergic-blockade on calf blood flow in hypertensive patients. Clin Physiol 1983; 3(1):35-42.

134. Lischner M, Lang R, Jutrin I et al. Atenolol vs. amiloride-hydrochlorothiazide in the treatment of mild to moderate hypertension: a double-blind, crossover, placebo- controlled study. Drug Intell Clin Pharm 1987; 21(1 Pt 1):43-46.

135. Maclean D, Mitchell ET, Lewis R et al. Comparison of once daily atenolol, nitrendipine and their combination in mild to moderate essential hypertension. Br J Clin Pharmacol 1990; 29(4):455-463.

136. Myers MG, de CJ, Myers MG, de Champlain J. Effects of atenolol and hydrochlorothiazide on blood pressure and plasma catecholamines in essential hypertension. Hypertension 1983; 5(4):591-596.

137. FDA drug approval package. Silver spring, MD: medical review for nebivolol (NEB305); c2007-2008. [accessed Jan 2013] Available at http://www.accessdata.fda.gov/drugsatfda_docs/nda/2007/021742s000TOC.cfm

138. Okawa KK, Okawa KK. Dose response studies of bevantolol in hypertensive patients. Angiology 1986; 37(3 Pt 2):233-238.

139. Papademetriou V, Hainer JW, Sugg J et al. Factorial antihypertensive study of an extended-release metoprolol and hydrochlorothiazide combination. Am J Hypertens 2006; 19(12):1217-1225.

140. Saunders E, Smith WB, DeSalvo KB et al. The efficacy and tolerability of nebivolol in hypertensive African American patients. J Clin Hypertens (Greenwich ) 2007; 9(11):866-875.

141. Seedat YK, Seedat YK. Trial of atenolol and chlorthalidone for hypertension in black South Africans. Br Med J 1980; 281(6250):1241-1243.

142. Stornello M, Valvo EV, Scapellato L, Stornello M, Valvo EV, Scapellato L. Comparative effects of enalapril, atenolol and chlorthalidone on blood pressure and kidney function of diabetic patients affected by arterial hypertension and persistent proteinuria. Nephron 1991; 58(1):52-57.

143. Stumpe K, Kolloch R, Mathieu M et al. A comparison of celiprolol and atenolol in the treatment of hypertension: a placebo controlled double blind study. BR J CLIN PRACT 1985; Supplement. 40:73-75.

144. Tonkin AL, Wing LM, Russell AE et al. Diltiazem and atenolol in essential hypertension: additivity of effects on blood pressure and cardiac conduction with combination therapy. J Hypertens 1990; 8(11):1015-1019.

145

145. Tseng C-D, Chiang F-T, Hsu K-L et al. Short-term efficacy and safety of bisoprolol in treatment of patients with mild-to-moderate essential hypertension - A two-center, double-blind study in Taiwan. Acta Cardiologica Sinica 1993; 9(3):155-160.

146. Van Bortel LM, Breed JG, Joosten J et al. Nebivolol in hypertension: a double-blind placebo-controlled multicenter study assessing its antihypertensive efficacy and impact on quality of life. J Cardiovasc Pharmacol 1993; 21(6):856-862.

147. van MT, Van Bortel LM, Smeets FA et al. Verapamil and nebivolol improve carotid artery distensibility in hypertensive patients. J Hypertens Suppl 1989; 7(6):S262-S263.

148. Van NL, Dupont AG, Vertommen C et al. A dose-response trial of nebivolol in essential hypertension. J Hum Hypertens 1997; 11(2):139-144.

149. Van NL, Taylor FR, Robertson JI, Van Nueten L, Taylor FR, Robertson JI. Nebivolol vs atenolol and placebo in essential hypertension: a double-blind randomised trial. J Hum Hypertens 1998; 12(2):135-140.

150. Vanhees L, Fagard R, Lijnen P et al. Effect of antihypertensive medication on endurance exercise capacity in hypertensive sportsmen. J Hypertens 1991; 9(11):1063- 1068.

151. Verdecchia P, Brignole M, Delfino G et al. Systolic time intervals as possible predictors of pressure response to sustained beta-adrenergic blockade in arterial hypertension. A within-patient, placebo-controlled study. Hypertension 1983; 5(1):140-146.

152. Verdecchia P, Gatteschi C, Benemio G et al. Duration of the antihypertensive action of atenolol, enalapril and placebo: a randomized within-patient study using ambulatory blood pressure monitoring. Int J Clin Pharmacol Ther Toxicol 1988; 26(11):570-574.

153. Weiss RJ, Weber MA, Carr AA et al. A randomized, double-blind, placebo-controlled parallel-group study to assess the efficacy and safety of nebivolol, a novel beta- blocker, in patients with mild to moderate hypertension.[erratum appears in J Clin Hypertens (Greenwich). 2007 Oct;9(10):806]. J Clin Hypertens (Greenwich ) 2007; 9(9):667-676.

154. Williams RL, Goyle KK, Herman TS et al. Dose-dependent effects of betaxolol in hypertension: a double-blind, multicenter study. J Clin Pharmacol 1992; 32(4):360- 367.

155. Wing LM, Chalmers JP, West MJ et al. Enalapril and atenolol in essential hypertension: attenuation of hypotensive effects in combination. Clin Exp Hypertens A 1988; 10(1):119-133.

146

156. Wald DS, Law M, Mills S, Bestwick JP, Morris JK, Wald NJ. A 16-week, randomized, double-blind, placebo-controlled, crossover trial to quantify the combined effect of an angiotensin-converting enzyme inhibitor and a beta-blocker on blood pressure reduction. Clinical.therapeutics 2008;30:2030-9.

157. Lancaster SG, Sorkin EM. Bisoprolol: a preliminary review of its pharmacodynamic and pharmacokinetics properties and therapeutics efficacy in hypertension and angina pectoris.. Drugs 1988;36:256-285.

158. Veverka A, Nuzum DS, Jolly JL. Nebivolol: A third generation beta adrenergic blocker. Ann Pharmacother 2006;40:1353-1360.

159. Musini VM. A systematic review of the blood pressure lowering efficacy of thiazide and loop diuretics in the treatment of primary hypertension [dissertation]. Vancouver, BC: University of British Columbia; 2000.

160. Zhang KL, Kim J, Wong GWK, Musini VM. Time course for blood pressure lowering of beta one selective blockers (Protocol). Cochrane Database of Systematic Reviews 2012, Issue 9. Art. No.: CD010077. DOI: 10.1002/14651858.CD010077.

161. Wu Q, Jiang WN, Wong GWK, Musini VM. Time course for blood pressure lowering of dual alpha and beta blockers (Protocol). Cochrane Database of Systematic Reviews 2012, Issue 8. Art. No.: CD010055. DOI: 10.1002/14651858.CD010055.

162. Zhang XY, Soufi S, Musini VM. Time course for blood pressure lowering of beta blockers with partial agonist activity (Protocol). Cochrane Database of Systematic Reviews 2012, Issue 8. Art. No.: CD010054. DOI: 10.1002/14651858.CD010054.

163. Wieseler B, Wolfram N, McGauran N, Kerekes MF, Vervolgyi V, et al. Completeness of Reporting of Patient-Relevant Clinical Trial Outcomes: Comparison of Unpublished Clinical Study Reports with Publicly Available Data. PLoS Med 2013, 10(10): e1001526. doi:10.1371/journal.pmed.1001526

164. Wong GWK, Laugerotte A, Wright JM. Blood pressure lowering efficacy of non- selective beta blockers for primary hypertension (Protocol). Cochrane Database of Systematic Reviews 2008, Issue 4. Art. No.: CD007452. DOI: 10.1002/14651858.CD007452.

165. Wong GWK, Laugerotte A, Wright JM. Blood pressure lowering efficacy of dual alpha and beta blockers for primary hypertension (Protocol). Cochrane Database of Systematic Reviews 2008, Issue 4. Art. No.: CD007449. DOI: 10.1002/14651858.CD007449.W

166. Wong GWK, Laugerotte A, Wright JM. Blood pressure lowering efficacy of partial agonist beta blockers for primary hypertension (Protocol). Cochrane Database of Systematic Reviews 2008, Issue 4. Art. No.: CD007450. DOI: 10.1002/14651858.CD007450.

147

167. Wong GWK, Laugerotte A, Wright JM. Blood pressure lowering efficacy of beta-1 selective beta blockers for primary hypertension (Protocol). Cochrane Database of Systematic Reviews 2008, Issue 4. Art. No.: CD007451. DOI: 10.1002/14651858.CD007451.

148

APPENDICES

Appendix A: Search strategies

Search strategy for MEDLINE 1. acebutolol.mp. 2. exp alprenolol/ 3. alprenolol.mp. 4. amosulalol.mp. 5. arotinolol.mp. 6. atenolol.mp. 7. befunolol.mp. 8. betaxolol.mp. 9. bevantolol.mp. 10. exp bisoprolol/ 11. bisoprolol.mp. 12. bopindolol.mp. 13. .mp. 14. bucumolol.mp. 15. bufetolol.mp. 16. bufuralol.mp. 17. bunitrolol.mp. 18. exp bupranolol/ 19. bupranolol.mp. 20. butofilolol.mp. 21. carazolol.mp. 22. exp carteolol/ 23. carteolol.mp. 24. carvedilol.mp. 25. exp celiprolol/ 26. celiprolol.mp. 27. cetamolol.mp. 28. cloranolol.mp. 29. .mp. 30. deacetylmetipranolol.mp. 31. .mp. 32. exp dilevalol/ 33. dilevalol.mp. 34. epanolol.mp. 35. esmolol.mp. 36. indenolol.mp. 37. .mp. 38. exp labetalol/ 39. labetalol.mp. 40. .mp. 41. exp levobunolol/

149

42. levobunolol.mp. 43. mepindolol.mp. 44. exp metoprolol/ 45. metoprolol.mp. 46. exp metipranolol/ 47. metipranolol.mp. 48. moprolol.mp. 49. exp nadolol/ 50. nadolol.mp. 51. nadoxolol.mp. 52. nebivolol.mp. 53. nifenalol.mp. 54. .mp. 55. exp oxprenolol/ 56. oxprenolol.mp. 57. exp penbutolol/ 58. penbutolol.mp. 59. exp pindolol/ 60. pindolol.mp. 61. exp practolol/ 62. practolol.mp. 63. pronethalol.mp. 64. exp propranolol/ 65. propranolol.mp. 66. exp sotalol/ 67. sotalol.mp. 68. sulfinalol.mp. 69. talinolol.mp. 70. tertatolol.mp. 71. tilisolol.mp. 72. exp timolol/ 73. timolol.mp. 74. toliprolol.mp. 75. xibenolol.mp. 76. or/1-75 77. exp hypertension/ 78. hypertens$.tw. 79. exp blood pressure/ 80. blood pressure.mp. 81. or/77-80 82. randomized controlled trial.pt. 83. controlled clinical trial.pt. 84. randomized.tw. 85. placebo.tw. 86. drug therapy/ 87. randomly.tw.

150

88. trial.tw. 89. groups.tw. 90. or/82-89 91. animals/ not (humans/ and animals/) 92. 90 not 91 93. 76 and 81 and 92

Search Strategy for EMBASE 1. acebutolol.mp. 2. amosulalol.mp. 3. arotinolol.mp. or exp arotinolol/ 4. exp alprenolol/ or alprenolol.mp. 5. atenolol.mp. 6. befunolol.mp. 7. betaxolol.mp. 8. bevantolol.mp. 9. bisoprolol.mp. or exp bisoprolol/ 10. bopindolol.mp. or exp bopindolol/ 11. bucindolol.mp. 12. bucumolol.mp. 13. bufetolol.mp. 14. bufuralol.mp. 15. bunitrolol.mp. 16. bupranolol.mp. or exp bupranolol/ 17. butofilolol.mp. 18. carazolol.mp. 19. carteolol.mp. or exp carteolol/ 20. carvedilol.mp. 21. celiprolol.mp. or exp celiprolol/ 22. cetamolol.mp. 23. cloranolol.mp. 24. cyanopindolol.mp. 25. deacetylmetipranolol.mp. 26. dihydroalprenolol.mp. 27. dilevalol.mp. 28. epanolol.mp. 29. esmolol.mp. 30. indenolol.mp. 31. iodocyanopindolol.mp. 32. labetalol.mp. or exp labetalol/ 33. landiolol.mp. 34. levobunolol.mp. or exp levobunolol/ 35. mepindolol.mp. 36. metoprolol.mp. or exp metoprolol/ 37. metipranolol.mp. 38. moprolol.mp.

151

39. exp nadolol/ or nadolol.mp. 40. nadoxolol.mp. 41. nebivolol.mp. 42. nifenalol.mp. 43. nipradilol.mp. 44. oxprenolol.mp. 45. penbutolol.mp. or exp penbutolol/ 46. pindolol.mp. or exp pindolol/ 47. practolol.mp. or exp practolol/ 48. pronethalol.mp. 49. exp propranolol/ or propranolol.mp. 50. proxodolol.mp. 51. sotalol.mp. or exp sotalol/ 52. sulfinalol.mp. 53. talinolol.mp. 54. tertatolol.mp. 55. tilisolol.mp. 56. exp timolol/ or timolol.mp. 57. toliprolol.mp. 58. xibenolol.mp. 59. or/1-58 60. exp hypertension/ 61. hypertens$.tw. 62. blood pressure.mp. 63. or/60-62 64. randomized controlled trial/ 65. crossover procedure/ 66. double-blind procedure/ 67. random$.tw. 68. (crossover$ or cross-over$).tw. 69. placebo$.tw. 70. (doubl$ adj blind$).tw. 71. assign$.tw. 72. allocat$.tw. 73. or/64-72 74. (animal$ not (human$ and animal$)).mp. 75. 73 not 74 76. 59 and 63 and 75

Search Strategy for CENTRAL #1 MeSH descriptor Adrenergic beta-Antagonists explode all trees #2 beta n2 (antagonist* or receptor* or adrenergic* block*):ti,ab,kw #3 acebutolol:ti,ab,kw #4 MeSH descriptor Alprenolol explode all trees #5 alprenolol:ti,ab,kw #6 amosulalol:ti,ab,kw

152

#7 arotinolol:ti,ab,kw in Clinical Trials #8 atenolol:ti,ab,kw in Clinical Trials #9 befunolol:ti,ab,kw #10 betaxolol:ti,ab,kw #11 bevantolol:ti,ab,kw #12 MeSH descriptor Bisoprolol explode all trees #13 bisoprolol:ti,ab,kw #14 bopindolol:ti,ab,kw #15 bucindolol:ti,ab,kw #16 bucumolol:ti,ab,kw #17 bufetolol:ti,ab,kw #18 bufuralol:ti,ab,kw #19 bunitrolol:ti,ab,kw #20 MeSH descriptor Bupranolol explode all trees #21 bupranolol:ti,ab,kw #22 butofilolol:ti,ab,kw #23 carazolol:ti,ab,kw #24 MeSH descriptor Carteolol explode all trees #25 carteolol:ti,ab,kw #26 carvedilol:ti,ab,kw #27 MeSH descriptor Celiprolol explode all trees #28 celiprolol:ti,ab,kw #29 cetamolol:ti,ab,kw #30 cloranolol:ti,ab,kw #31 cyanopindolol:ti,ab,kw #32 deacetylmetipranolol:ti,ab,kw #33 dihydroalprenolol:ti,ab,kw #34 dilevalol:ti,ab,kw #35 epanolol:ti,ab,kw #36 esmolol:ti,ab,kw #37 indenolol:ti,ab,kw #38 iodocyanopindolol:ti,ab,kw #39 MeSH descriptor Labetalol explode all trees #40 labetalol:ti,ab,kw #41 landiolol:ti,ab,kw #42 MeSH descriptor Levobunolol explode all trees #43 levobunolol:ti,ab,kw #44 mepindolol:ti,ab,kw #45 MeSH descriptor Metoprolol explode all trees #46 metoprolol:ti,ab,kw #47 MeSH descriptor Metipranolol explode all trees #48 metipranolol:ti,ab,kw #49 moprolol:ti,ab,kw #50 MeSH descriptor Nadolol explode all trees #51 nadolol:ti,ab,kw #52 nadoxolol:ti,ab,kw

153

#53 nebivolol:ti,ab,kw #54 nifenalol:ti,ab,kw #55 nipradilol:ti,ab,kw #56 oxprenolol:ti,ab,kw #57 MeSH descriptor Penbutolol explode all trees #58 penbutolol:ti,ab,kw #59 MeSH descriptor Pindolol explode all trees #60 pindolol:ti,ab,kw #61 MeSH descriptor Practolol explode all trees #62 practolol #63 pronethalol:ti,ab,kw #64 MeSH descriptor Propranolol explode all trees #65 propranolol:ti,ab,kw #66 MeSH descriptor Sotalol explode all trees #67 sotalol:ti,ab,kw #68 sulfinalol:ti,ab,kw #69 talinolol:ti,ab,kw #70 tertatolol:ti,ab,kw #71 tilisolol:ti,ab,kw #72 MeSH descriptor Timolol explode all trees #73 timolol:ti,ab,kw #74 toliprolol:ti,ab,kw #75 xibenolol:ti,ab,kw #76 (#1 OR #2 OR #3 OR #4 OR #5 OR #6 OR #7 OR #8 OR #9 OR #10 OR #11 OR #12 OR #13 OR #14 OR #15 OR #16 OR #17 OR #18 OR #19 OR #20 OR #21 OR #22 OR #23 OR #24 OR #25 OR #26 OR #27 OR #28 OR #29 OR #30 OR #31 OR #32 OR #33 OR #34 OR #35 OR #36 OR #37 OR #38 OR #39 OR #40 OR #41 OR #42 OR #43 OR #44 OR #45 OR #46 OR #47 OR #48 OR #49 OR #50 OR #51 OR #52 OR #53 OR #54 OR #55 OR #56 OR #57 OR #58 OR #59 OR #60 OR #61 OR #62 OR #63 OR #64 OR #65 OR #66 OR #67 OR #68 OR #69 OR #70 OR #71 OR #72 OR #73 OR #74 OR #75) #77 MeSH descriptor Hypertension explode all trees #78 hypertens*:ti,ab #79 MeSH descriptor Blood Pressure explode all trees #80 blood pressure:ti,ab,kw #81 (#77 OR #78 OR #79 OR #80) #82 (#76 AND #81)

154

Appendix B: Data extraction form

155

Appendix C: Table of included studies in non-selective beta blocker review

Study Descriptions

Chalmers 1976 Single centre, randomized, double blinded, placebo controlled,

[36] cross over study. 8-week run in period followed by four 8-week Methods treatment phases. No participant withdrew due to adverse

effects. 20/23 participants completed the study.

n=23

Age: 30-60

Participants DBP: 95mmHg - 120mmHg

Exclusion criteria: patients with pre-existing condition such as

obstructive airway diseases and diabetes.

Timolol 10mg TID, Hydrochlorothiazide 50mg QD,

Hydrochlorothiazide 50mg and Timolol 10mg combination, Interventions placebo.

Each treatment period lasted 8 weeks.

Supine, standing, casual SBP, DBP and mean BP Outcomes resting heart rate, plasma renin activity, WDAE

Dargie 1986 [37] Randomized, double blinded crossover study. Each period of Methods treatment lasted for 4 weeks.

14 patients (9 women, 5 men)

Participants DBP was 95-110mmHg on single drug therapy.

mean age 51.6 (30-65)

156

Study Descriptions

Verapamil 120mg TID

Propranolol 80mg TID Interventions Verapamil 120mg and propranolol 80mg TID

placebo

Outcomes Reported supine and erect SBP, DBP

De Plean 1981 Randomized, double blinded crossover study. Each treatment

[38] Methods period lasted for 4 weeks. BP was measured about 6hr after

drug intake. The mean of 2 or 3 readings is used.

16 patients entered the study. 10/16 completed the study. 1

patient withdrawn due to adverse effect.

Participants age: 30-56

Baseline: mean SBP 163.9, mean DBP 117.1, mean heart rate

88.8, mean body weight 83.2

80mg Penbutolol QD; 100 mg Hydrochlorothiazide QD or Interventions placebo

Supine and standing SBP, DBP, heart rate.

Outcomes Biochemical data.

Total withdrawal and WDAE.

Dupont 1984 [39] Randomized, double blinded, placebo controlled crossover Methods study. Each treatment periods lasted 4 weeks.

n=10

Participants age: mean= 54.4, 45-62

Diastolic ambulatory BP 95-115mmHg

Interventions Nadolol 80mg QD, placebo QD.

157

Study Descriptions

Outcomes Mean ABP; heart rate; Cardiac output; GFR; RBF; RVR

Fasano 1991 [40] Randomized double blinded placebo controlled parallel study.

Methods 4-week washout followed by 4-week treatment or placebo

period.

n=20 (12 male), with primary hypertension.

age 18-57 (mean =39)

Exclusion criteria: Secondary hypertension, cardiovascular Participants diseases, chronic diseases, metabolic diseases, history of severe

side effect with beta blockers, pregnancy, lactation or use of

oral contraceptive.

Interventions tertatolol 5mg QD or placebo

Supine and standing SBP, DBP, heart rate and cardiac Outcomes workload.

Ferrara 1987 [41] Randomized double blinded crossover study. Washout period Methods for 2 weeks followed by 3 weeks test period for each treatment.

n=7(3 female 4 male) with primary hypertension. Participants age 26-62

Interventions Propranolol 80 BID or placebo

Supine SBP, DBP and heart rate. Outcomes Glucose tolerance test, Insulin sensitivity.

Galloway 1976 Randomized double blinded placebo controlled crossover

[42] study. Eligible patients were seen in hypertension clinic 2 Methods weeks after discharge from hospital. Each treatment or placebo

period lasted for 4 weeks.

158

Study Descriptions

24 patients (6 male) aged 19-65, mean age 46.

Participants mean body weight 62.6kg

DBP higher than 90 but under 105 mmHg.

Propranolol 20mg TID; Propranolol 40mg TID; Propranolol Interventions 80mg TID; placebo

Outcomes Lying, standing, post-exercise SBP, DBP and heart rate.

Did not report on drop out or WDAE. However, author stated Note that no dose related adverse event was recorded.

Kelemen 1990 Randomized double blinded placebo controlled parallel study.

[43] After screening, patient entered a 4-week single blinded

placebo washout period. During which all antihypertensive Methods drugs were replaced by placebo. After washout, patients were

tested for baseline readings. Eligible patients then were

randomized to a double blinded treatment group for 11 weeks.

n=51, patient BP must be 90

Exclusion: patients with history of coronary disease, valvular

Participants diseases, renal, cerebral vascular, hepatic or neurological

diseases; history of alcohol or drug abuse or any

contraindication for propranolol, diltiazem or exercise.

Patients first started on 160mg propranolol, 240mg diltiazem or

placebo in the first week. Then they were force titrated to

Interventions 240mg propranolol or 360mg diltiazem per day or equal

amount of placebo tablets. After titration, all patients entered

a10-week exercise training.

159

Study Descriptions

Echocardiogram; resting supine and exercise SBP, DBP, heart Outcomes rate; Plasma lipid and lipoprotein.

Only exercise heart rate is reported. Resting heart rate is not Note reported.

Lepantalo 1985 After 4-week washout period, patients were randomly

[44], Lapantalo allocation to three 4-week treatment period. Patients would Methods 1983 [69] crossover to another treatment at the end of each treatment

period without washout.

n=34 (17 male)

Patients mean age is 54.1 yrs (52-68) with mild primary Participants hypertension.

Exclusion: patients with diabetes, peripheral arteries disease.

Interventions Propranolol 80mg BID, metoprolol 100mg BID, placebo.

BP, ankle arm systolic BP ratio, calf blood flow, peripheral Outcomes resistance.

heart rate was not reported. There is no drop out. The same Note data was published in Lepantalo 1983.

Malacco 1985 After 3 weeks of washout period, patients were randomized to

[45] three 4-week treatment periods. Each treatment periods were Methods separated by 7 days of washout period. BP was measured at the

end of each period.

160

Study Descriptions

n=12 (6 male)

SBP =>150mmHg and/or DBP=>100mmHg.

Age less than 60. Participants Exclusion: Congestive heart failure, lung diseases, ischemic

heart diseases, chronic alcoholism, diabetes, renal insufficiency

and bradycardia.

Interventions Propranolol 80mg BID, indenolol 60 BID, placebo

Outcomes MBP, heart rate, CO

Note Only heart rate were used in analysis

McCorvey 1993 Randomized double blinded placebo controlled crossover

[46] study. Eligible patients discontinued all previous

antihypertensive treatment and then randomized to treatment or

placebo immediately. All patients were forced titrated to Methods maximum dose 3 days after randomization. The dosage

remained the same for the duration of the next 4 week. Then

patients were crossover to other treatment or placebo until all

treatments or placebo has been administrated.

161

Study Descriptions

n=17

Patients with primary hypertension with documented elevated

BP for the previous 12 months.

Exclusion: history of stroke, MI, angina, hepatic dysfunction, Participants serum creatinin >1.5mg/dL, history of bronchospasm, AV

block greater than first degree, congestive heart failure, gout,

alcohol and drug abuse, patients who are unable to comply with

protocol.

Propranolol 120mg QD, hydrochlorothiazide 25mg QD,

Interventions enalapril 10mg QD or placebo for first 3 days. Then dose was

double for the reminder of the treatment period.

Outcomes BP, heart rate, QOL

McInnes 1985 Randomized, double blinded, placebo controlled crossover

[47] study. After 4-week of washout period, patients were Methods randomized to four 4-week treatment periods. BP was

measured at the end of each period.

n=14

Participants Hypertensive patients whom DBP= 95-110mmHg with single

drug therapy.

Verapamil 120mg TID, propranolol 80mg TID, verapamil Interventions 120mg + propranolol 80mg TID or placebo

SBP, DBP, heart rate, PR interval, end systolic dimension, end Outcomes diastolic deminsion, fractional shortening

162

Study Descriptions

Nicaise 1981 [48] Randomized double blinded placebo controlled crossover

study. After 1 month of placebo washout, patients were Methods randomized to treatment for 2 months and then crossovered to

the other treatment for another 2 months.

n=30 (8 males), mean age 79 (69-93) Participants SBP=>160mmHg and/or DBP =>95mmHg

Interventions (-)moprolol 75mg BID or placebo

Outcomes MAP, heart rate, dropouts and WDAE

7 people dropout, 5 WDAE (3 in placebo, 2 in moprolol). did Note not report SD

Oh 1985 [49] Randomized double blinded placebo controlled crossover

Methods study. After 2 weeks of washout, patients were randomized to

treatment or placebo for 4 weeks.

n=15

Participants Hypertensive Chinese patients with 144/94

Mean age =39.3 (23-57)

Interventions Propranolol 160mg QD, propranolol 320mg QD or placebo

Resting, exercise and after exercise SBP, DBP, heart rate, Outcomes adverse effect.

There were 4 groups (A, B, C, D), however, only group A data Notes were reported.

Pearson 1979 Randomized double blinded placebo controlled crossover

[50] Methods study. Without washout period, patients were randomized to

treatment upon entry. Each treatment period lasted for 6 weeks.

163

Study Descriptions

n=16 (8 males) mean age 54

DBP=>100mmHg with single drug therapy.

Participants exclusion: angina, MI, heart failure, stroke, COPD, gout,

women of childbearing potential or those using oral

contraceptive pills.

propranolol 80mg BID, tienilic acid 250mg OM, propranolol Interventions 80mg and tienilic acid 250mg combination, or placebo

Outcomes SBP, DBP, heart rate, adverse effect.

Petrie 1976 [51] Randomized double blinded placebo controlled crossover

Methods study. After 14 days of washout period, patients were

randomized to a sequence of 4 week treatments or placebo.

n=24 (13 males)

105

Exclusion: recent MI, evidence of heart failure, heart block, Participants gross ischemia, grade III or IV retinopathy, diabetes, gout,

impaired liver function, or creatinin clearance less than 60ml/

min or if they were on any other drug treatment.

methyldopa 250mg TID, practolol 200mg TID, propranolol

Interventions 80mg TID, methyldopa 250mg and propranolol 80mg TID,

methyldopa 250mg and practolol 200mg TID or placebo

Outcomes SBP, DBP, heart rate, QOL

SD were imputed using the average of other studies that Note provided SD

164

Study Descriptions

Petrie 1980 [52] Randomized double blinded placebo controlled crossover

study. After 4-week run in period, patients were randomized to Methods a sequence of 4 week treatments. There were no washout

period between treatments

n=23 (12 males)

Participants mean age is 40.9 (21-59)

DBP>90 mmHg

atenolol 100mg QD, sustained release oxprenolol 160mg QD, Interventions long acting propranolol 160mg QD or placebo

5 min lying and 2 min standing SBP, DBP, heart rate. Outcomes Measurement were made 20-22 hours after dose.

Ravid 1985 [53] Randomized double blinded placebo controlled crossover

Methods study. After 2 weeks of placebo run in period, patients were

randomly allocated to series of five 4-week treatments.

n=134 (66 males)

Mean age of men is 52 and mean age of women is 54.

Participants 95

Normal renal function, absent of liver diseases and gross

obesity.

Slow release propranolol 160mg QD, slow release oxprenolol

Interventions 160mg QD, atenolol 100mg QD, metoprolol 200mg QD or

placebo

Outcomes SBP, DBP, heart rate, adverse effects

165

Study Descriptions

Study reported male and female data separately. The data has Note greater than average SEM.

Richardson 1968 Randomized double blinded placebo controlled crossover

[54] Methods study. After a 1-week run in period, patients were randomized

to sequence of 5 week treatment periods.

n=19 (5 males) Participants Moderate severe hypertension.

Propranolol 40mg TID, chlorthalidone 50mg BID, combination Interventions of either drugs or placebo.

Outcomes SBP, DBP, HR, WDAE, side effects

Rosen 1994 [55] Randomized double blinded placebo controlled crossover

study. After 1 month single blind run in period, patients were Methods randomized to series of 1 month treatments. Each treatment

period were separated by 2 week washout period.

n=13 (100% male.)

Male patients with prior history of hypertension and sexual

dysfunction were recruited. Patients had 2 consecutive sitting Participants DBP between 90mmHg and 100mmHg. Patients with diabetes,

congestive heart failure, renal diseases, alcoholism, or other

medical or psychiatric disorders were excluded.

methyldopa 500mg BID, propranolol 80mg BID, atenolol Interventions 100mg QD, Dyazide 50/25 mg BID or placebo

Outcomes SBP, DBP, sleep study, sexual function, hormone study

Notes HR data is not reported.

166

Study Descriptions

Scheonberger Randomized double blinded placebo controlled parallel study.

1989 [56] After 4-weeks single blind run in period, patients were Methods randomized to 6 week of treatment, follow by 2 weeks of

tapering off period.

n=302 (160 males)

Adult patients with mild to moderate hypertension, defined as

supine DBP between 95mmHg and 115mmHg.

Exclusion criteria: pregnancy, childbearing potential,

significant hepatic or renal disorders, COPD, asthma, thyroid

Participants disorder, bradycardia, psychosis, use of drugs that interfere

with beta blockers 2 week before study, metabolic acidosis,

unstable diabetes, continue need for concomitant

antihypertensive medications or diuretics, heart failure, AV

block, conditions that will affect compliance, concomitant

administration of investigational drugs.

Interventions Penbutolol 10mg, 20mg, 40mg QD or placebo.

Supine and standing SBP, DBP and HR 24 hours after last Outcomes dose, WDAE, adverse effects.

Shukla 1979 [57] Randomized double blind placebo controlled crossover study.

Methods After 2 weeks of run in period, eligible patients were

randomized to series of 4 week treatments.

167

Study Descriptions

n=26

Patient with DBP between 90mmHg and 114mmHg were

eligible. Participants Patients with recent MI, heart failure, heart block, grade III/IV

retinopathy, impaired hepatic or renal functions, diabetes or

COPD were excluded.

Interventions Propranolol 60mg, 120mg, 180mg TID or placebo.

Outcomes SBP, DBP

Notes HR was not reported. No dropouts

Sica 2004 [58] Randomized double blind placebo controlled parallel study.

After 2-3 weeks of single blind placebo run in, eligible patients

were randomized to one of the treatments for 8 weeks. Patients

who were randomized to 120 mg or 160 mg would start at 80 Methods mg then force titrated to respective dose in the next 2 weeks.

Patients who were randomized to 640 mg/day group started at

160 mg/day and then forced titrated to 640 mg in the next 3

weeks.

n= 434 randomized, 356 completed. DBP between 96mmHg

and 114mmHg during 2 sitting.

Exclusion criteria: Pregnancy, lactating, renal artery stenosis,

Participants GI diseases, kidney or liver diseases, cardiac diseases within

the last 6 months, asthma, COPD, nonallergic bronchospasms,

insulin dependent diabetes, concomitant administration of

drugs that affects BP.

168

Study Descriptions

Propranolol 80mg QD, 120mg QD, 160mg QD, 640mg QD or Interventions placebo

Outcomes SBP, DBP, HR, plasma concentration, safety. van Herwaarden Randomized double blinded placebo controlled crossover 1977 [59] study. There was no run in period. Patients were randomized to

Methods four consecutive treatments, each lasting 4 weeks. At the end

of each treatment period, BP was measured 2 hr after the

morning dose.

n=8

Participants Patients with supine DBP after 15 mins rest between

100mmHg and 120mmHg were eligible.

Propranolol 80mg TID, metoprolol 100mg TID or placebo.

Interventions Each beta blocker treatment was alternated with placebo

treatment.

Outcomes SBP, DBP, HR, effect of adrenaline.

West 1980 [60] Randomized double blinded placebo controlled crossover

study. After a six week run in period, patients received, in Methods random order, 5 six-week test phase. The BP data was the

average of the lasted 4 weeks of each treatment period.

n=12

Participants Patients with untreated DBP between 95mmHg and 120mmHg

were eligible.

methyclothiazide 5mg QD, labetalol 300mg BID, propranolol Interventions 80mg BID, hydralazine 50mg BID or placebo

169

Study Descriptions

Outcomes SBP, DBP, HR, adverse effects, WDAE.

Notes SD were imputed

SBP: systolic blood pressure; DBP: diastolic blood pressure; HR: heart rate; WDAE: withdrawal due to adverse effects; AE: adverse effects; BID: twice daily; TID: three times daily: QOL: quality of life. QID: four times daily; QD: once daily; CVD: cardiovascular disease; ABPM: ambulatory blood pressure measurements; Cmax: Maximum serum concentration; CHF: congested heart failure; MI: myocardial infarction; COPD: chronic obstructive pulmonary diseases; SD: standard deviation; MBP: mean blood pressure; SEM: standard error of mean; AV block: atrioventricular block; RBF: renal blood flow; RVR: renal vascular resistance; GFR: Glomerular Filtration Rate

170

Appendix D: Table of included studies in dual receptor blocker review

Study Descriptions GSK B100 [75] Randomized, double blinded, placebo controlled. Patients

entered an 8-week single blinded placebo run in period.

Eligible patients then were randomized to one of the four Methods treatment groups. Subjects randomized to 12.5 mg and 25 mg

underwent a titration period for a week when patients were

force titrated to target doses.

Total number randomized were 480 with mean age of 56. Entry

criteria were sitting DBP between 95 mmHg to 115 mmHg.

Patients were excluded for secondary hypertension, Participants bradycardia, second or third degree AV block, sinus syndrome,

unstable angina, MI within 6 months, stroke within 6 months,

CHF, arrhythmias.

Carvedilol 6.25 mg BID, 12.5 mg BID, 25 mg BID, placebo Interventions BID.

Outcomes Trough SBP, trough DBP, trough HR, WDAE, AE

Authors mentioned that the measurements are trough, however

they did not specified how many hours after the last dose that Notes the measurements were made.

Study Descriptions GSK B101 [76] Randomized double blinded placebo controlled. Patients first

entered an 8-week placebo run in period for baseline

Methods assessments. Eligible patients were randomized to one of the

five treatment groups for 8 weeks. Subjects assigned to 12.5

mg or higher were force titrated until target dosage was met.

171

Study Descriptions Total number randomized was 520. Average age was 54.

Patients with sitting DBP between 95mmHg to 115mmHg.

Exclusion criteria: patients with secondary hypertension,

hypertensive retinopathy, bradycardia, 2nd or 3rd degree AV Participants block, sick sinus syndrome, unstable angina, recent MI or

stroke, CHF, arrhythmia, bronchospastic pulmonary diseases,

insulin dependent diabetes, liver, renal or biliary diseases, low

platelet count, low WBC counts.

carvedilol 6.25mg QD, 12.5mg QD, 25mg QD, 50mg QD, Interventions placebo.

Outcomes SBP, DBP, HR, at trough hour WDAE, AE.

The authors mentioned the measurements were trough but did

not state the hour from last dose when blood pressure were Notes measured.

Study Descriptions Frick 1976 [77] Randomized double blinded placebo crossover study. Patients

with DBP> 100 mmHg on repeated measuring were enrolled.

Methods All antihypertensive drugs were stopped for one to two months

before entry. Patients were placed on treatment for 4 weeks and

then crossover to other treatment.

N= 20, Baseline BP 161/113 mmHg, Patients with DBP > 100 Participants mmHg during the washout period were enrolled.

Interventions labetalol 100 mg TID, labetalol 200 mg TID or placebo

sitting SBP, DBP measured by mercury sphygmomanometer, Outcomes HR.

172

Study Descriptions Kane 1976 [78] Randomized double blinded placebo controlled crossover

study. All drugs were stopped for 2 weeks before entry. Part I

Methods of study, one tablet was taken twice daily, in part II, the dose

was doubled. Each part contained two 4 week period alternate

between treatment and placebo.

N=30, baseline BP 169/107 mmHg. Patients under 65 with Participants DBP between 95 to 120 mmHg were enrolled.

Interventions Labetalol 200 mg BID, 400 mg BID or placebo

sitting SBP, DBP measured by mercury sphygmomanometer, Outcomes HR.

Study Descriptions Lechi 1982 [79] Patients entered into a two weeks placebo washout period, and

Methods then eligible patients entered into crossover 4 treatment cycles,

each lasting 3 weeks.

32 patients (22 males) aged from 30 to 63, with WHO stage I

or II primary hypertension. Baseline BP 156/106 mmHg.

Exclusion: bradycardia, hypokalemia, body weight more than Participants 20% above ideal, pregnant women, gout, asthma, COPD,

collagen diseases, rheumatoid arthritis, diabetes mellitus,

neoplasm and liver diseases.

Labetalol 100 mg TID, chlorthalidone 10 mg TID, labetalol Interventions 100 mg + chlorthalidone 10 mg TID, placebo TID.

Outcomes SBP, DBP, HR, blood chemistry

Study Descriptions

173

Study Descriptions McPhillips 1988 Patients entered a 4-week placebo run in period. Then eligible

[80] patients were randomized to one of the parallel four treatment

groups for 4 week. All interventions were administrated double Methods blinded. At the end of the run in and treatment period, patients

were admitted to clinical research unit for a 30 hours

observation period.

92 patients screened and 44 patients qualified for treatment

period. 34 of the 44 eligible were male. mean age of all 4 Participants groups range from 44.3 to 56.1. Their baseline supine DBP was

99.1 mmHg.

Interventions 12.5mg QD, 25mg QD, 50mg QD carvedilol, placebo

Outcomes SBP, DBP, HR, Cmax, serum half life.

The baseline age of each group was different from each other. Notes Study Descriptions Rouffy 1986 [81] After a 30-day run in period, eligible patients were randomized Methods to a double blinded parallel treatment period for 3 months.

30 patients with mean age 45 range from 34-57 year old.

Baseline BP 174/98 mmHg.

Exclusion: CVD, taking lipid modifying drugs, contraindicated Participants for beta blockers, obesities, abnormal glycemic level,

consumption of more than 250g of alcohol per week and

patients whose life style change during the study.

Interventions Labetalol 200mg BID, placebo

Outcomes SBP, DBP, HR, lipid, lipoprotein

Study is in French. Notes

174

Study Descriptions Weber 2006 [82] After a 4-week run in period, eligible subjects were

randomized to one of the 4 treatment groups for 8 weeks. Methods Patients who were randomized to 40mg or 80mg groups were

force titrated to the dose within 4 weeks.

Total number of 337 subjects recruited. 20 mg n=87, 40 mg

Participants n=78, 80 mg n=88, placebo n=84. Baseline BP 150.7/99

mmHg.

20mg CR QD, 40mg CR QD, 80mg CR QD carvedilol and Interventions placebo

Office & ABPM SBP, DBP, HR (trough measurement at 20-24 Outcomes hr after last dose), safety and withdrawals.

Patients randomized to 80 mg/day carvedilol were only on

80mg for 2 weeks therefore the data did not met the inclusion Notes criteria.

SBP: systolic blood pressure; DBP: diastolic blood pressure; HR: heart rate; WDAE: withdrawal due to adverse effects; AE: adverse effects; BID: twice daily; TID: three times daily: QID: four times daily; QD: once daily; CVD: cardiovascular disease; ABPM: ambulatory blood pressure measurements; Cmax: Maximum serum concentration; CHF: congested heart failure; MI: myocardial infarction; COPD: chronic obstructive pulmonary diseases;

175

Appendix E: Table of included studies in partial agonist review

Study Descriptions

Chalmers 1976 Randomized double blinded placebo controlled crossover study Methods [89] with 8 week treatment periods.

N=21, 16 completed the study. Baseline DBP 95 to 120

Participants mmHg, age 30 to 60, primary hypertension and not COPD,

unstable hypertension, retinopathy or diabetes.

Pindolol 10 mg TID, Hydrochlorothiazide 50 mg QD, Interventions combination of the two, placebo

sitting, supine standing SBP, DBP, HR, lab, ECG, Chest x-ray, Outcomes side effect

Study Descriptions

Chalmers 1982 Randomized double blinded placebo controlled crossover study

[90] Methods with eight week washout and four eight week treatment

periods.

mean age 53±10, 7 men 9 women, with supine DBP between Participants 95 mmHg to 120 mmHg.

Pindolol 5 mg BID, indapamide 2.5 mg QD, indapamide and Interventions pindolol combo, placebo.

sitting, standing, supine SBP, DBP, HR, plasma renin and Outcomes aldosterone, plasma K, Cl, HCO3-, urine analysis, WDAE

Study Descriptions

Forette 1979 [91] Randomized double blinded placebo controlled crossover study

Methods with 10 week-treatment phase. 8 weeks on fixed dose then 2

week titration down period.

176

Study Descriptions

n= 34, all subjects were female, mean age 81 with SBP >160

Participants mmHg and DBP > 90 mmHg. Exclusion: renal insufficiency,

HR<50, Heart failure, AV block, asthma.

Acebutolol 200 mg BID, diuretic/acebutolol combination, Interventions placebo

Outcomes SBP, DBP, HR, serum drug level, plasma renin activity.

Study Descriptions

Hansson 1977 Randomized double blinded placebo controlled parallel study, Methods [92] with 12 week treatment period.

n=38, Patients with primary hypertension, average age 55.6, Participants baseline average BP 177/106.

Acebutolol 400 mg, 600 mg, 1200 mg QD with crossover style

Interventions at 4 week each. Placebo treatment was given for whole 12

week period.

Outcomes SBP, DBP, HR, side effects

Study Descriptions

Moleur 1988 [93] Randomized double blinded parallel group placebo controlled Methods 4-week duration.

n=117, 115 completed. Supine DBP between 90 to 114 mmHg

Participants at the end of run in period, excluded pregnancy and lactation,

secondary hypertension.

Interventions Bopindolol 0.5 mg, 1 mg and 2 mg QD, placebo

Outcomes SBP, DBP, HR, symptom questionnaire

Study Descriptions

177

Study Descriptions

Motolese 1975 Randomized double blinded parallel placebo controlled study Methods [94] with duration of 4 weeks.

n=329, baseline BP 189.5/113.3 mmHg. Patients age <65, with

SBP <250 mmHg and DBP between 100 and 130 mmHg.

Participants Excluded HF, asthma, bradycardia, AV conduction defect,

retinopathy, blood urea nitrogen exceeding 50 mg/ 100 ml,

serum creatinin greater than 1.5 mg/100 ml.

Oxprenolol 20, 40, 60, 80 mg QD or combination with Interventions hydrochlorothiazide or dihydralazine, placebo

Outcomes WDAE, supine and standing SBP, DBP, lab

Study Descriptions

Olkinuora 2006 Randomized double blind placebo controlled parallel study for Methods [95] 3 months.

total n=112, 56 randomized to beta blocker mono therapy or

placebo, baseline BP 159.5/94 mmHg, average age 61, with

systolic hypertension defined as SBP > 140 mmHg. Exclusion Participants criteria: BMI > 35, CHD, LVH, unstable asthma, newly

diagnosed type 2 diabetes, total serum cholesterol level > 9

mmol/L or SBP > 180 mmHg

Celiprolol 200 mg QD, simvastatin 10 mg QD, combination of Interventions the two QD, placebo

Outcomes SBP, DBP, total cholesterol, LDL, HDL, TG.

Study Descriptions

178

Study Descriptions

Salako 1979 [96] Randomized double blinded placebo controlled crossover study Methods with 8 week treatment period.

n=16, pre-treatment DBP between 95 to 120 mmHg. Patients

Participants had previously been treated with antihypertensive for 1 to 9

years. Exclusion: renal, liver disease, diabetes.

Interventions Alprenolol 200 mg BID, placebo

Outcomes SBP, DBP, HR, serum level, side effect, WDAE

Study Descriptions

Stumpe 1985 [97] Randomized double blind placebo controlled parallel study for Methods 4 weeks.

n= 232, 185 randomized to partial agonist mono therapy or

Participants placebo. Subject with supine DBP between 95 to 114 mmHg

during the washout period. Baseline BP 165/102 mmHg.

Celiprolol 200 mg, 400 mg, 600 mg, atenolol 50 mg QD, Interventions placebo

Outcomes supine SBP, DBP, HR

Study Descriptions

Tibblin 1969 [98] Randomized double blinded placebo controlled crossover study Methods with two 12-week treatment periods.

n=11, baseline BP 179/115 mmHg. Patients with supine SBP >

150 mmHg, and supine DBP > 100 mmHg. Exclusion asthma, Participants chronic bronchitis, heart failure, retinopathy or renal

hypertension.

Interventions alprenolol 100 mg QID, placebo

179

Study Descriptions

Outcomes Sitting, standing, supine SBP, DBP, blood and urine lab.

Study Descriptions

Trafford 1989 After single blind run in period, patients entered a randomized

[99] Methods double blinded 3 way crossover study with 4 week treatment

periods.

Age 20-65, sitting DBP between 94 to 116 mmHg during the 4

week run in period. Exclusion: DBP> 120mmHg, malignant or

labile hypertension, MI, heart failure, sinus bradycardia, 2nd or Participants 3rd degree heart block, arrhythmia, COPD, insulin dependent

diabetes, uncontrolled renal, hepatic, GI, endocrine diseases,

contraindicated for beta blockers.

Interventions Celiprolol 200 mg QD, 400 mg QD and placebo

Sitting SBP, DBP, HR, blood, clinical chemistry, liver Outcomes function, and urinalysis.

Study Descriptions

Vandongen 1986 Randomized double blinded placebo controlled cross-over Methods [100] study with no washout and 3 week treatment periods.

Participants Age 29-65, 10 men 3 women, with primary hypertension.

Interventions pindolol 5 mg BID, metoprolol 100 mg BID, placebo

Rest SBP, DBP, HR, post exercise SBP, DBP, HR, plasma Outcomes noradrenaline and adrenaline.

Notes Baseline information of patients was not given.

Study Descriptions

180

Study Descriptions

Watson 1980 Randomized double blinded placebo controlled crossover study Methods [101] with 4 week washout and three 4-week treatment periods.

n=13, patients with BP greater than 150/100 mmHg were Participants enrolled. Exclusion: organ damage or asthma.

Interventions Placebo BID, Acebutolol 400 QD, Acebutolol 200 mg BID

Outcomes resting SBP, DBP, HR, serum drug levels

SBP: systolic blood pressure; DBP: diastolic blood pressure; HR: heart rate; WDAE: withdrawal due to adverse effects; AE: adverse effects; BID: twice daily; TID: three times daily: QID: four times daily; QD: once daily; CVD: cardiovascular disease; ABPM: ambulatory blood pressure measurements; Cmax: Maximum serum concentration; CHF: congested heart failure; MI: myocardial infarction; COPD: chronic obstructive pulmonary diseases; SD: standard deviation; MBP: mean blood pressure; SEM: standard error of mean; AV block: atrioventricular block; RBF: renal blood flow; RVR: renal vascular resistance; GFR: Glomerular Filtration Rate

181

Appendix F: Table of included studies in beta-1 selective blocker review

Study Descriptions

Ades 1990 [105] Randomized double blinded placebo controlled parallel study, Methods 11 week treatment period.

N=30, 22 men, mean age =47, mean baseline BP 145/94 Participants mmHg, primary hypertension

Interventions Metoprolol 100 mg QD, propranolol 80 mg QD, placebo

SBP, DBP, HR, exercise training, cardiac output, Outcomes echocardiogram, muscle biopsy

Study Descriptions

Ameling 1991 Randomized, double blinded, placebo controlled, cross over

[106] study. Patients were seen for the first 2 weeks and BP was

measured in every visit. Patients who were eligible, were

Methods randomized after week 2. Each treatment period lasted for 4

weeks and then patients cross-over to another treatment for

another 4 weeks. BP was measured in 2 weeks interval during

the treatment period.

182

Study Descriptions

Previously undiagnosed patients who had mean office

DBP>95mmHg in 2 office visits, during the first 2 weeks, were

included into the study. 331 patients (188, 56.8% male) were

randomized to betaxolol or placebo. Mean age = 50. Baseline

SBP = 167.4mmHg, DBP=104.7mmHg, HR=80.3 beats/min Participants Exclusion criteria: age <20 or >70; SBP>190mmHg or

DBP>125mmHg; use of any medication that may affect BP;

pregnancy; use of oral contraceptive; heart failure; bradycardia;

chronic heptic, renal or metabolic diseases; bronchial asthma;

chronic pulmonary diseases.

Participants were randomized to betaxolol 20mg QD fixed dose Interventions (n= 163) or placebo (n=168) for 4 weeks, and then cross over.

Outcomes QOL, SBP, DBP, HR, WDAE

SD of BP measurements are not reported. SD was imputed Notes using the weighted mean SD of studies in this review.

Study Descriptions

Asmar 1991 Randomized double blinded placebo controlled crossover Methods [107] study. 4 week treatment period.

N=14, 7 men, mean age= 53, baseline BP 156/101, primary Participants hypertensive patients.

Interventions bisoprolol 10 mg QD, placebo

Outcomes Resting BP, HR, pulse wave velocity determinations

Study Descriptions

183

Study Descriptions

Baez 1986 [108] Randomized double blinded placebo controlled parallel study. Methods 3 week run in, 10 week treatment periods.

N=36, 24 men, age 33-68. Exclusion: hematological, renal, Participants hepatic, gastrointestinal, autoimmune or cardiac diseases.

Atenolol 50, 100 mg QD (only 50 mg data used), doxazosin 16 Interventions mg QD, placebo

Outcomes SBP, DBP, lab analysis

Study Descriptions

Bateman 1979 Randomized double blinded placebo controlled crossover Methods [109] study, 4 week treatment periods.

N=15, 11 men, age range 31-64. Primary hypertension, without Participants history of gout, COPD, diabetes, ischemic heart disease.

Atenolol 100 mg QD, chlorthalidone 25 mg QD, combination Interventions of two drugs, placebo

Outcomes SBP, DBP, ABPM, biochemical analysis.

Study Descriptions

Bengtsson 1976 Randomized double blinded placebo controlled crossover Methods [110] study, 4 week treatment period.

N=24, all women, mean age is 56, mean baseline BP 183/107 Participants mmHg. Primary hypertension with normal renal function.

Interventions Metoprolol 40 mg TID or placebo

Outcomes SBP, DBP, HR, lab parameters.

Study Descriptions

184

Study Descriptions

Berglund 1985 Randomized, double blinded, parallel study, 4 week wash out, Methods [111] 4 week treatment.

N=31, 24 male, mean age = 51.4, BP range 150-200/100-115 Participants mmHg.

Interventions Pafenolol 25 or 50 mg QD, placebo.

Outcomes Casual BP, HR, 24 hour BP,

Study Descriptions

Broekman 1992 Randomized double blinded crossover study. 6 week treatment Methods [112] period.

Participants N=13, all men, mean age 51.2.

Interventions bisoprolol 5 mg QD, placebo

Outcomes SBP, DBP, HR, sexual information,

Notes Group 1 only

Study Descriptions

Chan 1991 [113] Randomized double blinded placebo controlled parallel study. Methods 4 week run in, 4 week treatment period.

N=29, 7 men, mean age =46, mean baseline BP 163/106

Participants mmHg, primary hypertension, normal renal function, without

insulin dependent diabetes and secondary hypertension.

Interventions Nebivolol 5 mg QD, placebo

Outcomes BP, HR biochemistry, CXR, ECG, QOL

Study Descriptions

Chan 1992 [114] Randomized double blinded, placebo controlled parallel study. Methods 4 week run-in, 4 week treatment period.

185

Study Descriptions

N=32, 8 men, mean age= 47, mean baseline BP= 164/105.8

Participants mmHg. Primary hypertension, without renal disease, diabetes,

secondary hypertension.

Interventions Nebivolol 5 mg QD, placebo

Rest SBP, DBP, HR, plasma biochemistry, plasma renin and Outcomes hormone level.

Study Descriptions

Chrysant 1992 Randomized double blinded placebo controlled parallel study, Methods [115] 3 week run in, 4 week treatment period.

N=256, 120 men, mean age 53, mean baseline BP 152.5/103

mmHg. Primary hypertension without renal failure, CHF, lung Participants diseases, liver diseases, blood diseases, as well as women in

child bearing age were excluded.

Atenolol 25, 50 mg QD, hydrochlorothiazide 25 Interventions mg/triamterene 50 mg QD, their combinations, placebo

Outcomes SBP, DBP, HR, WDAE

Notes WDAE did not reported according to treatment groups.

Study Descriptions

Cilliers 1979 Randomized double blinded placebo controlled crossover Methods [116] study, 4 week treatment periods.

N=50, 29 men, mean age 47, mean baseline BP 165.8/107.7 Participants mmHg. Primary hypertension, never been treated patients.

Interventions Atenolol 100 mg QD, placebo

Outcomes SBP, DBP, HR, WDAE

186

Study Descriptions

Dahlof 1986 Multicenter, randomized, double blinded, parallel studies, 4 Methods [117] week run in, 4 week treatment

N=23, 12 men, mean age= 49, primary hypertensive, mean BP Participants 164/109 mmHg

Interventions Pafenolol 50 or 100 mg QD, placebo

Outcomes Rest and exercise BP, HR, dropouts

Study Descriptions

Davidov 1994 Multi-center, randomized double blinded, placebo controlled Methods [118] parallel study. 4-6 week run in period, 4 week treatment period.

N=276, 70% men, mean age = 53, mean baseline BP 149/100

Participants mmHg, primary hypertension without any condition that might

influence blood pressure.

Interventions Bisoprolol 5 mg, 10 mg, 20 mg QD, placebo

Outcomes SBP, DBP, HR

Study Descriptions

Deary 2001 [119] Randomized double blinded placebo controlled crossover Methods study, 2 week run in, 6 week treatment

N=34, 25 men, mean age= 47, DBP between 95 and 110 Participants mmHg in 3 readings in 3 months.

Bisoprolol 5 mg QD, amlodipine 5 mg QD, doxazosin 1-4 mg

Interventions QD, lisinopril 2.5-10 mg QD, bendrofluazide 2.5 mg QD,

placebo.

Outcomes Resting BP, QOL, ABPM, LV function.

Study Descriptions

187

Study Descriptions

Deary 2002 [120] Randomized, double blinded, placebo controlled, crossover Methods study. 2 week run-in, 6 week treatment

Participants N=30, 22 men, DBP > 95 mmHg after run in.

Bisoprolol 5 mg QD, amlodipine 5 mg QD, doxazosin 4 mg Interventions QD, lisinopril 10 mg QD, bendrofluazide 2.5 mg QD, placebo.

Outcomes BP, HR, Central BP, Plasma ANP, pulse wave analysis.

Study Descriptions

Dhakam 2008 Radomized double blinded placebo controlled crossover study, Methods [121] 2 week run in, 5 week treatment period.

N=16, 10 men, mean age 70, mean baseline BP 158/84 mmHg.

Participants Never treated subjects without secondary hypertension,

diabetes, renal impairment.

Interventions Atenolol 50 mg QD, nebivolol 5 mg QD, placebo

SBP, DBP, HR, aortic augmentation index, biochemical Outcomes analysis.

Notes Isolated systolic hypertension

Study Descriptions

Frishman 1995 Multi-centre, randomized, double blinded placebo controlled Methods [122] parallel study. 4-6 week run in period, 4 week treatment period.

N=509, 245 men, mean baseline BP = 151/100 mmHg, primary

Participants hypertension, no more than 10% below or 35% above ideal

weight.

Bisoprolol 5 mg QD, hydrochlorothiazide 25 mg QD, B5/H25 Interventions combination, placebo

188

Study Descriptions

Outcomes SBP, DBP, HR, WDAE

WDAE was unclear about which treatment group did dropout Notes patients belong.

Study Descriptions

Frishman 2006 Randomized double blinded placebo controlled parallel study, Methods [123] 4-5 week run in, 9 week treatment period.

N=1092, 57% men, mean age= 54, mean baseline BP

Participants 152.6/99.9 mmHg, primary hypertension without previous

cardiovascular event or contraindication to beta blockers.

Metoprolol 25, 100, 400 mg QD, felodipine 2.5, 10, 20 mg, any

Interventions of the combination of metoprolol and felodipine listed or

placebo

Outcomes Rest trough SBP, DBP, safety

Study Descriptions

Frisk-Holmberg Randomized double blinded placebo controlled crossover Methods 1985 [124] study, 3 week run in, 6 week treatment periods.

Participants N=8, all men, primary hypertension.

Interventions Atenolol 100 mg BID, placebo

Outcomes Exercise/rest blood pressure and heart rate, metabolic effect

Notes Did not report SBP, DBP.

Study Descriptions

Gostick 1977 Randomized double blinded placebo controlled crossover Methods [125] study, 12 week run in, 4 week treatment period each.

189

Study Descriptions

N=98, 84 were available for analysis, mean age 50.1 year,

Participants mean baseline BP 165.6/101.9 mmHg. Mild to moderate

hypertension

Interventions Atenolol 50, 100, 200 mg QD, placebo

Outcomes SBP, DBP, HR, blood urea, uric acid, serum electrolytes.

Study Descriptions

Gudbjornsdottir Randomized double blinded placebo controlled crossover Methods 1997 [126] study, 3 weeks run in, 6-week treatment period.

N=7, 4 men, mean age= 52, mean baseline BP 154/97. Primary

Participants hypertension diagnosed for at least 1 year, normal ECG, blood

count, serum electrolyte, normal liver and renal function.

Interventions metoprolol 100 mg BID, placebo

Outcomes SBP, DBP, blood glucose, hormone

Study Descriptions

Hansson 1975 Randomized double blinded placebo controlled parallel study,

[127] Methods 16 week treatment periods (data for week 4, 8 and 12 was

reported and used in analysis)

N=45, 26 men, mean age 46, mean baseline BP 169/106 Participants mmHg. Primary hypertension.

Interventions Atenolol 50 mg BID, placebo

Outcomes SBP, DBP, HR, WDAE

Study Descriptions

Himmelmann Randomized double blinded placebo controlled crossover Methods 1996 [128] study. 4 week run in, 4 week treatment period.

190

Study Descriptions

N=14=5, 12 men, mean age 50, 95

Interventions Nebivolol 5 mg QD, placebo

Outcomes SBP, DBP, HR, ABPM, PVR, blood chemistry.

Study Descriptions

Houston 1990 Randomized double blinded placebo controlled parallel study, Methods [129] 2-4 week run in, 8 week treatment period.

N=92, 61% men, mean age 48, mean baseline BP 140/96.5

Participants mmHg. Primary hypertension without renal, hepatic disease,

neoplastic or peripheral vascular diseases.

Atenolol 25, 50 mg BID (only 50 mg BID result was Interventions available), clonidine 0.1, 0.2 mg BID, placebo

Outcomes SBP, DBP, serum lipid,

Study Descriptions

Jaattela 1990 Randomized double blinded placebo controlled parallel study,

[130] 2 weeks run in, 4-week treatment period. This publication Methods contained two studies with two sets of patients, we combined

their data as one RCTs.

N=99, 45 men, mean age=59, mean baseline BP 169/101

Participants mmHg, primary hypertension, without heart failure, diabetes

and asthma.

Interventions Metoprolol 50 mg QD, placebo

Outcomes SBP, DBP, HR, response rate, safety

Study Descriptions

191

Study Descriptions

Jeffer 1977 [131] Randomized double blinded placebo controlled crossover Methods study, 4 week run in, 4 week treatment period.

N= 21, 8 men, mean age 53, mean baseline BP 161.7/111.5 Participants mmHg.

Interventions Atenolol 50, 100, 200 mg QD, placebo

Outcomes SBP, DBP, HR

Study Descriptions

Lacourciere 1994 Randomized double blinded placebo controlled parallel study. Methods [132] 8 week run-in, 12 week treatment period.

N=240 (226 completed), 95

Nebivolol 1, 5, 10 mg QD, Hydrochlorothiazide 12.5, 25, mg Interventions QD, every possible combination of the two drugs, placebo.

SBP, DBP, lipids, lipoprotein, apolipoprotein, routine lab Outcomes parameters.

Study Descriptions

Lepantalo 1983 Randomized double blinded placebo controlled crossover Methods [69] and study. 4 week run in, 4 week treatment period.

Lepantalo 1985 N=34, 17 men, mean age=54.1, mean baseline BP 160/100, Participants [44] primary hypertension.

Interventions Metoprolol 100 mg BID, propranolol 80 mg BID, placebo

Outcomes Rest, exercise SBP, DBP, PVR

Lepantalo 1985 was a duplicate publication of Lepantalo 1983. Notes Study Descriptions

192

Study Descriptions

Lischner 1987 Randomized double blinded placebo controlled crossover Methods [134] study, 2 week run in, 4 week treatment periods.

N=128, 60 men, mean age 54 for women, 52 for men, mean

Participants baseline BP 167/111.3 mmHg. Primary hypertension without

hepatic or renal diseases, obesity.

Atenolol 100 mg QD, amiloride 5 mg/hydrochlorothiazide 50 Interventions mg QD, placebo

Outcomes SBP, DBP, HR

Data for women and men were reported separately. We

Notes combined the data according to the recommendation of

Cochrane handbook.

Study Descriptions

Mclean 1990 Randomized double blinded placebo controlled parallel study, Methods [135] 3 week run in, 6 week treatment period.

N=136, primary hypertensive patients. Exclusion: renal or

Participants hepatic impairment, COPD, diabetes, peripheral vascular

diseases, heart failure, heart block, MI.

Atenolol 50 mg QD, nitrendipine 20 mg QD, Atenolol Interventions 50/nitrendipine 20, placebo

Outcomes SBP, DBP, adverse effects.

Study Descriptions

Myers 1983 [136] Randomized double blinded placebo controlled crossover Methods study, 6 week treatment period.

193

Study Descriptions

N=12, 8 men, mean age 52.8, mean baseline BP 173/109

mmHg. Primary hypertension, without AV block, history of Participants hepatic or renal diseases, diabetes, alcoholism, asthma, MI,

stroke, CHF, bradycardia.

Interventions Atenolol 100 mg QD, hydrochlorothiazide 50 mg QD, placebo

Outcomes SBP, DBP, HR, biochemical test.

Study Descriptions

NEB-305 [137] Multi-center, randomized, double blinded, placebo controlled, Methods parallel study. 12 week treatment period.

N=807, 53.5% men, mean age =53.4, mean baseline BP 151/99

Participants mmHg. Primary hypertension, without asthma and COPD,

renal and liver diseases, insulin dependent diabetes.

Interventions Nebivolol 5, 10, 20 mg QD, placebo

Outcomes Rest peak/trough SBP, DBP, WDAE, response rate.

Study Descriptions

Okawa 1986 Randomized double blinded placebo controlled parallel study.

[138] After 4-week placebo washout, patients were randomized to

treatment. During the first two weeks of treatment, patients Methods randomized to high dose were force titrated to respective

dosage. After that, dose were fixed for the duration of next 6

weeks.

n=139 (87 male) Participants 100mmHg

194

Study Descriptions

Bevantolol 50mg BID, 100mg BID, 150mg BID, 200mg BID Interventions or placebo

Outcomes SBP, DBP, HR, WDAE, adverse effect, clinical lab

Notes 3 WDAE in bevantolol, 0 in placebo

Study Descriptions

Papademetriou Multicenter, randomized, double blinded, placebo controlled, Methods 2006 [139] parallel study. 4-5 week run in, 8 week treatment period

N=1571, 51% men, mean age=53, mean baseline BP 151/100

Participants mmHg. Primary hypertension without heart failure, recent MI,

contraindication for beta blockers.

Metoprolol 25, 50, 100, 200 mg QD, Hydrochlorothiazide 6.25, Interventions 12.5, 25 mg QD, any combination of the two drugs, placebo

Outcomes Trough SBP, DBP, hematology, lipid, urinalysis, ECG.

Study Descriptions

Petrie 1976 [51] Randomized double blinded placebo controlled crossover

Methods study. After 14 days of washout period, patients were

randomized to a sequence of 4 week treatments.

n=24 (13 males)

105

Exclusion: recent MI, evidence of heart failure, heart block, Participants gross ischamia, grade III or IV retinopathy, diabetes, gout,

impaired liver function, or creatinin clearance less than 60ml/

min or if they were on any other drug treatment.

195

Study Descriptions

methyldopa 250mg TID, practolol 200mg TID, propranolol

Interventions 80mg TID, methyldopa 250mg and propranolol 80mg TID,

methyldopa 250mg and practolol 200mg TID or placebo

Outcomes SBP, DBP, HR, QOL

SD were imputed using the average of other studies that Notes provided SD

Study Descriptions

Petrie 1980 [52] Randomized double blinded placebo controlled crossover Methods study, 4 week run in, 4 week treatment periods.

N=23, 12 men, mean age 40.9, mean baseline BP 155/109.4 Participants mmHg.

Atenolol 100 mg QD, oxprenolol 160 mg QD, propranolol 160 Interventions mg QD, placebo

Outcomes SBP, DBP

Study Descriptions

Rosen 1994 [55] Randomized double blinded placebo controlled crossover Methods study, 4 week run in, 4 week treatment periods.

N=21, all men, age range 31-70, mean baseline BP 157/101.9 Participants mmHg.

Atenolol 100 mg QD, propranolol 80 BID, alpha-methyldopa

Interventions 500 mg BID, hydrochlorothiazide/triamterene 50/25 mg BID,

placebo

Outcomes SBP, DBP, sleep assessment, sexual function questionnaire.

Notes 8 of 21 patients withdraw from the study.

196

Study Descriptions

Saunders 2007 Multi-center, randomized, double blinded, placebo controlled, Methods [140] parallel study. 4 week run-in, 12 week treatment period.

N=300, 45.3% men, mean age = 50.9, mean baseline BP

Participants 152.2/100.2 mmHg. Primary hypertensive, BMI<40, no

cardiovascular condition or diabetes.

Interventions Nebivolol 2.5, 5, 10, 20, 40 mg QD, placebo

Outcomes Rest peak/trough SBP, DBP, HR, ECG, lab parameters

Multi-center, randomized, double blinded, placebo controlled, Notes parallel study. 4 week run-in, 12 week treatment period.

Study Descriptions

Seedat 1980 Randomized double blinded placebo controlled crossover Methods [141] study, 4 week treatment period.

N=24, Afriacans with primary hypertension. Exclusion criteria:

Participants heart failure, asthma, MI, heart block, pregnancy, impaired

renal and hepatic function.

Atenolol 100 mg QD, chlorthalidone 25 mg QD, combination Interventions of two drugs, placebo.

Outcomes SBP, DBP, HR, plasma renin activity, serum potassium

Study Descriptions

Stornello 1991 Randomized double blinded placebo controlled crossover Methods [142] study, 3 week run in, 6 week treatment periods.

N=12, 4 men, mean age 51.9, primary hypertension with

Participants NIDD, persisitent aluminuria, no evidence of UTI, other kidney

disease.

197

Study Descriptions

Atenolol 50 mg QD, enalapril 20 mg QD, chlorthalidone 12.5 Interventions mg QD, placebo

SBP, DBP, HR, renal vascular resistance, filtration fraction, Outcomes blood chemistry.

Study Descriptions

Stumpe 1985 Randomized double blinded placebo controlled parallel study, Methods [143] 4 week run in, 4 week treatment periods.

Participants N=232. Exclusion: history of heart disease or asthma.

Interventions Atenolol 50 mg QD, celiprolol 200, 400, 600 mg QD, placebo

Outcomes SBP, DBP

Study Descriptions

Tonkin 1990 Randomized double blinded placebo controlled crossover Methods [144] study, 4 week run in, 4 week treatment periods.

N=15, 7 men, median age 61, primary hypertension patients.

Participants Exclusion criteria: diabetes, cardiac disease, COPD, impaired

renal or hepatic function.

Atenolol 50 mg QD, diltiazem 60 mg QD, combination of two Interventions drugs, placebo

Outcomes SBP, DBP, ECG, hematology, adverse events.

Study Descriptions

Tseng 1993 [145] Randomized double blinded placebo controlled parallel study. Methods 2-4 week run in period, 3 week treatment period

N=65, 35 men, mean age= 51, mean baseline BP 160/101 Participants mmHg, primary hypertension without any co-morbidity.

198

Study Descriptions

Interventions Bisoprolol 5 mg, 10 mg QD or placebo

DBP, HR, response rate, Biochemical hematological change, Outcomes ECG.

Study Descriptions

Van Bortel 1993 Randomized double blinded placebo controlled crossover Methods [146] study. 4 week run in, 4 week treatment period.

N=114, 76 men, mean age=54, 95

Participants hypertension, normal liver and renal function, no history of

heart failure, severe vascular disease, uncontrolled diabetes.

Interventions Nebivolol 5 mg QD, placebo

Outcomes BP, HR, QOL, ECG, lab parameters

Notes Only 80 patients randomized to crossover group.

Study Descriptions

Van Herwaarden Randomized double blinded placebo controlled crossover Methods 1977 [59] study, 4 week treatment period.

Participants N=8, mean age=34.

Interventions Metoprolol 100 mg TID, propranolol 80 mg TID, placebo

Outcomes Rest and adrenaline infused SBP, DBP, HR

Study Descriptions

Van Merode Randomized, double blinded, placebo controlled crossover Methods 1989 [147] study. 4 week run in, 4 week treatment period.

N=29, 21 men, primary hypertension with normal kidney and Participants liver functions.

Interventions Nebivolol 5 mg QD, placebo

199

Study Descriptions

Outcomes SBP, DBP, pulse pressure, carotid diameter and compliance

Study Descriptions

Van Neuten 1997 Multi-centered, randomized, double blinded, placebo

[148] Methods controlled, parallel study. 4 week run in, 4 week treatment

period.

N=509, 53% men, baseline mean BP 158.4/102.7 mmHg.

Exclusion criteria: secondary hypertension, asthma and COPD,

Participants bradycardia, AV block, insulin dependent diabetes, MI or

stroke in past 6 months, heart failure, renal hepatic disease,

50% overweight base on BMI.

Interventions Nebivolol 0.5, 1, 2.5, 5, 10 mg QD, placebo

Rest peak/trough BP, ECG, lab parameters, symptom score, Outcomes WDAE

Study Descriptions

Van Neuten 1998 Randomized double blinded placebo controlled parallel study, Methods [149] 4-week run in, 4-week treatment period.

N=364, 191 men, mean age = 55, mean baseline BP 166.5/104

Participants mmHg. Primary hypertension, previously untreated or

responded poorly to previous treatment.

Interventions Nebivolol 5 mg QD, Atenolol 50 mg QD, placebo

SBP, DBP, adverse event, ECG, blood chemistry, hematology, Outcomes urine test.

Study Descriptions

200

Study Descriptions

Vanhees 1991 Randomized double blinded placebo controlled crossover Methods [150] study, 3 week run in, 3 week treatment periods.

Participants N=15, all men, mean age 39.4, mean baseline BP 166/93.

Atenolol 50 mg QD, verapamil 240 mg QD, enalapril 10 mg Interventions QD, matching placebo

Exercise tolerance, rest SBP, DBP, exercise parameters, urine Outcomes and blood analysis

Study Descriptions

Verdecchia 1983 Randomized double blinded placebo controlled crossover Methods [151] study, 2 week placebo 4 week treatment period.

N=20, 14 men, mean age =44, mean baseline BP 165.5/108.3, Participants primary hypertension with normal ECG.

Interventions Metoprolol 200 mg QD, placebo

Outcomes SBP, DBP, HR, systolic time interval.

Study Descriptions

Verdecchia 1988 Randomized double blinded placebo controlled crossover Methods [152] study, 3 week run in, 4 week treatment periods.

Participants N=12, 5 men, mean age 48.6, primary hypertension.

Interventions Atenolol 100 QD, placebo

Outcomes SBP, DBP, HR, ABPM, WDAE

Study Descriptions

Weiss 2007 [153] Multi-center, randomized double blinded, placebo controlled, Methods parallel study. 2 week run-in, 12-week treatment period.

201

Study Descriptions

N=909, 57% men, mean age = 54.7, mean baseline BP

Participants 153.1/99.5 mmHg. Primary hypertensive, BMI<35, no history

of cardiovascular disease or diabetes.

Interventions Nebivolol 1.25, 2.5, 5, 10, 20, 40 mg QD, placebo

Outcomes Rest peak/trough SBP, DBP, HR, safety,

Not all patients in 40 mg group took 40 mg, therefore this Notes group was excluded.

Study Descriptions

Williams 1992 Randomized double blinded placebo controlled parallel study.

[154] Patients entered a 3 to 4 week placebo run in period, which

they were instructed to take the capsule at 10AM every Methods morning. Patients who met the entry criteria were randomized

to one of the 4 treatment groups for another 5 weeks. After the

fixed dose period, there was a 1 week follow up period.

n=317

Participants Patients with mean DBP between 95mmHg and 110mmHg

during the run in period were eligible for randomization.

Interventions Betaxolol 5mg, 10mg, 20mg QD or placebo.

Outcomes SBP, DBP, HR, withdrawal, adverse effects.

Study Descriptions

Wing 1988 [155] Randomized double blinded placebo controlled crossover Methods study, 4 week run in, 4 week treatment periods.

N=16, 6 men, mean age 59. Primary hypertension without Participants target organ damage.

202

Study Descriptions

Atenolol 50 mg QD, enalapril 20 mg QD, atenolol/enalapril Interventions combination, placebo

Outcomes SBP, DBP, HR, blood chemical analysis.

SBP: systolic blood pressure; DBP: diastolic blood pressure; HR: heart rate; WDAE: withdrawal due to adverse effects; AE: adverse effects; BID: twice daily; TID: three times daily: QID: four times daily; QD: once daily; CVD: cardiovascular disease; ABPM: ambulatory blood pressure measurements; Cmax: Maximum serum concentration; CHF: congested heart failure; MI: myocardial infarction; COPD: chronic obstructive pulmonary diseases; SD: standard deviation; MBP: mean blood pressure; SEM: standard error of mean; AV block: atrioventricular block; RBF: renal blood flow; RVR: renal vascular resistance; GFR: Glomerular Filtration Rate

203